The Geological Record of Neoproterozoic Glaciations (Geological Society of London Memoir 36)

The Geological Record of Neoproterozoic Glaciations

Geological Society Memoirs The Geological Society of London Books Editorial Committee Chief Editor

Bob Pankhurst (UK) Society Books Editors

John Gregory (UK) Jim Griffiths (UK) John Howe (UK) Howard Johnson (UK) Rick Law (USA) Phil Leat (UK) Nick Robins (UK) Randell Stephenson (UK) Society Books Advisors

Eric Buffetaut (France) Jonathan Craig (Italy) Tom McCann (Germany) Mario Parise (Italy) Satish-Kumar (Japan) Gonzalo Veiga (Argentina) Maarten de Wit (South Africa)

IUGS/GSL publishing agreement This volume is published under an agreement between the International Union of Geological Sciences and the Geological Society of London and arises from IGCP project number 512. GSL is the publisher of choice for books related to IUGS activities, and the IUGS receives a royalty for all books published under this agreement. Books published under this agreement are subject to the Society’s standard rigorous proposal and manuscript review procedures.

It is recommended that reference to all or part of this book should be made in one of the following ways: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) 2011. The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36. Hoffman, P. F., Macdonald, F. A. & Halverson, G. P. 2011. Chemical sediments associated with Neoproterozoic glaciation: iron formation, cap carbonate, barite and phosphorite. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 67–80.

GEOLOGICAL SOCIETY MEMOIR NO. 36

The Geological Record of Neoproterozoic Glaciations EDITED BY

EMMANUELLE ARNAUD University of Guelph, Canada

GALEN P. HALVERSON McGill University, Canada and

GRAHAM SHIELDS-ZHOU University College London, UK

2011 Published by The Geological Society London

THE GEOLOGICAL SOCIETY The Geological Society of London (GSL) was founded in 1807. It is the oldest national geological society in the world and the largest in Europe. It was incorporated under Royal Charter in 1825 and is Registered Charity 210161. The Society is the UK national learned and professional society for geology with a worldwide Fellowship (FGS) of over 10 000. The Society has the power to confer Chartered status on suitably qualified Fellows, and about 2000 of the Fellowship carry the title (CGeol). Chartered Geologists may also obtain the equivalent European title, European Geologist (EurGeol). One fifth of the Society’s fellowship resides outside the UK. To find out more about the Society, log on to www.geolsoc.org.uk. The Geological Society Publishing House (Bath, UK) produces the Society’s international journals and books, and acts as European distributor for selected publications of the American Association of Petroleum Geologists (AAPG), the Indonesian Petroleum Association (IPA), the Geological Society of America (GSA), the Society for Sedimentary Geology (SEPM) and the Geologists’ Association (GA). Joint marketing agreements ensure that GSL Fellows may purchase these societies’ publications at a discount. The Society’s online bookshop (accessible from www.geolsoc.org.uk) offers secure book purchasing with your credit or debit card. To find out about joining the Society and benefiting from substantial discounts on publications of GSL and other societies worldwide, consult www.geolsoc.org.uk, or contact the Fellowship Department at: The Geological Society, Burlington House, Piccadilly, London W1J 0BG: Tel. þ 44 (0)20 7434 9944; Fax þ 44 (0)20 7439 8975; E-mail: [email protected]. For information about the Society’s meetings, consult Events on www.geolsoc.org.uk. To find out more about the Society’s Corporate Affiliates Scheme, write to [email protected]. Published by The Geological Society from: The Geological Society Publishing House, Unit 7, Brassmill Enterprise Centre, Brassmill Lane, Bath BA1 3JN, UK (Orders: Tel. þ44 (0)1225 445046, Fax þ44 (0)1225 442836) Online bookshop: www.geolsoc.org.uk/bookshop The publishers make no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility for any errors or omissions that may be made. # The Geological Society of London 2011. All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with the provisions of The Copyright Licensing Agency Ltd, Saffron House, 6 –10 Kirby Street, London EC1N 8TS, UK. Users registered with the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, USA: the item-fee code for this publication is 0435-4052/11/$15.00. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN 978-1-86239-334-9 Distributors For details of international agents and distributors see: www.geolsoc.org.uk/agentsdistributors Typeset by Techset Composition Ltd, Salisbury, UK Printed by CPI Antony Rowe, Chippenham, UK

Mexico Hut, 1978. Clockwise from the back left: Nick Cox (Motorboat Skipper and Engineer), Tim Druitt (Geologist), Robin (“Bruce”) Davies (Motorboat Skipper and Engineer), Brian Harland (Expedition Leader), James Carter ˚ lesund, Svalbard (undergraduate field assistant), Mike Hambrey (Geologist) and Paul Waddams (PhD Student), Ny-A

The editors dedicate this Memoir to Mike Hambrey and Brian Harland for their pioneering work in Neoproterozoic glacial geology and for their comprehensive 1981 volume Earth’s Pre-Pleistocene Glacial Record, which inspired the present work.

Contents List of Reviewers

ix

Introductory chapters ARNAUD, E., HALVERSON, G. P. & SHIELDS-ZHOU, G. The geological record of Neoproterozoic ice ages

1

HOFFMAN, P. F. A history of Neoproterozoic glacial geology, 1871– 1997

17

ARNAUD, E. & ETIENNE, J. L. Recognition of glacial influence in Neoproterozoic sedimentary successions

39

HALVERSON, G. P. & SHIELDS-ZHOU, G. Chemostratigraphy and the Neoproterozoic glaciations

51

HOFFMAN, P. F., MACDONALD, F. A. & HALVERSON, G. P. Chemical sediments associated with Neoproterozoic glaciation: iron formation, cap carbonate, barite and phosphorite

67

BAHLBURG, H. & DOBRZINSKI, N. A review of the Chemical Index of Alteration (CIA) and its application to the study of Neoproterozoic glacial deposits and climate transitions

81

EVANS, D. A. D. & RAUB, T. D. Neoproterozoic glacial palaeolatitudes: a global update

93

GREY, K., HILL, A. C. & CALVER, C. Biostratigraphy and stratigraphic subdivision of Cryogenian successions of Australia in a global context

113

CONDON, D. J. & BOWRING, S. A. A user’s guide to Neoproterozoic geochronology

135

GODDE´RIS, Y., LE HIR, G. & DONNADIEU, Y. Modelling the Snowball Earth

151

Africa SHIELDS-ZHOU, G. A., DEYNOUX, M. & OCH, L. The record of Neoproterozoic glaciation in the Taoude´ni Basin, NW Africa

163

MASTER, S. & WENDORFF, M. Neoproterozoic glaciogenic diamictites of the Katanga Supergroup, Central Africa

173

TAIT, J., DELPOMDOR, F., PRE´AT, A., TACK, L., STRAATHOF, G. & NKULA, V. K. Neoproterozoic sequences of the West Congo and Lindi/Ubangi Supergroups in the Congo Craton, Central Africa

185

HOFFMAN, P. F. Glaciogenic and associated strata of the Otavi carbonate platform and foreslope, northern Namibia: evidence for large base-level and glacioeustatic changes

195

PRAVE, A. R., HOFFMANN, K.-H., HEGENBERGER, W. & FALLICK, A. E. The Witvlei Group of East-Central Namibia

211

FRIMMEL, H. E. The Chameis Gate Member, Chameis Group, Marmora Terrane, Namibia

217

FRIMMEL, H. E. The Kaigas and Numees formations, Port Nolloth Group, in South Africa and Namibia

223

FRIMMEL, H. E. The Karoetjes Kop and Bloupoort formations, Gifberg Group, South Africa

233

Eurasia –Nubian Shield ALLEN, P. A., RIEU, R., ETIENNE, J. L., MATTER, A. & COZZI, A. The Ayn Formation of the Mirbat Group, Dhofar, Oman

239

ALLEN, P. A., LEATHER, J., BRASIER, M. D., RIEU, R., MCCARRON, M., LE GUERROUE´, E., ETIENNE, J. L. & COZZI, A. The Abu Mahara Group (Ghubrah and Fiq formations), Jabal Akhdar, Oman

251

MILLER, N. R., AVIGAD, D., STERN, R. J. & BEYTH, M. The Tambien Group, Northern Ethiopia (Tigre)

263

STERN, R. J., JOHNSON, P. R., ALI, K. A. & MUKHERJEE, S. K. Evidence for Early and Mid-Cryogenian glaciation in the Northern Arabian –Nubian Shield (Egypt, Sudan, and western Arabia)

277

CHUMAKOV, N. M. Glacial deposits of the Bokson Group, East Sayan Mountains, Buryatian Republic, Russian Federation

285

CHUMAKOV, N. M. The Neoproterozoic glacial formations of the North and Middle Urals

289

CONTENTS

vii

CHUMAKOV, N. M. Glacial deposits of the Nichatka Formation, Chara River basin and review of Upper Precambrian diamictites of Central Siberia

297

CHUMAKOV, N. M. Glacial deposits of the Baykonur Formation, Kazakhstan and Kyrgyzstan

303

CHUMAKOV, N. M., POKROVSKY, B. G. & MELEZHIK, V. A. The glaciogenic Bol’shoy Patom Formation, Lena River, central Siberia

309

SOVETOV, J. K. Late Cryogenian (Vendian) glaciogenic deposits in the Marnya Formation, Oselok Group, in the foothills of the East Sayan Range, southwestern Siberian Craton

317

MACDONALD, F. A. The Tsagaan Oloom Formation, southwestern Mongolia

331

MACDONALD, F. A. & JONES, D. S. The Khubsugul Group, Northern Mongolia

339

ETIENNE, J. L., ALLEN, P. A., LE GUERROUE´, E., HEAMAN, L., GHOSH, S. K. & ISLAM, R. The Blaini Formation of the Lesser Himalaya, NW India

347

ZHANG, Q.-R., CHU, X.-L. & FENG, L.-J. Neoproterozoic glacial records in the Yangtze Region, China

357

ZHU, M. & WANG, H. Neoproterozoic glaciogenic diamictites of the Tarim Block, NW China

367

North America MACDONALD, F. A. The Hula Hula Diamictite and Katakturuk Dolomite, Arctic Alaska

379

MACDONALD, F. A. & COHEN, P. A. The Tatonduk inlier, Alaska –Yukon border

389

HOFFMAN, P. F. & HALVERSON, G. P. Neoproterozoic glacial record in the Mackenzie Mountains, northern Canadian Cordillera

397

SMITH, M. D., ARNAUD, E., ARNOTT, R. W. C. & ROSS, G. M. The record of Neoproterozoic glaciations in the Windermere Supergroup, southern Canadian Cordillera

413

LINK, P. K. & CHRISTIE-BLICK, N. Neoproterozoic strata of southeastern Idaho and Utah: record of Cryogenian rifting and glaciation

425

LUND, K., ALEINIKOFF, J. N. & EVANS, K. V. The Edwardsburg Formation and related rocks, Windermere Supergroup, central Idaho, USA

437

MROFKA, D. & KENNEDY, M. The Kingston Peak Formation in the eastern Death Valley region

449

PETTERSON, R., PRAVE, A. R. & WERNICKE, B. P. Glaciogenic and related strata of the Neoproterozoic Kingston Peak Formation in the Panamint Range, Death Valley region, California

459

CARTO, S. L. & EYLES, N. The deep-marine glaciogenic Gaskiers Formation, Newfoundland, Canada

467

CARTO, S. L. & EYLES, N. The Squantum Member of the Boston Basin, Massachusetts, USA

475

South America CHEW, D. & KIRKLAND, C. The Chiquerı´o Formation, southern Peru

481

ALVARENGA, C. J. S., BOGGIANI, P. C., BABINSKI, M., DARDENNE, M. A., FIGUEIREDO, M. F., DANTAS, E. L., UHLEIN, A., SANTOS, R. V., SIAL, A. N. & TROMPETTE, R. Glacially influenced sedimentation of the Puga Formation, Cuiaba´ Group and Jacadigo Group, and associated carbonates of the Araras and Corumba´ groups, Paraguay Belt, Brazil

487

FIGUEIREDO, M. F., BABINSKI, M. & ALVARENGA, C. J. S. The Serra Azul Formation, Paraguay Belt, Brazil

499

GUIMARA˜ES, J. T., MISI, A., PEDREIRA, A. J. & DOMINGUEZ, J. M. L. The Bebedouro Formation, Una Group, Bahia (Brazil)

503

MISI, A., KAUFMAN, A. J., AZMY, K., DARDENNE, M. A., SIAL, A. N. & DE OLIVEIRA, T. F. Neoproterozoic successions of the Sa˜o Francisco Craton, Brazil: the Bambuı´, Una, Vazante and Vaza Barris/Miaba groups and their glaciogenic deposits

509

PEDROSA-SOARES, A. C., BABINSKI, M., NOCE, C., MARTINS, M., QUEIROGA, G. & VILELA, F. The Neoproterozoic Macau´bas Group, Arac¸uaı´ orogen, SE Brazil

523

ROCHA-CAMPOS, A. C., DE BRITO NEVES, B. B., BABINSKI, M., DOS SANTOS, P. R., DE OLIVEIRA, S. M. B. & ROMANO, A. Moema laminites: a newly recognized Neoproterozoic (?) glaciogenic unit, Sa˜o Francisco Basin, Brazil

535

viii

CONTENTS

UHLEIN, A., ALVARENGA, C. J. S., DARDENNE, M. A. & TROMPETTE, R. R. The glaciogenic Jequitaı´ Formation, southeastern Brazil

541

PAZOS, P. J., RAPALINI, A. E., BETTUCCI, L. S. & TO´FALO, O. R. The Playa Hermosa Formation, Playa Verde Basin, Uruguay

547

PECOITS, E., GINGRAS, M. K. & KONHAUSER, K. O. Las Ventanas and San Carlos formations, Maldonado Group, Uruguay

555

PAZOS, P. J. & RAPALINI, A. The controversial stratigraphy of the glacial deposits in the Tandilia System, Argentina

565

Europe HALVERSON, G. P. Glacial sediments and associated strata of the Polarisbreen Group, northeastern Svalbard

571

STOUGE, S., CHRISTIANSEN, J. L., HARPER, D. A. T., HOUMARK-NIELSEN, M., KRISTIANSEN, K., MACNIOCAILL, C. & BUCHARDT-WESTERGA˚RD, B. Neoproterozoic (Cryogenian–Ediacaran) deposits in East and North-East Greenland

581

RICE, A. H. N., EDWARDS, M. B., HANSEN, T. A., ARNAUD, E. & HALVERSON, G. P. Glaciogenic rocks of the Neoproterozoic Smalfjord and Mortensnes formations, Vestertana Group, E. Finnmark, Norway

593

STODT, F., RICE, A. H. N., BJO¨RKLUND, L., BAX, G., HALVERSON, G. P. & PHARAOH, T. C. Evidence of late Neoproterozoic glaciation in the Caledonides of NW Scandinavia

603

NYSTUEN, J. P. & LAMMINEN, J. T. Neoproterozoic glaciation of South Norway: from continental interior to rift and pericratonic basins in western Baltica

613

KUMPULAINEN, R. A. & GREILING, R. O. Evidence for late Neoproterozoic glaciation in the central Scandinavian Caledonides

623

KUMPULAINEN, R. A. The Neoproterozoic glaciogenic Lillfja¨llet Formation, southern Swedish Caledonides

629

ARNAUD, E. & FAIRCHILD, I. J. The Port Askaig Formation, Dalradian Supergroup, Scotland

635

PRAVE, A. R. & FALLICK, A. E. The Neoproterozoic glaciogenic deposits of Scotland and Ireland

643

Australia CALVER, C. R. Neoproterozoic glacial deposits of Tasmania

649

CORKERON, M. Neoproterozoic glacial deposits of the Kimberly Region and northwestern Northern Territory, Australia

659

GOSTIN, V. A., MCKIRDY, D. M., WEBSTER, L. J. & WILLIAMS, G. E. Mid-Ediacaran ice-rafting in the Adelaide Geosyncline and Officer Basin, South Australia

673

HILL, A. C., HAINES, P. W. & GREY, K. Neoproterozoic glacial deposits of central Australia

677

JENKINS, R. J. F. Billy Springs glaciation, South Australia

693

PREISS, W. V., GOSTIN, V. A., MCKIRDY, D. M., ASHLEY, P. M., WILLIAMS, G. E. & SCHMIDT, P. W. The glacial succession of Sturtian age in South Australia: the Yudnamutana Subgroup

701

WILLIAMS, G. E., GOSTIN, V. A., MCKIRDY, D. M., PREISS, W. V. & SCHMIDT, P. W. The Elatina glaciation (late Cryogenian), South Australia

713

Index

723

List of Reviewers Philip Allen Imperial College London, UK

Nick Eyles University of Toronto at Scarborough, Canada

Carlos Alvarenga University of Brasilia, Brazil

Ian Fairchild University of Birmingham, UK

Jose-Javier Alvaro Centro de Astrobiologia, Madrid, Spain

Mikhail Fedonkin Paleontological Institute, Moscow, Russia

John Arthurs Holywood, County Down, Ireland

Eric Font University of Lisbon, Portugal

Dov Avigad The Hebrew University of Jerusalem, Israel

Hartwig Frimmel University of Wu¨rzburg, Germany

Marly Babinsky University of Sao Paulo, Brazil

Claudio Gaucher Universidad de la Repu´blica, Uruguay

Heinrich Bahlburg Westfa¨lische Wilhelms-Universita¨t, Muenster, Germany

Vic Gostin University of Adelaide, Australia

Richard Bailey Northeastern University, USA

Dimitriy Grazhdankin Institute of Petroleum Geology and Geophysics, Novosibirsk, Russia

Dhiraj Banerjee University of Delhi, India Angelo Miguel Stipp Basei University of Sao Paulo, Brazil Julie Bartley State University of West Georgia Nicholas Beukes University of Johannesburgh, South Africa Knut Bjorlykke University of Oslo, Norway Robert Blodgett United States Geological Survey, USA Nick Butterfield Cambridge, UK Clive Calver Mineral Resources, Tasmania Nikolay Chumakov Geological Institute, Russian Academy of Sciences Allan Collins University of Adelaide, Australia Maurice Colpron Yukon Geological Survey, Canada Frank Corsetti University of Southern California, USA Lucieth Cruz Viera University of Sao Paulo, Brazil Bley Benjamin de Brito Neves University of Sao Paulo, Brazil Max Deynoux CNRS, Strasbourg, France Nicole Dobrzinsky Panterra Geoconsultants B. V., The Netherlands Eugene Domack Hamilton College, USA

Mike Hambrey Aberystwyth University, Wales, UK Richard Hanson Texas Christian University, USA Christoph Heubeck Freie Universita¨t Berlin, Germany Andrew Hill Centro de Astrobiologia, Madrid, Spain Rick Hiscott Memorial University, Canada Paul Hoffman Harvard University, USA/University of Victoria, Canada John Howe Scottish Marine Institute, UK Ganqing Jiang University of Las Vegas, USA Peter Johnson Washington, DC, USA Sandra Kamo University of Toronto, Canada A. F. Kamona University of Namibia, Namibia Alan Jay Kaufman University of Maryland, USA Brian Kendall Arizona State University, USA Andy Knoll Harvard University, USA Risto Kumpulainen Stockholm University, Sweden Alexandre Kuzmichev Geological Institute, Russian Academy of Sciences, Russia

James Etienne Neftex, UK

Anton Kuznetsov Institute of Precambrian Geology and Geochronology, St. Petersburg, Russia

David J. A. Evans Durham University, UK

Erwan Le Guerroue´ Universite´ de Rennes, France

x

LIST OF REVIEWERS

Zheng-Xiang Li Curtin University of Technology, Australia

Victor Ramos University of Buenos Aires, Argentina

John Lindsay Universities Space Research Association, USA

Pete Reid Petratherm, Australia

Paul Link Idaho State University, USA

Claudio Riccomini University of Sao Paulo, Brazil

Karen Lund United States Geological Survey, USA Francis Macdonald Harvard University, USA Adam Maloof Princeton University, USA Marcelo A. Martins Neto University of Ouro Preto, Brazil

Hugh Rice University of Vienna, Austria Ruben Rieu Repsol YPF, Exploration and Production, Spain Anthony Rocha-Campos University of Sao Paulo, Brazil

Andrey Maslov Institute of Geology and Geochemistry, Ekaterinburg, Russia

Julius Sovetov Institute of Oil and Gas Geology and Geophysics, Novosibirsk, Russia

Bill McClelland University of Idaho, USA

Anthony Spencer Statoil, Norway

Roy Miller Windhoek, Namibia

Doug Sprinkel Utah Geological Survey, USA

Aroldo Misi Federal University of Bahia, Brazil

Robert Strachan University of Portsmouth, UK

Malgorzata Moczydlowska-Vidal Uppsala University, Sweden

Nick Swanson-Hysell Princeton University, USA

Paul Myrow Colorado College, USA Afonso Noguiera Universidade Federal do Para´, Brazil Robert Pankhurst British Geological Survey, UK Sandra Passchier Montclair University, USA Pablo Pazos University of Buenos Aires, Argentina Vic Pease University of Stockholm, Sweden Sergei Pisarevsky The University of Western Australia, Australia Boris Pokrovsky Geological Institute, Moscow, Russia Michael Pope Washington State University, USA

Eric Thover University of Western Australia Elizabeth Turner Laurentian University, Canada Rob Van der Voo University of Michigan, USA Pat Vickers-Rich Monash University, Australia Hilmar von Eynatten University of Go¨ttingen, Germany Stephen Warren University of Washington, USA George Williams University of Adelaide, Australia Gary Yeo Denison Mines Corporation, Canada

Susannah Porter University of California-Santa Barbara, USA

Grant Young University of Western Ontario, Canada

Anthony Prave University of St. Andrews, UK

Chuanming Zhou Nanjing Institute of Geology and Paleontology, Chinese Academy of Science, China

Wolfgang Preiss Department of Primary Industries and Resources of South Australia, Australia

Udo Zimmermann University of Stavanger, Norway

Chapter 1 The geological record of Neoproterozoic ice ages EMMANUELLE ARNAUD1*, GALEN P. HALVERSON2,3 & GRAHAM SHIELDS-ZHOU4 1

School of Environmental Sciences, University of Guelph, Guelph, Ontario N1G 2W1, Canada

2

School of Earth and Environmental Sciences, The University of Adelaide, North Terrace, Adelaide, SA 5005, Australia

3

Present address: Department of Earth and Planetary Sciences, McGill University, 3450 University Street, Montreal, Quebec, H3A 2A7, Canada 4

Department of Earth Sciences, University College London, Gower Street, London WC1E 6BT, UK *Corresponding author (e-mail: [email protected])

The IUGS- and UNESCO-funded International Geoscience Programme Project #512 (Neoproterozoic Ice Ages) was conceived to contribute towards a global synthesis of current geological data on the number, duration, extent, causes and consequences of glacial episodes during the Neoproterozoic Era. IGCP 512 attracted more than 200 scientists from over 30 countries, many of whom provided their regional and specialist expertise on Neoproterozoic successions around the world to the realization of this volume. IGCP 512 focused on integrating various aspects of Neoproterozoic geology: geochronology, geochemistry, sedimentary geology, biostratigraphy, palaeomagnetism and economic geology. At its inaugural meeting on 27 August 2005 during the International Association of Sedimentology conference on glacial processes and products in Aberystwyth, Wales, IGCP 512 members decided to produce a volume that summarized existing data sets in a form similar to Earth’s Pre-Pleistocene Glacial Record by Hambrey & Harland (1981). An enormous amount of work has been carried out in the 12 years since the publication of Hoffman et al.’s (1998) paper on the Snowball Earth hypothesis for Neoproterozoic glaciation (Fairchild & Kennedy 2007). The Snowball Earth hypothesis and, more generally, Neoproterozoic climate, have been the topic of numerous special volumes, special sessions, a dedicated conference in Ascona (Switzerland) in 2006 (Shields 2006), and numerous documentaries. Motivated by this intense worldwide interest in the Neoproterozoic glaciations and an exploding body of research into the topic, this volume synthesizes the state-of-the-art in this now highly multidisciplinary research field. It is intended to facilitate the integration of data sets, inspire new research projects, and inform ongoing work into the definition and subdivision of the Neoproterozoic timescale, including selection of the Global Stratotype Section and Point (GSSP) for the base of the Cryogenian Period. Despite such lofty aims, any book such as this cannot claim to be complete, and there are indeed many gaps in our knowledge and also in this book’s coverage, some of which are outlined in this Introduction and throughout the volume.

Book organization, format and terminology This book contains ten introductory overview chapters followed by 60 site- or succession-specific chapters. The multidisciplinary overview chapters provide reviews of the study and interpretation of Neoproterozoic glaciations. The first chapter by Hoffman reviews the history of research (1871 – 1997) into late Precambrian glaciations from the first recorded discovery of Neoproterozoic glaciogenic rocks at Port Askaig on Islay, SW Scotland in 1871

to just before the renewal of interest in their significance after 1998. Arnaud & Etienne provide a ‘user’s guide’ to the identification of glacial influence in the rock record, with emphasis on the processes and sedimentary products found in various glaciated basins as well as some common issues encountered with determining the palaeoclimatic significance of commonly used indicators of glacial palaeoenvironmental conditions. Geochemistry, in particular, isotope chemostratigraphy, has been key to the recent revival of interest in Neoproterozoic climate change and underpins models of both glaciation and global correlations. Halverson & Shields-Zhou review Neoproterozoic chemostratigraphic records with a focus on how they have been applied to palaeoenvironmental studies and to constrain both the number and relative timing of glacial events. Hoffman et al. then review the occurrence of chemical sediments and their depositional environments associated with the Neoproterozoic glaciations. In particular, the role of the enigmatic, post-glacial ‘cap carbonates’ and the reoccurrence of iron formations after a billion-year hiatus are explored here. Climate affects how rocks are weathered, and this in turn influences the global carbon cycle, a key factor in modulating global climate change. Bahlburg & Dobrzinski offer in this regard a more specific review of the application of the Chemical Index of Alteration (CIA) to Neoproterozoic glacial deposits. If the Snowball Earth hypothesis remains contentious, the following chapter should at least dispel any doubt that Neoproterozoic glaciation reached the tropics. In their review of palaeolatitudes of Neoproterozoic glacial deposits, Evans & Raub contribute a comprehensive compilation of locations and available palaeomagnetic data for Neoproterozoic glacial deposits worldwide. Condon & Bowring also provide a much needed user’s guide to Neoproterozoic geochronology, with emphasis on the most commonly used techniques, their strengths and weaknesses and the source of uncertainties that should be considered when using these data to constrain the timing of climatic changes. In addition, Condon & Bowring provide a summary of the key geochronological constraints on the timing and duration of glaciations and other significant Neoproterozoic events related to isotopic excursions and biological evolution. Grey et al. tackle the Neoproterozoic fossil record, and in particular the question of how climate change affected life on Earth. In their detailed review, they formulate a Cryogenian biostratigraphy based on Australian data, which can potentially be used as a template for such studies, generally still in their infancy, in other regions and for other Proterozoic time periods. In the final overview chapter, Godde´ris et al. provide a critical review of the widely differing climate models that have been used to investigate the onset and melting of a Snowball Earth and highlight that whereas initiation of a Snowball Earth is

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 1 –16. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.1

2

E. ARNAUD ET AL.

not difficult to accomplish in many models, how the Earth escapes from the icy grip of a snowball requires more study from a modelling perspective. The remaining chapters, organized by current geography, address specific sites from around the world where Neoproterozoic deposits have been studied. The purpose of these chapters is to provide a summary of available data and key references; they are not comprehensive reviews. Each chapter was intended to follow a consistent format in order to cover concisely the following topics and to facilitate cross chapter comparisons: structural framework, regional stratigraphy, sedimentary characteristics of glacial strata and associated deposits, boundary relations with non-glacial strata, chemostratigraphy, palaeomagnetism and palaeogeography, geochronology and a discussion outlining interpretations that can be inferred from these data. The structural framework section was designed to include a description of the overall structural and tectonic setting (such as cratons, types of sedimentary basin, regional-scale folds and faults) as well as the degree to which the sections have been modified by post-depositional tectonism and metamorphism. Authors were also asked to discuss the history of basin development in the region. The stratigraphy section was meant to provide an overview of the relevant stratigraphic units with comments about any lateral variations. The glaciogenic and associated strata section was designed to include descriptions of typical sedimentary characteristics of the glaciogenic and associated strata such as ironstones and carbonates, with the nature of contact with the overlying and underlying

non-glacial units described in the following boundary relations section. The chemostratigraphy section was meant to include geochemical data within a stratigraphic context, including CIA values, d13C, d18O, d34S, 87Sr/86Sr, and other available isotopic data. Some authors also included analyses of Rare Earth Elements, whole rock geochemistry, major and trace elements, Nd data, and Ce, Eu and Ir anomalies. The other characteristics section was designed to cover economic deposits, biomarkers or any other notable feature not covered in the other sections. The palaeolatitude and palaeogeography section was primarily designed to discuss the palaeomagnetic data available for that succession in order to consider palaeogeographic location of the deposits at the time of deposition. Some authors also included provenance data in the context of palaeogeography based on lithology or geochemistry. The geochronological constraints section was designed to report any available radiometric or biostratigraphic data that constrains the minimum or maximum age of the succession. These dates could come from the glacial deposits themselves or from associated strata. Authors were also asked to discuss regional stratigraphic correlations based on radiometric data, chemostratigraphy, lithostratigraphy and/or biostratigraphy. Global stratigraphic correlations were discouraged as these are discussed in the introductory chapters by Condon & Bowring (geochronology) and Halverson & Shields-Zhou (chemostratigraphy). Lastly, the discussion section was designed to include interpretations of available datasets with respect to palaeoenvironmental conditions, timing of climate change and palaeogeography, showing how

Fig. 1.1. Map showing distribution of Neoproterozoic glaciogenic successions within Africa, as covered in this volume. (A) Taoudeni Basin, NW Africa; (B) Katanga Supergroup, central Africa; (C) West Congo and Lindi/Ubangi Supergroups, central Africa; (D) Otavi Group, northern Namibia; (E) Witvlei Group, East-Central Namibia; (F) The Chameis Gate Member, Namibia; (G) The Kaigas and Numees formations, South Africa and Namibia, (H) Karoetjes Kop and Bloupoort formations, South Africa.

Table 1.1. Summary of Neoproterozoic data sets from Africa Lead author

Glaciogenic strata

Structural framework

Chemostratigraphy d13C

d18O

d34S

Y

Y

87

Sr/86Sr

Foreland basin

Y

Synrift, foreland basin

Y

Passive margin

Y

Hoffman

Lower and Upper Diamictite Formations, West Congo Supergroup; diamictites of the Lindi/Ubangi Supergroup Chuos Fm., Ghaub Fm.

Passive margin

Y

Y

Prave Frimmel

Blaubeker Fm. Kaigas Fm., Numees Fm.

Y Y

Y

Y

Frimmel

Chameis Gate Mb., Chameis Group Karoetjes Kop Fm.; Bloupoort Fm.

Rift to passive Failed rift; passive; back-arc? Back-arc?

Y

Y

Y

Rift; pasive margin?

Y

Y

Y

Master

Tait

Frimmel

Economic

Palaeomagnetism

Other

Y

Ir

U–Pb detrital Scarce

Barite, BIF, petroleum source rocks

Y

Fe, Cu– Co, Zn– Pb–Cu –Ge– Ga, Cu –Au, Ba–Fe, Pb –Zn

Y

Y

Y

Y

Y

Y

[Fe, Mn, Ba], d11B, d44Ca

Y

Other

K– Ar, Rb– Sr

Ar/Ar

Y Y Y

Limited trace elements Limited trace elements

U– Pb direct Y

Y

Y

Geochronology

INTRODUCTION

Basal diamictite of the triad sequence, Taoudeni Basin, diamictite of the triad sequence, Volta Basin Grand Conglomerat Fm., Petit Conglomerat Fm.

Shields-Zhou

Geobiology

Y reset Pb, Zn

Y

Y Y

Pb/Pb, Ar/Ar Pb/Pb, Ar/Ar

Calcitic marble; Cu, Fe, Mn; Fe oxides

Note: ‘U –Pb direct’ means dates from volcanic intrusions or tuffs anywhere in the succession, not necessarily from the glaciogenic strata itself. ‘U– Pb detrital’ means dates from detrital zircons from within the succession. Data in presumed correlative successions were not included.

3

4

E. ARNAUD ET AL.

the tectonic, stratigraphic, sedimentological, isotopic, palaeomagnetic and geochronological data outlined in previous sections support such interpretations. Whereas authors clearly favoured certain interpretations, they were encouraged to discuss competing interpretations and continued controversies. In an attempt to avoid redundancy and maintain consistency, photographs were omitted from site-specific chapters, and are made available in a companion online photo atlas (http://neoproterozoic-glaciations.weebly. com). Although this format was generally followed, some flexibility was required to accommodate the complexity of certain areas and the widely differing perspectives that reflect the diversity of authors contributing to this volume. Authors were asked to avoid interpretive terms, such as tillite, varvite, and cap carbonate, in preference to descriptive terms, except in the discussion section where interpretations are presented. More specifically, the non-genetic term diamictite was meant to be used to describe poorly sorted materials that contain a mixture of gravel-, sand- and mud-sized particles. For example, the term tillite was meant to be used in the discussion section in referring to poorly-sorted materials that were demonstrably deposited by ice without subsequent disaggregation and flow. The use of the term varvites or varves for rhythmically laminated mudstone that record seasonal fluctuations in ice cover was discouraged considering the lack of chronological control on

Neoproterozoic successions and the inability to demonstrate seasonal cyclicity. In the discussion section, authors were asked, where possible, to distinguish between the cap carbonate sequence that overlies the glacial deposits and the cap dolostone that occupies the basal transgressive tract of the post-glacial sequence. Although the focus of the book is on Cryogenian glaciation, these glacial deposits and associated strata are best understood within the context of their overall stratigraphic successions. Therefore, several site chapters also review associated later Neoproterozoic, Ediacaran-age deposits to highlight the carbonate strata associated with the late Cryogenian glacial deposits and the evolution of early animals that followed the late Cryogenian and mid-Ediacaran glaciations. Every effort was made to include all the sites where Neoproterozoic glaciogenic successions have been studied. In some cases, multiple distinct glaciogenic successions within a single region were grouped into a single chapter, whereas in others, they were treated in separate chapters. For the Adelaide rift-basin in South Australia, where extensive work has been done on each of the older (Sturtian and equivalent) and younger (Elatina and equivalent) glacial successions, the reviews are split into two chapters – one for each glaciation. Unfortunately, not every site known to preserve evidence of Neoproterozoic glaciation is included in this volume. These

50

F

I G J E L K

H

D

40

30

O C

M

B

20

N

A

10

N 0 1000

0

2000

km 40

50

60

70

80

90

100

110

120

130

Fig. 1.2. Map showing distribution of Neoproterozoic glaciogenic successions within Eurasia and the Nubian Shield, as covered in this volume. (A) Ayn Fm., Oman; (B) Abu Mahara Group, Oman; (C) Tambien Group, northern Ethiopia; (D) northern Arabian– Nubian Shield (Egypt, Sudan and western Arabia); (E) Bokson Group, East Sayan Mountains, Russian Federation; (F) North and Middle Urals, Russian Federation; (G) Nichatka Fm., central Siberia; (H) Baykonur Fm., Kazakhstan and Kyrgystan; (I) Bol’shoy Patom Fm., central Siberia; (J) Marnya Fm., foothills of East Sayan mountains; (K) Tsagaan Oloom Fm., SW Mongolia; (L) Khubsugul Group, northern Mongolia; (M) Blaini Fm., NW India; (N) Yangtze region, China; (O) Tarim Block, NW China.

Table 1.2. Summary of Neoproterozoic data sets in Eurasia and the Nubian Shield Lead author

Glaciogenic strata

Structural framework

Chemostratigraphy d13C

Allen Allen Stern

Miller

Chumakov Chumakov

Chumakov Sovetov Macdonald Macdonald Etienne Zhang

Zhu

Zabit Fm., Kushatay Fm. (Bokson Gp.) Churochnaya Fm., Tany/ Koyva/Wil’va formations, Lower Starye Pechi Subformation Nichatka Fm. Baykonur Bol’shoy Patom Fm. Marnya Fm. (Oselok Gp.) Tsagaan Oloom Fm. Ongoluk Fm., Khesen Fm. (Khubsugul Gp.) Blaini Fm. Chang’an/Fulu Fm (Jiangkou Gp); Nantuo Fm. Polong Fm. Yutang/ Yulmeinak diamictite (Tarim Basin); Beiyixi, Altungol, Tereeken and Hankalchough diamictites (Tarim Block); and others

d34S

87

Sr/86Sr

Palaeomagnetism

Other

Geochronology U– Pb detrital

Rift Rift Arc–back-arc, oceanic basin

Y Y

CIA, MIA CIA, MIA REE, Ce, Eu

Arc, intra-oceanic platform

Y

Y

Foreland basin

Y

Y

Y

Passive margin

Y

Y

Y

Passive margin Unknown

Y Y

Y

Passive margin Foreland basin

Y Y

Passive margin? Rift?

Y Y

Passive margin Rift

Y Y

Foreland basin

Y

Y

Economic

Y, unreliable Y, unreliable

Y Y

BIF, Au, Cu, Zn, Ar,

CIA, PIA, d13C_TOC, Sr

Y

U– Pb direct

Other

Rb –Sr, K– Ar Y Y

Y

Bauxite, Phosphorite

Sm–Nd, Pb/ Pb, Rb– Sr Rb –Sr

Y

K –Ar, Sm– Nd, Rb –Sr, Pb/Pb

Y Y Y Y

Y

Y Y

Y Y Ce

Y Y, reset Y, overprinted

Y

K –Ar,Rb– Sr Ar/Ar

Y

Y

Phosphorite

Y Y

Vanadium, phosphorite

INTRODUCTION

Chumakov Chumakov

Ayn Fm., Shareef Fm. Ghubrah Fm., Fiq Fm. Meriti Gp.?, Mahd Gp.?, Atud & Nuwaybah diamictites Didikama Fm., Matheos Fm. (Tambien Gp.)

d18O

Geobiology

TOC

Y Y (Biomarker)

Y

Manganese, BIF

Unreliable Y

Y

Y

Y Y

Nd (tDM)

Y

Note: ‘U –Pb direct’ means dates from volcanic intrusions or tuffs anywhere in the succession, not necessarily from the glaciogenic strata itself. ‘U– Pb detrital’ means dates from detrital zircons from within the succession. Data in presumed correlative successions were not included.

5

6

E. ARNAUD ET AL.

lacunae are due either to the inability to find an appropriate author or because little or no work has been published in that region since the publication of Hambrey & Harland (1981). In such cases, readers are referred to that earlier volume for more information or, where appropriate, to more recent literature referenced in chapters covering neighbouring basins. Where possible, these overlooked successions were incorporated into chapters covering nearby regions or related deposits. Most are also listed in Evans & Raub’s review of the palaeomagnetic database.

The current knowledge base General trends Many of the successions and regions covered in this book were reviewed in Hambrey & Harland (1981). Since that time, the greatest advances in our knowledge have come from the fields of geochemistry, palaeomagnetism and geochronology, although palaeoenvironmental interpretations of many glaciogenic and overlying units have also been variably reinforced and challenged in recent years. One of the novel aspects of the Snowball Earth hypothesis was its integration of various geological data sets, specifically bringing attention to the geochemical signatures associated with the Neoproterozoic glacial deposits. As a result, geochemical analyses of associated carbonate rocks proliferated from the late 1990s onwards. Although the main focus has been in acquiring traditional stable isotope (namely d13C and d34S)

and 87Sr/86Sr data sets to reconstruct ocean compositions and terrestrial weathering processes (Halverson & Shields-Zhou 2011), these rocks have also proven fertile grounds for novel isotopic approaches such as Ca and B isotopes, and multiple O isotopes (e.g. Kasemann et al. 2004; Bao et al. 2008). In addition, the hypothesis generated renewed interest in the idea of global synchronous glaciation, its onset and initiation, which in turn prompted focused efforts to provide additional geochronological constraints. The development of new techniques such as the use of Re –Os in dating organic shales, the proliferation of SHRIMP and LA-ICP-MS analyses, and the refinement of the ID-TIMS U –Pb dating technique have simultaneously refined some age constraints and seriously challenged conventional age models for key Neoproterozoic successions, particularly in Australia. The glacial deposits themselves, as the most direct record of the global glaciations, have come under much greater scrutiny, and many units have been re-evaluated using facies and sequence stratigraphic analysis, incorporating advances in our understanding of glaciated basins made in the last 30 years. The role of tectonics in modulating climate change during this time period and the role of basin setting in controlling the preservation and nature of the sedimentary record has also been explored further (e.g. Eyles & Januszczak 2004; Allen 2007; Stern 2008). Some workers have focused on sites that previously had relatively little to no data, such as Mongolia, Alaska, Ethiopia, Egypt and western Arabia and some parts of Australia, Russian Federation and South America, while some of the classic Neoproterozoic sections have been the subject of additional study by several

Fig. 1.3. Map showing distribution of Neoproterozoic glaciogenic successions within North America, as covered in this volume. (A) Hula Hula diamictite, Arctic Alaska; (B) Tatonduk Inlier, Alaska–Yukon border; (C) Windermere Supergroup, Mackenzie Mountains, Canada; (D) Windermere Supergroup, southern Canadian Cordillera; (E) SE Idaho and Utah, USA; (F) Edwardsburg Fm., central Idaho, USA; (G) Kingston Peak Fm., eastern Death Valley region, USA; (H) Kingston Peak Fm., Panamint Range, USA; (I) Gaskiers Fm., Newfoundland, Canada; (J) Squantum Member, Boston Basin, USA.

Table 1.3. Summary of Neoproterozoic data sets in North America Lead author

Glaciogenic strata

Structural framework

Geochemistry d13C

d18O

d34S

Hula Hula Diamictite

Unknown

Y

Macdonald

Rapitan Gp.; Hay Creek Gp. Rapitan Gp.; Ice Brook Fm. (Stelfox Mb.) Toby Fm., Vreeland Fm (Windermere Supergroup) Pocatello Fm. (Idaho), Mineral Fork Fm. (Utah) Edwardsburg Fm., Moores Lake Fm. Kingston Peak Fm. (Eastern Death Valley) Kingston Peak Fm. (Panamint Range) Gaskiers Fm. Squantum Mb.

Rift to passive

Y

Rift to passive

Y

Y

Y

Rift; controversial upper WSG

Y

Y

Y

Rift to passive

Y

Hoffman

Smith

Link

Lund Mrofka

Petterson Carto Carto

Extensional (inferred) Extensional

Y

Rift

Y

Arc Extensional-rift, intra-arc

Y

Sr/86Sr

Economic

Palaeomagnetism

Other

Geochronology U –Pb detrital

U– Pb direct

Y

Y Y

Y

Y

d57Fe, Ce/Ce*, d44Ca, Elemental analysis, REE TOC, Mo, V/Cr

Y

Fe

Y

Cu

CIA, TOC, Nd, Sr

Y

Y (Rap. Gp.)

Y

Pb/Pb (detrital), Rb– Sr Ar/Ar Pb/Pb

Y

Y

Whole-rock geochem. Y

Other

Re–Os, Sm–Nd

Y

INTRODUCTION

Macdonald

87

Geobiology

Y Y

U, U– Th, Iron ore FeHR/FeT CIA

Y Y

Y Y

Y Y

Y Pb/Pb

Note: ‘U –Pb direct’ means dates from volcanic intrusions or tuffs anywhere in the succession, not necessarily from the glaciogenic strata itself. ‘U– Pb detrital’ means dates from detrital zircons from within the succession. Data in presumed correlative successions were not included.

7

8

E. ARNAUD ET AL.

research groups, notably Svalbard, northern Norway, Scotland/ Ireland, the North American Cordillera (Death Valley, Idaho and western Canada), Namibia, various regions in Australia, and China. As a result, some of these classic localities have become ‘type’ localities for investigating Neoproterozoic glaciations and their aftermath. With the increasingly global focus in research into the Neoproterozoic and ongoing efforts to define and normalize chronostratigraphic units in the Neoproterozoic, the terminology used to refer to Neoproterozoic glaciations and, implicitly, their age and global correlations has significantly evolved. For example, now largely absent from the international lexicon are such previously mainstream terms as Sinian, Vendian and Varangerian, which refer to overlapping subsets of Neoproterozoic time spanning glacial epochs. Because these intervals were defined based on regional stratigraphic sequences with little basis other than the occurrence of glacial deposits, their use was inconsistent and often confusing. New informal terms have emerged in their place and partly as a result of the intense efforts to determine the number of Neoproterozoic glacial epochs. Specifically, the Neoproterozoic

glaciations are now commonly identified by their relative timing and assumed correlation with well-studied localities. MidEdiacaran glacial deposits are widely referred to as Gaskiers, based on the eponymous glaciogenic formation on the Avalon Peninsula of southeastern Newfoundland, which occurs just below the first appearance of Ediacaran fossils c. 575 Ma (Narbonne & Gehling 2003). Older, Cryogenian-aged glacial deposits are commonly referred to as Sturtian (older) and Marinoan (younger) (e.g. Kennedy et al. 1998; Halverson et al. 2005) in an adaptation of terms originally intended for specific sequences in the Adelaidean rift-basin of South Australia (Preiss 2000). Despite their proliferation and appeal, the application of these terms usually hinges on inferred inter-regional correlations that are commonly controversial and often poorly backed up by radiometric data. Although biostratigraphy (Grey et al. 2011), chemostratigraphy (Halverson & Shields-Zhou 2011) and magnetic stratigraphy (Evans & Raub 2011) are increasingly able to resolve relative age assignments and correlations, we have favoured reference to the emerging Neoproterozoic timescale (i.e. Cryogenian and Ediacaran) over the terms Sturtian, Marinoan

Fig. 1.4. Map showing distribution of Neoproterozoic glaciogenic successions within South America, as covered in this volume. (A) Chiquerio Fm., Peru; (B) Puga Fm., Paraguay Belt, Brazil; (C) Serra Azul Fm., Paraguay Belt, Brazil; (D) Bebedouro Fm., Brazil; (E) Sa˜o Francisco Craton, Brazil; (F) The Macau´bas Group, SE Brazil; (G) Moema laminites, Sa˜o Francisco basin, Brazil; (H) Jequitaı´ Fm., SE Brazil; (I) Playa Hermosa Fm., Uruguay; (J) Last Ventanas and San Carlos formations, Uruguay; (K) Tandilia system, Argentina.

Table 1.4. Summary of Neoproterozoic data sets in South America Lead author

Glaciogenic strata

Structural framework

Chemostratigraphy d13C

d18O

Y

Y

Y

Y

d34S

87

Sr/86Sr

Chiquerı´o Fm.

Alvarenga Figueiredo

Puga Fm., Cuiaba´ Gp, Jacadigo Gp. Serra Azul Fm.

Guimaraes

Bebedouro Fm.

Extensional

Y

Y

Misi

Not well constrained

Y

Y

Rocha-Campos

Bambuı´ Gp., Una Gp., Vazante Gp., Vaza Barris/Miaba Gp. Serra do Catuni Fm., Nova Aurora Fm., Lower Chapada Acaua Fm. Moema Laminites

Uhlein

Jequitaı´ Fm.

Y

Pazos

Playa Hermosa Fm.

Pecoits

Las Ventanas Fm., San Carlos Fm. Sierra del Volca´n Fm.

Cratonic/ intracratonic? Not well constrained Strike–slip basin Not well constrained

Y

Pedrosa-Soares

Pazos

Extensional, not well constrained Extensional

Economic

Palaeomagnetism

Other

Fe, Cu, Au

Y

Y

Fe (BIF), Mn

Geochronology U –Pb detrital

U–Pb direct

Y

Y

Y

Y

Pb/Pb, Nd

Y

Ar/Ar, K– Ar, Sm–Nd Pb/Pb

Passive margin

Y (ltd)

Rift to passive margin

Intracratonic basin

Y

Y

Re– Os, Pb/ Pb, Sm–Nd

Y

Kaolinite (ceramic)

Major and trace elements, REE, Eu, Ce, Y

Barite nodules; phosphate Fe– Mn, Phosphorite, fluorite, barite, Pb–Zn, Zn Mn, Fe, diamonds

Other

INTRODUCTION

Chew

Geobiology

Y

Y

Rb– Sr, Pb/Pb

Y

Ar/Ar, K– Ar, Rb– Sr K– Ar, Rb –Sr, Ar/Ar Rb– Sr

Preliminary Y Y

Y Y?

Y

Note: ‘U –Pb direct’ means dates from volcanic intrusions or tuffs anywhere in the succession, not necessarily from the glaciogenic strata itself. ‘U– Pb detrital’ means dates from detrital zircons from within the succession. Data in presumed correlative successions were not included.

9

10

E. ARNAUD ET AL.

and Gaskiers in this volume in recognition that ages and correlations of the Neoproterozoic glacial deposits remain controversial. This said, the fact that the lower boundary of the newly defined Ediacaran Period is defined as the contact between the glaciogenic Elatina Fm. (¼Marinoan) and the overlying Nuccaleena cap carbonate (Knoll et al. 2006) is suggestive of the chronostratigraphic importance of at least one of these Neoproterozoic glaciations. The definition of this boundary was bolstered by several new, well placed and precise radiometric ages on basal Ediacaran cap dolostones (i.e. the lower, transgressive, dolomite facies of the cap carbonates; Hoffman et al. 2011) that strengthen the case for their global age equivalence. The trickle of new radiometric ages is generally tightening the age constraints on this late Cryogenian glaciation and the other glacial epochs and at the same time calibrating and helping to refine the secular evolution of seawater geochemical proxies (Halverson & Shields-Zhou 2011). New ages have demonstrated that some putative ‘Cryogenian’ glacial deposits are in fact post-Cryogenian in age, such as the Gaskiers unit in Newfoundland, while the purported ‘Cambrian’ glaciation in the Taoudeni Basin is now regarded unequivocally as Cryogenian in age. Recent U –Pb and increasingly Re –Os

data (Condon & Bowring 2011) confirm the division of Cryogenian glaciations into two main episodes (c. 720–c. 660 Ma and c. 650 –635 Ma), although older diamictites of unconfirmed origin may hint at localized earlier Cryogenian glaciation. In contrast to the now reasonably well constrained late Cryogenian glaciation, the wide range of dates obtained from older Cryogenian glacial deposits long thought to be correlative (mainly in Australia and the North American cordillera), have called into question their equivalence and have been used to argue against global glaciation. Whereas the available data allow for a single, long-lived earlier Cryogenian glaciation, it is clear that the implications and interpretations of these many new ages are contentious and rapidly evolving. Therefore, references throughout the book to early, earlier or middle Cryogenian glaciation should not be assumed to indicate synchronous glaciation. In recent years, with resource exploitation reaching unprecedented peaks, there has also been increasing interest in the economic potential of Neoproterozoic strata. Ediacaran rocks comprise petroleum source rocks in China and Oman, and this has triggered interest in correlative post-glacial successions in Africa, India, Australia, Brazil and Russia. Phosphorus resources have also become a vulnerable commodity worldwide as Moroccan deposits

Fig. 1.5. Map showing distribution of Neoproterozoic glaciogenic successions within Europe, as covered in this volume. (A) Polarisbreen Group, NE Svalbard; (B) NE Greenland; (C) Smalfjord and Mortensnes formations, Finnmark, Norway; (D) Caledonides, NW Scandinavia; (E) Moelv and Koppang formations, southern Norway; (F) Caledonides, central Scandinavia; (G) Lillfja¨llet Fm., southern Swedish Caledonides; (H) Port Askaig Fm., Scotland; (I) Scotland and Ireland.

Table 1.5. Summary of Neoproterozoic data sets in Europe Lead author

Glaciogenic strata

Structural framework

Chemostratigraphy d13C

Halverson

Stouge Rice

Nystuen Kumpulainen

Kumpulainen Arnaud Prave

Port Askaig Fm. Dalradian Supergroup overview-Port Askaig Fm.; Stralinchy-Reelan; MacDuff/Loch na Cille/Inishowen boulder beds

d34S

87

Sr/86Sr

Economic

Palaeomagnetism

Other

U–Pb detrital

D17O

Y

TC, TOC, TS

Y

Y/unreliable?

Y

Y

Y

Passive margin

Y (ltd)

Y

Rift; pericratonic shelf Rift to passive margin

Y (ltd)

Y

Thermally subsiding margin Rift to passive

Y

Rift to passive

Rift to passive margin? Extensional Rift

Y

Y

Y (ltd)

Geochronology

Lu–Hf

U –Pb direct

Other

Rb –Sr

Y

Y

Rb –Sr, Re– Os

INTRODUCTION

Stodt

Petrovbreen Mb., Wilsonbreen Fm. (Polarisbreen Group) Ulvesø Fm., Storeelv Fm. and Støvfanget Fm. Smalfjord Fm., Mortensnes Fm. Possible correlatives of the Mortensnes Fm. Moelv Fm., Koppang Fm. La¨ngmarkberg Fm., Risback Gp. Diamictite? Lillfja¨llet Fm.

d18O

Geobiology

Ar/Ar Y Y

Y Y

CIA

Y

Pb– Zn, magnetite Y

Y Y

Pb/Pb, Sm– Nd

Note: ‘U –Pb direct’ means dates from volcanic intrusions or tuffs anywhere in the succession, not necessarily from the glaciogenic strata itself. ‘U– Pb detrital’ means dates from detrital zircons from within the succession. Data in presumed correlative successions were not included.

11

12

E. ARNAUD ET AL.

become the only widely traded source of phosphorus for fertilizer. Ediacaran-age phosphorites of China are the only other worldclass phosphorite resource, sparking interest in possibly correlative, glacially associated phosphorites in West Africa, South America and Australia. In addition to being an oceanographic curiosity of the Cryogenian Period, the widespread deposition of iron and manganese deposits may also have economic implications for those regions where they are found (South America, northwestern Canada, India, China and Australia; Hoffman et al. 2011). Significant advances have been made in all regions since the publication of Hambrey & Harland (1981). However, with new data always come new questions and controversies. Some of the notable advances and contentious issues are highlighted below. Readers interested in specific glaciogenic successions are encouraged to read chapters from neighbouring sites as controversies are sometimes most apparent when viewing the geological record from a regional vantage point.

Africa Neoproterozoic glaciogenic successions reported from Africa are located primarily in the NW, central and southern regions, with a cluster of sites in Namibia (Fig. 1.1, Table 1.1). These occur in a wide range of tectonic settings, with few sites containing biostratigraphic or palaeomagnetic data. Geochemical and geochronological studies have been carried out at most sites, with the most complete data sets found in the Taoudeni Basin in NW Africa and the Otavi Group of Namibia. The glacial deposits in the former (Shields-Zhou et al. 2011) remain significant for being among the few to preserve unequivocal terrestrial glacial deposits, which are now known to be age-equivalent to other endCryogenian glaciogenic units around the world. Robust geochronological constraints spanning the two glaciogenic intervals of the Otavi Group are particularly notable. The Otavi Group has been the subject of intense study over the past two decades, and despite lingering controversy over interpretation of glacial deposits in the succession (Hoffman 2011), plays a central role in debates

over the Snowball Earth hypothesis. Prave et al. (2011) have provided a useful review of the hitherto poorly documented Witvlei Group in east-central Namibia (northern Kalahari craton), which provides a firmer basis for correlation with the numerous Neoproterozoic glacial deposits elsewhere on the Kalahari craton (see the chapters by Frimmel). Whereas few geochronological or geochemical data had been available from the West Congo belt until recently, new research has provided key data that strengthen glacial interpretations and correlations across the belt and elsewhere on the Congo craton (Tait et al. 2011), such as the Katangan Supergroup, best known for its world class Cu deposits (Master & Wendorff 2011).

Eurasia –Nubian Shield This region encompasses Neoproterozoic glaciogenic successions found throughout the Russian Federation, Mongolia, China and India, as well as those from Oman, the shores of the Red Sea and Ethiopia (Fig. 1.2, Table 1.2). Excellent preservation of Cryogenian glacial deposits in northern Oman have allowed detailed sedimentological and sequence stratigraphic analysis to be carried out over the last several decades. The results have shed light on the environmental conditions during Neoproterozoic glaciations, with some more recent geochemical and geochronological analyses contributing a relatively comprehensive data set for this region. Unfortunately, palaeomagnetic studies in Oman have thus far yielded no reliable palaeolatitudes. In contrast, studies in Ethiopia and the shores of the Red Sea have focused on geochemical and geochronological aspects of Neoproterozoic geology, making this region an ideal candidate for future work in sedimentology and stratigraphy to confirm glaciogenic conditions existed in these regions and maximize the impact of these new radiometric ages. The basins of the Arabian –Nubian Shield are unique in that they occur in arc or back-arc basins, a relatively rare tectonic setting for glaciated Neoproterozoic basins. Glaciogenic successions in the Russian Federation and neighbouring Mongolia, Kazakhstan and Kyrgyzstan are found within primarily passive margin and foreland basin settings, with

Fig. 1.6. Map showing distribution of Neoproterozoic glaciogenic successions within Australia, as covered in this volume. (A) Tasmania; (B) the Kimberly region and NW Northern Territory; (C) Mid-Ediacaran ice rafted deposits, Adelaide Geosyncline and Officer Basin; (D) Central Australia; (E) Billy Springs Fm.; (F) Yudnamutana Subgroup (Sturtian); (G) Elatina Fm.

Table 1.6. Summary of Neoproterozoic data sets in Australia Lead author

Calver

Corkeron

Hill

Jenkins Preiss

Williams

Cottons Breccia, Julius R. Mbr., Croles Hills Diamictite; Wedge R. beds, Cotcase Ck. Fm. Walsh Fm., Landrigan/ Egan Fms., Fargoo/ Moonlight Valley Fm. Bunyeroo Fm., Dey Dey Mudstone Glaciogenic deposits of Central Australia (Officer, Amadeus, Ngalia, Georgina Basins Billy Springs Fm. Yudnamutana Subgroup (Sturt and correlatives)

Yerelina Subgroup, Elatina Fm.

Structural framework

Chemostratigraphy d13C

d18O

d34S

87

Sr/86Sr

Rift, epicratonic basin

Y

Y

Y

Unknown

Y

Y

Unreliable

Thermally subsiding basin; platformal Intracratonic sag basin

Y

Geobiology

Y Y

Y

Rift

Y

Y

Unpub. Y

Calcic scheelite skarns

Unpublished

Y

Y

U–Pb direct

Other

Y

Re–Os, Nd–Sm

Y

Rb– Sr

Y

Y

Y

Rb– Sr

Y

Y

Y

Re–Os

Y Y

Geochronology U– Pb detrital

Y

Y

Rift complex Rift

Palaeomagnetism

Other

Atomic H/C ratios

Y

Economic

INTRODUCTION

Gostin

Glaciogenic strata

Y (ltd)

Elemental analysis, REE, Ironstone geochemistry Y

Y

Y

Rb– Sr, Re–Os

Y

Re–Os

Note: ‘U –Pb direct’ means dates from volcanic intrusions or tuffs anywhere in the succession, not necessarily from the glaciogenic strata itself. ‘U– Pb detrital’ means dates from detrital zircons from within the succession. Data in presumed correlative successions were not included.

13

14

E. ARNAUD ET AL.

geochemical data (primarily carbon, but also oxygen and strontium isotopes in some cases) as well as important biostratigraphic data. Interestingly, these studies have led to some contradictory stratigraphic correlations across neighbouring basins (see chapters on the Bokson Group and Khubsugul Basin). Chumakov (2011) suggests that two diamictite-bearing horizons within the Bokson Group in the East Sayan Mountains are likely Ediacaran in age based on biostratigraphic data, whereas Macdonald & Jones (2011) suggest that the two horizons in the neighbouring Mongolian basin correspond to an early and late Cryogenian glacial episode. Dispute also remains over the glacial origin of the younger Mongolian diamictite-bearing unit. Nonetheless, the tectonic setting of the region makes these basins suitable targets for future geochronological studies, which may help to resolve these stratigraphic inconsistencies. The chapter on the Marnya Fm. in the East Sayan Mountains (Sovetov 2011) is notable for its review of regional stratigraphic correlation and useful comparison of the Russian stratigraphic scheme with that of the more widely used International Subcommission on Stratigraphy (see also Table 27.2). Palaeogeographic reconstructions in northern Eurasia are hampered by the lack of reliable palaeomagnetic data, although ongoing studies in Mongolia show some promise. The Blaini Fm. of India did not receive much attention until the more recent studies reported in this volume, which include new sedimentological and geochemical data. In contrast, the glaciogenic successions of the Yangtze region (rift setting) and Tarim Block (foreland basin) have a long history of Neoproterozoic research that has resulted in extensive data sets for geochemistry, geobiology and biostratigraphy, palaeomagnetism and geochronology. Notably, a U– Pb zircon age from an ash bed in the Yangtze region has provided a definitive date for the end of the late Cryogenian glaciation (c. 635 Ma), whereas maximum and syn-glacial ages in China and elsewhere have shown that this late Cryogenian glaciation lasted ,10 Ma (Condon & Bowring 2011). In addition, palaeomagnetic data indicating low palaeolatitudes for the Nantuo diamictite have one of the highest reliability ratings of available palaeomagnetic data, as discussed by Evans & Raub (2011). Lastly, novel geobiological and biostratigraphic studies have demonstrated the preservation of animal embryos in the Ediacaran-age Doushantuo Fm. of the Yangtze region (Hagadorn et al. 2006).

South America Neoproterozoic glaciogenic successions in South America are primarily clustered in the Paraguay Belt and on the Sa˜o Franscisco craton of Brazil, with other sections reported from Uruguay, Argentina and Peru (Fig. 1.4, Table 1.4). The Araras Group in the northern Paraguay Belt, in particular, has been the subject of interdisciplinary studies and preserves some of the best evidence for Ediacaran glaciation in South America (Figueiredo et al. 2011). New entries to the South American database include the sections in Peru, Uruguay and Argentina as well as the Moema laminites in Brazil. Despite some new geochronological data, there remain significant challenges in regional stratigraphic correlation schemes, limited information regarding palaeolatitude, and uncertainty regarding tectonic setting and glacial origin of some of the units described herein. These challenges are manifested in significant controversies over the correlation and ages of Proterozoic glacial deposits across Brazil (Misi et al. 2011). Despite uncertainties in age, the correlations between the Bambuı´ and Una groups have been strengthened (Guimara˜es et al. 2011).

Europe The Neoproterozoic glaciogenic successions of Europe described here are primarily those found in extensional and passive margin basins of Scandinavia, Svalbard, Greenland and the British Isles (Fig. 1.5, Table 1.5). Although most of these sites were described in Hambrey & Harland (1981), new data have become available in recent decades, including additional geochemical data from most of these successions, additional information on associated carbonate strata and some additional geochronological constraints. Most successions in the region have been significantly impacted by Caledonian tectonics and hence lack reliable palaeomagnetic data directly related to the glacial deposits. Robust geochronological constraints are also largely lacking, making regional stratigraphic correlations relatively tenuous, despite some advances using chemostratigraphic data. Interestingly, ongoing studies in Greenland have led some researchers to propose that the Greenland and Svalbard sections are not as intimately connected palaeogeographically as was previously proposed.

North America Australia Many of the Neoproterozoic glaciogenic successions in North America are found along the length of the North American Cordillera in rift basins and passive margin settings, whereas others are found on exotic terranes in Arctic Alaska and along the eastern margin of North America (Fig. 1.3, Table 1.3). The Cordilleran sections generally have a long history of research, but have enjoyed renewed interest and have been the source of much new data. While many researchers have attempted to correlate these widely separated sections, recent studies have highlighted the possible diachroneity of Neoproterozoic glaciations and rifting related to Rodinian break-up. New additions to the North American database include the preliminary results of ongoing multidisciplinary studies in Alaska and the Yukon including geochronological and palaeomagnetic data (Macdonald 2011; Macdonald & Cohen 2011); new geochronological data from Idaho (Link & Christie-Blick 2011; Lund et al. 2011) and the Canadian Cordillera (Smith et al. 2011; Macdonald et al. 2010); detailed mapping and geochemical data from the Death Valley region; a better understanding of the tectonic development of the southern Cordillera margin based on work in Idaho, Utah and Death Valley; the analysis of a deepwater ‘cap’ carbonate in the southern Canadian Cordillera; and influential studies on the impressive Ediacaran biota and ocean redox in Newfoundland.

The Neoproterozoic glaciogenic successions in Australia covered in this volume are very similar to the entries found in Hambrey & Harland (1981) over 30 years ago (Fig. 1.6, Table 1.6). Recent advances include the collection of additional geochemical and biostratigraphic data, continued state-of-the-art palaeomagnetism research with confirmation of low-latitude glaciation, and additional geochronological constraints that have prompted abundant discussion about the duration and timing of the older Cryogenian glaciation. Entries for South Australia are focused on individual glacial events (Preiss et al. 2011; Williams et al. 2011), with new evidence for Ediacaran-aged glaciation within the Adelaidean rift-basin (Gostin et al. 2011; Jenkins 2011). Corkeron, Hill et al. and Calver provide timely and useful syntheses of the separate Neoproterozoic successions in the Kimberly region, ‘Centralian Superbasin’ and Tasmania, respectively.

Calibrating Neoproterozoic change One of the prominent developments in stratigraphy over the last decade has been the attempt to establish consistent, rock-based periods of geological time beneath the Cambrian Period. Neoproterozoic glaciation played a pivotal role in this when the Ediacaran

INTRODUCTION

Period, which was the first period of the geological timescale to be established for a hundred years, was defined as beginning directly after the widespread, end-Cryogenian (¼ Elatina or Marinoan) glaciation. Now, international efforts are under way to further subdivide the Ediacaran Period and define the underlying Cryogenian Period. In this endeavour, the integrated approach taken by many of the authors in the present volume will be key as a global stratigraphic framework can only be established in the Precambrian by meshing together sparse fossil information with high-resolution chemostratigraphy, strategically placed age constraints and lithostratigraphic markers. At present, the base of the Cryogenian Period is defined chronometrically at 850 Ma, but few age constraints exist to identify this level in sedimentary successions from around the world. Although, the base of a new chronostratigraphically defined Cryogenian Period is yet to be established, any GSSP is likely to be placed ‘beneath the oldest clearly glaciogenic deposits in a Neoproterozoic succession’ at an outcrop horizon with ‘proven potential for global C- and Sr-isotope stratigraphic correlation and preferably be amenable to microfossil biostratigraphy, isotope geochronology and other forms of global correlation such as magnetostratigraphy’ (Neoproterozoic Subcommission 2009 Annual Report). This emphasis on glacial strata will likely lead to a shorter Cryogenian Period more in line with its Phanerozoic counterparts, and with a base close in age to the Dunn et al. (1971) estimate of c. 750 Ma. The unique Neoproterozoic palaeoclimate, which underpins this volume, has already cemented the place of Neoproterozoic ice ages in the international geological timescale (Ogg et al. 2008) and will continue to be central in the subdivision of Proterozoic time.

Future work This memoir was inspired by Earth’s Pre-Pleistocene Glacial Record (Hambrey & Harland 1981), and it is hoped that it can serve similarly as a long-lasting reference work that provides useful summaries of most Neoproterozoic glacial successions, while also highlighting the state-of-the art in the study of these rocks. The large number of chapters and impressive list of contributors highlight the extraordinary interest in the Neoproterozoic and the vast amount of research carried out over the past 30 years. At the same time, the results of this research have revealed gaping holes in our understanding of the Neoproterozoic and aroused compelling new questions. Many glaciogenic units described in Hambrey & Harland (1981) have received little or no attention since that time, such as the widespread diamictites of the Sino-Korean craton, which are still of uncertain age and origin. Radiometric age constraints on most glacial units remain sparse or non-existent, which poses a persistent challenge to interregional correlations. Many important advances have been in the biostratigraphy of Cryogenian and Ediacaran successions, but most sections remain poorly studied. In addition, a welter of exciting new geochemical methods, such as the D47 palaeothermometer and transition-metal stable isotopes, await rigorous application to the Neoproterozoic. Clearly, much remains to be done, but in almost every region of the world there are now interdisciplinary teams of scientists at work in what has become a global endeavour not only to explore the extremes of past climate change, but also to evaluate the link between eukaryotic evolution and global glaciations. We wish to thank all the authors who made this collection possible, for agreeing to write these contributions amidst many other writing commitments, and patiently awaiting the completion of the volume. We particularly thank N. Chumakov and Q. R. Zhang for promptly contributing to the original template papers that were essential in establishing the format for all subsequent regional chapters. We are also especially grateful to P. Hoffman, F. Macdonald, M. Zhu, R. Stern and J. Tait for acquiescing to last-minute requests for contributions. We would also

15

like to thank all the reviewers (see list of reviewers) who provided meaningful and constructive comments and improved the individual chapters. I. Fairchild kindly provided the cover photo and the historical photograph for the dedication. E. Arnaud would also like to thank those who provided guidance in developing the instructions for authors with suggested content required of specific sections: namely D. Condon, N. Eyles, G. Halverson, B. MacGabhann, P. Martini, T. Raub, G. Shields-Zhou & R. van der Voo. The following regional coordinators also assisted in tracking down authors and names of Neoproterozoic glaciogenic successions in their regions: P. Allen, C. Calver, N. Chumakov, M. Corkeron, N. Eyles, V. Gostin, P. Hoffman, C. Hoffmann, G. Jiang, K. Kristiansen, R. Kumpulainen, P. Link, U. Linnemann, J. Meert, N. Miller, P. Pazos, T. Prave, Q. R. Zhang, H. Rice, T. Rocha Campos, S. Stouge, M. Wendorff and S. Xiao. We thank J. Howe for his contribution as the Society Books Editor and for his timely review of this chapter. Angharad Hills at the Geological Society of London Publishing House kept us on track and provided editorial assistance throughout the project. This memoir represents a contribution of the IUGS-UNESCO sponsored International Geoscience Programme (IGCP) Project # 512.

References Allen, P. A. 2007. The Huqf Supergroup of Oman: basin development and context for Neoproterozoic glaciation. Earth-Science Reviews, 84, 139– 185. Bao, H., Lyons, J. R. & Zhou, C. 2008. Triple oxygen isotope evidence for elevated CO2 levels after a Neoproterozoic glaciation. Nature, 453, 504–506. Chumakov, N. M. 2011. Glacial deposits of the Bokson Group, East Sayan Mountains, Buryatian Republic Russion Federation. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 285–288. Condon, D. J. & Bowring, S. A. 2011. A user’s guide to Neoproterozoic geochronology. In: Arnaud, E., Halverson, G. P. & ShieldsZhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 135– 149. Dunn, P. R., Thomson, B. P. & Rankama, K. 1971. Late Pre-Cambrian glaciation in Australia as a stratigraphic boundary. Nature, 231, 498– 502. Evans, D. A. D. & Raub, T. D. 2011. Neoproterozoic glacial palaeolatitudes: a global update. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 93– 112. Eyles, N. & Januszczak, N. 2004. Interpreting the Neoproterozoic glacial record; the importance of tectonics. In: Jenkins, A., McMenamin, M. A. S., McKey, C. P. & Sohl, L. (eds) The Extreme Proterozoic: Geology, Geochemistry and Climate. American Geophysical Union, Geophysical Monograph 146, 125–144. Fairchild, I. J. & Kennedy, M. J. 2007. Neoproterozoic glaciation in the Earth System. Journal of the Geological Society, London, 164, 895– 921. Figueiredo, M. F., Babinski, M. & Alvarenga, C. J. S. 2011. The Serra Azul Formation, Paraguay Belt, Brazil. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 499– 502. Gostin, V. A., McKirdy, D. M., Webster, L. J. & Williams, G. E. 2011. Mid-Ediacaran ice-rafting in the Adelaide Geosyncline and Officer Basin, South Australia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 673– 676. Grey, K., Hill, A. C. & Calver, C. 2011. Biostratigraphy and stratigraphic subdivision of Cryogenian successions of Australia in a global context. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 113– 134. Guimara˜es, J. T., Misi, A., Pedreira, A. J. & Dominguez, J. M. L. 2011. The Bebedouro Formation, Una Group, Bahia (Brazil). In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological

16

E. ARNAUD ET AL.

Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 503– 508. Hagadorn, J. W., Xiao, S. et al. 2006. Integrated X-ray insights into cellular and subcellular structures of Neoproterozoic animal embryos. Science, 314, 291–294. Halverson, G. P. & Shields-Zhou, G. 2011. Chemostratigraphy and the Neoproterozoic glaciations. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 51 –56. Halverson, G. P., Hoffman, P. F., Schrag, D., Maloof, A. C. & Rice, A. H. 2005. Towards a Neoproterozoic composite carbon isotope record. Geological Society of America Bulletin, 117, 1181–1207. Hambrey, M. J. & Harland, W. B. (eds) 1981. Earth’s Pre-Pleistocene Glacial Record. Cambridge Earth Science series. Cambridge University Press, Cambridge. Hoffman, P. F. 2011. A history of Neoproterozoic glacial geology, 1871– 1997. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 17 –37. Hoffman, P. F., Kaufman, A. J., Halverson, G. P. & Schrag, D. P. 1998. A Neoproterozoic snowball Earth. Science, 281, 1342– 1346. Hoffman, P. F., Macdonald, F. A. & Halverson, G. P. 2011. Chemical sediments associated with Neoproterozoic glaciation: iron formation, cap carbonate, barite and phosphorite. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 67 –80. Jenkins, R. J. F. 2011. Billy Springs Glaciation, South Australia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 693–699. Kasemann, S. A., Hawkesworth, C. J., Prave, A. R., Fallick, A. E. & Pearson, P. 2005. Boron and calcium isotope composition in Neoproterozoic carbonate rocks from Namibia: evidence for extreme environmental change. Earth and Planetary Sciences Letters, 231, 73 –86. Kennedy, M. J., Runnegar, B., Prave, A. R., Hoffmann, K. H. & Arthur, M. 1998. Two or four Neoproterozoic glaciations? Geology, 26, 1059–1063. Knoll, A. H., Walter, M. R., Narbonne, G. M. & Christie-Blick, N. 2006. The Ediacaran Period: a new addition to the geologic time scale. Lethaia, 39, 13 – 30. Link, P. K. & Christie-Blick, N. 2011. Neoproterozoic strata of Southeastern Idaho and Utah: record of Cryogenian rifting and glaciation. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 425–436. Lund, K., Aleinikoff, J. N. & Evans, K. V. 2011. The Edwardsburg Formation and related rocks, Windermere Supergroup, central Idaho, U.S.A. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 437– 447. Macdonald, F. A. 2011. The Hula Hula diamictite and Katakturuk Dolomite, Arctic Alaska. In: Arnaud, E., Halverson, G. P. & ShieldsZhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 379–387. Macdonald, F. A. & Cohen, P. A. 2011. The Tatonduk inlier, Alaska-Yukon border. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 389–396. Macdonald, F. A. & Jones, D. S. 2011. The Khubsugul Group, northern Mongolia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 339–345.

Macdonald, F. A., Schmitz, M. D. et al. 2010. Calibrating the Cryogenian. Science, 327, 1241–1243. Master, S. & Wendorff, M. 2011. Neoproterozoic glaciogenic diamictites of the Katanga Supergroup, Central Africa. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 173–184. Misi, A., Kaufman, A. J., Azmy, K., Dardenne, M. A., Sial, A. N. & de Oliveira, T. F. 2011. Neoproterozoic successions of the Sa˜o Francisco Craton, Brazil: The Bambuı´, Una, Vazante and Vaza Barris/Miaba groups and their glaciogenic deposits. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 509–522. Narbonne, G. M. & Gehling, J. G. 2003. Life after snowball: the oldest complex Ediacaran fossils. Geology, 31, 27 – 30. Ogg, J. G., Ogg, G. & Gradstein, F. M. 2008. The Concise Geological Time Scale. Cambridge University Press, Cambridge. Prave, A. R., Hoffmann, K.-H., Hegenberger, W. & Fallick, A. E. 2011. The Witvlei Group of East-Central Namibia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 211–216. Preiss, W. V. 2000. The Adelaide Geosyncline of South Australia and its significance in Neoproterozoic continental reconstruction. Precambrian Research, 100, 21 – 63. Preiss, W. V., Gostin, V. A., McKirdy, D. M., Ashley, P. M., Williams, G. E. & Schmidt, P. W. 2011. The glacial succession of Sturtian age in South Australia: the Yudnamutana Subgroup. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 701–712. Shields, G. A. 2006. Snowball Earth is dead! Long live Snowball Earth! Episodes, 29, 287–288. Shields-Zhou, G. A., Deynoux, M. & Och, L. 2011. The Record of Neoproterozoic Glaciation in the Taoude´ni Basin, NW Africa. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 163–171. Smith, M. D., Arnaud, E., Arnott, R. W. C. & Ross, G. M. 2011. The record of Neoproterozoic glaciation in the Windermere Supergroup, southern Canadian Cordillera. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 413– 423. Sovetov, J. K. 2011. Late Neoproterozoic (Vendian) glaciogenic deposits in the Marnya Formation, Oselok Group, in the foothills of the East Sayan Range, Southwestern Siberian craton. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 317–329. Stern, R. J. 2008. Neoproterozoic crustal growth: the solid Earth system during a critical episode of Earth history. Gondwana Research, 14, 33 – 50. Tait, J., Delpomdor, F., Pre´at, A., Tack, L., Straathof, G. & Kanda Nkula, V. 2011. Neoproterozoic Sequences of the West Congo and Lindi/Ubangi Supergroups in the Congo Craton, central Africa. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 185–194. Williams, G. E., Gostin, V. A., McKirdy, D. M., Preiss, W. V. & Schmidt, P. W. 2011. The Elatina glaciation (late Cryogenian), South Australia. In: Arnaud, E., Halverson, G. P. & ShieldsZhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 713–721.

Chapter 2 A history of Neoproterozoic glacial geology, 1871– 1997 PAUL F. HOFFMAN1,2 1

Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA, 02138, USA

2

School of Earth and Ocean Sciences, University of Victoria, Victoria, BC, V8W 2Y2, Canada (e-mail: [email protected]) Abstract: Neoproterozoic glacial records have been discovered on 23 palaeocontinents, their rate of discovery changing little since 1871. Yet, half of all the resulting publications appeared since 2000. The history of research before 1998 is described in five stages defined by publication spikes; subsequent work is not covered because historical perspective is lacking. In stage 1 (1871–1907), ‘Cambrian’ (now Neoproterozoic) glaciation was recognized successively in Scotland, Australia, India, Norway, Svalbard and China. Criteria for recognition included faceted and striated pebbles in matrix-supported conglomerates resting on ice-worn bedrock pavements. In stage 2 (1908–1940), Neoproterozoic glaciation was shown to have been widespread in Africa, Asia and the Americas. Major textbooks summarized these findings, but the rejection of continental drift (to account for late Palaeozoic glacial dynamics) put a chill on research. In stage 3 (1942– 1964), the occurrence of glacial deposits within carbonate successions, as well as nascent palaeomagnetic observations, suggested that Neoproterozoic glaciers reached sea-level at low palaeolatitudes, but the belated recognition of sediment gravity flowage caused glacial interpretations to be prematurely abandoned in key areas. In stage 4 (1965– 1981), the extent of Neoproterozoic glaciation was rethought in light of plate tectonics. Distinctive chemical sediments (iron + manganese formations and cap carbonates) were identified. In basic climate models, modest lowering of solar luminosity resulted in global glaciation due to ice-albedo feedback, and deglaciation due to greenhouse forcing resulted from silicate-weathering feedback in the carbon cycle. Neoproterozoic glacial geologists were blind to these ideas. In stage 5 (1982– 1997), reliable palaeomagnetic data combined with glacial marine sedimentation models confirmed that Neoproterozoic ice sheets reached sea level close to the palaeoequator.

The first Neoproterozoic glacial deposits were found in the SW of Scotland in 1871 and the rate of their discovery on 23 palaeocontinents and microcontinents was essentially linear until 1992 (Fig. 2.1). However, half of all the papers (n ¼ 811) written on them before the end of 2008 were published in the previous 10 years (Fig. 2.2). This surge in activity is the justification for the present volume, but this history ends at 1997. We lack historical perspective on the last decade and the endpoint eliminates the need to weigh any of the author’s own contributions on this topic. This chapter begins by placing the recognition of the Port Askaig boulder beds of Scotland as glaciogenic in its historical context, coming in 1871 on the heels of the discoveries of late Palaeozoic glacial deposits in India, Australia and South Africa. Their inspiration was the Pleistocene glacial controversy of 1837– 1865, which spawned both the orbital and ‘greenhouse’ theories of bidirectional climate change, and established the principal criteria for the recognition of past glaciation. The history is divided into five periods punctuated by spikes in publications (Fig. 2.2) associated with the 10th and 17th International Geological Congresses in Mexico City (1906) and Moscow (1937), respectively, the twin palaeoclimate conferences in Newcastle upon Tyne (1963) and Ko¨ln (1964), the IGCP Project 38 volume on Earth’s Pre-Pleistocene Glacial Record (Hambrey & Harland 1981), and the onset of the present surge triggered in 1998. The first period (1871 –1908) saw the recognition of ‘Lower Cambrian’ glaciogenic formations in SW Scotland (1871), South Australia (1884), NE Norway (1891), western Svalbard (1898), South China (1904) and NW India (1908). It ended with overviews of Late Palaeozoic and Lower Cambrian (‘possibly PreCambrian’) glaciations by Edgeworth David (1907a, b) at the 10th IGC, and the winning-over of a skeptical Geological Society in London by a fellow Australian (Howchin 1908). Faceted and striated clasts in diamictites were the decisive criteria for glaciogenesis, with a striated subglacial pavement beautifully exposed beneath ‘Reusch’s Moraine’ (Fig. 2.3) in Finnmark, Norway, for good measure. The second period (1909– 1941) saw Neoproterozoic glacial deposits recognized in southern, central and western Africa, central and northern Asia, eastern and western North America

and eastern South America (Fig. 2.1, Table 2.1). It was evident that glaciation ‘at or just before the beginning of the Cambrian’ had been more extreme than in the Pleistocene, possibly equal to the composite Late Palaeozoic glaciations of Gondwanaland. These findings were highlighted in a Symposium on Palaeozoic and Pre-Cambrian Climates at the 17th IGC (1937) in Moscow. Major textbooks made note of these developments (Ko¨ppen & Wegener 1924; Brooks 1922, 1926; Coleman 1926). In the third period (1942 – 1964), nascent evidence from palaeomagnetism and from the occurrence of glacial deposits atop thick carbonate-dominated successions in the North Atlantic region and within such successions in central and southern Africa implied that Neoproterozoic ice sheets had reached sea level in low palaeolatitudes. This inference met resistance and, after the discovery of turbidity currents as a cause of graded bedding (1950), some diamictites were reinterpreted as mass-flow deposits of non-glacial origin. These issues were discussed at major conferences on Palaeoclimates in Newcastle-upon-Tyne, England (1963), and Ko¨ln, Germany (1964). The radiation of macrofauna following soon after the ‘infra-Cambrian’ ice age was often remarked upon. The fourth period (1965– 1981) was notable for a detailed restudy of the classic Port Askaig Tillite, debate over the roles of true polar wander and plate tectonics in the distribution of glaciogenic deposits, and the recognition of widespread postglacial ‘cap dolomites’ and syn-glacial iron and manganese formations, features not known from Phanerozoic glaciations. Climate modelling received a kick-start when simple energybalance calculations suggested that the Earth could freeze over from pole to pole due to ice-albedo feedback in response to lowering of the solar constant by a few percent. Such a ‘white Earth’ disaster was shown to be self-reversing because of negative climate feedback associated with the geochemical cycle of carbon, but geologists were unaware of these developments. In the fifth period (1982 –1997), Neoproterozoic diamictites were increasingly interpreted in light of glacial marine sedimentation models and, combined with reliable palaeomagnetic data, proved that ice sheets had reached sea level close to the palaeoequator. The ‘Snowball Earth’ hypothesis, adapted from climate and planetary science, was advanced as a

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 17– 37. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.2

18

P. F. HOFFMAN

24 22 20 18 16 14 12 10 8 6 4 2 0

1880

1900

1920

1940

1960

1980

2000

Year of first report

1871 1884 1887 1891 1904 1910 1913 1915 1929 1930 1930 1933 1937 1950 1956 1959 1979 1981 1982 1984 1990 1992 2003 2009

overstep the bounds of prudent speculation in suggesting that those erratics are the reassorted materials of some great Northern Continent that has yielded to the ceaseless gnawing tooth of time, leaving scattered fragments as wreckage of its former greatness, and that the material of which the mass is composed have in time, deeper than we have hitherto suspected, been transported by the agency of ice.’

Laurentia (S) Australia India Baltica South China Avalonia Laurentia (N) Kalahari Congo Siberia Sao Francisco West Africa Tarim Cadomia Kazakhstan Amazonia Arequipa North China Arabia Rio Plata Mongolia Iran Nile ChukotkaArctic Alaska

James Thomson, F.G.S. 1871. On the stratified rocks of Islay. Report of the 41st Meeting of the British Association for the Advancement of Science, Edinburgh, John Murray, London, 110–111.

The Port Askaig boulder beds The first Neoproterozoic (then ‘Cambrian’) formation to be interpreted as glaciogenic (see quote above) was the Port Askaig, exposed on the east side of Islay (Thomson 1871, 1877), an island in the west of Scotland famous for its single malt whisky. The formation is a 750-m-thick succession of 47 mappable heterolithic-boulder diamictites with interbeds of crossbedded sandstone (Kilburn et al. 1965; Spencer 1971; Arnaud 2004). In his richly detailed sedimentological study, Spencer (1971) describes the Port Askaig as lying ‘at the same horizon, between two carbonate-rich formations, for 700 km from NE Scotland to western Ireland’. It is now thought to be the oldest of three distinct glaciogenic horizons within the Dalradian Supergroup (McCay et al. 2006) – older and younger Cryogenian and middle Ediacaran in age. James Thomson’s speculation (see above) about a ‘great Northern Continent’ as the source of the glacial debris turned out to be true (Cawood et al. 2003), although that continent did survive the ‘gnawing tooth of time’ after all. It is North America, or more precisely, its pre-Pangaean antecedent Laurentia. It was merely displaced from the Dalradian in the early Eocene, when Greenland separated from NW Europe. The Port Askaig was deposited at palaeolatitudes near 258S according to reliable palaeomagnetic data from mafic dykes, lavas and sills of the marginally older (723 –718 Ma) Franklin Igneous Suite of Arctic Laurentia (Evans 2000). It accumulated on a marine shelf, subject to repeated subaerial exposure (Spencer 1971; Johnston 1993), situated at the then southeastern margin

Fig. 2.1. Cumulative discovery of Neoproterozoic glaciogenic deposits by palaeocontinent. See Table 2.1 for locations and references. Note that Laurentia is counted twice because of the large geographical separation between southern (S) and northern (N) palaeohemisphere deposits of present eastern and western Laurentia, respectively.

parsimonious explanation for low-latitude glaciation, associated iron (+manganese) deposits and (later) cap carbonates. Initially ignored, the strong negative reaction to this hypothesis on the part of Neoproterozoic glacial sedimentologists will in future be seen as the most striking feature of the post-1998 period.

Prologue: discovery of the Port Askaig Tillite and its historical scientific context [On the stratified rocks of Islay.] ‘If . . . we compare the embedded boulders of granite [in the schist] with the granites found in situ throughout the Highlands, we feel the necessity of tracing them to another source, and hope we do not

1.0

60

60

Cumulative fractional growth 55

AGU GM146

0.8

(n = 811)

55

50

0.6

50

45

0.4

45

40

40 0.2

35

35

IGCP 38 Hambrey & Harland 1981

0.0

30

1880

1900

1920

1940

1960

1980

25

1880

1890

17th IGC Moscow 1937

15

10th IGC Mexico 1906

10

30 25

Newcastle and Koln “ conferences 1963-64

20 1870

2000

20 15 10

5 0 1890

5

1900

1910

1920

1930

1940

1950

1960

Fig. 2.2. Growth in the annual number of papers concerning Neoproterozoic glaciation, 1871–2008.

1970

1980

1990

2000

0 2010

HISTORY OF NEOPROTEROZOIC GLACIAL GEOLOGY

19

Late Palaeozoic glaciation and the theory of continental drift

Fig. 2.3. Iconic sketch of the glacially striated pavement beneath the end-Cryogenian Smalfjord diamictite (‘Reusch’s moraine’) at Bigganjargga, Varangerfjord, East Finnmark, northern Norway (from Reusch 1891).

of the palaeocontinent. The Port Askaig nicely illustrates the basic information needed to interpret Neoproterozoic glaciogenic deposits: their distribution, derivation, depositional environment, palaeogeographic setting and age.

Thomson’s (1871) interpretation of the Port Askaig came at a time of growing interest in pre-Pleistocene glaciation. A subglacial pavement overlain by diamictite in South Australia found in 1859 by A.C. Selwyn proved to be of Permo-Carboniferous age (Howchin 1912). A Permo-Carboniferous diamictite in the Talchir coal basin of NE India (Blanford et al. 1859) was shown to have been deposited by grounded, northward(!)-flowing glaciers (Fedden 1875; Koken 1907). The late Carboniferous Dwyka Tillite in Natal, South Africa, rests on a pavement scratched and grooved by onshore(!)-flowing glaciers (Sutherland 1870). Ironically, these findings had been inspired by the glacial interpretation of a Permian-age diamictite in England (Ramsay 1855) that proved to be non-glacial in origin. By 1907 it was evident that with continents in fixed relative positions, no amount of polar wander could prevent early Permian glaciation in the southern hemisphere from extending to within 108 latitude of the palaeoequator, while in the northern hemisphere tropical warmth stretched to the pole (Koken 1907; Irving 1956). ‘The Permian ice age poses an unsolvable problem to all models that do not dare to assume horizontal displacements of the continents’, wrote a 32-year-old German meteorologist (Wegener 1912). With the Atlantic and Indian oceans closed up, however, all the glacial action could be contained poleward of 458S palaeolatitude, comparable to the extent of northern hemisphere glaciation in the Pleistocene. Moreover, a radial pattern of flow could then be inferred, centred on south-central Africa (Martin 1981). ‘This would take everything mysterious away from the phenomenon’ (Wegener 1912). The distribution of Late Palaeozoic glacial deposits was prime motivation for Wegener’s theory of continental displacement, or drift (Martin 1981). Had he lived, vindication in the form of plate tectonics would have come to Wegener at the age of 87.

Table 2.1. First reported occurrences of Neoproterozoic glaciogenic deposits by palaeocontinent No.

Year

Palaeocontinent

1a 2 3 4 5 6 1b 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

1871 1884 1891 1904 1908 1910 1913 1915 1929 1930 1930 1933 1937 1950 1956 1959 1979 1981 1982 1984 1990 1992 2003 2009

Laurentia (south) Australia Baltica South China India Avalonia Laurentia (north) Kalahari Congo Siberia Sa˜o Francisco† West Africa Tarim Cadomia Kazakhstan Amazonia Arequipa North China Arabia Rio Plata Mongolia Iran Nile‡ Chukotka-Arctic Alaska

Continent Europe Australasia Asia North America* North America Africa

South America

*Before 1980, the Squantum Tillite was thought to be Late Palaeozoic in age. † Often considered part of the Congo palaeocontinent. ‡ Western margin of East African orogen.

Ref. Thomson (1871) Woodward (1884) Reusch (1891) Willis (1904) Holland (1908) Sayles & Laforge (1910) Hintze (1913) Rogers (1915) Beetz (1929) Nicolaev (1930) Moraes Rego (1930) Baud (1933), Furon (1933) Norin (1937) Wegmann et al. (1950) Nalivkin (1956) Maciel (1959) Caldas (1979) Mu (1981), Lu et al. (1985) Gorin et al. (1982) Spalletti & Del Valle (1984) Gibsher & Khomentovsky (1990) Hamdi (1992) Miller et al. (2003) Macdonald et al. (2009b)

20

P. F. HOFFMAN

The Pleistocene glacial controversy Interest in pre-Quaternary glaciation arose in the wake of the most acrimonious and far-reaching controversy in 19th-century geology, the brouhaha over the glacial theory for Pleistocene tills, erratic boulders and associated landforms (moraines, drumlins, eskers, kame-and-kettle, fluted ground, grooved bedrock, crag-and-tail, roches moutonne´es, whaleback rocks, cirques and hanging valleys, horns and areˆtes, U-shaped valleys and fjords). Glacial theory – that most of northern Europe and North America had been sculpted by enormous and dynamic ice sheets in the geologically recent past – was independently proposed by Esmark (1824) in Norway, Dobson (1925) in the USA, Venetz (1830) and Charpentier (1837) in Switzerland and Bernhardi (1832) in Germany. None of these papers – published in major journals by reputable professors (Esmark and Bernhardi), engineers (Venetz and Charpentier) and an industrialist (Dobson) – raised a ripple. No geologist ever overturned conventional wisdom with a single paper. Conventional theory held that these features were products of floods – meltwater floods resulting from dam bursts in the Alps (cf. de Saussure, von Humboldt, von Buch, de Beaumont) or iceberg-laden floods of Arctic origin in northern Europe and North America (cf. Sefstro¨m, Murchison, Hitchcock, Lyell). The controversy erupted when glacial theory was taken up by a young, ambitious, energetic and multilingual Swiss palaeoichthyologist (Agassiz 1837, 1840). A recent convert to glacial theory, Agassiz linked it to the extinction of the boreal megafauna (e.g. mammoth, mastodon, woolly rhinoceros, giant deer), a linkage easily disproved (Forbes 1846). He tied the end of the glacial period to the uplift of the Alps, exactly opposite to the view of Charpentier, his primary glaciological tutor, and laughable to those who knew that Alpine orogeny began in the Eocene. Of the geological establishment, only the former arch-Diluvialist William Buckland (Oxford University) was quickly won over to glacial theory. The climax of Agassiz’s campaign in support of glacial theory came in 1840 (Davies 1968), when he and Buckland toured the British Isles, finding varied and widespread evidence of glacial action (Agassiz 1842). The evidence included tills (boulder clays) with polished, faceted and striated clasts, moraines marking the limits of former glaciers, streamlined bedrock and perched terraces resulting from ice-dammed lakes. At first, Agassiz’s findings aroused great interest in the English-speaking world, but within months of his tour the glacial hypothesis was rejected by most geologists as unworkable (contrast Lyell 1840, 1841; Hitchcock 1841). Thereafter, glacial theory would languish until a new generation of Scottish geologists undertook systematic surveys that provided (to their own surprise) overwhelming evidence that glacial theory was correct after all (Ramsay 1860; Jamieson 1862, 1863, 1865; A. Geikie 1863, 1865; J. Geikie 1874). Agassiz had to wait 25 years for vindication. Had he lived, Jens Esmark would have been 99 years old. Lyell, who had ‘an undeniable penchant for the skilful defence of lost causes’ (Cunningham 1990, p. 247), never completely gave up on his decidedly non-uniformitarian iceberg-drift hypothesis. Conversely, it is insufficiently appreciated today that many of the criteria still used to distinguish ancient glaciogenic deposits were clearly spelled out over 170 years ago by Esmark, Dobson, Venetz, Bernhardi, Charpentier and Agassiz.

The problem of bidirectional climate change and the discovery of the ‘greenhouse’ effect The idea of a glacial period, or ‘Ice Age’ (Schimper 1837), was not controversial to geologists per se: characteristic fauna and flora now live c. 1000 km farther north of their occurrences in the ‘Drift’ (Smith 1836, 1839; Forbes 1846). However, to physicists

interested in climate, it came as a shock. They had assumed the the Earth’s climate was slowly cooling over time due to the dissipation of primordial heat in the Solar System according to the nebular hypothesis. This was consistent with undisputed palaeobotanical evidence (e.g. palm fronds in Switzerland) that most regions were warmer in the Cretaceous and early Cenozoic than they are today. If the climate could change dramatically in either direction, previously unknown factors must be at work in the climate system. It fell to the Irish physicist, mountaineer and orator John Tyndall (1861, 1863) to demonstrate experimentally that certain gaseous molecules in the atmosphere, notably water vapour and carbon dioxide (CO2), absorb infrared radiation (‘obscure rays’) emitted by the Earth but are transparent to sunlight. Fourier (1824) had previously mentioned the possibility of such a ‘greenhouse’ effect, but had rejected it as quantitatively unimportant. For Tyndall (1863, p. 204), water vapour provides ‘a blanket more necessary to the vegetable life of England than clothing is to man’. Radiative energy balance is maintained because the atmosphere ‘constitutes a local dam, by which the temperature of the Earth’s surface is deepened: the dam, however, finally overflows, and we give to space all that we receive from the Sun’ (Tyndall 1863, p. 205). Tyndall understood that water vapour cannot be the ultimate cause of climate change, because its concentration is itself a strong function of temperature. It amplifies any given climate change but cannot be its ultimate cause. On the other hand, the content of CO2 in the atmosphere, or any of the hydrocarbon gases, can vary independently. ‘It is not, therefore, necessary to assume alterations in the density and height of the atmosphere to account for different amounts of heat being preserved to the Earth at different times’, wrote Tyndall (1861, p. 277), ‘a slight change in its variable constituents would suffice for this. Such changes may in fact have produced all the mutations of climate which the researches of geologists have revealed’ [italics added]. Unknown to Tyndall, a prematurely deceased French ceramicist had already established the basis for the changes in CO2 that he required, through the geochemical cycle of carbon (Ebelmen 1845, 1847; see also Hunt 1880; Berner & Maasch 1996).

1871 –1908: pioneering discoveries Following the first identification of glaciogenic deposits now known to be Neoproterozoic in age in Scotland (Thomson 1871), similar discoveries were made in Australia (Woodward 1884), Norway (Reusch 1891), Svalbard (Garwood & Gregory 1898), South China (Willis 1904) and India (Holland 1908). The deposits in the extreme NE of Norway made the biggest initial impact. Hans Reusch (1852– 1922), then Director of the Norwegian Geological Survey, found them around the head of Varangerfjorden, not far from the Russian border. Reusch (1891) describes hills underlain by up to 50 m of non-stratified, matrix-supported conglomerate in which cobbles of Archaean gneiss and granite predominate, but which also carry smaller clasts of dolomite with facets and nonparallel striations identical to those found in Quaternary tills throughout Norway. On the coast of the fjord near the lighthouse at Bigganjargga, about 40 minutes’ walk east from Karlebotn, Reusch (1891) found an isolated ridge of diamictite c. 70 m long  8 m wide  3 m high (Bjørlykke 1967; Edwards 1975). The ridge belongs to the older (Smalfjord Formation) of two glaciogenic intervals in the region, and is onlapped and draped by shallow-marine sandstone. Just above the waterline of the fjord, the diamictite has eroded away to reveal a glacial pavement on crossbedded quartzite carrying glacial striations in two main orientations (Fig. 2.3). The less distinct (east – west) set of striations preserves lateral ridges and streaks of cataclasite, which suggest that the overlying diamictite is the melt-out tillite of a stagnant

HISTORY OF NEOPROTEROZOIC GLACIAL GEOLOGY

ice-cored moraine (Edwards 1975). ‘Reusch’s Moraine’ has been restudied by generations of geologists (Strahan 1897; Schiøtz 1898; Dal 1900; Holtedahl 1918; Rosendahl 1931, 1945; Føyn 1937; von Gaertner 1943; Crowell 1964; Reading & Walker 1966; Bjørlykke 1967; Edwards 1975, 1997; Jensen & WulffPedersen 1996; Rice & Hofmann 2000; Bestmann et al. 2006; Arnaud 2008) and, despite differences in interpretation, it remains an icon of Neoproterozoic glacial geology. Although they were not the first to recognize glacial action in ‘Cambrian’ (later Adelaidean) strata of South Australia, Walter Howchin in Adelaide and Edgeworth David in Sydney combined to bring its extent and importance to wide attention (Cooper 2009). Howchin (1901, 1903, 1908) traced the older of two glaciogenic intervals from its type section in the Sturt River gorge near Adelaide over an area of nearly 800  400 km2 (Sprigg 1986), suggesting that ‘the ice gathered on a plateau of comparatively low relief’ (David 1907b). The younger (Marinoan) glaciogenic interval was first recognized by Jack (1913; see also Mawson 1949b; Preiss 1987). At first, Howchin’s glacial interpretation was greeted skeptically in Australia, but his convincing photographs of striated clasts (Howchin 1908) were accepted without dissent by the Geological Society in London. If Howchin was responsible for much of the legwork in South Australia, David was intrumental in getting the word out overseas (Cooper 2009). Reports of the 10th IGC (Mexico City, 1906) include his 46-page synthesis (David 1907a), as well as two shorter papers (one in French) on pre-Cenozoic glacial epochs globally, with emphasis on the Australian glacial record. He also visited India, Great Britain and the USA during his 1906 travels and doubtless raised the awareness of ‘Cambrian’ glaciation. In the lower gorges of the Yangtze River in western Hubei Province, South China, Bailey Willis and Eliot Blackwelder described a body of ‘till’ of probable ‘Lower Cambrian’ age that carries heterogeneous boulders exhibiting polish and striae of ‘unquestionably glacial’ origin, based on examination of specimens by Quaternary glacial geologist T.C. Chamberlin (Willis 1904). The Nantuo tillite (Lee & Chao 1924) extends for 1200 km from NE to SW across the South China platform (Lee 1936; Lee & Lee 1940) and represents the younger of two discrete Cryogenian glaciations in the region (Wang et al. 1981). Three years later, a similar and likely correlative body (Jiang et al. 2003) was described in the Blaini Formation of the Lesser Himalaya, northern India (Holland 1908).

1909 – 1941: period of globalization The number of palaeocontinents subject to glaciation in what is now known to be Neoproterozoic time grew from five in 1908 to fourteen in 1937 (Fig. 2.1, Table 2.1), with new finds in eastern and western North America, central and northern Asia, eastern South America, and southern, central and western Africa. By the time of the 17th IGC in Moscow in 1937, Neoproterozoic glaciation was known on every continent except Antarctica. In 1913, a boulder slate at Big Cottonwood Canyon in the Wasatch Mountains of Utah was interpreted as glacial in origin (Hintze 1913; see also Blackwelder 1932). Earlier, the Squantum ‘Tillite’ member of the Roxbury Conglomerate in the Boston Basin of eastern Massachusetts was interpreted as glaciogenic (Sayles & LaForge 1910; Sayles 1914). This unit has a checquered history. As a result of miscorrelation, it was long considered to represent the only Carboniferous glaciation in the northern hemisphere. Its glacial origin was repeatedly questioned (e.g. Dott 1961) and later its age. Only after 1980 was its mid-Ediacaran age recognized through geological mapping, geochronology and palynology (Lenk et al. 1982; Thompson & Bowring 2000). In Ediacaran time, the Boston Basin was part of the ribbon continent Avalonia –Cadomia, located off North Africa. It

21

did not become part of Laurentia until the late Silurian (Wilson 1966). In southern Africa, glaciogenic diamictites of Neoproterozoic age were first recognized in the Numees Formation of northwesternmost South Africa and adjacent Namibia (Rogers 1915). ‘Anyone familiar with the southern Dwyka [Tillite] who was suddenly put down on the Numees beds would feel sure he was on the Dwyka again’ comments Rogers (1915, p. 89). In fact, distal facies of the Dwyka (Carboniferous) rest unconformably on the Numees, ‘which resembles the Dwyka tillite of the southern Karoo more closely than the Dwyka beds in the same neighbourhood do’ (Rogers 1915, p. 90). There were no new discoveries between 1915 and 1929 – who knows how many careers were cut down in World War I? – but four book-length syntheses of ancient climates appeared in rapid succession (Brooks 1922, 1926; Ko¨ppen & Wegener 1924; Coleman 1926). In the more theoretical of his books, the English meteorologist Brooks (1926) explicitly recognized and attempted to rationalize the existence of two distinct climate states, ‘glacial’ (eo-Cambrian, late Palaeozoic, late Cenozoic) and ‘nonglacial’ (most of Phanerozoic time). His names are preferable to the later terms ‘greenhouse’ and ‘icehouse’ used synonymously. Ko¨ppen & Wegener (1924) is justly famous – despite having never been translated into English – for its detailed treatment of Phanerozoic palaeoclimates in terms of continental drift, as well as for inserting the first of Milankovic’s graphs of northernhemisphere summer insolation calculated for different latitudes over the past 650 000 years. It devotes three pages to the Neoproterozoic (‘Kambrium’) glacial record. Coleman (1926), who had earlier proposed a glacial origin for the ‘slate conglomerates’ of Huronian age (early Palaeoproterozoic) in Ontario, Canada, when they were thought to be the oldest known sedimentary rocks (Coleman 1907), devotes a full chapter to glaciation ‘at or just before the beginning of the Cambrian’ (Coleman 1926). He concludes that the eo-Cambrian glaciation was more extreme than the Pleistocene and possibly equal to the composite late Palaeozoic glaciations, which were known to be diachronous, younging from west to east (Du Toit 1922). The same conclusion was reached by Kulling (1934), who updated the global picture (Table 2.1) at the end of a detailed account of Neoproterozoic stratigraphy in the NE of the Svalbard Archipelago of Arctic Europe. The pace of new discoveries picked up again from 1929 to 1937 with the recognition of glaciogenic formations in the Katangan Series of southern Zaire (Beetz 1929), beneath the Bambuı´ Group on the Sa˜o Francisco craton of central Brazil (de Moraes Rego 1930; see also Isotta et al. 1969), on the Yenisey Ridge at the western margin of the Siberian craton (Nikolaev 1930), in the Taoudeni Basin of the West African craton (Baud 1933; Furon 1933) and in the central Tien Shan of NW China (Norin 1937). A major symposium on Palaeozoic and Pre-Cambrian climates was held at the 17th IGC in Moscow in 1937 (published in 1940 as Report 6). Ten substantial papers describe Neoproterozoic glacial deposits in different parts of Asia, Africa, Australia, Europe and North America. Highlights include first-hand accounts of various pre-Dwyka glacial deposits throughout southern Africa (Gevers & Beetz 1940), the vast extent of the Nantuo Tillite in South China (Lee & Lee 1940) and the existence of glacial deposits in central Africa close to the equator (Davies 1940; Robert 1940). A consensus emerged that ‘eo-Cambrian’ glaciation was matched only in the late Palaeozoic, and that invertebrate animals arose following retreat of the ice sheets. However, the rejection of continental drift in the late 1920s following intense debate (Newman 1995; Oreskes 1999) had a chilling effect on pre-Pleistocene palaeoclimate research. If late Palaeozoic glaciation was allowed to remain ‘a hopeless riddle’, as Wegener put it, what chance was there to understand glacial periods twice their age?

22

P. F. HOFFMAN

1942– 1964: rebutting challenges The consensus that emerged from the Moscow Congress of 1937 carried over initially into the postwar period. In a 1948 Royal Society lecture in Sydney, Australia, the normally reserved Sir Douglas Mawson (Sprigg 1986) proclaimed that ‘the world must then have experienced its greatest Ice Age’ and speculated that prolonged genetic isolation during the period of refrigeration contributed to the subsequent Cambrian radiation of life (Mawson 1949a). However, Mawson was a fixist and therefore unduly impressed by the evidence for glaciation in equatorial Africa (Beetz 1929; Davies 1939, 1940). But what were mobilists to do? Plotting ‘infra-Cambrian’ glacial deposits on Pangaea reconstructions, assuming them to be closer to reality than the present geography, still left some deposits close to the palaeoequator (cf. Harland 1964a, b; Harland & Rudwick 1964). On the other hand, if the continents were tightly clustered, they might conceivably have crossed the poles diachronously in infra-Cambrian time due to polar wandering and/or continental drift. In that case, the extent of glacial deposits could far exceed the extent of ice at any time. What was needed were means of determining the palaeolatitudes of individual deposits, so that the ice extent inferred would not depend on correlation.

Palaeomagnetism and the meridional extent of Neoproterozoic ice sheets In the early 1950s, statistical corroboration of the geocentric – axial –dipole hypothesis by Jan Hospers (Frankel 1987; Irving 2008) had opened the way for palaeomagnetic and palaeoclimatic testing of polar wandering and continental drift (Irving 1956, 1959; Runcorn 1961). Wegener and Du Toit were soon vindicated, drift was unavoidable, and the revolution in the interpretation of ocean basins followed in 1962–1967. Palaeomagnetism offered a means by which the palaeolatitudes of ancient glacial deposits could be determined. Irving (1957a) had shown that the average magnetic inclination in Carboniferous glacial varves in Australia is subvertical, implying high-latitude deposition, consistent with Wegener’s (1912, 1929) assumption based on their glacial origin. Preservation of primary (depositional) natural remnant magnetization in red clastic sediments of Precambrian age having been demonstrated (Irving 1957b; Irving & Runcorn 1957), it fell to Harland & Bidgood (1959) to first attempt this method in infra-Cambrian glacial deposits. Their preliminary results (from a small subset of samples) implied equatorial palaeolatitudes for glaciogenic formations in southern Norway and East Greenland. Harland’s palaeomagnetic work lacked the sophistication and reach of Runcorn’s group three miles to the west (there was little contact between the two Cambridge University laboratories) and it would be 27 years before a truly reliable palaeomagnetic result was obtained for a Neoproterozoic periglacial marine formation (Embleton & Williams 1986). Palaeomagnetists have long been interested in the palaeolatitudes of climate-sensitive sedimentary facies (e.g. thick carbonates, evaporites, redbeds, aeolianites, tillites) as a test of the hypothesis that the time-averaged geomagnetic field closely approximated a geocentric axial dipole over Phanerozoic and Proterozoic times (Blackett 1961; Opdyke 1962; Briden & Irving 1964; Briden 1970; Evans 2006). Thick shallow-water carbonates are today found only equatorwards of 358 latitude (Rodgers 1957) and their zonal distribution did not change between glacial and non-glacial periods of the Phanerozoic (Briden 1970; Ziegler et al. 1984; Opdyke & Wilkinson 1990; Witzke 1990; Kiessling 2001). There are two reasons for this. First, the saturation state of seawater with respect to calcium carbonate or dolomite varies with the relative, not the absolute, temperature. Second, in glacial periods, higher alkalinity fluxes due to glacial erosion

in high latitudes are compensated by reduced weathering rates in the cooler tropics. The occurrence of glacial marine strata directly above or sandwiched between thick carbonate successions in different areas (e.g. Harland & Wilson 1956; Ziegler 1960; Katz 1961; Martin 1965a) suggested that, unless the glaciogenic and carbonate strata are everywhere separated by stratigraphic gaps of large magnitude, tidewater glaciers existed in the warmest parts of the surface ocean, consistent with nascent palaeomagnetic evidence for low palaeolatitudes (Girdler 1964). If glaciers occurred in relatively warm areas, including marine platforms where no mountains existed, higher latitudes and elevations must also have been glaciated. One is compelled to infer glaciation at all palaeolatitudes, independent of any assumption with respect to correlation (Harland 1964a, b). This is the crux of the Neoproterozoic glacial conundrum.

The Newcastle and Ko¨ln palaeoclimate conferences of 1963 – 1964 The spectre of glaciation at low palaeolatitudes caused some to question the glacial origin of Neoproterozoic ‘pebbly mudstones’ (Crowell 1957; Schermerhorn & Stanton 1963; Schermerhorn 1974). Since the recognition of turbidity current as a cause of graded bedding (Kuenen & Migliorini 1950; Natland & Kuenen 1951), mass-flows loomed as alternatives to glaciers in some areas. This would not, however, account for the disproportionate occurrence of diamictites in late Neoproterozoic times. Martin et al. (1985) suggested that mass-flows at low palaeolatitudes might have been triggered by glacioeustatic changes driven by high-latitude glaciation, but late Palaeozoic glaciation did not result in low-latitude diamictites comparable in setting and extent to those of the late Neoproterozoic. These were among the issues discussed at a pair of palaeoclimate conferences that proved to be a watershed for Neoproterozoic glacial geology (Fig. 2.2). The NATO Palaeoclimates Conference held at Newcastle-upon-Tyne in January 1963 was the first to highlight the topic since the Moscow symposium in 1937. Girdler (1964) reviewed the palaeolatitudes of continents during preMesozoic glacial periods according to existing palaeomagnetic data. The Permo-Carboniferous and Huronian (early Palaeoproterozoic) glaciations occurred at high palaeolatitudes, affirming the data, but middle and low palaeolatitudes were registered for the eo-Cambrian glaciations. There were three possible explanations (Girdler 1964): the (eo-Cambrian) palaeomagnetic data were faulty, the rocks are not glacial in origin, or the whole globe was glaciated. An entire session was devoted to the recognition of glacial sediments and of till-like deposits of non-glacial origin (Schwarzbach 1964a; Crowell 1964; Heezen & Hollister 1964). In the Discussion, geologist Wally Pitcher and others pushed back against non-glacial interpretations, citing detailed stratigraphic studies on the Port Askaig boulder beds (Kilburn et al. 1965). Harland (1964a) provided a global synthesis of Precambrian glacial deposits and their stratigraphic settings, concluding that infra-Cambrian glaciation had extended into the marine environment at low palaeolatitudes, unlike any younger ice age. An early tectonic ‘mobilist’, his views were informed by extensive fieldwork on the diamictite-bearing Hecla Hoek succession in Svalbard (e.g. Harland & Wilson 1956; Wilson & Harland 1964). He urged that such a glaciation provided an exceptional basis for global temporal correlation. Rudwick (1964) discussed the glacial aftermath as an ecological cradle for the Cambrian radiation. Four months after the Newcastle meeting, another global synthesis of Palaeozoic and Precambrian glaciation appeared, graced with fine images of Neoproterozoic glacial debris from central Africa (Cahen 1963). In March 1964, Manfred Schwarzbach convened an international symposium on palaeoclimates in Ko¨ln, simultaneous with the publication in English translation of his book Climates

HISTORY OF NEOPROTEROZOIC GLACIAL GEOLOGY

of the Past (Schwarzbach 1964b). The book has a chapter on eo-Cambrian glaciation, the subject of 7 of 37 papers from the Ko¨ln symposium published in the Geologische Rundschau. Harland (1964b) again summarized the global infra-Cambrian glacial record, and Chumakov (1964, 1992) reviewed the growing number of Precambrian tilloids in the (European) USSR. Spjeldnaes (1964, p. 38) affirmed the glacial origin of strata in different parts of Norway and Svalbard, noting that the eo-Cambrian deposits had been studied ‘with the same methods, in the same regions and by the same scientists as have the more easily interpreted Quaternary deposits’. Perhaps the defining summation of the infra-Cambrian glacial problem from this period is Harland & Rudwick (1964) in the popular science magazine The Scientific American – which is still worth reading.

1965 – 1981: in the wake of the plate tectonic revolution Plate tectonics revolutionized the Earth sciences, and Neoproterozoic glacial geology was no exception. It freed the distribution of ancient glacial deposits from the assumption of fixed relative positions of the continents, as foreseen by Harland (1964a) and (earlier) Wegener (1912), just as the concept of polar wandering (Gold 1955) had freed them from fixed position with respect to the poles. Equally important, plate tectonics provided a basis for making sense of geochemical cycles, and thereby to understand the controls on and changes to the chemical compositions of the oceans and atmosphere (Siever 1968). This opened up the possibility of understanding secular changes in atmospheric CO2 concentration – the dream of Ebelmen (1845, 1847), Tyndall (1861, 1863) and Chamberlin (1898, 1899) – and consequent changes in climate through ‘greenhouse’ radiative forcing. This led to renewed interest during the period 1965 –1981 in certain chemical sediments characteristically associated with Neoproterozoic glaciogenic deposits (banded iron- and manganese-formations, and post-glacial ‘cap’ carbonates). Finally, the period saw the publication of an influential textbook (Sugden & John 1976) that did much to invigorate the study of processes of glacial erosion, sedimentation and landscape development.

Centenary of Thomson’s (1891) study of the Port Askaig Tillite There could have been no finer centennial celebration than the publication of Anthony (Tony) Spencer’s Geological Society Memoir on Late Pre-Cambrian Glaciation in Scotland (Spencer 1971). It remains the gold standard in Neoproterozoic glacial sedimentology. The centenary was also marked by papers symbolizing the disunity over the hot-button issue of diachronous or synchronous glaciation (Crawford & Daily 1971; Dunn et al. 1971). If the continents were tightly clustered, it was conceivable that continental drift or polar wandering could move them all at different times through the polar region, as was the case for Gondwanaland in the Palaeozoic (Du Toit 1922; Wegener 1929; Crowell & Frakes 1970). However, drift alone could not account for Palaeozoic glaciations, because there were long stretches of Palaeozoic time when Gondwanaland was situated over the South Pole yet no large ice sheets existed.

Associated chemical sediments: banded iron- and manganese-formations and cap carbonates Sedimentary iron- and manganese-formations (BIFs) were the focus of worldwide mineral exploration for post-World War II reconstruction. Major deposits in SW Brazil (Dorr 1945; Almeida 1946; Walde et al. 1981), NW Canada (Ziegler 1960;

23

Young 1976; Yeo 1981), Namibia (Martin 1965a) and Australia (Whitten 1970) turned out to be intimately associated with Cryogenian glaciogenic diamictites. None is closely associated with volcanics. They are the only large-scale BIFs in the stratigraphic record younger than the Palaeoproterozoic (1.9 Ga) deposits around the Superior craton of North America. Martin (1965b) suggested that ‘this peculiar combination of sediments’ might be attributed to ‘oxygen deficiency in stagnating bottom waters caused by an ice cover’ (p. 116). The presence of a generally rather thin but remarkably continuous layer of carbonate, directly above Neoproterozoic diamictites, was occasionally noted by early workers (David 1907b; Norin 1937; Robert 1940; Mawson 1949b). These descriptions now become more specific, noting that the ‘capping dolomite’ is a distinctive pink or less commonly cream colour, invariably laminated or thin-bedded, and persists over basement highs where the diamictite cuts out (Dow 1965; see also Biju-Duval & Gariel 1969; Dunn et al. 1971; Rankama 1973; Deynoux & Trompette 1976; Plummer 1978; Williams 1979). Williams (1979) was the first to report stable isotope values (measured by Karlis Muehlenbachs) for ‘cap dolostones’ (and the first to use the term), from the Kimberleys of Western Australia. He notes that their d18O is similar to other Neoproterozoic dolostones, but their d13C is moderately depleted. Assuming the dolomite to be primary or early diagenetic, he concludes that ‘abrupt climatic warming at the close of late Precambrian glacial epochs is implied’, consistent with the d18O data (Williams 1979).

Climate models and the ‘white Earth’ instability The geological literature in this period was fixated on the use or misuse of glaciation for the division of Neoproterozoic time (e.g. Harland 1964a; Crawford & Daily 1971; Dunn et al. 1971; Rankama 1973; Schermerhorn 1977). The cause of glaciation was barely mentioned – Harland (1964a) suggested diminished solar forcing and Williams (1972, 1974, 1975) appealed to large variations in the Earth’s orbital obliquity (see below). Climate physicists, however, made a startling discovery when they used radiative enery-balance equations to calculate surface temperatures as a function of latitude and variable solar forcing, with simple parameterizations of meridional heat transport and feedbacks due to snow, ice and clouds. When they reduced solar irradiance by a few percent, surface temperatures fell below freezing everywhere due to runaway snow and ice-albedo feedback (Eriksson 1968; Sellers 1969, 1990; Budyko 1969; see also North 1990). Although decadal to millenial solar variability is only c. 0.1%, solar luminosity is thought to have slowly increased by 25–30% since 4.5 Ga due to the production of helium in the Sun’s core. The energy-balance calculations were therefore assumed to be in error because they predicted that the Earth should have been permanently frozen: irradiance roughly 25% higher than present is required to overcome the albedo of an icecovered planet, which reflects over 60% of the sunlight it receives. This problem was a stimulus for the nascent science of climate modelling. However, the ‘white Earth’ instability (Wetherald & Manabe 1975) turned out to be a robust feature not only of simple energy-balance models but also of most atmospheric general-circulation models. In 1981, three planetary scientists proposed a solution to the ‘white Earth’ problem (Walker et al. 1981). They appealed to the geochemical cycle of carbon (which is not accommodated in physical climate models because of its timescale of c. 106 years). CO2 is supplied to the ocean and atmosphere by metamorphicvolcanic outgassing and is consumed by silicate rock weathering. The latter is temperature-dependent because of the direct effect of temperature on reaction kinetics and also because moisture in wet regions, where most weathering occurs, increases with temperature (Clausius – Clapeyron relationship). The

24

P. F. HOFFMAN

temperature-dependence of weathering provides a negative climate feedback (abiotic Gaia) that acts as a natural thermostat. If global temperatures rise (or fall) for any reason, CO2 is consumed at a faster (or slower) rate, thereby limiting the temperature change. In effect, the feedback slowly adjusts the atmospheric CO2 concentration to maintain a balance between CO2 sources and sinks. As solar luminosity increased over geological time, atmospheric CO2 concentration adjusted downwards, keeping surface temperatures within the range suitable for life. This simple hypothesis not only went a long way towards solving the ‘faint young Sun’ paradox (Sagan & Mullen 1972), it rationalized long-term climate change while explaining why such changes were limited in magnitude. In the penultimate paragraph of their classic paper, Walker et al. (1981) suggested that silicate weathering feedback had been responsible for averting a ‘white Earth’. They added, however, that ‘if a global glaciation were to occur, the rate of silicate weathering should fall nearly to zero, and carbon dioxide should accumulate in the atmosphere at whatever rate it is released from volcanoes. Even the present rate of release would yield 1 bar of carbon dioxide in only 20 Ma. The resultant large greenhouse effect should melt the ice cover in a geologically short period of time’ (Walker et al. 1981). If a ‘white Earth’ did occur, it would be self-reversing. Accordingly, its occurrence in deep time could not be ruled out a priori. This, in a nutshell, is the climatological concept behind the ‘Snowball Earth’ hypothesis (Kirschvink 1992). Walker et al. (1981) made no reference to ancient glaciation, Neoproterozoic or otherwise. This was not because they were unaware of geology. On the contrary, the ‘faint young Sun’ problem was brought into focus for Walker by his participation in the Precambrian Palaeobiology Research Group (PPRG) organized by micropalaeontologist J. William Schopf. (Third author Kasting and palaeomagnetist Joe Kirschvink were PPRG participants later in the decade.) By 1981, Harland’s great infraCambrian glaciation had fallen off the radar screen. Palaeomagnetism, on which high hopes were pinned, had encountered problems. The foremost was the susceptibility of sediments to lowtemperature chemical remagnetization (e.g. McCabe & Elmore, 1989). Overcoming these problems would require time-consuming stepwise chemical and thermal demagnetization, studies of magnetic mineralogy, ‘field tests’ to constrain the age of remnant magnetic components, and more sensitive magnetometers.

Earth’s Pre-Pleistocene Glacial Record volume In 1981, Cambridge University Press published Earth’s PrePleistocene Glacial Record (Hambrey & Harland 1981), a product of Project 38 (Pre-Pleistocene Tillite Project) of the International Geological Correlation Programme. This sought-after volume contains 58 unified descriptions of Neoproterozoic glaciogenic formations worldwide, and is a tribute to Harland’s vision and Hambrey’s dedication. It was synthesized in Harland (1983) and Hambrey & Harland (1985). An additional global synthesis of Neoproterozoic glaciogenic deposits was published by Chumakov (1981).

1982– 1997: the gathering storm There were many developments during this period – burgeoning information regarding glaciomarine processes and deposits, widespread use of carbon isotopes as a tool for correlation, acquisition of reliable palaeomagnetic constraints on palaeolatitudes of proximal glaciomarine deposits, increased awareness of the sedimentological peculiarites of post-glacial ‘cap’ carbonates, and growing interest in causal mechanisms for low-latitude glaciation.

Glaciomarine deposits are more likely than their terrestrial counterparts to be preserved in the geological record because they are less susceptible to erosion, but ice-proximal processes in the marine environment are difficult to study because they occur underwater and often under ice. Nevertheless, much has been gleaned from geomorphic and seismic surveys of polar continental shelves backed up by drilling programmes (e.g. Elverhøi 1984; Powell 1984, 1990; Alley et al. 1989; Barrett 1989; Boulton 1990; King et al. 1991; Powell & Domack 1995). A growing emphasis on glaciomarine processes and deposits is reflected in many books written during this period (e.g. Molnia 1983; Drewry 1986; Dowdeswell & Scourse 1990; Anderson & Ashley 1991; Hambrey 1994; Benn & Evans 1998; Anderson 1999). Neoproterozoic glacial geology benefited enormously from these developments, although the complexity of facies architecture poses a challenge given the limited three-dimensional exposure of ancient deposits and weak or non-existent constraints on rates of accumulation. Beginning in the mid-1950s, oxygen isotopes have been used as a basic tool for stratigraphic correlation of Quaternary marine carbonates and for estimating seawater temperatures and global ice-sheet volumes. Unfortunately, Neoproterozoic carbonates do not preserve their original mineralogy, and their oxygen isotope compositions are seriously compromised by aqueous diagenesis. However, carbonate rocks buffer the isotopic composition of the relatively low concentration of carbon in aqueous solution. Consequently, the carbon isotopic compositions of carbonate rocks resist alteration. Beginning in the mid-1980s, carbon isotopes were increasingly used for stratigraphic correlation of Neoproterozoic carbonates and to a lesser extent for organic-rich shales (Knoll et al. 1986; Knoll & Walter 1992; Kaufman et al. 1997). Studies show that Neoproterozoic glaciogenic formations are typically bracketed by negative isotopic excursions. The excursions are reproduceable and approximately five times larger in magnitude than spatial or depth variations in the modern ocean. Because atmospheric CO2 concentrations were likely higher in the Neoproterozoic than today because of the dimmer Sun, a larger reservoir of dissolved inorganic carbon should have damped spatial and depth variations of d13C in Neoproterozoic oceans. Therefore, the observed isotopic excursions are most easily explained as secular variations. Their regular occurrence directly before and after glaciogenic sedimentation suggests that glaciation was roughly synchronous in many areas, even if not every isotopic excursion is associated with glaciation. For those who accepted these basic principles, the arguments for diachronous glaciation receded as the body of Neoproterozoic carbon isotope data grew. The first robust palaeomagnetic support for glaciation at sea level in low palaeolatitudes came from the Elatina Formation in South Australia (Embleton & Williams 1986). The formation is a glaciomarine unit comprised of reddish sandstone, siltstone and diamictite (Lemon & Gostin 1990; Williams et al. 2008), and a 10-m-thick interval of rhythmically laminated siltstone has been the subject of repeated palaeomagnetic studies. The palaeopole determined by Embleton & Williams (1986) implies a palaeolatitude of c. 58 and, although no field test was performed, the result is considered reliable because the natural remanent magnetization is stably carried by detrital haematite. At the time of the study, the laminations were considered to be annual varves, and the rhythmic bundles of 12–14 laminae were thought to represent sunspot cycles (Williams 1981; Williams & Sonett 1985). Later, they were reinterpreted as tidal rhythmites (Williams 1989), supporting a glaciomarine origin for the associated diamictites. Field tests were subsequently carried out using soft-sediment folds (Schmidt et al. 1991; Schmidt & Willliams 1995) and polarity reversals (Sohl et al. 1999), which confirmed the primary nature of the remanent magnetization and the palaeolatitude of 7.58 or less. A somewhat larger palaeolatitude of c. 158 was obtained in a recent palaeomagnetic study (Raub & Evans

HISTORY OF NEOPROTEROZOIC GLACIAL GEOLOGY

2006) of the cap dolostone (Nuccaleena Formation) above the Elatina Formation, implying a certain amount of inclination flattening during compaction of the Elatina siltstone, but this hardly alters the thrust of the result because temperatures are broadly uniform across the tropics.

Cap carbonates Arguably the first person to fully appreciate that the basal Ediacaran (Marinoan) cap carbonate was not only ‘an important stratigraphic marker’ but also ‘a perplexing and paradoxical lithostratigraphic unit’ (Table 2.2) was Aitken (1991), a leading Rocky Mountain carbonate stratigrapher. Concerning the cap dolostone atop the younger glaciogenic unit (Stelfox Member of the Icebrook Formation) in the Mackenzie Mountains of NW Canada, Aitken (1991) stressed that the aphanitic microcrystalline dolostone was not deposited as fine-grained mud, but as silt and sand-sized peloids. Ubiquitous subparallel lamination is defined by oscillatory variation in peloid size. His photographs show well-sorted, coarse-grained peloids in layers only one or two peloids deep. Where coarse peloids form the base of a layer, they commonly rest on facets, apparently resulting from abrasion. The layers are not laterally continuous, and low-angle crosslamination is ubiquitous. The lamination is clearly mechanical in origin, interspersed locally with microbially stabilized masses of micropeloidal stromatolite (James et al. 2001). Aitken (1991) argued that the intrastratal ‘tepees’ for which the unit was informally named (Eisbacher 1981) are not true peritidal tepees as defined by Assereto & Kendall (1977; see also

25

Kendall & Warren 1987), which result from layer-parallel expansive growth associated with cementation under alternating vadose and phreatic conditions. The cap ‘tepees’ are not associated with breccias or void-filling cements, as are peritidal tepees, and in plan view their crestlines are always linear and parallel, not polygonal like true tepees (Assereto & Kendall 1977). Bundles of laminae may exhibit onlap or offlap relations on ‘tepee’ flanks. These metre-scale trochoidal bedforms were later analysed as giant wave ripples (Allen & Hoffman 2005). Large-scale stromatolites, within which are spaced, micriteor cement-filled tubes or gutters that maintain a palaeovertical (‘geoplumb’) orientation irrespective of the inclination of laminations in the host stromatolite, were described from cap dolostones in California (Cloud et al. 1974; Wright et al. 1978), Namibia (Hegenberger 1987) and later in Alaska, Brazil, Canada and Mongolia (Fig. 2.3). Aitken (1991) identified crystal fans and pseudo-stromatolites in limestone overlying the cap dolostone, correctly interpreting them as pseudomorphic after aragonite sea-floor cements. His photograph of ‘blades of calcite’ is from a unique layer of sea-floor barite (BaSO4) cement, variably calcitized, which occurs in the top ,10 cm of the cap dolostone for .150 km along a strike southeastward from the location of Aitken’s (1991) photograph. Thus, cap dolostones and overlying limestones display a panoply of idiosyncratic features (Fig. 2.3) that generally occur in a consistent vertical sequence (Hoffman et al. 1987). Aitken (1991) offered no genetic explanation for cap carbonates. Eyles (1993) suggested that they represent detrital rock ‘flour’, as proposed earlier for carbonate layers within glaciogenic diamictites (Fairchild 1983). A detrital origin could hardly

Table 2.2. Idiosyncratic sedimentary and early diagenetic features in cap dolostones Palaeocontinent Amazonia Arabia Arctic Alaska Australia Australia Australia Australia Baltica Congo Congo Congo India Kalahari Laurentia Laurentia Laurentia Laurentia Laurentia Laurentia Mongolia Mongolia South China Tarim West Africa West Africa West Africa

Cap dolostone (ref.) Mirassol d’Oeste (1) Hadash (2) Nularvik (3) Mount Doreen (4) Nuccaleena (5) Cumberland Creek (6) Lower Ranford (7) Lower Nyborg (8) Keilberg (9) Calcaire Rose (10) C1 (11) Upper Blaini (12) Dreigratberg (13) Ravensthroat (14) Noonday (15) Lower Canyon (16) Lower Dracoisen (17) Cranford (18) Hard Luck (19) Ol (20) Baxha (21) Lower Doushantuo (22) Lower Zhamoketi (23) Oued Djouf (24) Amogjar (25) Mid Sud-Banboli (26)

SCC – – – p p – – – p – – – – p – – – p p p – p – p – p

TPB – – – – – – – – – – – – – – – – – p p

DGB

– – p

– – – – – – – – – – – – – – – – – – – – p p

– p p p

– – p p

PEL

LAC

GWR

TBS

p

p

p

p

– p p

– p p

– p

– p

– p p

– p p p p p

– p

– p

– – – – – p

– – – p p p p p p

– – – p p p

– – – p p p

– p – – p

– – – – p

– – – – – –

– – – – – –

– – p – – – p p p – p – – p – p p – p p

– p – – – – – p

SFB – – – p – – – – – – – – – p – – – – – – p – – – – –

Abbreviations: SCC, sheet-crack cements; TPB, tepee breccia; DGB, diagenetic barite; PEL, peloids; LAC, low-angle cross-laminae; GWR, giant wave ripples; TBS, tube biostrome; SFB, seafloor barite. References: (1) Nogueira et al. (2003); (2) Allen et al. (2004); (3) Macdonald et al. (2009b); (4) Kennedy (1996); (5) Plummer (1978); (6) Calver & Walter (2000); (7) Corkoran (2007); (8) Edwards (1984); (9) Hoffman et al. (2007); (10) Cahen & Lepersonne (1981); (11) Cahen (1950); (12) Kaufman et al. (2006); (13) Macdonald et al. (2011); (14) James et al. (2001); (15) Corsetti & Grotzinger (2005); Corsetti & Kaufman (2005); Wright et al. (1978); (16) Hambrey & Spencer (1987); (17) Halverson et al. (2004); (18) McCay et al. (2006); (19) F. A. Macdonald, pers. comm.; (20) Macdonald et al. (2009a); (21) F. A. Macdonald & D.S. Jones, pers. comm.; (22) Jiang et al. (2006); (23) Xiao et al. (2004); (24) Bertrand-Sarfati et al. (1997); (25) Shields et al. (2007); (26) Ne´de´lec et al. (2007).

26

P. F. HOFFMAN

account for the systematic variations in d13C observed in cap dolostones (Kennedy et al. 1998; Hoffman et al. 2007). Kaufman et al. (1993) and Grotzinger & Knoll (1995) related cap carbonates to the overturn of alkalinity-charged, isotopically depleted, bottom water after a lengthy period of ocean stagnation. This suggestion requires that primary production be maintained in the absence of upwellings, which is the overwhelmingly predominant nutrient flux in the modern ocean. Upwellings would not be necessary if organic matter did not settle, but then alkalinity would not build up in the deep nor would a vertical isotopic gradient develop. This and the physical implausibility of prolonged (.10 000 years) ocean stagnation (in the absence of an ice cover) make the overturn hypothesis unconvincing, despite its popular appeal. Kennedy (1996) proposed that cap carbonates represent nonskeletal analogues of carbonates deposited (in low latitudes) during times of Quaternary deglaciation according to the ‘coral reef’ hypothesis (Berger 1982; Opdyke & Walker 1992). This results from the hypsometry of ocean basins. When sea level is lowered due to ice-sheet growth, the area of shallow shelves and inland seas is disproprotionately reduced. These areas are favoured for carbonate burial because the sediment never encounters undersaturated deepwaters. When these areas are lost, all carbonate production occurs in the open ocean and the alkalinity of deepwaters increases. Upon deglaciation, alkalinityrich waters flood the shelves, depositing carbonate and releasing CO2 to the atmosphere. The process can account for much of the glacial –interglacial variation in pCO2 (Opdyke & Walker 1992) and thus provides a positive climate feedback. This mechanism must have contributed to cap carbonates, but the estimated average thickness of carbonate produced (Ridgwell et al. 2003) is only approximately one-tenth of the average thickness of cap dolostones of 18.5 m (Hoffman et al. 2007, Table 2.1). The coral reef hypothesis does not explain why cap carbonates are only associated with the ultimate deglaciation, and do not accompany glacial –interglacial cycles within the glacial period where these are recognized (Leather et al. 2002; Allen et al. 2004). If cap carbonates were eroded during glacial readvances, they would be present as clasts in glacial deposits, which they are not. Nearly 70 papers concerning cap carbonates appeared in the decade after 1997. This reflects the central but contentious role they have assumed in the ongoing controversy over the nature of Cryogenian glaciation. Arguably they are the most richly enigmatic horizons in the entire stratigraphic record.

Causative theories for Neoproterozoic glaciation Walker et al. (1981) provided the conceptual basis for a selfreversing global glaciation from the point of the initial ‘white Earth’ (ice-albedo) instability. They did not speculate on how the climate ever reached that point because they credited silicateweathering feedback with its avoidance. However, a number of causative theories (not mutually exclusive) were proposed between 1971 and 1997 to explain Neoproterozoic glaciation (Table 2.3). Carbonate burial. Roberts (1971, 1976) was struck by the fact that the Cryogenian glaciation was preceded by thick successions of shallow-water carbonate strata in the North Atlantic (Greenland, Svalbard and the British Isles), western North America and central Africa. He postulated a period of anomalous carbonate burial in which atmospheric CO2 was sequestered, producing an ‘anti-greenhouse’ effect. He did not give any reason why this might have occurred. It would require a rise in the global rate of silicate weathering relative to the rate of CO2 outgassing (see below; Continental distribution and Continental break-up).

Table 2.3. Causative theories for Neoproterozoic glaciation Astronomical theories 1. Solar variation 3. Large obliquity 5. Ice-ring collapses 13. Impact ejecta 4. Giant molecular clouds Oceanographic theories 2. Carbonate burial 7. Ocean stagnation 10. Hypsometric effect 11. Organic burial Geodynamic theories 6. Continental distribution 9. Continental break-up 15. Basalt weathering 16. True polar wander Biological theories 8. Biocatalysed weathering 12. Methane destruction 14. Methane substitution

Harland (1964a) Williams (1972, 1975), Jenkins (2000), Donnadieu et al. (2002) Sheldon (1984) Bendtsen & Bjerrum (2002), Fawcett & Boslough (2002) Pavlov et al. (2005) Roberts (1971) Kaufman et al. (1993), Grotzinger & Knoll (1995) Kennedy (1996), Ridgwell et al. (2003) Kaufman et al. (1997) Marshall et al. (1988), Worsley & Kidder (1991), Kirschvink (1992), Schrag et al. (2002) Eyles (1993), Young (1995a), Donnadieu et al. (2004a) Godde´ris et al. (2003) Li et al. (2004) Carver & Vardavas (1994), Lenton & Watson (2004), Kennedy et al. (2006) Pavlov et al. (2000, 2003), Catling et al. (2001), Claire et al. (2006) Halverson et al. (2002), Schrag et al. (2002)

Large orbital obliquity. The Australian sedimentologist George

E. Williams always had an eye on ‘the big picture’ and possessed an early appetite for planetary orbital mechanics. Impressed by the prevalence of varves (i.e. rhythmic lamination, supposedly seasonal) and other seasonal indicators such as polygonal sand wedges associated with Neoproterozoic glaciogenic strata (Chumakov 1968; Spencer 1971), he proposed that the Earth’s obliquity (i.e. the angle between the equatorial and ecliptic planes) oscillated between large (e.g. Neoproterozoic) and small (e.g. Phanerozoic) values (Williams 1972). When the angle was large, the seasonal cycle was strong everywhere; when small, the seasons did not differ greatly in low latitudes (except in strongly monsoonal areas). In addition, as the angle increased, the meridional insolation gradient declined and actually reversed (i.e. greater annual insolation at the poles than at the equator) whenever the angle exceeded 548. Williams (1972, 1974, 1975) inferred that when the obliquity was very large, low latitudes were subject to glaciation preferentially. As low latitudes cover more area than high latitudes, there should be more ice (and higher planetary albedo) during periods of low-latitude glaciation than when ice is limited to the poles. Accordingly, glacial periods occur when obliquity is large and when there is strong seasonality everywhere (contrary to the Wegener – Milankovic hypothesis that ice-sheet growth is favoured by cool summers and low seasonality). Under large obliquity, glaciation at the poles is precluded by intense summertime insolation. The large obliquity hypothesis was strengthened by the discovery of 2.5-m-deep sand wedges associated with the Elatina Formation (Williams & Tonkin 1985), which, combined with palaeomagnetic evidence (Embleton & Williams 1986), indicates strong seasonality at low palaeolatitude (Williams 1993; but see Maloof et al. 2002). In the Phanerozoic, deep periglacial sand wedges occur only in middle and high latitudes, where seasonal forcing is strong (Lachenbruch 1962). With large obliquity, seasonality is strong at all latitudes. Like any theory, large obliquity was not without difficulties. Although a large (and chaotic) obliquity could have been imparted

HISTORY OF NEOPROTEROZOIC GLACIAL GEOLOGY

27

on the early Earth by stochastic accretion (Laskar & Robutel 1993), obliquity cannot oscillate between large and small angles because of ‘entrapment’ at small angles by gravitational attraction between the Moon and the Earth’s equatorial bulge (Laskar et al. 1993). Therefore, the theory was revised such that obliquity was persistently large until Ediacaran times, when it was rapidly reduced to small angles before the Cambrian (Williams 1993, 2000). This proved difficult to replicate in models or justify from orbital theory (Ne´ron de Surgy & Laskar 1997; Williams et al. 1998; Hoffman & Maloof 1999; Pais et al. 1999; Donnadieu et al. 2002; Levrard & Laskar 2003). Moreover, large obliquity over most of Precambrian time is at odds with a growing body of palaeomagnetic evidence that pre-Ediacaran evaporites and carbonate-dominated successions formed at latitudes inconsistent with a reversed climatic gradient (Park 1994; Buchan et al. 2001; Evans 2006; Maloof et al. 2006). However, decisive falsification of the large obliquity hypothesis, glaciation at high palaeolatitude, has yet to be observed.

ocean largely covered by pack ice, perhaps with ‘warm tropical “puddles” in the sea of ice, shifting slightly from north to south with the seasons’ (Kirschvink 1992). At the 1989 PPRG meeting at UCLA (Maugh 1989), he proposed that the ‘white Earth’ disaster had actually occured in the Neoproterozoic, perhaps repeatedly, and that each pan-glacial period was abruptly terminated when the slow build up of CO2 reached a critical level, as deduced by Walker et al. (1981). He attributed the deposition of banded iron-formation during Neoproterozoic glaciation to ocean stagnation and deepwater anoxia, as previously proposed by Martin (1964b), and he predicted that the abrupt climate switches accompanying glaciation and deglaciation should be represented in widely separate areas by lithologically similar strata, a result of the global scale of the climatic fluctuations (Kirschvink 1992). He called the glacial state a ‘Snowball’ Earth (Maugh 1989) to highlight the central role of planetary albedo in the phenomenon. It is an evocative description of how the late Cryogenian Earth might have looked from outer space.

Ice-ring collapses. Richard P. Sheldon was a sedimentary phosphorite specialist with the United States Geological Survey who developed an ‘outlandish hypothesis’ (Sheldon 1984) for Neoproterozoic glaciation following a field excursion in northern Mongolia, where phosphorites, carbonates and glacial marine strata occur in close stratigraphic proximity. He speculated that ice rings denser than those of Saturn orbited the Archaean Earth. In the Proterozoic, sublimation thinned the rings and they became more discrete. As the Moon pulled away from the Earth, their orbits slowly decayed and they successively entered the atmosphere. As each ring approached the Earth, its shadow caused an ice age. Upon entering the atmosphere and before the next ring approached, the shadow was lost and greenhouse gases were added, causing a warm interglacial. Once the final ring had fallen, the tropics were shadowless for the first time. Ocean overturning driven by the steepened climatic gradient led to phosphogenesis and organic diversification. Sheldon (1984) noted that seasonality would be greatest in glacial times, consistent with Williams’ (1993) observation, because the rings cast shadows only on the winter hemisphere. Again, this is contrary to the Wegener – Milankovic theory of ice ages, which posits that ice sheets grow when summers (not winters) are cool and seasonality is weak (Ko¨ppen & Wegener 1924; Milankovic 1941).

Impact ejecta. Impact ejecta figured in discussions of Neoprotero-

zoic glaciation in two completely different ways. They were proposed as an alternative emplacement mechanism for diamictite, implying that the low-latitude and carbonate-associated glacial problem was a chimera (Oberbeck et al. 1993; Rampino 1994). This was a small reenactment of the pebbly mudflow challenge of 1957 –1974. Impact ejecta aprons do in fact share many diagnostic features with glaciogenic diamictites (e.g. unsorted debris, faceted and striated erratic stones, outsized dropstones). However, continuous ejecta blankets extend only about one craterradius from the crater rim, regardless of crater size (Melosh 1989). Therefore, an improbably large flux of big impacts (roughly one Cretaceous –Tertiary sized impact every 100 ka for 200 Ma) would be required to account for the distribution of Neoproterozoic diamictites, given the low probability of their exposure at the Earth’s surface today. It was later proposed on the basis of modelling experiments that large impacts could have caused Neoproterozoic glaciation (Bendtsen & Bjerrum 2002; Fawcett & Boslough 2002). The models suggest that short-term (,10 year) shielding of sunlight by ejecta from a Cretaceous –Tertiary sized impact could cause the ocean surface to freeze over if seawater was as cold as today, unlike the Cretaceous ocean, which was 8– 128 warmer than present.

Continental distribution. Atmospheric pCO2 will adjust downwards

in response to an equatorward migration of the continents, because weathering rates are greatest in the tropics (Marshall et al. 1988; Worsley & Kidder 1991). Global cooling is therefore consistent with sedimentological and palaeomagnetic evidence for an unusual preponderance of low-latitude continents in the Neoproterozoic, ‘a situation that has not been encountered at any subsequent time in Earth history’ (Kirschvink 1992). Equatorward displacement of the continents also changes the planetary albedo: the tropical ocean is a strong absorber of sunlight, unlike land areas (minimally vegetated in the Neoproterozoic) and the fog-bound polar oceans, which are relatively good reflectors (Kirschvink 1992). Moreover, a positive albedo feedback would result from the enlargement of tropical land area that would accompany any glacioeustatic fall in sea level (Kirschvink 1992). A palaeomagnetist, Kirschvink (1992) accepted the low-latitude result for the Elatina glaciation (Embleton & Williams 1986) but rejected the large-obliquity hypothesis because of the cooccurrence of Neoproterozoic glacial deposits and thick carbonatedominated successions. If the meridional climatic gradient were reversed, the carbonate belts would move from low latitudes to the poles, ‘where the glaciers (in Williams’ model) should not encounter them’ (Kirschvink 1992). He was therefore compelled to assume that if ice sheets reached sea level close to the equator, higher latitudes must have been glaciated as well, wherever ablation did not exceed precipitation. He envisaged an

Ocean stagnation. As the patterns of C and Sr isotope variations associated with Neoproterozoic and Phanerozoic glaciations are distinct (Kaufman et al. 1993), the cause of glaciation might also be different. Kaufman et al. (1993) and Grotzinger & Knoll (1995) attempted to genetically link the isotopic patterns, glaciation and the associated Fe- and Mn-ore formations and ‘cap’ carbonates in a conceptual model involving prolonged ocean stagnation followed by overturn. They hypothesized that when the ocean was stagnant, organic export and subsequent respiration in anoxic deep waters would progressively enrich the surface waters in 13C, while simultaneously creating an isotopically depleted, bicarbonate-enriched reservoir at depth (Kaufman et al. 1993; Grotzinger & Knoll 1995). In their model, the attendant transfer of CO2 from the atmosphere to the deep ocean contributes to global cooling. Ultimately, the growth of sea ice triggers the formation of cold, saline, deep water, and the meridional overturning circulation is reestablished. Upwelling of alkalinityladen deepwater leads to the precipitation of isotopically depleted cap carbonates and the simultaneous release of CO2 to the atmosphere, which melts the ice (Kaufman et al. 1993; Grotzinger & Knoll 1995). The parsimony of the model temporarily outshone its flawed foundation. Short of covering the ocean with ice, physical stagnation (as distinct from dynamic stratification) on geological timescales is implausible because the energy driving the upwelling

28

P. F. HOFFMAN

flux (the rate-limiting step in the overturning circulation) derives from winds and tides (Wunsch 2002). Winds and tides are not so easily turned off. Moreover, if the surface ocean was ever deprived of nutrients upwelled from depth, primary production would crash, depriving the deep of its alkalinity pump and the isotopic gradient of its driver. To its credit, the overturn hypothesis at least lined up all the key observations in the same viewfinder: glaciation, isotopes, iron formations and cap carbonates.

only a handful of the highest peaks (.4.0 km above sea level) on the Ethiopean Dome were glaciated at the Last Quaternary Maximum (Osmaston 2004). For Neoproterozoic ice sheets to have reached sea level at the same latitudes implies a drastically colder climate.

Biocatalysed weathering. Carver & Vardavas (1994) constructed a

Figure 2.2 clearly shows that publications concerning Neoproterozoic glaciation ‘took off’ in the mid-1990s and possibly peaked in 2007. There is little doubt that the galvanizing factor was the Snowball Earth hypothesis (Kirschvink 1992; Hoffman & Schrag 2002). In the first four years after its publication, Kirschvink (1992) had been cited only three times, and favourably only once (Klein & Beukes 1993). However, as soon as it was advocated prominently (Hoffman et al. 1998; Hoffman & Schrag 2000) and in the public arena (Walker 2003), the reaction was swift. This is not the place to review all the papers that have appeared since 1998: more time and more space are needed. Opinions about the nature of Neoproterozoic glaciation are now at a stalemate. Most workers agree that one, and possibly two, Cryogenian glaciations occurred simultaneously on virtually all palaeocontinents (Evans 2000; Halverson 2006; Hoffman & Li 2009; but see Eyles & Januszczak 2003 for a contrary view). The basic argument is that if ice sheets reached sea level in the warmest areas (i.e. close to the palaeoequator and where thick non-skeletal carbonates accumulated), then colder areas must have been frozen as well. This argument makes no a priori assumption about correlation, but correlation follows from the premise. The stalemate concerns the state of the ocean: was it largely ice-covered as assumed in the Snowball Earth hypothesis (Kirschvink 1992; Hoffman & Schrag 2002), or substantially ice-free as posited in the so-called slushball solution (Hyde et al. 2000; Peltier et al. 2004; but see Bendtsen 2002; Voigt et al. 2011)? Because of subduction, direct evidence from the ocean floor has been eliminated, or survives only from post-Cryogenian (Ediacaran), regional-scale glaciation (Kawai et al. 2008). Those who favour the snowball model point to its ability to account for synglacial Fe- and Fe –Mn deposits, post-glacial cap carbonates, and isotopic evidence for highly elevated pCO2 during and after glaciation (Bao et al. 2008, 2009). The slushball solution does not account for these features (Pollard & Kasting 2005). Those who favour the slushball solution point to evidence that the snowball hypothesis is incompatible with the fossil record (Knoll 2003; Xiao 2004; Corsetti et al. 2006; Moczydlowska 2008) and also with the sedimentology of glacial deposits (e.g. McMechan 2000; Condon et al. 2002; Allen & E´tienne 2008). The biotic argument presupposes a knowledge of the limits to survival of species. The argument from sedimentology assumes knowledge of when the deposits were formed because the snowball hypothesis predicts that glaciation was ‘polar’ in character at the beginning and ‘temperate’ or ‘Alaskan’ at the end. Studies of modern ice-sheet stability show that outlet glaciers are buttressed by ice shelves (Dowdeswell et al. 2000; De Angelis & Skvarca 2003; Nick et al. 2009): iceshelf removal triggers ice-sheet deflation, with potential for ice-sheet collapse. On a Snowball Earth, the ice shelf is global. Iceshelf removal, the first step in snowball deglaciation, should trigger rapid deflation and collapse of low-latitude ice sheets. Much of the glacial sediment record on continental margins will date from this period. These deposits will bear evidence of open water (e.g. iceberg rafting and dumping, wave ripples) because open water then existed. Sedimentology is thus a clumsy means of testing the snowball hypothesis because of uncertainty over which stage of the glacial cycle is stratigraphically preserved (Hoffman 2005). Moreover, sediment fluxes cannot be estimated because chronology is lacking. Even varve chronology (De Geer 1912), from which the duration of the Holocene was estimated

time-dependent model of the Earth’s mean surface temperature controlled by the geochemical cycle of carbon (Walker et al. 1981). They parameterized CO2 outgassing as having declined rapidly before and more slowly after 3.5 Ma. The solar flux rose almost linearly, and the consumption of CO2 through silicate weathering was shaped by the rapid growth of 90% of the present continental crust between 3.5 and 2.0 Ga, and by stepwise increases in weathering efficiency (i.e. the rate of CO2 consumption for any given pCO2) resulting from biological colonizations of the land (Schwartzman & Volk 1991). The first colonization (microbial) is assumed to have occurred before 3.5 Ga. The second (organisms unspecified) supposedly occurred between 1.2 and 0.7 Ga, and the third in the middle Phanerozoic in response to the rise of vascular plants. The observed enrichment in 13C of most pre-Ediacaran carbonates (e.g. Knoll et al. 1986) is cited in support of the middle (and largest) step in weathering efficiency (Carver & Vardavas 1994), although the isotopic pattern requires an increase in fractional organic burial as opposed to overall carbon burial. Increased Neoproterozoic biocatalysed weathering receives some support from micropalaeontology (Horodyski & Knauth 1994), molecular divergence analysis (Heckman et al. 2001) and clay mineralogy (Kennedy et al. 2006), but remains controversial. The modelled temperature curves (Carver & Vardavas 1994) feature two relatively cold intervals in Earth history, early Palaeoproterozoic in response to continental growth and mid-Neoproterozoic in response to the prescribed increase in biocatalysed weathering. Continental break-up. Supercontinents tend to be dry because most

land is far from the ocean. Upon fragmentation, new continental margins are created and land overall is brought closer to the source of moisture. As silicate-weathering is catalysed by moisture as well as by temperature (Walker et al. 1981), continental break-up should result in lower pCO2 and a colder global climate. Cooling will be most severe if rifting occurs in the tropics, and least so for rifting in the polar regions where chemical weathering rates are low. Continental break-up in tropical Pangaea was followed by the coolest period (Late Jurassic –Early Cretaceous) of the Mesozoic era (Frakes et al. 1992); break-up in the polar North Atlantic by the warmest period (Early Eocene) of the Cenozoic (Zachos et al. 2001). This is consistent with Cryogenian cooling, related to the break-up and dispersal of equatorial Rodinia beginning c. 800 Ma (Li et al. 2008). These ideas have recently been tested with a simplified atmospheric general circulation model coupled to a model of the geochemical cycle of carbon (Donnadieu et al. 2004a, b). It is therefore difficult to accept the argument that Neoproterozoic glaciation could be explained in terms of a close spatial (as well as temporal) association with rift valleys and rifted margins, without recourse to extreme climate states (Eyles 1993; Young 1995a; Eyles & Januszczak 2003). [In Reply to a Comment, Young (1995b) wrote that he did not intend to imply a genetic relationship, but then his paper should have been titled, ‘Was the preservation of Neoproterozoic glacial deposits on the margins of Laurentia related to the fragmentation of two supercontinents?’.] The Late Quaternary glacial maxima were arguably as cold as any time in the Phanerozoic, but moraines on the mountains of the East African rift system do not extend below c. 3.5 km above sea level (Osmaston 2004). In the Red Sea rift,

Epilogue: the long road to consensus

Ramsay Jamieson A. Geikie

HISTORY OF NEOPROTEROZOIC GLACIAL GEOLOGY

Agassiz

10

Rejection

1.0 Revival

0.8 0.6 0.4

Venetz Bernhardi

Esmark

Playfair

Hutton

5

Hall MacCulloch Buch

Early Martyrs

29

0.2 0

0 1800

1820

1840 Year (AD)

with a fair degree of accuracy, is closed to us because of the relatively low palaeolatitudes of most Neoproterozoic deposits, where seasonality is weak. The decisive falsifying test is chronometric: any glaciation that is short-lived (,5 Ma) cannot have been a snowball glaciation because of the time required to accumulate enough CO2 to overcome the ‘white Earth’s’ albedo. It is the duration of the glaciation, not that of particular glacial deposits, that provides the test. The tightest existing constraint on the duration of the end-Cryogenian (Marinoan) glaciation is 19.3 + 4.2 Ma (Condon et al. 2005; Zhang et al. 2008). The older Cryogenian (Sturtian) glaciation is virtually unconstrained, whereas the midEdiacaran (Gaskiers) glaciation could not have been a snowball glaciation because of its short duration (Hoffman & Li 2009). If the snowball hypothesis is not falsified geochronologically, how and when might a consensus on the nature of Cryogenian glaciation come about? The history of the Pleistocene glacial controversy (Fig. 2.4) may offer a preview. Agassiz’s outspoken advocacy of the glacial theory of Esmark, Venetz and Bernhardi engendered much excitement, but failed to achieve consensus. In fact, with few exceptions (James Smith, Charles Maclaren, Robert Buckland, Robert Chambers and, on-and-off, Charles Darwin), British geologists soon turned hostile to the glacial theory and remained so from 1842 until 1860. One can track the fortunes of the theory through the writings of Charles Lyell (1841, 1845, 1851, 1852, 1855, 1857, 1863, 1865), or the Anniversary addresses of the President to the Geological Society, London (see also Woodward 1907; Davies 2007). Then, in the first half of the 1860s, everything changed. This was the result of comprehensive regional studies by a new generation of Scottish geologists led by Ramsay (1860), Jamieson (1862, 1863, 1865) and the Geikie brothers (1863, 1874) in what may be termed the ‘Scottish glacial revival’. When they began their studies, they had assumed the glacial theory was false because that was the prevailing view. Their own work, systematic and wide-ranging, left them no choice but to conclude that Agassiz had been correct after all. From their time onwards, critics of the glacial theory became the exception rather than the majority. One is tempted to think that consensus might be reached faster in the early 21st century than in the mid-19th century. This is doubtful. It is clear from the literature that Victorian geologists actually read each others’ papers. No one could claim that this is true today. Will consensus be achieved through some conceptual or technical breakthough? I doubt it: there have been several already and consensus is no closer than at the millenium. Was consensus on Pleistocene glaciation brought about by Jamieson’s (1882) solution to the submergence problem, which was that each ice

1860

1880

Fig. 2.4. Bibliographic history of the Pleistocene (‘Newer Pliocene’) glacial controversy. Number of papers (black) and books (grey) per year concerning the ‘Great Northern Drift’. Early Martyrs refer to the essentially correct but ignored glacial theories of Esmark (1824), Dobson (1925), Venetz (1830) and Bernhardi (1832). Agassiz refers to the time of tumult associated with the championing of the glacial theory by Agassiz (1837, 1840, 1842). Rejection was a 20-year period of eclipse of glacial theory. Revival marks the widespread acceptance of glacial theory following the studies of Ramsay (1860), Jamieson (1862, 1863, 1865) and Geikie (1863).

age was accompanied by submergence of the land, followed by slow reemergence after the ice disappeared? This problem (based on marine fossils within tills that are now elevated above sea level) was central to the interpretation of the Drift (e.g. Smith 1836). Agassiz had offered no explanation for it, there being no marine fossils in the Swiss glacial deposits. Geophysicists attributed ice-age submergence to the gravitational ‘pull’ of a gigantic polar ice cap on the adjacent seas (Adhe´mar 1842; Croll 1875), a theory easily accommodated by those like Lyell and Murchison who believed that the drift came by way of iceberg-laden floods of Arctic origin. In contrast, Shaler (1847) and Jamieson (1865) suggested, and later demonstrated with geological evidence (Jamieson 1882), that submergence was caused by deflection of the lithosphere under the load of the ice sheet (glacioisostasy). However, Jamieson’s (1882) theory of submergence came after broad acceptance of the glacial theory and cannot therefore have been its cause. Consensus on the glacial theory came about because of the overwhelming weight of evidence of the same kind as that obtained in skeletal form by Agassiz (1842). If consensus on the Snowball Earth hypothesis ultimately rests on the weight of evidence of the kind identified by Kirschvink (1992) (palaeomagnetic evidence for low-latitude glaciation at sea level, syn-glacial sedimentary Fe2O3 and MnO2 ore formations, and global evidence of abrupt climate switching (cap carbonates)) a positive outcome appears more and more probable in the end because that evidence has only become stronger. Must we also wait for two decades or more for consensus to emerge? Long delays between initial discussion of radical hypotheses and their broad acceptance is a general phenomenon in science (e.g. heliocentrism, biological evolution, continental drift, the big bang), so we must suppose there is a reason for the delays. The simplest one is that no consensus can emerge until after the original antagonists, tethered to hasty judgements, have passed from the scene. Given the lengthening of productive lifespans, we could be in for a long wait. Whatever the final outcome concerning the state of the Cryogenian glacial ocean, it is apparent that a third climate state needs to be added to the two first recognized by Brooks (1926): ‘non-glacial’ (no ice sheets), ‘glacial’ (polar to temperate ice sheets) and ‘pan-glacial’ (ice sheets at all latitudes). As a comparative newcomer to Neoproterozoic glacial studies, I am grateful for reviews of a draft of this paper by M. Hambrey and two anonymous reviewers. I also thank Barry Cooper for a preprint of his paper in Earth Sciences History on early contributions from South Australia. However, I bear sole responsibility for the views expressed here. This work was supported by the US National Science Foundation (grant EAR-0417422), the Earth System Evolution

30

P. F. HOFFMAN

Project of the Canadian Institute for Advanced Research (CIFAR) and Harvard University. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Adhe´mar, J. A. 1842. Re´volutions de la Mer. Fain et Thunot, Paris. Agassiz, L. 1837. Discours prononce´ a` l’ouverture des se´ances de la Socie´te´ Helve´tique des Sciences Naturelles, a` Neuchaˆtel, le 24 juillet 1837. Actes de la Socie´te´ Helve´tique des Sciences Naturelles (22e`me session, Neuchaˆtel, 24 – 26 juillet 1837), 5– 32. (Reprinted in English: Upon glaciers, moraines, and erratic blocks. Edinburgh New Philosophical Journal, 24, 364– 383, 1838.) Agassiz, L. 1840. E´tudes sur les glaciers. Neuchaˆtel et Soleure. (Reprinted in English: Studies on glaciers, preceded by the Discourse of Neuchaˆtel. Hafner, New York, 1967.) Agassiz, L. 1842. The glacial theory and its recent progress. New Edinburgh Philosophical Journal, 33, 217–283. Aitken, J. D. 1991. The Ice Brook Formation and post-Rapitan, Late Proterozoic glaciation, Mackenzie Mountains, Northwest Territories. Geological Survey of Canada, Bulletin, 404, 43. Allen, P. A. & Hoffman, P. F. 2005. Extreme winds and waves in the aftermath of a Neoproterozoic glaciation. Nature, 433, 123– 127. Allen, P. A. & Etienne, J. L. 2008. Sedimentary challenge to Snowball Earth. Nature Geoscience, 1, 817–825. Allen, P., Leather, J. & Brasier, M. D. 2004. The Neoproterozoic Fiq glaciation and its aftermath, Huqf Supergroup of Oman. Basin Research, 16, 507 – 534. Alley, R. B., Blankenship, D. D., Rooney, S. T. & Bentley, C. R. 1989. Sedimentation beneath ice shelves – the view from ice stream B. Marine Geology, 85, 101– 120. Almeida, F. F. M. 1946. Origem dos mine´rios de ferro y manganes de Urucum (Corumba´, Estado Mato Grosso). Boletim de Divisa˜o de Geologia e Mineralogia, 119, 58. Anderson, J. B. 1999. Antarctic Marine Geology. Cambridge University Press, Cambridge. Anderson, J. B. & Ashley, G. M. (eds) 1991. Glacial Marine Sedimentation: Paleoclimatic Significance. Geological Society of America Special Paper, 261, Boulder, CO. Arnaud, E. 2004. Giant cross-beds in the Neoproterozoic Port Askaig Formation, Scotland: implications for snowball Earth. Sedimentary Geology, 165, 155–174. Arnaud, E. 2008. Deformation in the Neoproterozoic Smalfjord Formation, northern Norway: an indicator of glacial depositional conditions? Sedimentology, 55, 335– 356. Assereto, R. L. A. M. & Kendall, C. G. St. C. 1977. Nature, origin and classification of peritidal tepee structures and related breccias. Sedimentology, 24, 153–210. Bao, H., Lyons, J. R. & Zhou, C. 2008. Triple oxygen isotope evidence for elevated CO2 levels after a Neoproterozoic glaciation. Nature, 452, 504– 506. Bao, H., Fairchild, I. J., Wynn, P. M. & Spo¨tl, C. 2009. Stretching the envelope of past surface environments: Neoproterozoic glacial lakes from Svalbard. Science, 323, 119–122. Barrett, P. J. (ed.) 1989. Antarctic Cenozoic History from the CIROS-1 Drillhole, McMurdo Sound. DSIR Bulletin, 245, Science Information Publishing Center, Wellington, New Zealand. Baud, L. 1933. Le conglome´rat argilo-calcareux dans la re´gion de Kayes et de Bafoulabe´ et sa position stratigraphique. Compte Rendu, 197, 172– 173. ¨ ber das Wahrscheinlich altcambrische oder jungproterBeetz, W. 1929. U ozoische Alter der Glazialschichten an der Basis des KundelunguSystems in Katanga und am unteren Kongo. Neues Jahrbuch fu¨r Mineralogie, Geologie und Pala¨ontologie, Beilageba¨nde, Abteiling B, 61, 61– 82. Bendtsen, J. 2002. Climate sensitivity to changes in solar insolation in a simple coupled climate model. Climate Dynamics, 18, 595–609. Bendtsen, J. & Bjerrum, C. J. 2002. Vulnerability of climate on Earth to sudden changes in insolation. Geophysical Research Letters, 29, doi: 10.1029/2002GL014829.

Benn, D. I. & Evans, D. J. A. 1998. Glaciers and Glaciation. Arnold Publishers, London. Berger, W. H. 1982. Increase of carbon dioxide in the atmosphere during deglaciation: The coral reef hypothesis. Naturwissenschaften, 69, 87 – 88. Berner, R. A. & Maasch, K. A. 1996. Chemical weathering and controls on atmospheric O2 and CO2: fundamental principles were enunciated by J. J. Ebelmen in 1845. Geochimica et Cosmochimica Acta, 60, 1633– 1637. Bertrand-Sarfati, J., Flicoteaux, R., Moussine-Pouchkine, A. & Aı¨t Kaci Ahmed, A. 1997. Lower Cambrian apatitic stromatolites and phospharenites related to the glacio-eustatic cratonic rebound (Sahara, Algeria). Journal of Sedimentary Research, 67, 957– 974. Bernhardi, A. 1832. Wie kamen die aus dem Norden stammenden Felsbruchstu¨cke und Geshiebe, welche man in Norddeutschland und den benachbarten La¨ndern findet, an ihre gegenwa¨rtigen Fundorte? (How did the rock fragments and boulders of northern origin found in northern Germany and neighboring countries get to their present positions?) Jahrbuch fu¨r Mineralogie, Geognosie, Geologie, und Petrefaktenkunde, 3, 257– 267. Bestmann, M., Rice, A. H. N., Langenhorst, F., Grasemann, B. & Heidelbach, F. 2006. Subglacial bedrock welding associated with glacial earthquakes. Journal of the Geological Society, London, 163, 417– 420. Biju-Duval, B. & Gariel, O. 1969. Nouvelles observations sur les phe´nome`nes glaciaires ‘E´ocambriens’ de la bordure nord de la syne´clise de Taoudeni, entre le Hank et le Tanezrouft, Sahara occidental. Palaeogeography, Palaeoclimatology, Palaeoecology, 6, 283– 315. Blackett, P. M. S. 1961. Comparison of ancient climates with the ancient latitudes deduced from rock magnetic measurements. Proceedings of the Royal Society of London, Series A, 263, 1 –30. Blackwelder, E. 1932. An ancient glacial formation in Utah. Journal of Geology, 40, 289–304. Blanford, W. T., Blanford, H. F. & Theobald, W. 1859. On the geological structure and relations of the Talcheer Coal Field, in the District of Cuttack. Memoirs of the Geological Survey of India, 1, 33 – 89. Boulton, G. S. 1990. Sedimentary and sea level changes during glacial cycles and their control on glacimarine facies architecture. In: Dowdeswell, J. A. & Scourse, J. D. (eds) Glacimarine Evironments: Processes and Sediments. Geological Society, London, Special Publications, 53, 15– 52. Briden, J. C. 1970. Palaeolatitude distribution of precipitated sediments. In: Runcorn, S. K. (ed.) Palaeogeophysics. Academic Press, London, 437– 444. Briden, J. C. & Irving, A. 1964. Palaeolatitude spectra of sedimentary palaeoclimatic indicators. In: Nairn, A. E. M. (ed.) Problems in Palaeoclimatology. John Wiley & Sons, New York, 199–224. Brooks, C. E. P. 1922. The Evolution of Climate. Ernest Benn, London. Brooks, C. E. P. 1926. Climate Through the Ages. R.V. Coleman, New York. Buchan, K. L., Ernst, R. E., Hamilton, M. A., Mertanen, S., ˚ . 2001. Rodinia: the evidence from Pesonen, L. J. & Elming, S.-A integrated palaeomagnetism and U–Pb geochronology. Precambrian Research, 110, 9– 32. Budyko, M. I. 1969. The effect of solar radiation variations on the climate of the Earth. Tellus, 21, 611–619. Bjørlykke, K. 1967. The Eocambrian Reusch moraine at Bigganjargga and the geology of Varangerfjord, northern Norway. Norsk Geologiske Undersøkelse, 251, 18 –44. Cahen, L. 1950. Le Calcaire de Sekelolo, le complexe tillitique et la dolomie rose C1 dans l’anticlinal de Congo de Kati (Bas-Congo). Annales de la Muse´e Royale Congo Belge, Sciences Ge´ologiques, 7, 1 –55. Cahen, L. 1963. Glaciations anciennes et de´rive des continents (Ancient glaciations and continental drift). Annales de la Socie´te´ Ge´ologique de Belgique, 86, 19– 83. Cahen, L. & Lepersonne, J. 1981. Proterozoic diamictites of Lower Zaire. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s PrePleistocene Glacial Record. Cambridge University Press, Cambridge, 153– 157.

HISTORY OF NEOPROTEROZOIC GLACIAL GEOLOGY

Caldas, J. 1979. Evidencias de una glaciacio´n Precambriana en la costa sur del Peru´. Segundo Congreso Geolo´gico Chileno, Arica, Vol. J, 29 – 37. Calver, C. R. & Walter, M. R. 2000. The late Neoproterozoic Grassy Group of King Island, Tasmania: correlation and palaeogeographic significance. Precambrian Research, 100, 299– 312. Carver, J. H. & Vardavas, I. M. 1994. Precambrian glaciations and the evolution of the atmosphere. Annales Geophysicae, 12, 674– 682. Catling, D. C., Zahnle, K. J. & McKay, C. P. 2001. Biogenic methane, hydrogen escape, and the irreversible oxidation of early Earth. Science, 293, 839– 843. Cawood, P. A., Nemchin, A. A., Smith, M. & Loewy, S. 2003. Source of the Dalradian Supergroup constrained by U– Pb dating of detrital zircon and implications for the East Laurentian margin. Journal of the Geological Society, London, 160, 231–246. Chamberlin, T. C. 1898. The ulterior basis of time divisions and the classification of geological history. Journal of Geology, 6, 449– 462. Chamberlin, T. C. 1899. An attempt to frame a working hypothesis of the cause of glacial periods on an atmospheric basis. Journal of Geology, 7, 545–584. Charpentier, J. de. 1837. Some conjectures regarding the great revolutions which have so changed the surface of Switzerland, and particularly that of the Canton of Vaud, as to give rise to its present aspect. Edinburgh New Philosophical Journal, 22, 27 –36. Chumakov, N. M. 1964. Pra¨kambrische tillit-a¨hnliche Gesteine der Sowjetunion. Geologische Rundschau, 54, 83 –102. Chumakov, N. M. 1968. On the character of the Late Precambrian glaciation of Spitsbergen (in Russian). Doklady Akademiya Nauk USSR, Geological Series, 180, 1446–1449. Chumakov, N. M. 1981. Upper Proterozoic glaciogenic rocks and their stratigraphic significance. Precambrian Research, 15, 373–395. Chumakov, N. M. 1992. The problems of old glaciations (pre-Pleistocene glaciogeology in the USSR). Soviet Scientific Reviews, Section G Geology, 1, 1– 208. Claire, M. W., Catling, D. C. & Zahnle, K. J. 2006. Biogeochemical modelling of the rise in atmospheric oxygen. Geobiology, 4, 239– 269. Cloud, P., Wright, L. A., Williams, E. G., Diehl, P. & Walter, M. R. 1974. Giant stromatolites and associated vertical tubes from the upper Proterozoic Noonday Dolomite, Death Valley region, eastern California. Geological Society of America Bulletin, 85, 1869– 1882. Coleman, A. P. 1907. A Lower Huronian ice age. American Journal of Science, 23, 187– 192. Coleman, A. P. 1926. Ice Ages: Recent and Ancient. MacMillan, New York. Condon, D. J., Prave, A. R. & Benn, D. I. 2002. Neoproterozoic glacialrainout intervals: observations and implications. Geology, 30, 35– 38. Condon, D., Zhu, M., Bowring, S. A., Wang, W., Yang, A. & Jin, Y. 2005. U– Pb ages from the Neoproterozoic Doushantuo Formation, China. Science, 308, 95– 98. Cooper, B. 2009. Snowball Earth: the early contribution from South Australia. Earth Sciences History, 29, 121– 145. Corkoran, M. 2007. ‘Cap carbonates’ and Neoproterozoic glacigenic successions from the Kimberley region, north-west Australia. Sedimentology, 54, 871– 903. Corsetti, F. A. & Grotzinger, J. P. 2005. Origin and significance of tube structures in Neoproterozoic post-glacial cap carbonates: example from Noonday Dolomite, Death Valley, United States. Palaios, 20, 348– 363. Corsetti, F. A. & Kaufman, A. J. 2005. The relationship between the Neoproterozoic Noonday Dolomite and the Ibex Formation: new observations and their bearing on ‘snowball Earth’. Earth-Science Reviews, 73, 63 –78. Corsetti, F. A., Olcott, A. N. & Bakermans, C. 2006. The biotic response to Neoproterozoic snowball Earth. Palaeogeography, Palaeoclimatology, Palaeoecology, 232, 114– 130. Crawford, A. R. & Daily, B. 1971. Probable non-synchroneity of Late Precambrian glaciations. Nature, 230, 111– 112. Croll, J. 1875. Climate and Time in Their Geological Relations: A Theory of Secular Changes of the Earth’s Climate. Appleton & Co., New York.

31

Crowell, J. C. 1957. Origin of pebbly mudstones. Geological Society of America Bulletin, 68, 993–1010. Crowell, J. C. 1964. Climatic significance of sedimentary deposits containing dispersed megaclasts. In: Nairn, A. E. M. (ed.) Problems in Palaeoclimatology. Interscience, London, 86– 99. Crowell, J. C. & Frakes, L. A. 1970. Phanerozoic glaciations and the causes of ice ages. American Journal of Science, 268, 193– 224. Cunningham, F. F. 1990. James David Forbes, Pioneer Scottish Glaciologist. Scottish Academic Press, Edinburgh. Dal, A. 1900. Geologiske iagttagselser omkring Varangerfjorden. Norges Geologiske Undersøkelse, 28, 1– 16. David, T. W. E. 1907a. Conditions of climate at different geological epochs, with special reference to glacial epochs. 10th International Geological Congress, Mexico City, 1906, Compte Rendu, 437–482. David, T. E. W. 1907b. Some problems of Australian glaciations. Report of the Australasian Association for the Advancement of Science, 11, 457– 465. Davies, G. L. 1968. The tour of the British Isles made by Louis Agassiz in 1840. Annals of Science, 24, 131– 146. Davies, G. L. H. 2007. Whatever is under the Earth: the Geological Society of London, 1807– 2007. Geological Society, London. Davies, K. A. 1939. The glacial sediments of Bunyoro, N.W. Uganda. Bulletin of the Geological Survey of Uganda, 3, 20 –37. Davies, K. A. 1940. The glacial series of Bunyoro, north Uganda. 17th International Geological Congress, Moscow, 1937, 6, 115– 119. De Angelis, H. & Skvarca, P. 2003. Glacier surge after ice shelf collapse. Science, 299, 1560– 1562. De Geer, G. 1912. A geochronology of the last 12,000 years. 11th International Geological Congress, Stockholm, 1910, 1, 241– 253. Deynoux, M. & Trompette, R. 1976. Discussion: Late Precambrian mixtites: glacial and/or nonglacial? Dealing especially with the mixtites of West Africa. American Journal of Science, 276, 1302– 1315. Dobson, P. 1925. Remarks on Bowlders. American Journal of Science and Arts, 10, 217. Donnadieu, Y., Ramstein, G., Fluteau, F., Besse, J. & Meert, J. 2002. Is high obliquity a plausible cause for Neoproterozoic glaciations? Geophysical Research Letters, 29, doi: 10.1029/2002GL015902. Donnadieu, Y., Godde´ris, Y., Ramstein, G., Ne´de´lec, A. & Meert, J. 2004a. A ‘snowball Earth’ climate triggered by continental break-up through changes in runoff. Nature, 428, 303– 306. Donnadieu, Y., Ramstein, G., Fluteau, F., Roche, D. & Gonopolski, A. 2004b. The impact of atmospheric and oceanic heat transport on the sea-ice instability during the Neoproterozoic. Climate Dynamics, 22, 293– 306. Dorr, J. V. H. 1945. Manganese and iron deposits of Morro do Urucum, Mato Grosso, Brazil. United States Geological Survey Bulletin, 946-A, 1 –47. Dott, R. H. Jr. 1961. Squantum ‘tillite’, Massachusetts – evidence of glaciation or subaqueous mass movements? Geological Society of America Bulletin, 72, 1289– 1305. Dow, D. B. 1965. Evidence of a Late Pre-Cambrian glaciation in the Kimberley Region of Western Australia. Geological Magazine, 102, 407– 419. Dowdeswell, J. A. & Scourse, J. D. (eds) 1990. Glacimarine Environments: Processes and Sediments. Geological Society, London, Special Publications, 53. Dowdeswell, J. A., Whittington, J. A., Jennings, A. E., Andrews, J. T., Mackensen, A. & Marienfield, P. 2000. An origin for laminated glacimarine sediments through sea-ice build-up and supressed iceberg rafting. Sedimentology, 47, 557– 576. Drewry, D. 1986. Glacial Geologic Processes. Edward Arnold, London. Dunn, P. R., Thomson, B. P. & Rankama, K. 1971. Late Pre-Cambrian glaciation in Australia as a stratigraphic boundary. Nature, 231, 498– 502. Du Toit, A. L. 1922. The Carboniferous glaciation of South Africa. Transactions of the Geological Society of South Africa, 24, 188– 227. Ebelmen, J.-J. 1845. Sur les produits de la de´composition des especes mine´rales de la famille silicates. (On the products of the weathering of silicate minerals.) Annales des Mines, 7, 3– 66. Ebelmen, J.-J. 1847. Sur la de´composition des roches. (On the weathering of rocks.) Annales des Mines, 12, 627– 654.

32

P. F. HOFFMAN

Edwards, M. B. 1975. Glacial retreat sedimentation in the Smalfjord Formation, Late Precambrian, north Norway. Sedimentology, 22, 75 –94. Edwards, M. B. 1984. Sedimentology of the Upper Proterozoic glacial record, Vestertana Group, Finnmark, North Norway. Norges Geologiske Undersøkelse Bulletin, 394, 76. Edwards, M. B. 1997. Discussion of glacial or non-glacial origin of the Bigganjargga tillite, Finnmark, northern Norway. Geological Magazine, 134, 873– 876. Eisbacher, G. H. 1981. Sedimentary tectonics and glacial record in the Windermere Supergroup, Mackenzie Mountains, northwestern Canada. Geological Survey of Canada Paper, 80-27, 40. Elverhoi, A. 1984. Glacigenic and associated marine sediments in the Weddell Sea, Fjords of Spitsbergen and the Barents Sea: a review. Marine Geology, 57, 53– 88. Embleton, B. J. J. & Williams, G. E. 1986. Low latitude of deposition for late Precambrian periglacial varvites in South Australia: implications for palaeoclimatology. Earth and Planetary Science Letters, 79, 419– 430. Eriksson, E. 1968. Air– ocean – icecap interactions in relation to climatic fluctuations and glaciation cycles. Meteorological Monographs, 8, 68 –92. Esmark, J. 1824. Bidrag til vor jordklodes historie. (Contribution to the history of our Earth.) Magazin for Naturvidenskaberne, 2(1), 29 –54. (Reprinted in English: Remarks tending to explain the geological history of the Earth. Edinburgh New Philosophical Journal, 2, 107–121, 1827.) Evans, D. A. D. 2000. Stratigraphic, geochronological, and paleomagnetic constraints upon the Neoproterozoic climatic paradox. American Journal of Science, 300, 347– 433. Evans, D. A. D. 2006. Proterozoic low orbital obliquity and axial-dipolar geomagnetic field from evaporite palaeolatitudes. Nature, 444, 51 –55, doi: 10.1038/nature05203. Eyles, N. 1993. Earth’s glacial record and its tectonic setting. EarthScience Reviews, 35, 1– 248. Eyles, N. & Januszczak, N. 2003. ‘Zipper-rift’: a tectonic model for Neoproterozoic glaciations during the breakup of Rodinia after 750 Ma. Earth-Science Reviews, 65, 1– 73. Fairchild, I. J. 1983. Effects of glacial transport and neomorphism on Precambrian dolomite crystal sizes. Nature, 304, 714– 716. Fawcett, P. J. & Boslough, M. B. E. 2002. Climatic effects of an impact-induced equatorial debris ring. Journal of Geophysical Research, 107(D15), doi: 10.1029/2001JD001230. Fedden, F. G. S. 1875. On the evidence of ‘ground-ice’ in tropical India, during the Ta´lchı´r period. Records of the Geological Survey of India, 8, 16 –18. Forbes, E. 1846. On the connexion between the distribution of the existing fauna and flora of the British Isles, and the geological changes which have affected their area, especially during the epoch of the Northern Drift. Memoirs of the Geological Survey of Great Britain, 1, 336– 403. Fourier, J. 1824. Remarques ge´ne´rales sur les tempe´ratures du globe terrestre et des espaces plane´taires. Annales de Chimie et de Physique (2nd series), 27, 136–167. Translated as, General remarks on the temperature of the terrestrial globe and the planetary spaces. American Journal of Science and Arts, 32, 1– 20, 1837. Føyn, S. 1937. The Eocambrian series of the Tana district, northern Norway. Norsk Geologisk Tidsskrift, 17, 65– 164, 4 plates, 1:250,000 scale map. Frakes, L. A., Francis, J. E. & Syktus, J. I. 1992. Climate Modes of the Phanerozoic. Cambridge University Press, Cambridge, 274. Frankel, H. 1987. Jan Hospers and the rise of paleomagnetism. Eos, Transactions of the American Geophysical Union, 68, 577, 579– 581. Furon, R. 1933. Observations sur la stratigraphie de l’ouest africain (Mauritanie et Soudan). Compte Rendu, 196, 1905–1906. Gaertner, H. R. von. 1943. Bemerkungen u¨ber den Tillit von Bigganjargga am Varangerfjord. Geologische Rundschau, 34, 226–231. Garwood, E. J. & Gregory, J. W. 1898. Contribution to the glacial geology of Spitsbergen. Quarterly Journal of the Geological Society, London, 54, 197–225. Geikie, A. 1863. On the phenomena of the glacial drift of Scotland. Transactions of the Geological Society of Glasgow, 1, 190.

Geikie, A. 1865. The Scenery of Scotland, Viewed in Connexion with its Physical Geology. MacMillan, London. Geikie, J. 1874. The Great Ice Age. D. Appleton & Co., New York, 545. Gevers, T. W. & Beetz, W. 1940. Pre-Dwyka glacial periods in southern Africa. 17th International Geological Congress, Moscow, 1937, 6, 65 – 98. Gibsher, A. S. & Khomentovsky, V. V. 1990. The section of the Tsagaan Olum and Bayan Gol Formations of the Vendian – Lower Cambrian in the Dzabkhan zone of Mongolia. In: Khomentovsky, V. V., Gibsher, A. S. & Karlova, G. A. (eds) The Late Precambrian and Early Paleozoic of Siberia. Institut Geologii I Geofiziki, Sibirskoe Otdelenie, Akademiya Nauk SSSR, Novosibirsk, 79 –91. Girdler, R. W. 1964. The palaeomagnetic latitudes of possible ancient glaciations. In: Nairn, A. E. M. (ed.) Problems in Palaeoclimatology. Interscience, London, 119–149. Godde´ris, Y., Donnabieu, Y. et al. 2003. The Sturtian ‘snowball’ glaciation: fire and ice. Earth and Planetary Science Letters, 211, 1– 12. Gold, T. 1955. Instability of the Earth’s axis of rotation. Nature, 175, 526– 529. Gorin, G. E., Racz, L. G. & Walter, M. R. 1982. Late Precambrian – Cambrian sediments of Huqf Group, Sultanate of Oman. American Association of Petroleum Geologists Bulletin, 66, 2609– 2627. Grotzinger, J. P. & Knoll, A. H. 1995. Anomalous carbonate precipitates: is the Precambrian the key to the Permian? Palaios, 10, 578– 596. Halverson, G. P. 2006. A Neoproterozoic chronology. In: Xiao, S. & Kaufman, A. J. (eds) Neoproterozoic Geobiology and Paleobiology. Springer, Dordrecht, 231– 271. Halverson, G. P., Hoffman, P. F., Schrag, D. P. & Kaufman, J. A. 2002. A major perturbation of the carbon cycle before the Ghaub glaciation (Neoproterozoic) in Namibia: prelude to snowball Earth? Geophysics, Geochemistry, Geosystems, 3, doi: 10.1029/ 2001GC000244. Halverson, G. P., Maloof, A. C. & Hoffman, P. F. 2004. The Marinoan glaciation (Neoproterozoic) in northeast Svalbard. Basin Research, 16, 297– 324. Hambrey, M. J. 1994. Glacial Environments. CRC Press, Boca Raton, FL. Hambrey, M. J. & Harland, W. B. 1981. Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge. Hambrey, M. J. & Harland, W. B. 1985. The Late Proterozoic glacial era. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 255– 272. Hambrey, M. J. & Spencer, A. M. 1987. Late Precambrian glaciation of central East Greenland. Meddelelser om Grønland, Geoscience, 19, 1 –50. Hamdi, B. 1992. Late Precambrian glacial deposits in central Iran. 29th International Geological Congress, Kyoto, 1992, Abstracts, 2, 263. Harland, W. B. 1964a. Evidence of late Precambrian glaciation and its significance. In: Nairn, A. E. M. (ed.) Problems in Palaeoclimatology. Interscience, London, 119– 149. Harland, W. B. 1964b. Critical evidence for a great infra-Cambrian glaciation. Geologische Rundschau, 54, 45 –61. Harland, W. B. 1983. The Proterozoic glacial record. In: Medaris, L. G., Jr., Byers, W. C., Mickelson, D. M. & Shanks, W. C. (eds) Proterozoic Geology. Geological Society of America, Memoir, 161, 279– 288. Harland, W. B. & Bidgood, E. T. 1959. Palaeomagnetism in some Norwegian Sparagmites and the Late Pre-Cambrian ice age. Nature, 184, 1860– 1862. Harland, W. B. & Rudwick, M. J. S. 1964. The great infra-Cambrian ice age. Scientific American, August, 42 –49. Harland, W. B. & Wilson, C. B. 1956. The Hecla Hoek Succession in Ny Friesland, Spitsbergen. Geological Magazine, 93, 2265– 2286. Heckman, D. S., Geiser, D. M., Eldell, B. R., Stauffer, R. L., Kardos, N. L. & Hedges, S. B. 2001. Molecular evidence for the early colonization of land by fungi and plants. Science, 293, 1129–1133. Hegenberger, W. 1987. Gas escape structures in Precambrian peritidal carbonate rocks. Communications of the Geological Survey of Namibia, 3, 49 –55. Heezen, B. C. & Hollister, C. 1964. Turbidity currents and glaciation. In: Nairn, A. E. M. (ed.) Problems in Palaeoclimatology. Interscience, London, 99– 108.

HISTORY OF NEOPROTEROZOIC GLACIAL GEOLOGY

Hintze, F. F. 1913. A contribution to the geology of the Wasatch Mountains, Utah. Annals of the New York Academy of Sciences, 23, 85 – 143. Hitchcock, E. 1841. Diluvium, or drift. In: Final Report on the Geology of Massachusetts. J.S. & C. Adams, Amherst, MA, 350– 406; postscript, 3a –11a. Hoffman, P. F. & Maloof, A. C. 1999. Glaciation: the snowball theory still holds water. Nature, 397, 384. Hoffman, P. F. & Schrag, D. P. 2000. Snowball Earth. Scientific American, 282, 68 –75. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129– 155. Hoffman, P. F., Kaufman, A. J., Halverson, G. P. & Schrag, D. P. 1998. A Neoproterozoic snowball Earth. Science, 281, 1342–1346. Hoffman, P. F., Halverson, G. P., Domack, E. W., Husson, J. M., Higgins, J. A. & Schrag, D. P. 2007. Are basal Ediacaran (635 Ma) post-glacial ‘cap dolostones’ diachronous? Earth and Planetary Science Letters, 258, 114– 131. Holland, T. H. 1908. On the occurrence of striated boulders in the Blaini Formation of Simla, with a discussion of the geological age of the beds. Records of the Geological Survey of India, 37, 129– 135. Holtedahl, O. 1918. Varangerhalvøn: Strøet omkring bunden av Varangerfjord. Norsk Geologiske Undersøkelse, 84, 148– 173. Hoffman, P. F. 2005. 28th DeBeers Alex. Du Toit Memorial Lecture: on Cryogenian (Neoproterozoic) ice-sheet dynamics and the limitations of the glacial sedimentary record. South African Journal of Geology, 108, 557– 576. Hoffman, P. F. & Li, Z. X. 2009. A palaeogeographic context for Neoproterozoic glaciation. Palaeogeography, Palaeoclimatology, Palaeoecology, 277, 158– 172. Horodyski, R. J. & Kauth, L. P. 1994. Life on land in the Precambrian. Science, 263, 494– 498. Howchin, W. 1901. Preliminary note on the existence of glacial beds of Cambrian age in South Australia. Transactions of the Royal Society of South Australia, 21, 74 –86. Howchin, W. 1903. Report on South Australian Glacial Investigation Committee. Report of the Australasian Association for the Advancement of Science, 9, 194– 200. Howchin, W. 1908. Glacial beds of Cambrian age in South Australia. Quarterly Journal of the Geological Society, London, 64, 234– 259. Howchin, W. 1912. Australian glaciations. Journal of Geology, 20, 193– 227. Hunt, T. S. 1880. The chemical and geological relations of the atmosphere. American Journal of Science, 19, 349– 363. Hyde, W. T., Crowley, T. J., Baum, S. K. & Peltier, W. R. 2000. Neoproterozoic ‘snowball Earth’ simulations with a coupled climate/icesheet model. Nature, 405, 425–429. Irving, E. 1956. Palaeomagnetic and palaeoclimatological aspects of polar wandering. Geofisica Pura e Applicada, 33, 23 – 41. Irving, E. 1957a. Directions of magnetization in the Carboniferous glacial varves of Australia. Nature, 180, 280– 281. Irving, E. 1957b. Analysis of the palaeomagnetism of the Torridonian sandstone series of north-west Scotland. Philosophical Transactions of the Royal Society, London, Series A, 250, 83 –89. Irving, E. 1959. Palaeomagnetic pole positions: a survey and analysis. Geophysical Journal of the Royal Astronomical Society, 2, 51 –79. Irving, E. 2008. Jan Hospers’s key contributions to geomagnetism. Eos, Transactions, American Geophysical Union, 89, 457–458. Irving, E. & Runcorn, S. K. 1957. The origin of the palaeomagnetism of the Torridonian sandstones of north-west Scotland. Philosophical Transactions of the Royal Society, London, Series A, 250, 100– 110. Isotta, C. A., Rocha-Campos, A. C. & Yoshida, R. 1969. Striated pavement of the Upper Pre-Cambrian glaciation in Brazil. Nature, 222, 466– 468. Jack, R. L. 1913. The Mount Grainger Goldfield. Report of the Geological Survey of South Australia, 1, 1 –24. James, N. P., Narbonne, G. M. & Kyser, T. K. 2001. Late Neoproterozoic cap carbonates: Mackenzie Mountains, northwestern Canada: precipitation and global glacial meltdown. Canadian Journal of Earth Sciences, 38, 1229–1262.

33

Jamieson, T. F. 1862. On the ice-worn rocks of Scotland. Quarterly Journal of the Geological Society of London, 18, 164– 184. Jamieson, T. F. 1863. On the parallel roads of Glen Roy, and their place in the history of the glacial period. Quarterly Journal of the Geological Society of London, 19, 235– 259. Jamieson, T. F. 1865. On the history of the last geological changes in Scotland. Quarterly Journal of the Geological Society of London, 21, 161– 203. Jamieson, T. F. 1882. On the cause of the depression and re-elevation of the land during the glacial period. Geological Magazine, 9, 400– 407, 457– 466. Jenkins, G. S. 2000. Global climate model high-obliquity solutions to the ancient climate puzzles of the faint-young-Sun paradox and lowlatitude Proterozoic glaciation. Journal of Geophysical Research, 105, 7357–7370. Jensen, P. A. & Wulff-Pedersen, E. 1996. Glacial or non-glacial origin for the Bigganjargga tillite, Finnmark, northern Norway. Geological Magazine, 133, 137– 145. Jiang, G., Sohl, L. E. & Christie-Blick, N. 2003. Neoproterozoic stratigraphic comparison of the Lesser Himalaya (India) and Yangtse clock (south China): paleogeographic implications. Geology, 31, 917– 920. Jiang, G., Kennedy, M. J., Christie-Blick, N., Wu, H. & Zhang, S. 2006. Stratigraphy, sedimentary structures, and textures of the late Neoproterozoic Doushantuo cap carbonate in South China. Journal of Sedimentary Research, 76, 978– 995. Johnston, J. D. 1993. Ice wedge casts in the Dalradian of South Donegal – evidence for subaerial exposure of the Boulder Bed. Irish Journal of Earth Sciences, 12, 13– 26. Katz, H. R. 1961. Late Precambrian to Cambrian stratigraphy in East Greenland. In: Raasch, G. O. (ed.) Geology of the Arctic, Vol 1. University of Toronto Press, Toronto, 299– 328. Kaufman, A. J., Jacobsen, S. B. & Knoll, A. H. 1993. The Vendian record of Sr and C isotopic variations in seawater: implications for tectonics and paleoclimate. Earth and Planetary Science Letters, 120, 409–430. Kaufman, A. J., Knoll, A. H. & Narbonne, G. M. 1997. Isotopes, ice ages, and terminal Proterozoic Earth history. Proceedings of the National Academy of Sciences USA, 94, 6600–6605. Kaufman, A. J., Jiang, G., Christie-Blick, N., Banerjee, D. M. & Rai, V. 2006. Stable isotope record of the terminal Neoproterozoic Krol platform in the Lesser Himalayas of northern India. Precambrian Research, 147, 156–185. Kawai, T., Windley, B. F., Terabayashi, M., Yamamoto, H., Isozaki, Y. & Maruyama, S. 2008. Neoproterozoic glaciation in the mid-oceanic realm: an example from hemi-pelagic mudstones on Llanddwyn Island, Anglesey, UK. Gondwana Research, 14, 105– 114. Kendall, C. G. St. C. & Warren, J. 1987. A review of the origin and setting of tepees and their associated fabrics. Sedimentology, 34, 1007–1027. Kennedy, M. J. 1996. Stratigraphy, sedimentology, and isotopic geochemistry of Australian Neoproterozoic postglacial cap dolostones: deglaciation, d13C excursions, and carbonate precipitation. Journal of Sedimentary Research, 66, 1050– 1064. Kennedy, M., Droser, M., Mayer, L. M., Pevear, D. & Mrofka, D. 2006. Late Precambrian oxygenation: inception of the clay mineral factory. Science, 311, 1446–1449. Kennedy, M. J., Runnegar, B., Prave, A. R., Hoffmann, K.-H. & Arthur, M. A. 1998. Two or four Neoproterozoic glaciations? Geology, 26, 1059– 1063. Kilburn, C., Pitcher, W. S. & Shackleton, R. M. 1965. The stratigraphy and origin of the Portaskaig Boulder Bed series (Dalradian). Geological Journal, 4, 343– 360. Kiessling, W. 2001. Paleoclimatic significance of Phanerozoic reefs. Geology, 29, 751– 754. King, L. H., Rokoengen, K., Fader, G. B. J. & Gunleiksrud, T. 1991. Till-tongue stratigraphy. Geological Society of America Bulletin, 103, 637–659. Kirschvink, J. L. 1992. Late Proterozoic low-latitude glaciation: the snowball Earth. In: Schopf, J. W. & Klein, C. (eds) The Proterozoic Biosphere. Cambridge University Press, Cambridge, 51 –52.

34

P. F. HOFFMAN

Klein, C. & Beukes, N. J. 1993. Sedimentology and geochemistry of the glacigenic Late Proterozoic Rapitan iron-formation in Canada. Economic Geology, 88, 542– 565. Knoll, A. H. 2003. Life on a Young Planet, the First Three Billion Years of Evolution on Earth. Princeton University Press, Princeton. Knoll, A. H. & Walter, M. R. 1992. Latest Proterozoic stratigraphy and Earth history. Nature, 356, 673–678. Knoll, A. H., Hayes, J. M., Kaufman, A. J., Swett, K. & Lambert, I. B. 1986. Secular variation in carbon isotope ratios from Upper Proterozoic successions of Svalbard and East Greenland. Nature, 321, 831– 838. Koken, 1907. Indisches Perm und die Permische Eiszeit. (Indian Permian and the Permian ice age.) Neues Jahrbuch fu¨r Mineralogie, Geologie, und Pala¨ontologie, Festband 1907, 446–545, 19 plates (including map). Ko¨ppen, W. & Wegener, A. 1924. Die Klimate der geologischen Vorzeit. (Climates of the geological past.) Gebru¨der Borntraeger, Berlin. Kuenen, P. H. & Migliorini, C. I. 1950. Turbidity currents as a cause of graded bedding. Journal of Geology, 58, 91 –127. Kulling, O. 1934. The Hecla Hoek Formation round Hinlopenstredet. Geografiska Annaler, 14, 161–253. Lachenbruch, A. 1962. Mechanics of thermal contraction cracks and ice-wedge polygons in permafrost. Special Paper, Geological Society of America, 70, 1 –69. Laskar, J. & Robutel, P. 1993. The chaotic obliquity of the planets. Nature, 361, 608–612. Laskar, J., Joutel, F. & Robutel, P. 1993. Stabilization of the Earth’s obliquity by the Moon. Nature, 361, 615– 617. Leather, J., Allen, P. A., Brasier, M. D. & Cozzi, A. 2002. Neoproterozoic snowball Earth under scrutiny: evidence from the Fiq glaciation of Oman. Geology, 30, 891–894. Lee, Y. Y. 1936. The Sinian glaciation in the lower Yangtze valley. Bulletin of the Geological Society of China, 15, 131–134. Lee, J. S. & Chao, Y. T. 1924. Geology of the gorge district of the Yangtze (from Ichang to Tzekuei) with special reference to the development of the gorges. Bulletin of the Geological Society of China, 3, 351– 391. Lee, J. S. & Lee, Y. Y. 1940. Sinian glaciation of China. 17th International Geological Congress, Moscow, 1937, Report, 6, 33 – 41. Lemon, N. M. & Gostin, V. A. 1990. Glacigenic sediments of the late Proterozoic Elatina Formation and equivalents, Adelaide Geosyncline, South Australia. In: Jago, J. B. & Moore, P. S. (eds) The Evolution of a Late Precambrian–Early Paleozoic Rift Complex: the Adelaide Geosyncline. Geological Society of South Australia Special Publication, Adelaide, 16, 149–163. Lenk, C., Strother, P. K., Kaye, C. A. & Barghoorn, E. S. 1982. Precambrian age of the Boston Basin: new evidence from microfossils. Science, 216, 619–620. Lenton, T. M. & Watson, A. J. 2004. Biotic enhancement of weathering, atmospheric oxygen and carbon dioxide in the Neoproterozoic. Geophysical Research Letters, 31, L05202, doi:10.1029/2003GL018802. Levrard, B. & Laskar, J. 2003. Climate friction and the Earth’s obliquity. Geophysical Journal International, 154, 970–990. Li, Z. X., Evans, D. A. D. & Zhang, S. 2004. A 908 spin on Rodinia: possible causal links between the Neoproterozoic supercontinent, superplume, true polar wander and low-latitude glaciation. Earth and Planetary Science Letters, 220, 409–421. Li, Z. X., Bogdanova, S. V. et al. 2008. Assembly, configuration, and break-up history of Rodinia: a synthesis. Precambrian Research, 160, 179– 210. Lu, S. N., Ma, G. G., Gao, Z. J. & Lin, W. X. 1985. Primary research on the glaciogenic rocks of the Late Precambrian in China. In: Precambrian Geology Committee (eds) Precambrian Geology, No. 1, The Collected Works on the Late Precambrian Glaciogenic Rocks of China. Geology Publication House, Beijing, 1– 86. Lyell, C. 1840. On the geological evidence of the former existence of glaciers in Forfarshire. Edinburgh New Philosophical Journal, 30, 199–202. Reprinted in Philosophical Magazine (2nd series), 18, 579– 591, 1841. Lyell, C. 1841. Elements of Geology, 2nd edn. Hilliard, Gray, and Co., Boston.

Lyell, C. 1845. Travels in North America, 1841– 2, with Geological Observations on the United States, Canada, and Nova Scotia (2 vol.). Wiley & Putnam, New York, 231. (Reprinted in 1978 by Arno Press, New York.) Lyell, C. 1851. Manual of Elementary Geology, 3rd edn. John Murray, London. Lyell, C. 1852. Manual of Elementary Geology, 4th edn. John Murray, London. Lyell, C. 1855. Manual of Elementary Geology, 5th edn. John Murray, London. Lyell, C. 1857. Manual of Elementary Geology, 6th edn. John Murray, London. Lyell, C. 1863. The Geological Evidence of the Antiquity of Man. Reprinted unabridged in 2004 by Dover Publications, Mineola, NY. Lyell, C. 1865. Elements of Geology, 6th enlarged edn. John Murray, London. Macdonald, F. A., Jones, D. S. & Schrag, D. P. 2009a. Stratigraphic and tectonic implications of a newly discovered glacial diamictitecap carbonate couplet in southwestern Mongolia. Geology, 37, 123– 126. Macdonald, F. A., McClelland, W. C., Schrag, D. P. & Macondald, W. P. 2009b. Neoproterozoic glaciation on a carbonate platform margin in Arctic Alaska and the origin of the North Slope subterrane. Geological Society of America Bulletin, 121, 448–473. ¨ . & Schrag, Macdonald, F. A., Strauss, J. V., Rose, C. V., Duda´s, F. O D. P. 2011. Stratigraphy of the Port Nolloth Group of Namibia and South Africa and implications for the age of Neoproterozoic iron formations. American Journal of Science, 310, 862– 888. Maciel, P. 1959. Tilito Cambriano(?) no Estado de Mato Grosso. Sociedad Brasilieras Geologia Boletino, 8, 3– 49. Maloof, A. C., Kellogg, J. B. & Anders, A. M. 2002. Neoproterozoic sand wedges: crack formation in frozen soils under diurnal forcing during a snowball Earth. Earth and Planetary Science Letters, 204, 1 –15. Maloof, A. C., Halverson, G. P., Kirschvink, J. L., Schrag, D. P., Weiss, B. P. & Hoffman, P. F. 2006. Combined paleomagnetic, isotopic and stratigraphic evidence for true polar wander from the Neoproterozoic Akademikerbreen Group, Svalbard. Geological Society of America Bulletin, 118, 1099– 1124. Marshall, H. G., Walker, J. C. G. & Kuhn, W. R. 1988. Long-term climate change and the geochemical cycle of carbon. Journal of Geophysical Research, 93, 791– 801. Martin, H. 1965a. The Precambrian Geology of South West Africa and Namaqualand. Precambrian Research Unit Bulletin, 1, University of Cape Town, South Africa. Martin, H. 1965b. Beobachtungen zum Problem der jungpra¨kambrischen Glazialen Ablagerungen in Su¨dwestafrika (Observations concerning the problem of the late Precambrian glacial deposits in South West Africa). Geologische Rundschau, 54, 115– 127. Martin, H. 1981. The late Palaeozoic Gondwana glaciation. Geologische Rundschau, 70, 480 – 496. Martin, H., Porada, H. & Walliser, O. H. 1985. Mixtite deposits of the Damara sequence, Namibia, problems of interpretation. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 159–196. Maugh, T. H. II. 1989. Super ice age gave life on Earth growing pains. Los Angeles Times, September 7, 1, 3, 28. Mawson, D. 1949a. The Late Precambrian ice age and glacial record of the Bibliando dome. Journal and Proceedings of the Royal Society of New South Wales, 82, 150– 174. Mawson, D. 1949b. The Elatina glaciation: a third recurrence of glaciation evidenced in the Adelaide system. Transactions of the Royal Society of South Australia, 73, 117–121. McCabe, C. & Elmore, R. D. 1989. The occurrence and origin of Late Paleozoic remagnetization in the sedimentary rocks of North America. Reviews of Geophysics, 27, 471–494. McCay, G. A., Prave, A. R., Alsop, G. I. & Fallick, A. E. 2006. Glacial trinity: Neoproterozoic Earth history within the British –Irish Caledonides. Geology, 34, 909–912, doi: 10.1130/G22694A.1 McMechan, M. E. 2000. Vreeland Diamictites – Neoproterozoic glaciogenic slope deposits, Rocky Mountains, northeast British Columbia. Bulletin of Canadian Petroleum Geology, 48, 246– 261.

HISTORY OF NEOPROTEROZOIC GLACIAL GEOLOGY

Melosh, H. J. 1989. Impact Cratering: A Geologic Process. Oxford University Press, New York. Milankovic, M. 1941. Canon of Insolation and the Ice-Age Problem. (English translation published by Zavod za Udzbenike i Nastavna Sredstva, Belgrade, 634, 1998.) Miller, N. R., Alene, M. et al. 2003. Significance of the Tambien Group (Tigrai, N. Ethiopia) for Snowball Earth events in the Arabian-Nubian Shield. Precambrian Research, 121, 263–283. Moczydlowska, M. 2008. The Ediacaran microbiota and the survival of Snowball Earth conditions. Precambrian Research, 167, 1 –15. Molnia, B. F. 1983. Glacial– Marine Sedimentation. Plenum Press, New York. Moraes Rego, L. F. de. 1930. Glaciac¸a˜o eopaleozo´ica no centro do Brazil (Glaciation in the earliest Palaeozoic of central Brazil). Anais Academia Brasileira de Cieˆncias, 2, 109–112. Mu, Y. 1981. Luoquan Tillite of the Sinian System in China. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 402– 413. Nalivkin, D. V. 1956. Ucheniye o fatsiyak. Geograficheskiye usloviya obrazovaniya osadkov, 2 (Science of facies. Geographical environment of sedimentation). Izvestiya Akademia Nauk SSSR, Moscow. Natland, M. L. & Kuenen, P. H. 1951. Sedimentary history of the Ventura Basin, California and the action of turbidity currents. In: Turbidity Currents and the Transportation of Coarse Sediments into Deep Water. Special Publication 2, Society of Economic Paleontologists and Mineralogists, Tulsa, OK, 76 –107. Ne´de´lec, A., Affaton, P., France-Lanord, C., Charrie`re, A. & Alvaro, J. 2007. Sedimentology and chemostratigraphy of the Bwipe Neoproterozoic cap dolostones (Ghana, Volta Basin): a record of microbial activity in a peritidal environment. Comptes Rendus Geoscience, 339, 223– 239. Ne´ron de Surgy, O. & Laskar, J 1997. On the long term evolution of the spin of the Earth. Astronomy and Astrophysics, 318, 975–989. Newman, R. P. 1995. American instransigence: the rejection of continental drift in the great debates of the 1920’s. Earth Sciences History, 14, 62 – 83. Nick, F. M., Vieli, A., Howat, I. M. & Joughin, I. 2009. Large-scale changes in Greenland outlet glacier dynamics triggered at the terminus. Nature Geoscience, 2, 110– 114. Nikolaev, J. 1930. The glacial deposits (tillites) of Lower Cambrian age in the Yenissei Range. Bulletin of the Geological and Prospecting Service of the U.S.S.R., 49(7), 1 – 15. Nogueira, A. C. R., Riccomini, C., Sial, A. N., Moura, C. A. V. & Fairchild, T. R. 2003. Soft-sediment deformation at the base of the Neoproterozoic Puga cap carbonate (southwestern Amazon craton, Brazil): confirmation of rapid icehouse to greenhouse transition in snowball Earth. Geology, 31, 613– 616. Norin, E. 1937. Geology of the western Quruq Tagh, eastern Tien Shan. Reports of the Sino-Swedish Expedition III. Geology. Bokfo¨rlags Aktiebolaget Thule, Stockholm. North, G. R. 1990. Multiple solutions in energy balance climate models. Palaeogeography, Palaeoecology, Palaeoclimatology (Global and Planetary Change Section), 82, 225– 235. Oberbeck, V. R., Marshall, J. R. & Aggarwal, H. 1993. Impacts, tillites, and the breakup of Gondwanaland. Journal of Geology, 101, 1 – 19. Opdyke, B. N. & Walker, J. C. G. 1992. Return of the coral reef hypothesis: basin to shelf partitioning of CaCO3 and its effect on atmospheric CO2. Geology, 20, 733–736. Opdyke, B. N. & Wilkinson, B. H. 1990. Palaeolatitude distribution of Phanerozoic marine ooids and cements. Palaeogeography, Palaeoclimatology, Palaeoecology, 78, 135–148. Opdyke, N. D. 1962. Palaeoclimatology and continental drift. In: Runcorn, S. K. (ed.) Continental Drift. Academic Press, New York, 41– 65. Oreskes, N. 1999. The Rejection of Continental Drift. Oxford University Press, New York. Osmaston, H. 2004. Quaternary glaciations in the East African mountains. In: Ehlers, J. & Gibbard, P. L. (eds) Quaternary Glaciations – Extent and Chronology, Part III. Elsevier, Amsterdam, 139–150. Pais, M. A., Le Moue¨l, J. L., Lambeck, K. & Poirier, J. P. 1999. Late Precambrian paradoxical glaciation and obliquity of the Earth – a

35

discussion of dynamical constraints. Earth and Planetary Sciences, 174, 155–171. Park, J. K. 1994. Palaeomagnetic constraints on the position of Laurentia from middle Neoproterozoic to Early Cambrian times. Precambrian Research, 69, 95– 112. Pavlov, A. A., Kasting, J. F., Brown, L. L., Rages, K. A. & Freedman, R. 2000. Greenhouse warming by CH4 in the atmosphere of early Earth. Journal of Geophysical Research, 105, 11 981– 11 990. Pavlov, A. A., Hurtgen, M. T., Kasting, J. F. & Arthur, M. A. 2003. Methane-rich Proterozoic atmosphere? Geology, 31, 87– 90. Pavlov, A. A., Toon, O. B., Pavlov, A. K., Bally, J. & Pollard, D. 2005. Passing through a giant molecular cloud: ‘snowball’ glaciations produced by interstellar dust. Geophysical Research Letters, 32, L03705, doi: 10.1029/2004GL021890. Peltier, W. R., Tarasov, L., Vettoretti, G. & Solheim, L. P. 2004. Climate dynamics in deep time: modeling the ‘snowball bifurcation’ and assessing the plausibility of its occurrence. In: Jenkins, G. S., McMenamin, M. A. S., McKey, C. P. & Sohl, L. (eds) The Extreme Proterozoic: Geology, Geochemistry, and Climate. Geophysical Monograph, American Geophysical Union, Washington, DC, 146, 107–124. Plummer, P. S. 1978. Note on the palaeoenvironmental significance of the Nuccaleena Formation (upper Precambrian), central Flinders Ranges, South Australia. Journal of the Geological Society of Australia, 25, 395– 402. Pollard, D. & Kasting, J. F. 2005. Snowball Earth: a thin-ice solution with flowing glaciers. Journal of Geophysical Research, 110, C07010, doi: 10.1029/2004JC002525. Powell, R. D. 1984. Glacimarine processes and inductive lithofacies modelling of ice shelf and tidewater glacier sediments based on Quaternary examples. Marine Geology, 57, 1– 52. Powell, R. D. 1990. Glacimarine processes at grounding-line fans and their growth to ice-contact deltas. In: Dowdeswell, J. A. & Scourse, J. D. (eds) Glacimarine Environments: Processes and Sediments. Geological Society, London, Special Publication, 53, 53 –73. Powell, R. D. & Domack, E. W. 1995. Modern glaciomarine environments. In: Menzies, J. (ed.) Glacial Environments, Vol. 1. Modern Glacial Environments: Processes, Dynamics and Sediments. Butterworth-Heinemann, Oxford, 445– 486. Preiss, W. V. 1987. The Adelaide Geosyncline: Late Proterozoic stratigraphy, sedimentation, palaeontology and tectonics. Geological Survey of South Australia Bulletin, 53, 438. Rampino, M. R. 1994. Tillites, diamictites, and the ballistic ejecta of large impacts. Journal of Geology, 102, 439– 456. Ramsay, A. C. 1855. Angular, subangular, polished and striated fragments and boulders in the Permian breccia of Shropshire, Worcestershire, etc.; and on the probable existence of glaciers and icebergs in the Permian epoch. Quarterly Journal of the Geological Society of London, 11, 185– 205. Ramsay, A. C. 1860. The Old Glaciers of Switzerland and North Wales. Longman, Green, Longman & Roberts, London. Rankama, K. 1973. The Late Precambrian glaciation, with particular reference to the Southern Hemisphere. Journal and Proceedings, Royal Society of New South Wales, 106, 89 –97. Raub, T. D. & Evans, D. A. D. 2006. Magnetic reversals in basal Ediacaran cap carbonates: a critical review. Eos, Transactions of the American Geophysical Union, Joint Assembly Supplement, 87, Abstract GP41– 02. Reading, H. G. & Walker, R. G. 1966. Sedimentation of Eocambrian tillites and associated sediments in Finnmark, northern Norway. Palaeogeography, Palaeoclimatology, Palaeoecology, 2, 177– 212. Reusch, H. 1891. Skuringmærker og morængrus eftervist i Finnmarken fra en periode meget ældre end ‘istiden’ (Glacial striae and boulderclay in Norwegian Lapponie from a period much older than the last ice age). Norges Geologiske Undersøkelse, 1, 78 –85 (English summary, 97– 100). Rice, A. H. N. & Hofmann, C.-C. 2000. Evidence for a glacial origin of Neoproterozoic III striations at Oaibaccannjar’ga, Finnmark, northern Norway. Geological Magazine, 137, 355– 366. Ridgwell, A. J., Kennedy, M. J. & Caldeira, K. 2003. Carbonate deposition, climate stability, and Neoproterozoic ice ages. Science, 302, 859– 862.

36

P. F. HOFFMAN

Robert, M. 1940. La glaciation du Kundelungu au Katanga (Congo Belge). 17th International Geological Congress, Moscow, 1937, Report 6, 99 – 113. Roberts, J. D. 1971. Late Precambrian glaciation: an anti-greenhouse effect? Nature, 234, 216. Roberts, J. D. 1976. Late Precambrian dolomites, Vendian glaciation, and synchroneity of Vendian glaciations. Journal of Geology, 84, 47 – 63. Rodgers, J. 1957. The distribution of marine carbonate sediments: a review. In: Le Blanc, R. J. & Breeding, J. G. (eds) Regional Aspects of Carbonate Deposition. Society of Economic Paleontoliogists and Mineralogists (SEPM) Special Publication No. 5, 2 – 14, Tulsa, Oklahoma. Rogers, A. W. 1915. The geology of part of Namaqualand. Transactions of the Geological Society of South Africa, 18, 72 –101, 14 plates. Rosendahl, H. 1931. Bidrag til Varangernesets geologi. Norsk Geologisk Tidsskrift, 25, 327– 349. Rosendahl, H. 1945. Prekambrium-Eokambrium i Finnmark. Norsk Geologisk Tidsskrift, 25, 327–349. Rudwick, M. J. S. 1964. The infra-Cambrian glaciation and the origin of the Cambrian fauna. In: Nairn, A. E. M. (ed.) Problems in Palaeoclimatology. Interscience, London, 150–155, 184– 185. Runcorn, S. K. 1961. Climatic change through geological time in light of the palaeomagnetic evidence for polar wandering and continental drift. Quarterly Journal of the Royal Meteorological Society, 87, 282– 313. Sagan, C. & Mullen, G. 1972. Earth and Mars: evolution of atmospheres and surface temperatures. Science 177, 52 –26. Sayles, R. W. 1914. The Squantum Tillite. Bulletin of the Museum of Comparative Zoology at Harvard College, 56, 141– 175, 12 plates. Sayles, R. W. & LaForge, L. 1910. The glacial origin of the Roxbury Conglomerate. Science, 32, 723–724. Schermerhorn, L. J. G. 1974. Late Precambrian mixtites: glacial and/or non-glacial? American Journal of Science, 274, 673– 824. Schermerhorn, L. J. G. 1977. Late Precambrian dolomites, Vendian glaciation, and synchroneity of Vendian glaciations: a discussion. Journal of Geology, 85, 247– 250. Schermerhorn, L. J. G. & Stanton, W. I. 1963. Tilloids in the West Congo geosyncline. Quarterly Journal of the Geological Society of London, 119, 201–241. ¨ ber die Eiszeit. (On the Ice Age.) Actes Socie´te´ Schimper, K. 1837. U Helvetique des Sciences Naturelles, Neuchaˆtel, 38– 51. Schmidt, P. W. & Williams, G. E. 1995. The Neoproterozoic climatic paradox: equatorial palaeolatitude for Marinoan glaciation near sea level in South Australia. Earth and Planetary Science Letters, 134, 107– 124. Schmidt, P. W., Williams, G. E. & Embleton, B. J. J. 1991. Low palaeolatitude of Late Proterozoic glaciation: early timing of remnance in Haematite of the Elatina Formation, South Australia. Earth and Planetary Science Letters, 105, 355–367. Schiøtz, O. E. 1898. Om Dr Reusch’s pra¨glaciale skuringdmerker. Nyt Magazin Naturvissenschaften, 36, 1 –10. Schrag, D. P., Berner, R. A., Hoffman, P. F. & Halverson, G. P. 2002. On the initiation of a snowball Earth. Geophysics, Geochemistry, Geosystems, 3, doi: 10.1029/2001GC000219. Schwartzman, D. W. & Volk, T. 1991. Biotic enhancement of weathering and surface temperatures since the origin of life. Palaeogeography, Palaeoecology, Palaeoclimatology, 90, 357– 374. Schwarzbach, M. 1964a. Criteria for the recognition of ancient glaciations. In: Nairn, A. E. M. (ed.) Problems in Palaeoclimatology. Interscience, London, 81 –85. Schwarzbach, M. 1964b. Climates of the Past. D. Van Nostrand Co., London. (English translation of the 2nd revised edn of Das Clima der Vorzeit. Ferdinand Enke Verlag, Stuttgart, 1961.) Sellers, W. D. 1969. A global climatic model based on the energy balance of the Earth-atmosphere system. Journal of Applied Meteorology, 8, 392–400. Sellers, W. D. 1990. The genesis of energy balance modeling and the cool Sun paradox. Palaeogeography, Palaeoecology, Palaeoclimatology (Global and Planetary Change Section), 82, 217–224. Shaler, N. S. 1847. Preliminary report on the secent changes of level on the coast of Maine: with reference to their origin and relation to

other similar changes. Memoirs of the Boston Society of Natural History, 2, 321–323, 335– 340. Sheldon, R. P. 1984. Ice-ring origin of the Earth’s atmosphere and hydrosphere and late Proterozoic –Cambrian phosphogenesis. Special Publication, Geological Survey of India, 17, 17 – 21. Shields, G. A., Deynoux, M., Strauss, H., Paquet, H. & Nahon, D. 2007. Barite-bearing cap dolostone of the Taoude´ni Basin, northwest Africa: sedimentary and isotopic evidence for methane seepage after a Neoproterozoic glaciation. Precambrian Research, 154, 209– 235. Siever, R. 1968. Sedimentological consequences of a steady-state ocean-atmosphere. Sedimentology, 11, 5 –29. Smith, J. 1836. On indications of changes in the relative levels of sea and land in the West of Scotland. Proceedings of the Geological Society, London, 2, 427–429. Smith, J. 1839. On the climate of the newer pliocene tertiary period. Proceedings of the Geological Society of London, 3, 118– 119. Sohl, L. E., Christie-Blick, N. & Kent, D. V. 1999. Paleomagnetic polarity reversals in Marinoan (ca 600 Ma) glacial deposits of Australia: implications for the duration of low-latitude glaciation in Neoproterozoic time. Geological Society of America Bulletin, 111, 1120– 1139. Spalletti, L. & Del Valle, A. 1984. Las diamictitas del sector oriental de Tandilia: characteres sedimentolo´gicos y origen. Revista de la Asociacio´n Geolo´gica Argentina, 39, 188– 206. Spencer, A. M. 1971. Late Pre-Cambrian glaciation in Scotland. Geological Society of London Memoir, 6, 99. Spjeldnaes, N. 1964. The Eocambrian glaciation in Norway. Geologische Rundschau, 54, 24 –45. Sprigg, R. C. 1986. The Adelaide Geosyncline: a century of controversy. Earth Sciences History, 5, 66– 83. Strahan, A. 1897. On glacial phenomena of Palaeozoic age in the Varanger Fiord. Quarterly Journal of the Geological Society, London, 53, 137– 146, Discussion, 153–156. Sugden, D. E. & John, B. S. 1976. Glaciers and Landscape: A Geomorphological Approach. Edward Arnold, London. Sutherland, P. C. 1870. Notes on an ancient boulder-clay of Natal. Quarterly Journal of the Geological Society of London, 26, 514– 517. Thompson, M. D. & Bowring, S. A. 2000. Age of the Squantum ‘tillite’, Boston Basin, Massachusetts: U– Pb zircon constraints on terminal Neoproterozoic glaciation. American Journal of Science, 300, 630– 655. Thomson, J. 1871. On the stratified rocks of Islay. Report of the 41st Meeting of the British Association for the Advancement of Science, Edinburgh. John Murray, London, 110– 111. Thomson, J. 1877. On the geology of the island of Islay. Transactions of the Geological Society of Glasgow, 5, 200–222. Tyndall, J. 1861. On the absorbtion and radiation of heat by gases and vapours, and on the physical connexion of radiation, absorbtion, and conduction – the Bakerian Lecture. Philosophical Magazine (series 4), 22, 169– 194, 273–285. Tyndall, J. 1863. On the radiation through the Earth’s atmosphere. Philosophical Magazine (series 4), 25, 200–206. Venetz, I. 1830. Sur l’extension qu’il pre´sume que les glaciers avaient autrefois, et sur leur retraits dans leurs limites actuelles. Actes Socie´te´ Helvetique des Sciences Naturelles, Lausanne, 31. Voigt, A., Abbot, D. S., Pierrehumbert, R. T. & Marotzke, J. 2011. Initiation of a Marinoan Snowball Earth in a state-of-the-art atmosphere-ocean general circulation model. Climate of the Past, 7, 1– 15. Walde, D. H. G., Gierth, E. & Leonardos, O. H. 1981. Stratigraphy and mineralogy of the manganese ores of Urucum, Mato Grosso, Brazil. Geologische Rundschau, 70, 1077– 1085. Walker, G. 2003. Snowball Earth: The Story of the Great Global Catastrophe that Spawned Life As We Know It. Crown Publishers, New York, 269. Walker, J. C. G., Hays, P. B. & Kasting, J. F. 1981. A negative feedback mechanism for the long-term stabilization of Earth’s surface temperature. Journal of Geophysical Research, 86(C10), 9776– 9782. Wang, Y., Lu, S. & Gao, Z. 1981. Sinian tillites of China. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 386–401.

HISTORY OF NEOPROTEROZOIC GLACIAL GEOLOGY

Wegener, A. 1912. Die Entstehung der Kontinente. Petermanns Geographische Mitteilungen, 58, 185–195, 253– 256, 305–309. (Reprinted in English translation: The origins of the continents. Journal of Geodynamics, 32, 29– 63, 2001.) Wegener, A. 1929. Die Enstehung der Kontinente und Ozeane, 4th edn. Friedrich Vieweg & Sohn, Braunschweig. (Reprinted in English translation by John Biram: The Origin of Continents and Oceans. Dover Publications, New York, 246, 1966.) Wegmann, C. E., Dangeard, L. & Graindor, M. J. 1950. Sue quelques caracte`res remarquables de la formation pre´-cambrienne connue sous le nom de Poudinage de Granville. Compte Rendus, 230, 979– 980. Wetherald, R. T. & Manabe, S. 1975. The effects of changing the Solar constant on the climate of a general circulation model. Journal of the Atmospheric Sciences, 32, 2044–2059. Whitten, G. F. 1970. The investigation and exploitation of the Razorback Ridge iron deposit. Geological Survey of South Australia Reports of Investigations, 33, 165. Williams, D. M., Kasting, J. F. & Frakes, L. A. 1998. Low-latitude glaciation and rapid changes in the Earth’s obliquity explained by obliquity – oblateness feedback. Nature, 396, 453–455. Williams, G. E. 1972. Geological evidence relating to the origin and secular rotation of the Solar system. Modern Geology, 3, 165– 181. Williams, G. E. 1974. Discussion of Late Precambrian glacial climate and the Earth’s obliquity. Journal of the Geological Society, London, 130, 599– 601. Williams, G. E. 1975. Late Precambrian glacial climate and the Earth’s obliquity. Geological Magazine, 112, 441–544. Williams, G. E. 1979. Sedimentology, stable-isotope geochemistry and palaeoenvironment of dolostones capping late Precambrian glacial sequences in Australia. Journal of the Geological Society of Australia, 26, 377–386. Williams, G. E. 1981. Sunspot periods in the late Precambrian glacial climate and solar – planetary relations. Nature, 291, 624–628. Williams, G. E. 1989. Late Precambrian tidal rhythmites in South Australia and the history of the Earth’s rotation. Journal of the Geological Society, London, 146, 97 –111. Williams, G. E. 1993. History of the Earth’s obliquity. Earth-Science Reviews, 34, 1 –45. Williams, G. E. 2000. Geological constraints on the Precambrian history of Earth’s rotation and the Moon’s orbit. Reviews of Geophysics, 38, 37 – 59. Williams, G. E. & Sonett, C. P. 1985. Solar signature in sedimentary cycles from the late Precambrian Elatina Formation, Australia. Nature, 318, 523– 527. Williams, G. E. & Tonkin, D. G. 1985. Periglacial structures and palaeoclimatic significance of a late Precambrian block field in the Cattle Grid copper mine, Mount Gunson, South Australia. Australian Journal of Earth Sciences, 32, 287–300. Williams, G. E., Gostin, V. A., McKirdy, D. M. & Preiss, W. V. 2008. The Elatina glaciation, late Cryogenian (Marinoan Epoch), South Australia: sedimentary facies and palaeoenvironments. Precambrian Research, 163, 307– 331. Willis, B. 1904. Geological research in eastern Asia. Carnegie Institution of Washington, Yearbook, 3, Washington, DC, 275–291.

37

Wilson, C. B. & Harland, W. B. 1964. The Polarisbreen Series and other evidences of late pre-Cambrian ice ages in Spitsbergen. Geological Magazine, 101, 198– 219. Wilson, J. T. 1966. Did the Atlantic close and then reopen? Nature, 211, 676– 681. Witzke, B. J. 1990. Palaeoclimatic constraints for Palaeozoic palaeolatitudes of Laurentia and Euramerica. In: McKerrow, W. S. & Scotese, C. R. (eds) Palaeozoic Palaeogeography and Biogeography. Geological Society of London Memoirs, 12, 57 –73. Woodward, H. B. 1907. The History of the Geological Society of London. Geological Society, London, 336. Reprinted in 1978 by Arno Press, New York. Woodward, H. P. 1884. Report on the range to the east of Farina. Parliamentary Papers for 1884. Government of South Australia, 40, 1– 5. Worsley, T. R. & Kidder, D. L. 1991. First-order coupling of paleogeography and CO2, with global surface temperature and its latitudinal contrast. Geology, 19, 1161– 1164. Wright, L., Williams, E. G. & Cloud, P. 1978. Algal and cryptalgal structures and platform environments of the late pre-Phanerozoic Noonday Dolomite, eastern California. Geological Society of America Bulletin, 89, 321–333. Wunsch, C. 2002. What is the thermohaline circulation? Science, 298, 1179–1180. Xiao, S. 2004. Neoproterozoic glaciations and the fossil record. In: Jenkins, G. S., McMenamin, M. A. S., McKey, C. P. & Sohl, L. (eds) The Extreme Proterozoic: Geology, Geochemistry, and Climate. Geophysical Monograph, American Geophysical Union, Washington, DC, 146, 199–214. Xiao, S., Bao, H. et al. 2004. The Neoproterozoic Quruqtagh Group in eastern Chinese Tianshan: evidence for a post-Marinoan glaciation. Precambrian Research, 130, 1– 26. Yeo, G. M. 1981. The Late Proterozoic Rapitan glaciation in the northern Cordillera. In: Campbel, F. H. A. (ed.) Proterozoic Basins of Canada. Geological Survey of Canada Paper, 81– 10, 25 –46. Young, G. M. 1976. Iron-formation and glaciogenic rocks of the Rapitan Group, Northwest Territories, Canada. Precambrian Research, 3, 137– 158. Young, G. M. 1995a. Are Neoproterozoic glacial deposits preserved on the margins of Laurentia related to the fragmentation of two supercontinents? Geology, 23, 153–156. Young, G. M. 1995b. Reply: Are Neoproterozoic glacial deposits preserved on the margins of Laurentia related to the fragmentation of two supercontinents? Geology, 23, 1054– 1055. Zachos, J., Pagani, M., Sloan, L. & Thomas, E. 2001. Trends, rhythms, and aberrations in global climate 65 Ma to present. Science, 292, 686– 693. Zhang, S., Jiang, G. & Han, Y. 2008. The age of the Nantuo Formation and Nantuo glaciation in South China. Terra Nova, 20, 289– 294. Ziegler, P. A. 1960. Fru¨hpala¨ozoische Tillite in o¨stlichen YukonTerritorium (Kanada). Eclogae Geologicae Helveticae, 52, 735– 741. Ziegler, A. M., Hulver, M. L., Lottes, A. L. & Schmachtenberg, W. F. 1984. Uniformitarianism and paleoclimates: inferences from the distribution of carbonate rocks. In: Brenchley, P. J. (ed.) Fossils and Climate. John Wiley & Sons, New York, 3 –25.

Chapter 3 Recognition of glacial influence in Neoproterozoic sedimentary successions EMMANUELLE ARNAUD1* & JAMES L. ETIENNE2 1

School of Environmental Sciences, University of Guelph, Guelph, Ontario, N1G 2W1, Canada 2

Neftex Petroleum Consultants Ltd, 97 Milton Park, Abingdon, Oxfordshire OX14 4RY, UK *Corresponding author (e-mail: [email protected])

Abstract: This chapter provides an overview and key references of glacial processes and resulting sedimentary products in subglacial, terrestrial proglacial and glaciomarine or glaciolacustrine settings. These settings are characterized by a wide variety of processes ranging from subglacial lodgement and deformation, ice-push and sediment remobilization, which in turn result in a wide range of products such as diamictite, conglomerate, sandstone, siltstone and mudstone. The sedimentary record of proglacial settings exhibits the most lateral and vertical variability due to the dynamic nature of ice margins and the most direct record of climatic fluctuations. Many Neoproterozoic successions, however, preserve glaciomarine deposits that can provide a more continuous and high-resolution (though indirect) record of change. This chapter will enable the reader to identify features that may be used to infer a glacial influence on the formation of ancient deposits. The chapter also outlines some of the important issues that require consideration when evaluating palaeoclimatic models for Neoproterozoic sedimentary successions. These include the equivocal significance of most commonly used proxies such as occurrence of diamictite, outsized clasts in laminated sediments, clast characteristics, lithostratigraphic trends and sequence boundaries. Careful analysis of multiple lines of sedimentary evidence, together with other proxies of climatic changes, can yield meaningful reconstructions and provide a basis for testing palaeoclimate models for this time period. A summary table outlining the characteristics of diamictite with different depositional origins is also included in order to assist with the interpretation of the Neoproterozoic sedimentary record.

The ability to recognize glacial deposits, or a glacial influence on sedimentation, has long been a subject of debate (Schermerhorn 1974; Deynoux & Trompette 1976; Jensen & Wulff-Pedersen 1996), and yet is critical for the accurate reconstruction of palaeoenvironmental conditions during past ice ages. This chapter summarizes the historical development of the terminology and approach used in identifying deposits of glacial origin, the characteristics of glacial environments in terms of processes, the resulting sedimentary facies and stratal geometries, as well as some key points to consider when reconstructing palaeoenvironmental conditions in ancient successions. The chapter is far from exhaustive, and many others have previously covered these topics in more detail elsewhere. Rather, it is meant to provide an overview with key references for workers who focus on other aspects of Neoproterozoic geology and may not necessarily be familiar with glacial deposits. As the focus of this book is on Neoproterozoic ice ages, characteristics related to micro- or macrofauna and biogenic indicators associated with glacial settings are not discussed. Similarly, geomorphic features are only briefly mentioned, as most Neoproterozoic successions do not have the prerequisite outcrop exposure or preservation potential of landforms found in more recent glacial settings.

Historical development The terminology used to refer to glacial deposits has evolved over the years, with an increasing appreciation of the processes occurring in glaciated basins and a desire to refine palaeoenvironmental interpretations to include more information about the severity of climatic conditions, the nature of the ice mass, the record of associated sea-level changes and depositional setting (Flint et al. 1960; Carey & Ahmad 1961; Dreimanis 1978; Hambrey & Harland 1978; Boulton & Deynoux 1981; Boulton 1990; Eyles 1993; Hambrey & Glasser 2003). The criteria used to identify glacial deposits in ancient sedimentary successions has also evolved with this refined understanding of processes and a recognition of the non-uniqueness of many of the sedimentary characteristics that were once thought diagnostic of glacial deposits (Crowell 1964; Harland 1964; Harland et al. 1966; Schermerhorn 1974;

Flint 1975; Hambrey & Harland 1978; Boulton & Deynoux 1981; Dreimanis & Schluchter 1985; Chumakov 1992; Eyles 1993; Crowell 1999). This improved understanding has occurred as a result of the emergence of modern sedimentological techniques, which focus on process and products, and the application of facies and basin analysis (Reading 1978), which have enhanced our ability to reconstruct palaeoenvironmental conditions from ancient sedimentary successions (see Eyles & Januszczak 2004 for a historical review). For the purpose of this book, we encouraged authors describing deposits from around the world to use the non-genetic term diamictite to describe poorly sorted materials that contain a mixture of gravel-, sand- and mud-sized particles (Fig. 3.1). Terms such as ‘tillite’ were to be reserved for the interpretation section where poorly sorted materials could be shown to have been deposited directly by ice, without significant subsequent disaggregation and flow. In describing reworked glacial deposits, the interpretive term ‘glaciogenic’ was recommended as a modifier (e.g. glaciogenic debris flow). Authors were also encouraged to use ‘outsized clast’ or ‘lonestone’ rather than the term ‘dropstone’ in their descriptions, leaving the latter word to the discussion where a dropstone origin could be interpreted based on deformation of underlying laminae and onlapping overlying laminae. The use of terms ‘varvite’ or ‘varve’ to describe laminated fine-grained sediment in glaciogenic successions was discouraged, considering the lack of radiometric control required to confirm annual depositional cycles. These are best described as ‘rhythmically laminated’ when a cyclical nature can be identified or simply ‘laminated’ when the stacking pattern appears random. Readers are referred to Hambrey & Glasser (2003) for an up-to-date discussion of terminology related to glacial sediments and Evans et al. (2006) for a more specific discussion of subglacial till classification.

Characteristics of glacial environments Deposition occurs within a number of different settings in glaciated basins, including subglacial, proglacial, glaciolacustrine and glaciomarine environments. Each of these depositional settings has characteristic processes, deposits and landforms

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 39– 50. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.3

40

E. ARNAUD & J. L. ETIENNE

Fig. 3.1. Classification of poorly sorted sediments according to Hambrey & Glasser (2003). Reproduced with permission from Springer.

(Menzies 1995, 1996; Benn & Evans 1998, 2010; Evans 2005). Sedimentary analysis of Neoproterozoic successions can thus provide much-needed palaeoenvironmental information when this variability is taken into consideration. In contrast to more recent glacial deposits, landforms tend not to be preserved or are poorly exposed in ancient successions, although there are exceptions. An understanding of the processes, together with the resulting deposits and their three-dimensional distribution in these settings, is thus critical for palaeoenvironmental reconstruction. Glaciomarine successions tend to be the most common in the Neoproterozoic geological record, as marine basins provide accommodation space away from subsequent ice advances and erosion, allowing preservation of thick packages of glaciogenic sediments (Carey & Ahmad 1961; Nystuen 1985; Eyles 1993). Although these settings tend to provide the longest and highest resolution records of Neoproterozoic climate change, the depocentres tend to be dislocated from the glaciated shelves, which acted as the sediment supply pathway, complicating efforts to reconstruct the distribution of palaeo-ice masses. An appreciation of other glacial settings is important as some successions will record conditions at the basin margin, close to land-based glaciers (Rieu et al. 2006; Arnaud 2008). Basin margin successions such as these are particularly useful in terms of palaeoclimate reconstruction as they can provide information regarding the nature of the ice mass (stagnant, active or fast-moving) and allow better control on the relative timing of the changes associated with that ice margin. They may also provide a better sense of the temporal relationship between glacial conditions and palaeoenvironmental perturbations inferred from geochemical data that have recently been so central to Neoproterozoic research. A discussion of the various settings in glaciated basins follows with particular emphasis on the nature of these environments, the processes that operate within them and the resulting sedimentary products. References provided herein are by no means exhaustive, but were selected as they focus on sedimentary characteristics and emphasize their connection to processes, thus facilitating the reconstruction of palaeoenvironmental conditions from the sedimentary record. Readers are referred to the ‘glaciers online’ site of Alean & Hambrey (2008) as well as Eyles (1993), Crowell (1999), Hambrey & Glasser (2003) and Etienne et al. (2007) for helpful photographs and additional general discussions of criteria used in the recognition of ancient glacial deposits.

Subglacial settings Conditions in subglacial settings are primarily controlled by the relationship between pressure and temperature, which ultimately controls the transition between solid and liquid phases and the alternation between melting and regelation in the basal traction zone. In other instances, ice will melt and refreeze in response to broader temperature fluctuations. These subglacial conditions in turn control the ability of the ice to erode, transport and deposit debris. The thermal characteristics of ice masses thus have a direct impact upon the accumulation of subglacial debris, with polythermal (variable distribution of warm and cold-based ice) and temperate glaciers (basal ice close to pressure melting point) accumulating the most significant volumes of debris. Although cold-based glaciers have been shown to entrain, transport and deposit sediment, they generally accumulate significantly less debris. In some cases, subglacial lakes may form, as shown by recent inventories of subglacial lakes in Antarctica (Siegert et al. 2005). Satellite data have demonstrated significant ice elevation changes associated with these subglacial lakes, and these have been attributed to major discharge events and reorganization of subglacial meltwater (Wingham et al. 2006; Fricker et al. 2007). Subglacial processes that are most relevant for palaeoenvironmental reconstructions include lodgement or plastering of debris onto the substrate, deformation of soft-sediment substrates, hydrofracturing of substrate (Boulton & Caban 1995), plucking of angular material from hardrock substrates as a result of freeze and thaw, abrasion of both the debris within the ice and the underlying substrate, and meltwater erosion. Meltwater processes are particularly significant in temperate glacial settings. The meltwater can be found in subglacial distributary channels, which comprise part of the drainage system of the glacier, or as a thin film at the base of the glacier, which helps facilitate movement. These processes are responsible for subglacial deposits such as lodgement and deformation tills (Boyce & Eyles 2000; Evans et al. 2006), commonly characterized by diamicton (the unlithified equivalent of diamictite), in addition to gravel and sand deposited in subglacial meltwater tunnels (Fig. 3.2; Table 3.1). Whereas deposition related to meltout or undermelting of ice in a subglacial setting is theoretically possible, studies have shown that it is very difficult to distinguish these deposits from other subglacial deposits considering the high likelihood of subsequent overprinting by

RECOGNITION OF GLACIAL DEPOSITS

41

Fig. 3.2. Photographs of typical glacial sedimentary indicators in Neoproterozoic successions. (a) Mud-rich diamictite with extra-basinal clasts; (b) outsized clast showing clear piercing and onlap relationship typical of dropstones; (c) striated pavement, Varangerfjorden, Norway; (d) striated clast; (e) bullet-shaped clast; (f) faceted clast. It should be noted that none of these features are exclusive to glacial conditions, so careful analysis of other features is necessary to confirm a glacial origin. See text for discussion.

deformation and lodgement processes (Evans et al. 2006). Subglacial processes also result in striated pavements, and clasts that show evidence of transport and erosion in the basal traction zone such as striations, faceting, chattermarks and a bullet shape (also referred to as stoss and lee clasts, flat-iron clasts or bullet-shaped boulders; Fig. 3.2, Table 3.1). Care must be taken to distinguish between striations arising from subglacial erosion and those that are related to slickenside development associated with faulting or pressure solution. Diamictite units may exhibit a gradational change from deformed and sheared at the base through chaotically stratified to massive at the top, indicative of the increase in stress with proximity to the ice – sediment interface (Boulton & Hindmarsh 1987; Hart & Boulton 1991; Boulton 1996; Benn & Evans 1996). Other sedimentary characteristics that can be indicative of subglacial settings are boulder pavements (boulders preferentially aligned one behind the other) and soft-sediment deformation features such as boudinage, tectonic lamination, rooted recumbent folding and inclusions of underlying material that exhibit shear stress in a preferred direction (Table 3.1; Hart

& Boulton 1991; Hart & Roberts 1994; McCarroll & Rijsdijk 2003; Roberts & Hart 2005; Arnaud 2008). Clastic dykes may also develop in areas where a frozen substrate and the overlying ice mass result in high pore water pressures (Boulton & Caban 1995; Le Heron & Etienne 2005). Preferred orientation of the a-axes of clasts (fabric) has also been used to discriminate between various depositional origins of diamictite (Domack & Lawson 1985; Hart 1994; Benn & Evans 1996); however, others have argued that clast fabric is not always indicative of a particular depositional setting and that other indicators of glacial conditions are needed to confirm interpretations (Bennett et al. 1999; Hambrey & McKelvey 2000). Preferred clast orientations can be very useful in determining direction of transport, which may in turn facilitate palaeoenvironmental discrimination (Table 3.1). Clast fabric analyses are somewhat more difficult for Neoproterozoic successions that have undergone multiple phases of postdepositional deformation. Finally, it is difficult to say what sedimentary expression subglacial lakes will have in the rock record and how their existence may be inferred without the

42

Table 3.1. Characteristics of diamictite units of different origin Types of diamictite

Textural features

Compositional features

Sedimentary structures

Deformation

Poorly sorted Boulder pavements Evidence of glacial transport on particle surface† Textural immaturity Wide range of stone roundness Micromorphological characteristics associated with high stress conditions Strong a-axis fabrics (lodgement)

Glaciotectonic Massive or Mixed provenance deformation such as chaotically (intra- and shear structures (rooted stratified extrabasinal clasts) folds, boudinage, Gradational upward Compositional attenuated bedding, change from highly immaturity tectonic laminae), deformed to Inclusions/ intraclasts nappes, imbricate massive or rafts of other structures, thrust or sediment types that Preferred orientation reverse faulting, shear of shear structures exhibit ductile planes and decollement deformation surfaces. Clastic dykes

Meltout tillite §

Poorly sorted Localized sorting from meltwater reworking Evidence of glacial transport on particle surface such as striations Textural immaturity Wide range of stone roundness

Crudely stratified Mixed provenance (discontinuous (intra- and layers of sorted extrabasinal clasts) sediments Compositional or draped layers immaturity over large clasts) Soft sediment or soft bedrock inclusions or intraclasts that exhibit no deformation

‘Rainout’ deposits

Poorly sorted Evidence of glacial transport on particle surface† Textural immaturity Wide range of stone roundness

Grounding ice berg Massive to crudely Mixed provenance structures and ice-keel stratified (intra- and furrows extrabasinal clasts) Discontinuous stringers of sand Compositional and gravel (boulder immaturity pavements) from Clast clusters from winnowing iceberg dumping Diamict pellets from Variable clast content from changing frozen debris on sedimentation rate icebergs or supply of Dropstones ice-rafted debris

Sharp Erosional or deformed Basal shear plane Grooves or deformation in sub-till sediments

Extensional faulting from Conformable Sharp melting of buried ice Convoluted bedding and Undisturbed substrate folding

Gradational

Geometry

Thin (,m to m) and tabular Laterally extensive (hundreds of km2) Discontinuous if close to ice margin

Rapid lateral facies changes

Variable thickness Tabular to blanket-like geometry, draping underlying topography Laterally continuous (distal) to rapid lateral facies changes (proximal)

Associated deposits

Useful references

Well bedded conglomerate and sandstone, commonly cross-stratified and exhibiting cut and fill geometry (glaciofluvial outwash) Thick deposits of windblown silt (loess), sandstone wedges, convoluted bedding from cryoturbation and in situ breccia‡ (periglacial) Laminated or massive mudstone with/without outsized clasts (lacustrine) Well bedded conglomerate and sandstone, commonly cross-stratified and exhibiting cut and fill geometry (glaciofluvial outwash) Laminated or massive mudstone with or without outsized clasts from ponded water or lake deposits Lodgement till from ice advance onto shelf Sediment gravity flow deposits Laminated mudstone with/ without outsized clasts Subaquatic fan deposits in proximal settings Pelagic clastic or carbonate sediments

Benn & Evans (1996) Boulton (1972) Boyce & Eyles (2000) McCarroll & Rijsdijk (2003) Hart & Roberts (1994) Menzies (2000) Evans et al. (2006)

Boulton (1972) Lawson (1982) McDonald & Shilts (1975) Evans et al. (2006)

Eyles et al. (1985) Thomas & Connell (1985)

E. ARNAUD & J. L. ETIENNE

Lodgment & deformation tillite*

Basal contact

Poorly sorted Evidence of glacial transport on particle surface* if glaciogenic source Textural maturity depending on source Wide range of stone roundness depending on source Subaqueous debris flows Poorly sorted Evidence of glacial transport on particle surface† Textural immaturity Wide range of stone roundness Terrestrial debris flows

Massive to chaotically stratified

Flow nose Basal shearing structures Convoluted bedding or folding

Erosional with rip-up clasts

Extent (m– km)

Lodgement or meltout tills

Lawson (1982) Benvenuti & Martini (2002) Colella & Prior (1990)

Massive to chaotically stratified Graded bedding Coarse-tail grading

Flow noses Basal shear structures Folding and convoluted bedding Load casts, water escape structures

Erosional with rip-up clasts Conformable when hydroplaning has occurred

Moderate to thick Tabular to lenticular Channelized Extent (m to 10s of km)

Lodgement tills ‘Rainout’ deposits Laminated mudstone with/ without outsized clasts Subaquatic fan deposits Pelagic clastic or carbonate sediments Other sediment gravity flow deposits (e.g. turbidites)

Nardin et al. (1979), Visser (1983), Prior et al. (1984), Postma et al. (1988), Hart & Roberts (1994), Mohrig et al. (1999), Eyles & Eyles (2000), Hambrey & McKelvey (2000), Laberg & Vorren (2000), Mulder & Alexander (2001), McCarroll & Rijsdijk (2003)

Note: No one of these features are sufficient for identification of a glacial or glacially influenced origin and any one of the features may be absent. *We combine these two types of till, as lodgement and deformation processes will often both affect individual deposits (see discussion in Evans et al. 2006); Evans et al. (2006) argued that meltout till generated in subglacial settings could similarly not be easily distinguished). † Evidence of glacial transport: striations, chattermarks, polish, grooves, facets, bullet shape (also referred to as ‘stoss and lee’ or ‘flat iron’ clasts). ‡ These sedimentary characteristics are not unique to periglacial settings; their origin may be difficult to confirm even with careful stratigraphic and sedimentological consideration. § Although meltout of debris is theoretically possible in subglacial settings, their identification in the rock record seems unlikely (Evans et al. 2006). Here, we only consider meltout tillite in the context of proglacial terrestrial settings (see text for discussion).

RECOGNITION OF GLACIAL DEPOSITS

Local (non-glacial) or mixed (glaciogenic source) provenance Compositional maturity depending on source Inclusions or rafts of other sediment types that exhibit ductile deformation Local (non-glacial) or mixed (glaciogenic source) provenance Compositional maturity depending on source Inclusions or rafts of other sediment types that exhibit ductile deformation, some can be very large (10s of m in size)

43

44

E. ARNAUD & J. L. ETIENNE

benefit of the geomorphic context that is available in modern environments. Studies to date have been based on remote sensing data, so very little is known about the types of deposits within these lakes and their longer-term preservation potential. Their unequivocal identification in the rock record would appear difficult, as deposits would look very similar to glaciomarine successions recording advances and recessions of ice. In Neoproterozoic successions, biota are not available to distinguish between the lacustrine and marine settings. In terms of three-dimensional geometry, lodgement and deformation tills tend to be tabular, relatively thin (on the scale of less than a metre to several metres) and can cover extensive areas (typically hundreds of km2). Basal contacts can be erosional or show evidence of deformation and incorporation of materials from the underlying substrate. Conglomerate and sandstone from subglacial tunnels will have sharp outer contacts and a pipe-like geometry in cross-section. The nature of the associated sediments will depend on whether or not this ice mass overrode land or a marine shelf. Terrestrial subglacial deposits may be associated with coarse-grained glaciofluvial outwash sediments and other terrestrial proglacial sediments such as glaciolacustrine facies (Boulton 1972), whereas glaciomarine subglacial deposits will be associated with subaqueous proglacial sediments, tidal rhythmites, calcareous, biogenic or clastic fine-grained sediments deposited by suspension settling and/or a range of fine- to coarsegrained sediment gravity flow deposits (e.g. turbidites and hyperconcentrated flows; Domack 1988; Cowan & Powell 1990; Anderson & Ashley 1991; Fairchild 1993; Dowdeswell et al. 1996, 1998; Anderson 1999; Taylor et al. 2002).

Terrestrial proglacial settings Conditions in proglacial settings are primarily controlled by the proximity to the ice margin and whether or not the ice margin is stagnant, actively advancing or surging. Proglacial processes include in situ meltout of debris, remobilization of debris downslope, reworking by meltwater and melting of buried stagnant ice (Boulton 1972; Lawson 1982). These processes typically result in highly irregular topography, ponding of meltwater and development of proglacial lakes. The entire proglacial succession may be deformed and cannibalized by subsequent ice advances that bulldoze, thrust and re-work the deposits. Such events are known from modern environments (Huddart & Hambrey 1996) and the ancient glacial geological record (Le Heron et al. 2005). Stagnating ice margins will often have multiple recessional moraines and abundant supraglacial debris with buried ice. In contrast, active ice margins will be characterized by multiple push moraines and ridges associated with the fluctuating ice margin (Boulton et al. 1999; Bennett 2001; Bennett et al. 2004). Highly variable depositional conditions in terrestrial proglacial environments (ponding, meltwater flow, bulldozing and remobilization of sediments) tend to result in complex lithofacies assemblages with a high degree of lateral variability. These assemblages may include diamictite (derived from meltout, subaerial debris flow and bulldozing and deformation of glaciogenic debris; Table 3.1), conglomerate, sandstone, siltstone and mudstone (meltwater processes) and laminated mudstone and siltstone (ponding) in these settings. Clasts may show evidence of recent transport in the basal traction zone such as faceting, striations, chattermarks and bullet-shape, but surface features rarely survive fluvial transport. Deformation structures ranging in scale from centimetres to hundreds of metres can be quite prevalent in these proglacial settings (Benn & Evans 1996; Phillips et al. 2002). Extensional faulting and ductile folding of sediments associated with in situ melting of ice can be found in all facies types. In addition, reverse or thrust faults, shear planes and low-angle de´collement surfaces, nappes and complex folding and shear structures can be found resulting from bulldozing by ice.

The wide range in depositional conditions associated with terrestrial proglacial environments often leads to an inherently complex distribution of sediment types and three-dimensional geometries that vary from irregular to channelized and lenticular with common intraformational unconformities and sudden bed termination. Basal contacts are variable from sharp to irregular and deformed or erosional to conformable and draping over irregular topography. Associated sediments are likely to be subglacial and glaciofluvial outwash deposits.

Glaciolacustrine and glaciomarine settings Conditions in glacially influenced water bodies are primarily controlled by their proximity to the ice margin, whether or not the margin terminates in the water body, the nature of the ice margin (i.e. tidewater glacier or floating ice shelf), the thermal regime of the ice mass (e.g. temperate v. polar) and the characteristics of the water body (lake v. ocean; deep sea, shelf, fjord or shoreline; Smith & Ashley 1985; Eyles et al. 1985; Powell & Elverhoi 1989; Dowdeswell & Scourse 1990; Anderson & Ashley 1991; Dowdeswell et al. 1998; Evans & Pudsey 2002). Lakes that are dammed by ice or sediment (such as moraines) are also regulated by dam stability (Tweed & Russell 1999). Glacial processes dominate in settings that are close to the ice margin (,kms), whereas normal lacustrine/marine processes dominate in settings that are distal to the ice margin (.tens of km). Where ice terminates in a water body (ice-contact lakes or tidewater glacier marine settings), depositional settings are affected by melt-out of subglacial and englacial debris, slumping off the ice margin, iceberg calving, efflux from subglacial and englacial meltwater conduits and sediment input from surface meltwater drainage. These processes supply sediment to grounding line fan complexes and more distal areas. In contrast, water bodies fed by glacial meltwater are primarily influenced by diurnal and seasonal fluctuations in meltwater discharge and seasonal ice cover. Meltwater processes are very important in settings influenced by temperate glaciers and limited to non-existent in polar settings. The configuration of the basin will influence the types of non-glacial processes that also contribute to the distribution and nature of deposits found in these settings. For example, lakes may experience thermal stratification and seasonal mixing, whereas deposition in oceans will be affected by tides and salinity. Sediment gravity flows resulting from high sedimentation rates are common in most water bodies but may be more prevalent in steep-sided lakes or fjords. However, it is important to note that subaqueous glaciogenic debris flows may have run-out distances of hundreds of kilometres, even on shallow slopes (Dowdeswell et al. 1996, 1998; Taylor et al. 2002). Oscillatory waves will affect shallow water settings, while storm waves will affect both shallow and shelf settings. Processes include rapid deposition of coarse-grained sediments at the mouth of subglacial tunnels, density underflows and suspension-settling of fine-grained sediments related to buoyant meltwater plumes and sub-ice shelf melting, winnowing or reworking by sub-ice shelf or open water currents as well as bulldozing, lodgement and deformation of sediments related to ice ´ Cofaigh & Dowdeswell 2001; Evans advances (Lønne 1995; O et al. 2005; McKay et al. 2008; Naish et al. 2009). Remobilization of debris by gravity flows is common in these settings and results from high sedimentation rates, oversteepening along the edges of grounding-line fans, and normal sediment-transfer processes at the shelf-slope break or along steep-sided fjords. Gravity flows may also be initiated directly by subaqueous iceberg calving, or as a result of fluctuating water depths associated with calving or abrupt lake drainage events. Icebergs can have an impact on both proximal and distal settings through iceberg roll-over and dumping, meltout of entrained debris and scouring and bulldozing when grounded in shallow water (Thomas & Connell 1985; Gilbert

RECOGNITION OF GLACIAL DEPOSITS

1990; Woodworth-Lynas & Guigne 1990). Deposition of coldwater carbonates is also common in glacially influenced polar settings. Ikaite (hydrated calcium carbonate) is one such mineral that forms in near-freezing water temperatures. Unstable at higher temperatures, ikaite has poor preservation potential and typically undergoes replacement by calcite. Where the original crystal structure is preserved, these pseudomorphs are called glendonites and are an important palaeoenvironmental indicator of cold-water conditions (James et al. 2005; Domack et al. 2007; Halverson 2011). Processes such as those identified above often lead to a predominance of diamictite and mudstone in glacially influenced lacustrine/marine settings, although conglomerate and sandstone can also be found associated with grounding-line fans and subglacial tunnel outflow as well as stratified facies related to sub-ice shelf sedimentation or reworking by sub-ice shelf or open water currents (Eyles et al. 1985; Lønne 1995; Nemec et al. 1999; Barrett 2007; McKay et al. 2008). In open water conditions where icebergs are calving or in sub ice-shelf settings, suspension settling of finegrained sediment and rainout of ice-rafted debris can lead to the formation of diamictite (Table 3.1) or mudstone with outsized clasts. Where outsized clasts rafted from icebergs deform underlying laminated sediments (truncating laminae, loading sediment to develop compaction-related folds and/or water escape structures) they may be referred to as dropstones (Fig. 3.2; Thomas & Connell 1985). Material of pebble size or greater is generally considered as a good proxy for ice-rafting, although some care is needed in interpretation (see discussion below). In some cases, frozen masses of ice-rafted debris are released from icebergs into water bodies and result in till pellets or inclusions of diamict within either fine-grained sediments or a diamict with slightly different characteristics (Gilbert 1990). These can be distinguished from intraclasts resulting from erosion at the base of sediment gravity flows by closely examining the type of materials and their relationship to underlying materials, the nature of the outer boundary of these inclusions (sharp and irregular for erosional intraclast v. diffuse for till pellets) as well as the basal contact of the unit in which they are found. Ice advance onto the continental shelf can also result in the deposition of lodgement or deformation tillite (Table 3.1), as demonstrated by overconsolidated diamicton displaying shear structures, clastic dykes and associated subglacial landforms such as grooves, lineations and drumlins in recent continental shelf sediments of Antarctica (Camerlenghi et al. 2001; ´ Cofaigh et al. 2002; Evans et al. 2005; Naish et al. 2009). DiaO mictite may of course also form as a result of sediment gravity flows (Table 3.1; Wright & Anderson 1980; Eyles & Eyles 2000; Mulder & Alexander 2001). Rhythmically laminated mudstone may result from tidal pumping in ice-proximal marine settings (Smith et al. 1990) or from seasonal discharge fluctuations in non-ice contact lakes (Leonard 1986). The latter are referred to as varves in more recent deposits, but the use of this term is discouraged in Neoproterozoic successions as a seasonal control cannot be clearly demonstrated due to limited geochronological constraints. Although Evans & Pudsey (2002) suggested that iceshelf deposits could be distinguished from tidewater glacier deposits by the absence of meltwater-derived facies, similar stratified sediments have been reported from ice-shelf settings in Antarctica and attributed to grounding line zones (Evans et al. 2005; McKay et al. 2008; Naish et al. 2009), making it unlikely that the two will be easy to distinguish in ancient glaciogenic successions in the absence of biogenic facies, which are sensitive to ice cover, or contextual knowledge of the ice margin available in more recent glaciomarine environments. Laminated or massive fine-grained sediments with rare clasts dominate in ice distal settings where typical marine processes dominate and suspension settling from icebergs provide a marginal contribution to sedimentation. Boulder pavements may develop on marine shelves as a result of winnowing of diamict and overriding by ice during sea-level lowstands (Eyles 1994). Deformation may

45

be found in ice-proximal grounding line fan deposits as a result of slumping and bulldozing by ice, in remobilized glaciogenic debrisflow deposits along the continental slope and in other deposits where iceberg grounding has occurred (Table 3.1; WoodworthLynas & Guigne 1990; Hart & Roberts 1994).

The record of climatic changes over time in glacial environments Ancient glaciogenic sedimentary successions contain a record of fluctuating climatic conditions over time (Boulton 1990). Ice ages typically consist of interstadials (relatively warmer) and stadials (relatively colder) periods, which are recorded in the partial advance and recession of ice and changes in the lateral and vertical distribution of sediments associated with these fluctuating environments. Such fluctuations have been documented in several Neoproterozoic successions (Allen et al. 2004; Arnaud & Eyles 2006), as well as in more recent glacial successions (Domack 1983; Barrett 2007; Naish et al. 2009). Changes from glacial to non-glacial periods will similarly be recorded at the margins of marine basins. Here, glacioisostacy (crustal movements associated with ice loading and unloading) and glacioeustasy (changes in global ocean volume associated with changes in terrestrial ice volume) will combine with local tectonics and sediment supply to influence relative sea level and ultimately the nature of stratal geometries (Ravna˚s & Steel 1998; Powell & Cooper 2002; Allen 2007). Glacio-eustatic variability can lead to significant correlative global cycles of deposition. Glaciations affect various parts of the world and can be expressed quite differently depending on the thermal characteristics of the ice mass mentioned above. These characteristics can vary spatially and temporally. Today, temperate glaciers (i.e. warm ice close to melting point) are found in Alaska, polythermal glaciers (mixture of warm and cold ice) in Greenland and polar glaciers (cold ice well below melting point) in Antarctica. However, these thermal characteristics can change over time, as shown by the sedimentary record of the Ross Ice Shelf, which shows temperate conditions in the past compared to the polar conditions of today (Naish et al. 2009).

Reconstructing Neoproterozoic palaeoenvironmental conditions: key considerations The sedimentary record is at times difficult to interpret as many, if not most, sedimentary characteristics are not unique to specific environments. The following discussion highlights key points to consider when reconstructing palaeoenvironmental conditions from Neoproterozoic successions, focusing on indicators typically used to infer glacial conditions.

The palaeoclimatic significance of diamictite Diamictite can result from both glacial and non-glacial processes (Table 3.1). In glaciated settings, diamictite results from lodgement and deformation of subglacial sediments, in situ meltout of ice and release of poorly sorted basal, englacial or supraglacial debris, slumping and reworking of glaciogenic debris in proglacial, glaciolacustrine or glaciomarine settings and from ‘rainout’ of fine-grained sediment and ice-rafted debris. In non-glaciated settings, diamictite can form as a result of terrestrial or subaqueous debris flows (hyperconcentrated flows) or as a result of impact ejecta (Fisher 1971; Nardin et al. 1979; Rampino 1994; MartinsNeto 1996; Sohn 2000; Weaver et al. 2000; Mulder & Alexander 2001; Benvenuti & Martini 2002). Studies of modern glaciated settings and of sediment gravity flows and their deposits have greatly improved our ability to link depositional processes with their

46

E. ARNAUD & J. L. ETIENNE

sedimentary products. However, careful consideration of sedimentary characteristics, geometry, as well as lateral and vertical distribution of diamictite and associated deposits, is needed to distinguish between these various depositional origins (Arnaud & Eyles 2006; Table 3.1). For example, a diamictite formed as a result of ‘rainout’ may have a gradational basal contact, whereas those related to sediment gravity flows or subglacial lodgement tend to have sharp conformable or erosional basal contacts (Table 3.1). In a marine setting, diamictite showing draped upper contact may be attributed to a sediment gravity flow rather than subglacial processes. Diamictite formed as a result of subglacial lodgement may exhibit upwardly increasing deformation (Benn & Evans 1996), in contrast to deposits of sediment gravity flows where the maximum deformation will occur at the base of the deposit (Nardin et al. 1979; Mulder & Alexander 2001). Associated facies play a particularly key role in distinguishing between terrestrial or marine settings in these ancient successions where faunal indicators are lacking. Careful analysis is particularly important in considering ancient glaciogenic successions, as many of them were deposited in tectonically active settings where diamictite may record tectonically induced sediment instability (Krzyszkowski 1993; Tanaka & Maejima 1995) rather than renewed glacial conditions. Sediment– landform associations may not be available in these basins to better constrain a glacial influence. Even with the most careful examination, the depositional origin of diamictite will sometimes be difficult to determine unequivocally as indicators are sometimes absent or obscured by post- depositional tectonism or weathering. This is particularly significant considering that Neoproterozoic deposits will have been subjected to a long history of post-depositional modification with sometimes very limited and poor preservation in structurally complex settings. Weathering is problematic when looking for surface markings on carbonate clasts (e.g. striations, faceting or bullet-shape) within Neoproterozoic diamictite, a common problem considering that many Neoproterozoic glaciogenic successions overlie carbonate strata and have incorporated these softer lithologies. In other situations, interpretation will be confounded by the fact that the poorly sorted debris has been affected by a number of different processes, and therefore retains a composite sedimentary signature of these processes. In their comprehensive review of subglacial processes and products, Evans et al. (2006) suggested that it will be difficult to distinguish between different types of tillites in ancient glacial deposits, as diamictite will ultimately record a composite of these subglacial processes and the dynamic nature of subglacial systems as they change over time and space. In another example, glaciomarine deposits could be subsequently over-ridden by ice, preserving some characteristics of both glaciomarine and subglacial processes and thus making it very difficult to identify whether these represent open glaciomarine or subglacial conditions. Another equally difficult situation is where large-scale controls on the sedimentary succession can leave similar signatures in the rock record such as diamictites formed in offshore Alaska during the late Miocene (Eyles et al. 1991), which were affected by both climatic and tectonic controls and Neoproterozoic volcaniclastic deposits in the Arabian –Nubian shield (Miller et al. 2009), which may be related to glaciogenic and/or volcaniclastic processes. Debates over the palaeoclimatic significance of the Neoproterozoic diamictite in Namibia, for example, revolve in part around the relative importance of tectonic and climatic controls on sedimentation and the different processes ascribed to these deposits. It is important to acknowledge these uncertainties where relevant, because very different palaeoenvironmental conditions are implied depending on the depositional process responsible for the formation of diamictite. In places where it is possible, careful consideration of the sedimentary characteristics and spatial geometry of the diamictite and its associated deposits can greatly refine

our depositional models by distinguishing between different types of depositional process, glacial setting (e.g. terrestrial proglacial v. glaciomarine/glaciolacustrine) and thermal regime, which can in turn contribute much needed constraints on evaluating palaeoclimate models that have been proposed for this time period (Condon et al. 2002; Leather et al. 2002; Allen et al. 2004; Arnaud & Eyles 2006; Rieu et al. 2006; Allen & Etienne 2008).

The palaeoclimatic significance of outsized clasts in bedded sediments Outsized clasts in laminated sediments have often been interpreted as ice-rafted debris or ‘dropstones’ and a clear indication of glacial conditions. Piercing or disturbance of underlying laminae and draping of sediment on top of the outsized clasts are probably the best criteria to imply that the clast was dropped from floating ice (Thomas & Connell 1985). Without this deformation and onlap relationship, the outsized clast may in fact be an outrunner in a succession of turbidites and other sediment gravity flows (Postma et al. 1988) or a clast surrounded by lamination resulting from post-depositional compaction or deformation. Once a ‘dropstone’ origin has been established, care must still be taken in terms of inferring glacial conditions from the presence of these outsized clasts. Ice-rafted debris may be derived from seasonal ice unrelated to any ice margin, proximal icebergs or far-travelled icebergs (Gilbert 1990). Distinction of seasonal ice, proximal icebergs and far-travelled icebergs is very unlikely in ancient successions. It is therefore important to recognize that the occurrence of dropstones can have varying palaeoclimatic significance (cold climate if resulting from seasonal ice, locally glaciated or glaciated some distance away if resulting from icebergs). This is an important consideration when using the sedimentary record to test competing palaeoclimate models for this time period.

The palaeoclimatic significance of clast characteristics Clast characteristics suggesting transport in the basal traction zone of glaciers are also often used to infer glacial conditions. These include the presence of striations, chattermarks, faceting, bullet-shaped boulders or flat iron clasts and extra-basinal provenance. It is important to note however, that although these confirm a glaciogenic influence on transport history, contextual analysis is required to ensure that the clasts have not been re-worked into younger deposits lacking a direct glacial influence. Although it is true that such surface markings are unlikely to survive with significant transport in meltwater glaciofluvial systems, many glaciomarine deposits (which tend to be the most common in ancient successions) may not have undergone significant fluvial reworking. This is particularly relevant when considering the amount of time elapsed between the deposition of a diamictite thought to record glacial conditions and its associated ‘cap’ carbonate thought to record warmer post-glacial conditions in Neoproterozoic successions.

The palaeoclimatic significance of stratigraphic trends and sequence boundaries Many studies of glacially influenced successions have equated alternating beds of diamictite and other sediments as the product of multiple advances and recessional phases of ice margins and multiple diamictite-bearing intervals of successions separated by thick sections of non-diamictite bearing sediments as representing discrete ice ages. Although this may turn out to be true (Naish et al. 2009), such interpretations must be made with caution considering the various depositional origins that can be ascribed to diamictite, the possibility that the succession at times is too distal to the ice margin to preserve any clear indicator of glacial conditions (e.g. relatively high-order transgressive or highstand systems tracts

RECOGNITION OF GLACIAL DEPOSITS

characterized by ice-distal mudstones in glaciomarine settings with limited or no icebergs), and the lack of geochronological control to constrain the amount of time represented by those thick non-diamictite bearing successions. The lateral and vertical heterogeneity of lithofacies within glacially influenced successions can also be highly complex. In both terrestrial and shallow marine (neritic) environments, regional sequence boundaries may be developed as a result of subglacial erosion either directly by ice via processes of regelation, abrasion from subglacially entrained debris, or by meltwater drainage under normal or elevated hydrostatic pressures. Multiple glacial advance-recessional cycles across marine shelf areas may result in stacked sequence boundaries, with lowstand packages of sediment in slope to bathyal environments providing the only continuous record of glacial – interglacial cyclicity. A cryptic shallow marine record provided by multistorey stacking of subglacial drainage conduits such as tunnel valleys may also be preserved, although these are not well documented from Neoproterozoic successions (see Ordovician examples in Le Heron et al. 2009). More localized disconformities may develop in marine environments as a result of iceberg keel-scour and processes of submarine mass-wasting related to high glaciomarine sedimentation rates or local tectonic activity. In terrestrial proglacial environments, the suite of erosion processes associated with braided fluvial systems, glaciotectonic deformation, and incision generated by glacial lake outburst floods or jo¨kulhlaups resulting from subglacial eruptions can all lead to local or regional sequence boundary development. This picture can become even more complicated where thin cold-based ice occurs in accumulation areas and pre-existing sediments such as periglacial deposits can be preserved with little or no modification, effectively leading to a correlative surface lacking erosion. In addition, as ancient glaciogenic successions tend to be preferentially preserved in basins that were tectonically active, the evolution of the basin must be well constrained in order to assess the impacts on accommodation space, sediment supply, gross-scale stratigraphic architecture and characteristics of sequence boundaries (Allen 2007). The points above highlight some of the considerations when applying a lithostratigraphic or sequence stratigraphic analysis within glacially influenced successions.

Conclusions Criteria for the recognition of glacial deposits in Neoproterozoic sedimentary successions have developed over the last 50 years with an increasing appreciation for the different kinds of settings and processes that occur within glaciated basins. The analysis of modern glacial settings and processes has allowed us to better recognize the complexity of glaciated basins and the various challenges in confirming a glacial influence on deposition. This includes a better appreciation of the processes occurring in glaciomarine settings as well as the difficulty in distinguishing the nature of floating ice producing dropstones (local iceberg, distant iceberg or sea ice), the depositional origin of diamictite (glacial, glacially influenced or non-glacial) and the controls on large-scale stratigraphic trends in facies successions (alternating lithofacies assemblages associated with autocyclicity or eustatic trends of transgression and regression). Poorly sorted deposits (diamictite), outsized clasts in laminated lithofacies (dropstones), high lateral facies variability in proximal settings and the presence of clasts that show evidence of subglacial transport are common features in glacial settings and these remain common indicators used in the study of ancient glaciogenic successions. However, recent studies have shown that such sedimentary characteristics can have widely different climatic implications such that an understanding of the processes and products of glacial settings and sediment gravity flows as well as careful

47

evaluation of sedimentary characteristics, geometry of deposits, deformation styles and associated facies are needed to identify a glaciogenic source and reconstruct the extent and nature of glacial conditions at the time of deposition. E. Arnaud’s research in glacial sedimentology is funded by the Natural Sciences and Engineering Research Council. J. L. Etienne’s work follows research experience on modern, Pleistocene and Neoproterozoic glacial deposits funded under NERC Studentship NER/S/A/2000/03690/ and Schweizerischer NationalFonds Grant 103502. E. Arnaud thanks Carolyn Eyles (McMaster University). J.L. Etienne thanks Mike Hambrey, Neil Glasser (University of Aberystwyth), Dan Le Heron (Royal Holloway) and Philip Allen (Imperial College London). Constructive reviews from David Evans and Eugene Domack greatly improved the manuscript and are much appreciated. This work represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Alean, J. & Hambrey, M. J. 2008. ‘Earth’s glacial record’ in Glaciers Online. (Available at http://www.swisseduc.ch/glaciers/earth_icy_ planet/glaciers15-en.html.) Allen, P. A. 2007. The Huqf Supergroup of Oman: basin development and context for Neoproterozoic glaciation. Earth-Science Reviews, 84, 139– 185. Allen, P. A. & Etienne, J. L. 2008. Sedimentary challenge to snowball Earth. Nature Geoscience, 1, 817– 825. Allen, P. A., Leather, J. & Brasier, M. D. 2004. The Neoproterozoic Fiq glaciation and its aftermath, Huqf Supergroup of Oman. Basin Research, 16, 507– 534. Anderson, J. B. 1999. Antarctic Marine Geology. Cambridge University Press, Cambridge. Anderson, J. B. & Ashley, G. M. (eds) 1991. Glacial marine sedimentation; paleoclimatic significance. Geological Society of America Special Paper, 261, 232. Arnaud, E. 2008. Deformation in the Neoproterozoic Smalfjord Formation, northern Norway: an indicator of glacial depositional conditions? Sedimentology, 55, 335–356. Arnaud, E. & Eyles, C. H. 2006. Neoproterozoic environmental change recorded in the Port Askaig Formation, Scotland: climatic vs tectonic controls. Sedimentary Geology, 183, 99 –124. Barrett, P. J. 2007. Cenozoic climate and sea level history from glacimarine strata off the Victoria Land coast, Cape Roberts Project, Antarctica. In: Hambrey, M. J., Christoffersen, P., Glasser, N. F. & Hubbard, B. (eds) Glacial Sedimentary Processes and Products. International Association of Sedimentologists Special Paper, 39, 259– 287. Benn, D. I. & Evans, D. J. A. 1996. The interpretation and classification of subglacially-deformed materials. Quaternary Science Reviews, 15, 23– 52. Benn, D. I. & Evans, D. J. A. 1998. Glaciers and Glaciation. Arnold Publishers, London. Benn, D. I. & Evans, D. J. A. 2010. Glaciers and Glaciation, 2nd edn. Hodder, London. Bennett, M. R. 2001. The morphology, structural evolution and significance of push moraines. Earth Science Reviews, 53, 197–236. Bennett, M. R., Waller, R. I., Glasser, N. F., Hambrey, M. J. & Huddart, D. 1999. Clast fabrics: genetic fingerprint or wishful thinking? Journal of Quaternary Science, 14, 125–135. Bennett, M. R., Huddart, D. et al. 2004. Sedimentary and tectonic architecture of a large push moraine: a case study from Hagafjellsjokull-Eystri, Iceland. Sedimentary Geology, 172, 269– 292. Benvenuti, M. & Martini, I. P. 2002. Analysis of terrestrial hyperconcentrated flows and their deposits. Special Publications International Association of Sedimentologists, 32, 167– 193. Boulton, G. S. 1972. Modern arctic glaciers as depositional models for former ice sheets. Journal of the Geological Society, London, 128, 361– 393. Boulton, G. S. 1990. Sedimentary and sea-level changes during glacial cycles and their control on glacimarine facies architecture. In: Dowdeswell, J. A. & Scourse, J. D. (eds) Glacimarine

48

E. ARNAUD & J. L. ETIENNE

Environments: Processes and Sediments. Geological Society, London, Special Publications, 53, 15– 53. Boulton, G. S. 1996. Theory of glacial erosion, transport and deposition as a consequence of subglacial sediment deformation. Journal of Glaciology, 140, 43 –62. Boulton, G. S. & Caban, P. 1995. Groundwater flow beneath ice sheets: Part II — its impact on glacier tectonic structure and moraine formation. Quaternary Science Reviews, 14, 563– 587. Boulton, G. S. & Deynoux, M. 1981. Sedimentation in glacial environments and the identification of tills and tillites in ancient sedimentary sequences. Precambrian Research, 15, 397– 422. Boulton, G. S. & Hindmarsh, R. C. A. 1987. Sediment deformation beneath glaciers: rheology and sedimentological consequences. Journal of Sedimentary Research, 92, 9059–9082. Boulton, G. S., Van Der Meer, J. J. M., Beets, D. J., Hart, J. K. & Ruegg, G. H. J. 1999. The sedimentary and structural evolution of a recent push moraine complex: Holmstrombreen, Spitsbergen. Quaternary Science Reviews, 18, 339–371. Boyce, J. I. & Eyles, N. 2000. Architectural element analysis applied to glacial deposits: internal geometry of a late Pleistocene till sheet, Ontario, Canada. Geological Society of America Bulletin, 112, 98 –118. Camerlenghi, A., Domack, E. et al. 2001. Glacial morphology and post glacial contourites in northern Prince Gustav Channel (NW Weddell Sea, Antartica). Marine Geophysical Researches, 22, 417– 443. Carey, S. W. & Ahmad, N. 1961. Glacial marine sedimentation. In: Raasch, G. O. (ed.) Geology of the Arctic. Volume II. University of Toronto Press, Toronto, 865– 894. Chumakov, N. M. 1992. The Problems of Old Glaciations (PrePleistocene Glaciogeology in the USSR). Harwood Academic Publishers, New York. Colella, A. & Prior, D. B. (eds) 1990. Coarse-Grained Deltas. International Association of Sedimentologists Special Publication, 10. Blackwell Scientific Publications, Oxford. Condon, D. J., Prave, A. R. & Benn, D. J. 2002. Neoproterozoic glacialrainout intervals: Observations and implications. Geology, 30, 35 –38. Cowan, E. A. & Powell, R. D. 1990. Suspended sediment transport and deposition of cyclically interlaminated sediment in a temperate glacial fjord, Alaska, U.S.A. In: Dowdeswell, J. A. & Scourse, J. D. (eds) Glacimarine Environments: Processes and Sediments. Geological Society, London, Special Publications, 53, 75– 89. Crowell, J. C. 1964. Climatic significance of sedimentary deposits containing dispersed megaclasts. In: Nairn, A. E. M. (ed.) Problems in Palaeoclimatology. Interscience Publishers, London, 86 – 99. Crowell, J. C. 1999. Pre-Mesozoic Ice Ages: their Bearing on Understanding the Climate System, 192. Geological Society of America, Memoir, 106. Deynoux, M. & Trompette, R. 1976. Late Precambrian mixtites: glacial and/or non-glacial? Dealing especially with the mixtites of West Africa. American Journal of Science, 276, 1302–1324. Domack, E. W. 1983. Facies of late Pleistocene glacial-marine sediments on Whidbey Island, Washington: an istostatic glacial marine sequence. In: Molnia, B. F. (ed.) Glacial-Marine Sedimentation. Plenum, New York, 535–570. Domack, E. W. 1988. Biogenic facies in the Antarctic glaciomarine environment: basis for a polar glaciomarine summary. Palaeogeography, Palaeoclimatology, Palaeoecology, 63, 357– 372. Domack, E. W. & Lawson, D. E. 1985. Pebble fabric in an ice-rafted diamicton. Journal of Geology, 93, 577– 591. Domack, E. W., Halverson, G. P., Willmott, V., Leventer, A., Brachfield, S. & Ishman, S. 2007. Spatial and temporal distribution of ikaite crystals in Antarctic glacial marine sediments. United States Geological Survey and The National Academies, USGS OF-2007-1047, Extended Abstract 015, 5. Dowdeswell, J. A. & Scourse, J. D. (eds) 1990. Glaciomarine Environments: Processes and Sediments. Geological Society, London, Special Publications, 53. Dowdeswell, J. A., Elverhøi, A. & Spielhagen, R. 1998. Glacimarine sedimentary processes and facies on the polar North Atlantic Margins. Quaternary Science Reviews, 17, 243–272.

Dowdeswell, J. A., Kenyon, N. H., Elverhøi, A., Laberg, J. S., Hollender, F.-J., Mienert, J. & Siegert, M. J. 1996. Large-scale sedimentation on the glacier-influenced Polar North Atlantic margins: Long-range side-scan sonar evidence. Geophysical Research Letters, 23, 3535– 3538. Dreimanis, A. 1978. The problems of waterlain tills. In: Schluchter, C. (ed.) Moraines and Varves. A. A. Balkema Publishers, Rotterdam, 167– 177. Dreimanis, A. & Schluchter, C. 1985. Field criteria for the recognition of till and tillite. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 7– 14. Etienne, J. L., Allen, P. A., Rieu, R. & Le Guerroue´, E. 2007. Neoproterozoic glaciated basins: a critical review of the Snowball Earth hypothesis by comparison with Phanerozoic basins. In: Hambrey, M. J., Christoffersen, P., Glasser, N. F. & Hubbard, B. (eds) Glacial Sedimentary Processes and Products. International Association of Sedimentologists Special Publication, 39, 343–399. Evans, D. J. A. (ed.) 2005. Glacial Landsystems. Oxford University Press. Evans, J. & Pudsey, C. J. 2002. Sedimentation associated with Antarctic Peninsula ice shelves: implications for palaeoenvironmental reconstructions of glaciomarine sediments. Journal of the Geological Society, London, 159, 233–237. Evans, D. J. A., Phillips, E. R., Hiemstra, J. F. & Auton, C. A. 2006. Subglacial till: Formation, sedimentary characteristics and classification. Earth Science Reviews, 78, 115– 176. Evans, J., Pudsey, C. J., O’Cofaigh, C., Morris, P. & Domack, E. 2005. Late Quaternary glacial history, flow dynamics and sedimentation along the eastern margin of the Antarctic Peninsula Ice Sheet. Quaternary Science Reviews, 24, 741– 774. Eyles, C. H. 1994. Intertidal boulder pavements in the northeastern Gulf of Alaska and their geological significance. Sedimentary Geology, 88, 161– 173. Eyles, N. 1993. Earth’s glacial record and its tectonic setting. EarthScience Reviews, 35, 1 –248. Eyles, C. H. & Eyles, N. 2000. Subaqueous mass flow origin for Lower Permian diamictites and associated facies of the Grant Group, Barbwire Terrace, Canning Basin, Western Australia. Sedimentology, 47, 343– 356. Eyles, N. & Januszczak, N. 2004. ‘Zipper-rift’: a tectonic model for Neoproterozoic glaciations during the breakup of Rodina after 750 Ma. Earth-Science Reviews, 65, 1– 73. Eyles, C. H., Eyles, N. & Lagoe, M. B. 1991. The Yakataga Formation: a six million year record of temperate glacial marine sedimentation in the Gulf of Alaska. In: Anderson, J. B. & Ashley, G. M. (eds) Glacial Marine Sedimentation: Paleoclimatic Significance. Geological Society of America Special Paper, 261, 159– 180. Eyles, C. H., Eyles, N. & Miall, A. D. 1985. Models of glaciomarine sedimentation and their application to the interpretation of ancient glacial sequences. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 15 –84. Fairchild, I. 1993. Balmy shores and icy wastes: The paradox of carbonates associated with glacial deposits in Neoproterozoic times. Sedimentology Review, 1, 1– 16. Fisher, R. V. 1971. Features of coarse-grained, high concentration fluids and their deposits. Journal of Sedimentary Petrology, 41, 916– 927. Flint, R. F. 1975. Features other than diamicts as evidence of ancient glaciations. In: Wright, A. E. & Morsley, F. (eds) Ice Ages: Ancient and Modern. Seelhouse Press, Liverpool, 121–135. Flint, R. F., Sanders, J. E. & Rodgers, J. 1960. Diamictite, a substitute term for symmictite. Bulletin of the Geological Society of America, 71, 1809– 1810. Fricker, H. A., Scambos, T., Bindschadler, R. & Padman, L. 2007. An active subglacial water system in West Antarctica mapped from space. Science, 315, 1544–1548. Gilbert, R. 1990. Rafting in glacimarine environments. In: Dowdeswell, J. A. & Scourse, J. D. (eds) Glacimarine Environments: Processes and Sediments. Geological Society, London, Special Publications, 53, 105– 121. Halverson, G. P. 2011. Glacial sediments and associated strata of the Polarisbreen Group, Northeastern Svalbard. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological

RECOGNITION OF GLACIAL DEPOSITS

Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 571–579. Hambrey, M. J. 1994. Glacial Environments. UCL Press, London. Hambrey, M. J. & Glasser, N. F. 2003. Glacial sediments: processes, environments and facies. In: Middleton, G. V. (ed.) Encyclopedia of Sediments and Sedimentary Rocks. Kluwer Academic Publishers, Dordrecht, 316–331. Hambrey, M. J. & Harland, W. B. 1978. Analysis of Pre-Pleistocene glacigenic rocks: aims and problems. In: Schluchter, C. (ed.) Moraines and Varves. A. A. Balkema Publishers, Rotterdam, 271– 275. Hambrey, M. J. & McKelvey, B. 2000. Neogene fjordal sedimentation on the western margin of the Lambert Graben, East Antarctica. Sedimentology, 47, 577– 607. Harland, W. B. 1964. Critical evidence for a great infra-cambrian glaciation. Geologisches Rundschau, 54, 45– 61. Harland, W. B., Herod, K. N. & Krinsley, D. H. 1966. The definition and identification of tills and tillites. Earth-Science Reviews, 2, 225– 256. Hart, J. K. 1994. Till fabric associated with deformable beds. Earth Surface Processes and Landforms, 19, 15 – 32. Hart, J. K. & Boulton, G. S. 1991. The interrelation of glaciotectonic and glaciodepositional processes within the glacial environment. Quaternary Science Reviews, 10, 335– 350. Hart, J. K. & Roberts, D. H. 1994. Criteria to distinguish between subglacial glaciotectonic and glaciomarine sedimentation, I. Deformation styles and sedimentology. Sedimentary Geology, 91, 191– 213. Huddart, D. & Hambrey, M. J. 1996. Sedimentary and tectonic development of a high-arctic, thrust-moraine complex: Comfortlessbreen, Svalbard. Boreas, 25, 227– 243. James, N. P., Narbonne, G. M., Dalrymple, R. W. & Kyser, C. 2005. Glendonites in Neoproterozoic low-latitude interglacial sedimentary rocks, northwest Canada: insights on the Cryogenian ocean and Precambrian cold carbonates. Geology, 33, 9 – 12. Jensen, P. A. & Wulff-Pedersen, E. 1996. Glacial or non-glacial origin of the Bigganjargga tillite, Finnmark, northern Norway. Geological Magazine, 133, 137– 145. Krzyszkowski, D. 1993. Pleistocene glaciolacustrine sedimentation in a tectonically-active zone, Kleszczow Graben, central Poland. Sedimentology, 40, 623– 644. Laberg, J. S. & Vorren, T. O. 2000. Flow behaviour of the submarine glacigenic debris flows on the Bear Island Trough Mouth Fan, western Barents Sea. Sedimentology, 47, 1105–1117. Lawson, D. E. 1982. Mobilization, movement and deposition of active subaerial sediment flows, Matanuska Glacier, Alaska. Journal of Geology, 90, 279–300. Leather, J., Allen, P. A., Brasier, M. D. & Cozzi, A. 2002. Neoproterozoic snowball Earth under scrutiny: evidence from the Fiq glaciation of Oman. Geology, 30, 891– 894. Le Heron, D. P. & Etienne, J. L. 2005. A complex subglacial clastic dyke swarm, So´lheimajo¨kull, southern Iceland. Sedimentary Geology, 181, 25 –37. Le Heron, D. P., Craig, J. & Etienne, J. L. 2009. Ancient glaciations and hydrocarbon accumulations in North Africa and the Middle East. Earth Science Reviews, 93, 47 –76. Le Heron, D. P., Sutcliffe, O. E., Whittington, R. J. & Craig, J. 2005. The origins of glacially-related soft-sediment deformation structures in Upper Ordovician glaciogenic rocks: implication for ice sheet dynamics. Palaeogeography, Palaeoclimatology, Palaeoecology, 218, 75 – 103. Leonard, E. M. 1986. Varve studies at Hector Lake, Alberta, Canada, and the relationship between glacial activity and sedimentation. Quaternary Research, 25, 199– 214. Lønne, I. 1995. Sedimentary facies and depositional architecture of icecontact glaciomarine systems. Sedimentary Geology, 98, 13 –43. Martins-Neto, M. A. 1996. Lacustrine fan-deltaic sedimentation in a Proterozoic rift basin: the Sopa-Brumadinho Tectonosequence, southeastern Brazil. Sedimentary Geology, 106, 65 –96. McCarroll, D. & Rijsdijk, K. F. 2003. Deformation styles as a key for interpreting glacial depositional environments. Journal of Quaternary Science, 18, 473–489.

49

McDonald, B. C. & Shilts, W. W. 1975. Interpretation of faults in glaciofluvial sediments. In: Jopling, A. V. & McDonald, B. C. (eds) Glaciofluvial and Glaciolacustrine Sedimentation. SEPM Special Publication, 123–131. McKay, R. M., Dunbar, G. B., Naish, T. R., Barrett, P. J., Carter, L. & Harper, M. 2008. Retreat history of the Ross Ice Sheet (Shelf) since the last glacial maximum from deep-basin sediment cores around Ross Island. Palaeogeography, Palaeoclimatology, Palaeoecology, 260, 245– 261. Menzies, J. (ed.) 1995. Modern Glacial Environments: Processes, Dynamics and Sediments. Butterworth-Heinemann, Oxford. Menzies, J. (ed.) 1996. Past Glacial Environments: Sediments, Forms and Techniques. Butterworth-Heinemann, Oxford. Menzies, J. 2000. Microstructures within diamictites of the Lower Gowganda Formation (Huronian) near Elliot Lake, Ontario. Journal of Sedimentary Research, 35, 210– 216. Miller, N. R., Stern, R. J., Avigad, D., Beyth, M. & Schilman, B. 2009. Cryogenian slate-carbonate sequences of the Tambien Group, northern Ethiopia (I): pre-sturtian chemostratigraphy and correlations. Precambrian Research, 170, 129– 156. Mohrig, D., Elverhoi, A. & Parker, G. 1999. Experiments on the relative mobility of muddy subaqueous and subaerial debris flows and their capacity to remobilize antecedent deposits. Marine Geology, 154, 117–129. Moncrieff, A. M. 1989. Classification of poorly sorted sedimentary rocks. Sedimentary Geology, 65, 191– 194. Mulder, T. & Alexander, J. 2001. The physical character of subaqueous sedimentary density flows and their deposits. Sedimentology, 48, 269– 299. Naish, T., Powell, R. et al. 2009. Obliquity-paced Pliocene West Antarctic ice sheet oscillations. Nature, 458, 322–328. Nardin, T. R., Hein, F. J., Gorsline, D. S. & Edwards, B. D. 1979. A review of mass movement processes, sediment and acoustic characteristics and contrasts in slope and base of slope systems versus canyon-fan basin floor systems. In: Doyle, L. J. & Pilkey, O. H. (eds) Geology of Continental Slopes. Society of Economic Paleontologists and Mineralogists, Special Publication, 61– 73. Nemec, W., Lønne, I. & Blikra, L. H. 1999. The Kregnes moraine in Gauldalen, west central Norway: anatomy of a Younger Dryas proglacial delta in a palaeofjord basin. Boreas, 28, 254–476. Nystuen, J. P. 1985. Facies and preservation of glaciogenic sequences from the Varanger ice age in Scandinavia and other parts of the North Atlantic Region. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 209–229. ´ Cofaigh, C. & Dowdeswell, J. A. 2001. Laminated sediments in glaO cimarine environments: diagnostic criteria for their interpretation. Quaternary Science Reviews, 20, 1411– 1436. ´ Cofaigh, C., Pudsey, C. J., Dowdeswell, J. A. & Morris, P. 2002. O Evolution of subglacial bedforms along a paleo-ice stream, Antarctic Peninsula continental shelf. Geophysical Research Letters, 29, 1199–1203. Phillips, E. R., Evans, D. A. J. & Auton, C. A. 2002. Polyphase deformation at an oscillating ice margin following the Loch Lomond Readvance, central Scotland, U.K. Sedimentary Geology, 149, 157–182. Postma, G., Nemec, W. & Kleinspehn, K. L. 1988. Large floating clasts in turbidites: a mechanism for their emplacement. Sedimentary Geology, 58, 47 –61. Powell, R. D. & Cooper, J. M. 2002. A glacial sequence stratigraphic model for temperate glaciated continental shelves. In: Dowdeswell, ´ Cofaigh, C. (eds) Glacier-Influenced Sedimentation on J. A. & O High Latitude Continental Margins. Geological Society, London, Special Publications, 203, 215–244. Powell, R. D. & Elverhoi, A. (eds) 1989. Modern glacimarine environments: Glacial and marine controls of modern lithofacies and biofacies. Marine Geology, 85, 416. Prior, D. B., Bornhold, B. D. & Johns, M. W. 1984. Depositional characteristics of a submarine debris flow. Journal of Geology, 92, 707– 727. Rampino, M. R. 1994. Tillites, diamictites and ballistic ejecta of large impacts. The Journal of Geology, 102, 439–456.

50

E. ARNAUD & J. L. ETIENNE

Ravna˚s, R. & Steel, R. J. 1998. Architecture of marine rift-basin successions. American Association of Petroleum Geologists Bulletin, 82, 110– 146. Reading, H. G. 1978. Sedimentary Environments and Facies, 1st edn. Blackwell Scientific Publications, Oxford. Rieu, R., Allen, P. A., Etienne, J. L., Cozzi, A. & Weichert, U. 2006. A Neoproterozoic glacially influenced basin margin succession and ‘atypical’ cap carbonate associated with bedrock paleovalleys, Mirbat area, southern Oman. Basin Research, 18, 471– 496. Roberts, D. H. & Hart, J. K. 2005. The deforming bed characteristics of a stratified till assemblage in north East Anglia, U.K.: investigating controls on sediment rheology and strain signatures. Quaternary Science Reviews, 24, 123–140. Schermerhorn, L. J. G. 1974. Late Precambrian mixtites: glacial and/or non-glacial. American Journal of Science, 274, 673–824. Siegert, M. J., Carter, S., Tabacco, I., Popov, S. & Blankenship, D. D. 2005. A revised inventory of Antarctic subglacial lakes. Antarctic Science, 17, 453–460. Smith, N. D. & Ashley, G. M. 1985. Proglacial lacustrine environments. In: Ashley, G. M., Shaw, J. & Smith, N. D. (eds) Glacial Sedimentary Environments. SEPM Short Course Notes, 16, 135– 216. Smith, N. D., Phillips, A. C. & Powell, R. D. 1990. Tidal drawdown: A mechanism for producing cyclic sediment laminations in glaciomarine deltas. Geology, 18, 10 –13. Sohn, Y. K. 2000. Depositional process of submarine debris flows in the Miocene fan deltas, Pohang Basin, S.E. Korea, with special reference to flow transformation. Journal of Sedimentary Research, 70, 491– 503. Tanaka, J. & Maejima, W. 1995. Fan-delta sedimentation on the basin margin slope of the Cretaceous, strike-slip Izumi Basin, southwestern Japan. Sedimentary Geology, 98, 205– 213.

Taylor, J., Dowdeswell, J. A., Kenyon, N. H. & O’Cofaigh, C. 2002. Late Quaternary architecture of trough-mouth fans: debris flows and suspended sediments on the Norwegian margin. In: Dowdeswell, ´ Cofaigh, C. (eds) Glacier-Influenced Sedimentation on J. A. & O High Latitude Continental Margins. Geological Society, London, Special Publications, 203, 55– 71. Thomas, G. S. P. & Connell, R. J. 1985. Iceberg drop, dump and grounding structures from Pleistocene glacio-lacustrine sediments, Scotland. Journal of Sedimentary Petrology, 55, 243– 249. Tweed, F. S. & Russell, A. J. 1999. Controls on the formation and sudden drainage of glacier-impounded lakes: implications for jo¨kulhaup characteristics. Progress in Physical Geography, 23, 79 – 110. Visser, J. N. J. 1983. The problems of recognizing ancient subaqueous debris flow deposits in glacial sequences. Transactions of the Geological Society of South Africa, 86, 127– 135. Weaver, P. P. E., Wynn, R. B., Kenyon, N. H. & Evans, J. 2000. Continental margin sedimentation, with special reference to the north-east Atlantic margin. Sedimentology, 47(Suppl. 1), 239– 256. Wingham, D. J., Siegert, M. J., Sheperd, A. & Muir, A. S. 2006. Rapid discharge connects Antarctic subglacial lakes. Nature, 440, 1033– 1036. Woodworth-Lynas, C. M. T. & Guigne, J. Y. 1990. Iceberg scours in the geological record: examples from glacial Lake Agassiz. In: Dowdeswell, J. A. & Scourse, J. D. (eds) Glaciomarine Environments: Processes and Sediments. Geological Society, London, Special Publications, 53, 217– 223. Wright, R. & Anderson, J. B. 1980. The importance of sediment gravity flow to sediment transport and sorting in a glacial marine environment: Eastern Weddell Sea, Antarctica. Geological Society of America Bulletin, 93, 951–963.

Chapter 4 Chemostratigraphy and the Neoproterozoic glaciations GALEN P. HALVERSON1,3* & GRAHAM SHIELDS-ZHOU2 1

School of Earth and Environmental Sciences, The University of Adelaide, North Terrace, Adelaide, SA 5005, Australia 2

Department of Earth Sciences, University College London, Gower Street, London, WD1E 6BT, UK

3

Present address: Department of Earth and Planetary Sciences, McGill University, 3450 University Street, Montreal, QC, H3A 2A7, Canada *Corresponding author (e-mail: [email protected]) Abstract: Although the pre-glacial Proterozoic isotopic record is poorly constrained, it is apparent that the chemical and isotopic composition of the oceans began to change during the early to mid-Neoproterozoic and experienced considerable fluctuations alongside climatic instability during much of the subsequent Cryogenian and Ediacaran periods. The earliest known large negative d13C excursion appears to post-date 811 Ma and fluctuations became progressively more extreme, culminating in the late-Ediacaran ‘Shuram– Wonoka’ anomaly. The negative excursions are commonly associated with pre-glacial and post-glacial times, while extremely high d13C values are characteristic of strata between glaciations. The precise causal mechanism for these excursions is subject to debate. Seawater 87Sr/86Sr rose during the Neoproterozoic, with abrupt increases following deglaciation consistent with enhanced weathering rates. Reported marine sulphate and pyrite d34S data exhibit marked variation through this interval, although the changes are not always consistent within or between sedimentary successions of equivalent age. Iron-speciation studies indicate that much of this variation was caused by fluctuating and low sulphate concentrations in seawater, which at times led to the build-up of ferruginous conditions in the ocean. The application of chemostratigraphy to understanding and correlating the Neoproterozoic glaciations evokes considerable controversy, and many questions persist regarding the reliability and calibration of the d13C, 87Sr/86Sr and d34S record. Nevertheless, the individual glaciations appear to be characterized by distinct combined chemostratigraphic signatures, in large part due to the generally increasing strontium isotope composition of seawater through the Neoproterozoic Era.

Chemical stratigraphy has enjoyed widespread application to the study of the Neoproterozoic sedimentary record, in particular with regard to the number, correlation, causes and consequences of glaciations during this era. Owing to the scarcity of biostratigraphically useful fossils, it has become the tool of choice for global stratigraphic correlation of Neoproterozoic strata (Kaufman & Knoll 1995; Walter et al. 2000). In their broad survey of carbonates, Schidlowski et al. (1975) pioneered the application of carbon-isotope stratigraphy to the Precambrian sedimentary record. Williams (1979) subsequently documented negative d13C signatures associated with post-glacial carbonates in Australia. Knoll et al. (1986) later focused attention on the Neoproterozoic Era with coupled carbonate (d13Ccarb) and total organic carbon (d13Corg) carbon-isotope records from carbonate-dominated successions in Svalbard and East Greenland that showed large and in-phase variations ostensibly linked to glacial episodes. This study complemented contemporaneous investigations that similarly showed large fluctuations in the d13C composition of ancient marine carbonates spanning the Precambrian –Cambrian boundary (Hsu et al. 1985; Tucker 1986; Magaritz et al. 1986). These studies bolstered the argument that ancient carbonates and sedimentary organic matter could be used as proxies for the carbon-isotopic composition of the seawater in which they formed. Inasmuch, they opened the door to carbon-isotope chemostratigraphy as a commonplace tool for basin- and global-scale correlation and as a measure of the behaviour of the exogenic carbon cycle, in particular as it related to extreme climate variability and biospheric change (e.g. Derry et al. 1992; Kaufman & Knoll 1995). Notwithstanding challenges to the integrity of carbon-isotope signatures as proxies for seawater composition (e.g. Frimmel 2009; Knauth & Kennedy 2009; Derry 2010) and poor radiometric age control on many Neoproterozoic successions, carbon-isotope stratigraphy is a powerful tool for establishing a chronological framework for this interval, during which extensive carbonates were deposited but for which biostratigraphy is of limited (but

increasing) use (Knoll & Walter 1992; Knoll 2000). Multiple negative carbon-isotope excursions and intervening intervals of sustained high d13Ccarb appear to correlate globally (e.g. Kaufman et al. 1997; Halverson et al. 2005). However, because there are also multiple glaciations, each apparently associated with negative carbon-isotope anomalies, these excursions are non-unique and robust correlations, in the absence of firm radiometric ages, require additional data. Like carbon isotopes, sulphur-isotope data on both pyrite and sulphate show impressive variability spanning the Neoproterozoic glaciations (Gorjan et al. 2000; Hurtgen et al. 2002). However, the strontium-isotope record is the most useful in discriminating between the various negative carbon-isotope anomalies because 87 Sr/86Sr increases throughout most of the Neoproterozoic (Shields 1999; Melezhik et al. 2001; Halverson et al. 2007) such that each of the prominent carbon-isotope anomalies appears to be associated with distinct strontium-isotope signatures. Many other chemostratigraphic methods have been applied to the Neoproterozoic sedimentary record. Iron speciation data have revealed remarkable fluctuations in the redox state of seawater throughout the Neoproterozoic, in particular spanning glaciations and leading up to the first appearance of the Ediacaran fauna (Canfield et al. 2007, 2008). Post-glacial carbonates have also proven to be exceptional archives of the highly unusual environmental conditions during recovery from the Cryogenian glaciations, preserving in some cases extraordinary geochemical anomalies in non-traditional geochemical proxies, such as calcium and boron isotopes (Kasemann et al. 2005) and D17O (Bao et al. 2008). For a more extensive review of these and other chemostratigraphic methods applied to the Neoproterozoic sedimentary record, the reader is referred to Halverson et al. (2010). A detailed account of the application of the chemical index of alteration (CIA) to the Neoproterozoic glacial record is presented in this volume by Bahlburg & Dobrzinski (2011), and chemostratigraphic and geochronological data from individual sedimentary successions are summarized in reviews throughout

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 51– 66. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.4

52

G. P. HALVERSON & G. SHIELDS-ZHOU

this volume. The aim of this chapter is to provide a brief review of the carbon-, sulphur- and strontium-isotope systems and the ironspeciation proxy and a synthesis of published data bracketing the Neoproterozoic glaciations, with a focus on the structure of these records and their temporal relation to glaciation.

plausibility of sufficient oxidants being available to generate the accompanying highly negative d13Ccarb anomalies (Bristow & Kennedy 2008), the undeniable lack of correlation in certain data sets requires a mechanistic explanation.

Sulphur isotopes Carbon isotopes Under normal conditions and at the scale of resolution appropriate in Precambrian chemostratigraphy, the carbon-isotope compositions of unaltered marine carbonates (d13Ccarb) precipitated in equilibrium with seawater closely approximate the composition of the total dissolved inorganic carbon (DIC) pool (Hayes et al. 1999). If it is assumed that the carbonates retain their primary signatures after burial, then carbon-isotope chemostratigraphy has diverse applications, such as establishing global correlations (e.g. Kaufman et al. 1993, 1997), regional studies of basin architecture and carbonate sequence stratigraphy (Halverson et al. 2002; Cozzi et al. 2004; Jiang et al. 2007), modelling of the global carbon cycle (Derry et al. 1992), and as a constraint on mechanisms for the causes of Neoproterozoic glaciation (Kaufman et al. 1997; Halverson et al. 2002; Schrag et al. 2002) and post-glacial carbonate deposition (Kaufman et al. 1997; Hoffman et al. 1998; Kennedy et al. 2001; Higgins & Schrag 2003; Shields 2005). Debate continues as to whether Neoproterozoic marine carbonates faithfully preserve their primary depositional signatures, particularly where they are highly 13C-depleted (Bristow & Kennedy 2008; Frimmel 2009; Knauth & Kennedy 2009; Derry 2010). Nevertheless, the general consensus is that, by and large, ancient marine carbonates do preserve seawater compositions, so secular trends in d13Ccarb can be interpreted in terms of inputs of carbon to and outputs from the global ocean (Holser 1997). The record of the carbon-isotope composition of sedimentary organic matter (d13Corg), while much less applied in the Neoproterozoic, is often used where carbonates are not sufficiently abundant to construct a d13Ccarb profile. The analysis of d13Corg has recently surged due to the hypothesis that some of the extreme negative d13Ccarb anomalies might be the result of partial oxidation of an enormous reactive organic carbon reservoir in the deep ocean (Rothman et al. 2003). Organic carbon data are intrinsically more difficult to acquire than inorganic carbon data, and the connection between extant kerogen (or total organic carbon, TOC) and original biomass is inevitably blurred by diagenesis, thermal alteration of the organic matter (Hayes et al. 1983; Kaufman et al. 1991; Hayes et al. 1999) and input of detrital kerogen. Nevertheless, early stratigraphically constrained Neoproterozoic d13Corg data sets (e.g. Knoll et al. 1986; Kaufman et al. 1997) showed broad correlation with contemporaneous d13Ccarb, suggesting that it could be used in parallel with or in place of (where primary carbonates are unavailable) inorganic carbon data as a proxy for seawater compositions, albeit with the caveat that the net fractionation between the original DIC reservoir and extracted kerogen is inherently variable. A compilation of organic carbon data by Hayes et al. (1999) suggested an average difference between contemporaneous carbonates and organic matter (1TOC) of c. 30‰ in the Neoproterozoic, but with large fluctuations (+10‰) associated with d13Ccarb anomalies, in particular those spanning glaciations. More recent and detailed data sets from Ediacaran-aged sediments have revealed sustained intervals where d13Ccarb and d13Corg are decoupled or even anti-correlated (Calver 2000; Fike et al. 2006; McFadden et al. 2008). Swanson-Hysell et al. (2010) recently argued, based on new, paired data from Australia, that decoupling of these two proxies began sometime in the Cryogenian. Decoupling between these two proxies is a prediction of the Rothman et al. (2003) hypothesis, because the implied oceanic organic-carbon reservoir would be much larger than the DIC reservoir. While questions remain as to the

Sedimentary sulphur-isotope ratios are measured on sedimentary sulphates or sulphides. Sulphur-isotope data on sulphates can be recovered, with varying degrees of reliability, from evaporites, barite, phosphorites and carbonates (as carbonate-associated sulphate, CAS). Sulphate-bearing marine minerals commonly do not form in isotopic equilibrium with seawater, so additional arguments need to be marshalled to support the presumption that they record seawater sulphate isotopic compositions. The d34S values of marine evaporite minerals may be higher or lower than contemporaneous seawater depending on the extent of basin restriction and connection to meteoric influence. Stratiform barite d34S values are also prone to deviate from seawater, although lowermost values are generally close approximations to ocean composition (Shields 2007). Phosphorite can also show higher d34S values than seawater due to Rayleigh fractionation during bacterial sulphate reduction (e.g. Hough et al. 2006), and the same may also be true of some diagenetic carbonate rocks (e.g. Marenco et al. 2008a). An additional problem with carbonate- (and phosphate-) associated sulphate sulphur-isotopic analyses derives from the inadvertent incorporation of contaminant sulphate during the dissolution process before analysis (Marenco et al. 2008b). Unlike marine sulphate minerals, sulphides (typically pyrite) record the net fractionation imparted during bacterial sulphate reduction (BSR), plus additional fractionation effects contributed by disproportionation reactions during oxidative recycling of sulphide (Canfield & Teske 1996; Detmers et al. 2001). Sulphate and pyrite d34S trends broadly mirror one another, but pyrite is generally much more variable, attributable to the fact that reduction, oxidation and disproportionation reactions commonly occur during early diagenesis, where distinct local effects, such as pore water sulphate concentrations, availability of appropriate substrates for BSR, and physiology of the BSR population (Detmers et al. 2001) strongly influence the final sulphur-isotope signature. Sulphur isotopes have been widely applied to problems in Precambrian Earth history. For example, a gradual increase in the spread of d34Spyrite in shales (d34Ssulphate – d34Spyrite isotopic discrimination) broadly coincides with separate geological and geochemical evidence for a Neoproterozoic rise in the oxygenation of the Earth’s environment (Canfield & Teske 1996; Fike et al. 2006). Originally, this isotopic shift was linked to the evolutionary radiation of sulphide-oxidizing bacteria (Canfield & Teske 1996), but recent evidence puts this evolutionary event much earlier (Johnston et al. 2005). Instead, this isotopic shift could have been caused by a rise in the importance of sulphide oxidation and disproportionation reactions during early diagenesis, which may be indirectly linked to oxygenation through seafloor redox changes. The d34S pyrite compilation is also the basis for the model that the global deep ocean became dominantly euxinic, rather than oxygenated, as massive banded iron-formation deposition came to a close in the late Palaeoproterozoic (Poulton et al. 2004). The Neoproterozoic is characterized by extraordinary fluctuations in both d34Spyrite and d34Ssulphate in the Neoproterozoic, which are closely coupled to glacial events (Halverson & Hurtgen 2007). However, the extent to which these variations truly reflect global seawater sulphate compositions and environmental conditions is poorly established. Highly precise measurement of all four sulphur isotopes (32S, 33S, 34S, 36S) has shown resolvable and consistent variations in the parameters D33S and D36S, which can reflect both massindependent (MIF) and mass-dependent fractionation (MDF) processes. MIF of sulphur isotopes has been applied to reconstructing

CHEMOSTRATIGRAPHY AND THE NEOPROTEROZOIC GLACIATIONS

the evolution of atmospheric oxygen levels (Farquhar et al. 2000), but has not been systematically applied to the Neoproterozoic, when atmospheric O2 concentrations likely exceeded the low level (,1025 present levels) required to suppress MIF fractionation in the atmosphere (Pavlov & Kasting 2002). On the other hand, variations in D33S and D36S resulting from MDF record information about specific sulphur transformations and pathways that cannot be gleaned from traditional measurements of 2 S/4S alone (Johnston et al. 2005; Ono et al. 2006). Although not reviewed here, the technique of measuring quadruple sulphur isotopes promises to yield significant insight into behaviour of the sulphur cycle leading up to and following the Neoproterozoic glaciations.

Iron speciation The impressive range of d34Spyrite and d34Ssulphate values that are typical of the Neoproterozoic are commonly interpreted in terms of a low sulphate ocean (Halverson & Hurtgen 2007). This interpretation is consistent with the observation that seawater during the Neoproterozoic was occasionally rich in dissolved ferrous iron, based on the occurrence of iron formation. Ironspeciation studies shed further light on the redox state of the oceans and are being carried out increasingly on Neoproterozoic successions (Canfield et al. 2008). Redox conditions in the water column affect the speciation of iron in marine sediments. The highly reactive iron pool (FeHR) is that which is geochemically available during diagenesis and includes iron bound in carbonates, iron oxides and oxyhydroxides, and iron sulphides (Canfield 1989). Sediments deposited under an oxic water column tend to contain less reactive iron as a fraction of total sedimentary iron (FeHR/FeT) than those deposited under an anoxic water column, with a proposed cut-off at 0.38 (Raiswell et al. 1988; Lyons & Severmann 2006). Canfield et al. (2008) further proposed that the measure of pyrite-bound iron v. total reactive iron (FePy/ FeHR) can be used to distinguish iron-rich anoxic v. euxinic water columns, with sediments exhibiting FePy/FeHR . 0.8 being classified as euxinic. Neoproterozoic sediments deposited beneath a storm wave base typically exhibit high FeHR/FeT ratios (Fig. 4.1), indicating that anoxia was common in the deeper marine environment. However, sulphidic conditions are so far only indicated during the pre-glacial Neoproterozoic and early Cambrian (Canfield et al. 2008), suggesting ferruginous rather than sulphidic (euxinic) anoxia during the interval of climatic fluctuations. Because ferruginous conditions can only arise when the molar flux of FeHR to the deep ocean is greater than half the flux of sulphide, their reappearance during the Neoproterozoic after an absence of more than a billion years is consistent with a lowsulphate ocean reservoir.

Strontium isotopes Over timescales of .106 years, the 87Sr/86Sr ratio of seawater records the relative strontium fluxes from continental weathering and hydrothermal input, superimposed on the long-term increase in 87 Sr/86Sr as a result of radioactive decay of 87Rb (Edmond 1992). As such, marine 87Sr/86Sr is a useful measure of evolving tectonic and long-term climatic regimes. Nevertheless, the ensemble of driving mechanisms for the variations in seawater 87Sr/86Sr spanning the Neoproterozoic remains controversial. It is doubtful that seawater 87Sr/86Sr can be simplified to a simple balance between the hydrothermal and continental inputs, but rather must also be influenced by continental configuration and CO2 outgassing rates (Halverson et al. 2007; Shields 2007). Strontium-isotope stratigraphy is based on the presumption that seawater was always isotopically homogeneous with respect to Sr

53

because of its long ocean residence time. Although isotopic analysis is now standard procedure, samples need to be carefully screened because of the risk of contaminating the sample with diagenetically altered or detrital materials. This risk is reduced for Phanerozoic studies for which microscopically well-preserved low-Mg calcite tests (e.g. foraminifera, brachiopods and belemnites) can be analysed. However, for much of the Precambrian, it is necessary to identify least altered components of the bulk carbonate rock (e.g. Derry et al. 1992; Kaufman et al. 1993). Strontium isotope analysis comprises the following six steps, of which the first four vary considerably between research groups: (i) sample selection using petrographic and geochemical criteria; (ii) physical extraction from the host sample; (iii) chemical pretreatment to remove contaminant Sr ions; (iv) sample dissolution; (v) chemical separation of the element Sr; (vi) mass spectrometric analysis. Studies show that least altered ratios can most easily be obtained by targeted microsampling of fine-grained calcite (micrite), which is free of later recrystallization and dolomitization, although early diagenetic cements may also be reliable tracers of ocean composition (e.g. Kaufman et al. 1993; Fairchild et al. 2000). Chemical pre-leaching helps remove contaminant Rb and radiogenic Sr (Gorokhov et al. 1995), while subsequent incomplete dissolution in weak acids such as acetic acid has been shown to limit inadvertent dissolution of clay mineral-bound Sr (Bailey et al. 2000). The Neoproterozoic strontium-isotope record is typically reconstructed from calcite that has relatively high strontium and low rubidium concentrations or otherwise exhibits evidence of limited exchange with diagenetic fluids, as measured, for example, in Mn/Sr ratios (Brand & Veizer 1981; Jacobsen & Kaufman 1999). Although carbonate originally precipitated as aragonite is intrinsically more suitable due to its high original Sr content (Kulp et al. 1952), the abundance of metastable carbonate precipitates in the Precambrian indicates that some apparently pristine components could be the result of diagenetic replacement (Fairchild et al. 2000). Because alteration usually increases the 87 Sr/86Sr of carbonates (Banner & Hanson 1990), the lowest measured values within a suite of rocks are typically inferred to most closely approximate the composition of the seawater from which it precipitated. However, this ought not merely to be assumed, as post-depositional fluids are not always more radiogenic than coeval seawater. With this in mind, least-altered ratios are best extrapolated from diagenetic trends, for example by plotting Mn/Sr, Mg/Ca, 1/Sr or d18O against measured 87Sr/86Sr. Recent compilations of the 87Sr/86Sr evolution of Neoproterozoic oceans share the common, dominant feature of ratios rising from as low as c. 0.7055 in the early Neoproterozoic to as high as 0.7085 in the Ediacaran Period (Jacobsen & Kaufman 1999; Walter et al. 2000; Melezhik et al. 2001; Halverson et al. 2007a; Shields 2007). This feature alone makes strontium-isotope chemostratigraphy a key tool in establishing correlations and relative age in otherwise poorly radiometrically dated Neoproterozoic successions. Additional detail in the strontium-isotope record remains controversial due to the relatively poor resolution of the Neoproterozoic record and variable quality of samples used in compilations (Melezhik et al. 2001). However, an apparently reproducible finer-scale structure in the record is emerging with the promise of improving the Neoproterozoic chronology and the link between coupled tectonic– climatic processes. Importantly, and as will be discussed further below, the seawater strontiumisotope record appears to be distinct, spanning each of the Neoproterozoic glacial epochs (Table 4.1).

The pre-glacial Neoproterozoic The most characteristic features of the Neoproterozoic chemostratigraphic record are the generally high d13Ccarb values that prevail throughout most of the era and the episodic negative anomalies

54

G. P. HALVERSON & G. SHIELDS-ZHOU

Fig. 4.1. Working chemostratigraphic compilations of d13Ccarb, 87Sr/86Sr, d34Ssulphate, d34Spyrite and iron-speciation (FeHR/FeT) data for the Neoproterozoic (modified after Halverson et al. 2005, 2007a, 2010; Halverson 2006; Canfield et al. 2008). This compilation does not include all available high-quality data due to the difficulty of integrating data that have poor and variable age constraints (see text for discussion). The names shown vertically in the top of the figure are the informal names of the principal negative anomalies, as well as the distinct positive shift in d13Ccarb recorded in the lower Little Dal Group (Halverson 2006) and the ‘Keele Peak’ positive anomaly (Kaufman et al. 2007). Note that there are conflicting sulphur-isotope data, in particular for the late Ediacaran Period (e.g. Ries et al. 2009). A low of 87Sr/86Sr in the earliest Neoproterozoic (dashed oval) is based on data from Gorokhov et al. (1995). Iron-speciation data (FeHR/FeT) are from Canfield et al. (2008), replotted as five-point running averages, and from Nagy et al. (2009). Open boxes and diamonds are additional d34Ssulphate and iron-speciation data from the Chuar Group as published in Johnston et al. (2010). Principle U–Pb geochronological constraints on the d13Ccarb record are shown in triangles at the top of the compilation: (1) an 811.5 + 0.3 Ma tuff in the Fifteenmile Group, Yukon Territory (Macdonald et al. 2010b); (2) a 760 + 1 Ma tuff in the Ombombo Subgroup, NW Namibia (Halverson et al. 2005); (3) a 716.5 + 0.2 Ma tuff in the lower Rapitan Group, Yukon Territory (Macdonald et al. 2010b); (4) an ash bed in the Ghaub Formation in northern Namibia dated at 635 + 0.6 Ma (Hofmann et al. 2005) and an ash bed in the lower Duoshantuo Formation in South China dated at 635.2 + 0.6 Ma (Condon et al. 2005) bracket the age of the Cryogenian –Ediacaran boundary; (5) a 542.0 + 0.6 Ma ash bed at the Precambrian –Cambrian boundary in Oman (Amthor et al. 2003); and several slightly older ages (overlapping triangles) from the latest Ediacaran of the Nama Group, S Namibia (Grotzinger et al. 1995). Shaded boxes and lines show the inferred range of glacial events, with the gradational shading in the middle Cryogenian box reflecting the wide range of ages for middle Cryogenian glaciation (e.g. Allen & Etienne 2008).

CHEMOSTRATIGRAPHY AND THE NEOPROTEROZOIC GLACIATIONS

that punctuate it (Knoll et al. 1986; Kaufman & Knoll 1995; Shields & Veizer 2002). These basic observations raise several questions. First, when did the average values near 0‰ typical of the Mesoproterozoic (Brasier & Lindsay 1995) give way to the average values of þ5‰ (Halverson et al. 2005) characteristic of the Neoproterozoic? Second, did the increased variability in d13Ccarb accompany the higher average values and, if not, when did this begin? Finally, what drove these fundamental shifts in the behaviour of the exogenic carbon cycle? These questions are difficult to answer because of the poor carbon-isotope coverage and radiometric ages spanning the Mesoproterozoic– Neoproterozoic transition. Bartley et al. (2001) and Frank et al. (2003) argued that the increase in the amplitude and frequency of fluctuations in d13Ccarb began sometime in the middle to late Mesoproterozoic, based on data from Siberia and NE Canada, respectively. Halverson (2006) argued that sustained d13Ccarb  5‰ did not start until the early Neoproterozoic based on data from the Platform Assemblage of the Little Dal Group in NW Canada, where the lowermost data show a distinct ramp from 0 to 5‰ (Fig. 4.1), followed by generally high (5‰) values up to the onset of the Bitter Springs anomaly (see below). However, the age and correlation of the lowermost Little Dal Group and its equivalence to other rocks within Succession B in NW Canada is not known and only constrained to be between c. 1004 (Rainbird et al. 1996) and 811 Ma (Macdonald et al. 2010b). It is therefore unclear whether the lower Little Dal positive d13C shift (Fig. 4.1) truly does record the first highly 13 C-enriched seawater in the Neoproterozoic, or what the precise age of this event might be. Much better constrained now is the timing of presumably the first major negative d13Ccarb excursion in the Neoproterozoic: the Bitter Springs anomaly (Table 4.1; Halverson et al. 2005). Although older negative d13Ccarb excursions have been reported (Bartley et al. 2001; Shields 2007), these are generally from poorly dated successions and may be the result of alteration. For example, 13C-depleted I6 unit rocks of the Atar Group, Mauritania (Shields 2007), although now known to be c. 1.1 Ga in age (Rooney et al. 2010), are likely to have undergone significant isotopic alteration (Fairchild et al. 1990). The Bitter Springs negative anomaly was first fully documented in the Bitter Springs Formation in the Amadeus basin of central Australia (Hill et al. 2000). It was subsequently recorded in more detail in Svalbard (Halverson et al. 2005), where it is clearly defined by reciprocal d13C shifts of .8‰ that perfectly coincide with rare exposure surfaces in the Akademikerbreen Group. The anomaly also appears to occur in the upper Little Dal Group (Upper Carbonate Formation) in the Mackenzie Mountains (Halverson 2006). These global correlations are reinforced by additional geochemical data (Halverson et al. 2007a), namely 87Sr/86Sr (see below) and an updated and high-resolution d13Ccarb record through the Bitter Springs Formation (Swanson-Hysell et al. 2010), as well as the coincidence of the shifts with prominent sequence boundaries. Both features have been interpreted to relate to large true polar wander episodes (Maloof et al. 2006). The Bitter Springs anomaly (Table 4.1; Fig. 4.1) has now been reproduced across northwestern Canada: within the middle Lower Tindir Group (Tatunduk Inlier), upper Fifteenmile Group (Coal Creek Inlier), and the Wynniatt Formation of the Shaler Supergroup (Victoria Island; Macdonald et al. 2010b; Jones et al. 2010), although in this latter case only the negative shift is preserved and this is recorded in sediments intruded by mafic sills, which means it may be an overprint related to contact metamorphism rather than a primary signal. A likely correlative has also been documented in the Tambien Group of northern Ethiopia (Alene et al. 2006; Miller et al. 2009). Macdonald et al. (2010b) produced a precise U – Pb zircon age of 811.51 + 0.25 Ma on a tuff c. 50 m beneath the negative shift presumed to define the base of the anomaly in upper Fifteenmile Group, thus constraining the onset of the Bitter Springs anomaly to c. 810 Ma and

55

providing a more robust basis for global correlations. At the upper boundary of the Bitter Springs anomaly, d13Ccarb shifts back to values exceeding 5‰. The d13Ccarb of seawater at the time of the putative and possibly regional Kaigas (Bayisi) glaciation at c. 750– 740 Ma (Frimmel et al. 1996; Xu et al. 2009) is not well known. Negative d13C values occur in the lower Wallekraal Formation, which overlies the Kaigas Formation (Fo¨lling & Frimmel 2002) but no large d13C anomaly occurs between the Bitter Springs anomaly and the Islay anomaly in Svalbard (Fig. 4.1). Instead, d13Ccarb values remain generally high until the onset of the Islay anomaly (Prave et al. 2009), which appears to precede the first major Cryogenian glaciation (Fig. 4.1) and is discussed in more detail in the following section. The strontium-isotope record through the early Neoproterozoic shown in Figure 4.1 is largely derived from data from Svalbard and northwestern Canada, even though other presumably primary values have been obtained from other successions such as the Bitter Springs Formation. As noted by Bartley et al. (2001), seawater 87Sr/86Sr values reach a nadir around the Mesoproterozoic –Neoproterozoic transition, presumably related to the final assembly of Rodinia. The least radiogenic latest Mesoproterozoic 87 Sr/86Sr values reported by Bartley et al. (2001) hover around 0.7055, which is virtually the same as a value of 0.7056 from limestones of the Atar Group, Taoudeni Basin (Veizer et al. 1983), now dated at 1.1 Ga (Rooney et al. 2010) and the lowest values in the lower Little Dal Group (Halverson et al. 2007a). However, Gorokhov et al. (1995) reported values from pristine limestones of the presumed early Neoproterozoic Burovaya Formation (Turukhansk uplift, NW Siberia) as low as 0.7052. Thus, 0.7055 is a plausible approximation for seawater 87Sr/86Sr at the Mesoproterozoic –Neoproterozoic boundary, but probably does not represent the nadir before the protracted Neoproterozoic rise in 87Sr/86Sr. Strontium-isotope ratios had risen to 0.7063 by the onset of the Bitter Springs anomaly (Fig. 4.1). Although a more discreet and smaller shift to slightly higher ratios accompanies the negative d13C shift at the base of the anomaly, 87Sr/86Sr returns to c. 0.7063 at the end of the Bitter Springs anomaly, before apparently increasing to as high as 0.7070 in the Backlundtoppen Formation in Svalbard (prior to the Islay anomaly). No intervening data from this transition have been documented. As will be discussed below, this middle Neoproterozoic 87Sr/86Sr peak is followed by a temporary return to values as low as 0.7063. The sulphur-isotope record for the early Neoproterozoic is relatively sparse, despite the fact that voluminous evaporites were deposited in Australia and NW Canada c. 830 to 800 Ma (Evans 2006). Indeed, the compilation in Figure 4.1 does not include any data older than c. 820 Ma, although this lacuna is in part an artefact of poor control on ages and correlations, which limit the data that can be used. d34Ssulphate from evaporites in the Gillen Member of the Bitter Springs Formation (i.e. before the Bitter Springs anomaly) ranges from 15 to 21‰ (Gorjan et al. 2000; Hill et al. 2000). From the available data, it appears that early Neoproterozoic seawater d34Ssulphate values do not depart greatly from the Gillen Member values (Fig. 4.1), averaging c. 20‰ through the middle Neoproterozoic, close to average values for the Cretaceous and Cenozoic (Paytan et al. 1998, 2004; Kampschulte & Strauss 2004), but distinct from the highly 34S-enriched values that characterize the early and late Ediacaran and early Palaeozoic (Shields et al. 2004; Kampschulte & Strauss 2004; Halverson & Hurtgen 2007). Although most d34Ssulphate data of the composite curve for this time interval are derived from CAS measurements on Svalbard samples (Halverson et al. 2010), available data from evaporites spanning the Bitter Springs anomaly in the Amadeus Basin (Gorjan et al. 2000; Hill et al. 2000) yield similar values. Pyrite data in this record are also sparse and similarly derived mainly from Svalbard (Halverson et al. 2010) and the Bitter Springs

56

G. P. HALVERSON & G. SHIELDS-ZHOU

Table 4.1. Summary of the d13Ccarb and 87Sr/86Sr signatures associated with the major Neoproterozoic negative carbon isotope anomalies (Fig. 4.1) and an abridged compilation of locations where they are well preserved (for the Maieberg and Rasthof anomalies, only key locations where the anomalies are preserved in great detail are given). PB, Paraguai Belt; UTG, Upper Tindir Group Shuram-Wonoka anomaly d13Ccarb Sr/86Sr

87

Minimum: , –10‰; protracted recovery 0.7081– 0.7084

Region

Unit

Reference

S Namibia Oman S China S Australia N Norway SW USA SE Siberia

Nama Gp Shuram/Kufai Fm Doushantuo Fm Wonoka Fm Nyborg Fm Johnnie Fm Nikol’skaya/Chenchinskaya

Maieberg anomaly d13Ccarb 87 Sr/86Sr

Minimum: –4 to – 6‰, gradual to abrupt return to near 0‰ values 0.7072 at base, increases upsection to 0.7078– 0.7080

Region

Unit

NW Namibia S Namibia Oman NW China S China SW Mongolia E Svalbard Scotland NE Alaska SW USA Brazil (PB) SE Siberia

Maieberg Fm Bloeddrif Mb Hadash/Masirah Bay Tureeken Fm Doushantuo Fm Oi Mb Dracoisen Fm Cranford Ls Katakturuk Dolomite (K2) Noonday Dolomite Araras Fm Barakunskaya Fm

Trezona anomaly d13Ccarb 87 Sr/86Sr

Minimum: –6 to – 8‰, increases before glaciation 0.7072– 0.7074

Region

Workman et al. (2002) Burns & Matter (1993), Le Guerroue´ et al. (2006) Jiang et al. (2007), McFadden et al. (2008) Pell et al. (1993), Calver (2000) Halverson et al. (2005) Corsetti & Kaufman (2003) Pokrovskii et al. (2006), Melezhik et al. (2009)

Reference Hoffman et al. (1998) Fo¨lling & Frimmel (2002) Le Guerroue´ et al. (2006) Xiao et al. (2004) Jiang et al. (2007), McFadden et al. (2008) and others Macdonald et al. (2009a) Kaufman et al. (1997), Halverson et al. (2004) McCay et al. (2006) Macdonald et al. (2009b) Prave (1999) Nogueira et al. (2007) Pokrovskii et al. (2006), Melezhik et al. (2009)

Unit

Reference

NW Namibia South Australia N Norway Scotland NE Alaska Mackenzie Mts

Ombaatjie Fm Trezona Fm Grasdal Fm Craignish/Ardrishaig Katakturuk Dolomite (K1) Keele Fm

Rasthof anomaly d13Ccarb 87 Sr/86Sr

Minimum: –4 to 0‰, rises rapidly 0.7067– 0.7069 at base, increases up section to 0.7072– 0.7073

Region

Halverson et al. (2002) McKirdy et al. (2001) Halverson et al. (2005) Prave et al. (2009) Macdonald et al. (2009b) Halverson et al. (2005)

Unit

Reference

NW Namibia SW Mongolia S Australia Scotland NE Alaska Mackenzie Mts SW USA

Rashtof Fm Base Tayshir Mb Tapley Hill Fm Bonahaven Fm* Katakturuk Dolomite (K1) Twitya Fm. base Beck Spring Fm

Islay anomaly d13Ccarb 87 Sr/86Sr

Minimum: –5 to – 9‰, return to .0‰ before glaciation 0.7065– 0.7067, declines upsection to 0.7064

Region Tasmania E Svalbard

Hoffman et al. (1998), Yoshioka et al. (2003) Shields et al. (1997), Macdonald et al. (2009a) McKirdy et al. (2001) Brasier & Shields (2000), McCay et al. (2006), Prave et al. (2009) Macdonald et al. (2009b) Hoffman & Schrag (2002) Corsetti & Kaufman (2003)

Unit Black River Fm Elbobreen Fm

Reference Calver (1998) Halverson et al. (2004) (Continued)

CHEMOSTRATIGRAPHY AND THE NEOPROTEROZOIC GLACIATIONS

57

Table 4.1. Continued Scotland

Islay/Lossit Fm

E Greenland Victoria Island Mackenzie Mts Alaska-Yukon SW USA

Bed Group 19 Killian Formation Copper Cap Fm Upper Dolomite (LTG) Upper Beck Spring Fm

Bitter Springs anomaly d13Ccarb 87 Sr/86Sr

Minimum: –5 to 0‰, bracketed by sharp positive and negative shift 0.7063–0.7065

Region N Ethiopia Central Australia Western Australia E Svalbard Scotland Mackenzie Mts Yukon Alaska-Yukon

Brasier & Shields (2000), Prave et al. (2009), McCay et al. (2006), Sawaki et al. (2010b) Fairchild et al. (2000) Macdonald et al. (2010b) Halverson (2006) Macdonald et al. (2010a) Corsetti & Kaufman (2003)

Unit Tambien Group Bitter Springs Formation Hussar Formation Grusdievbreen/Svanbergfjellet Ballachulish Sbgp Upper Carbonate Fm Upper Fifteenmile Gp (PF1) Upper Shale (LTG)

Reference Alene et al. (2006) Hill et al. (2000) Hill (2005) Halverson et al. (2005) Prave et al. (2009) Halverson (2006) Macdonald et al. (2010b) Macdonald et al. (2010a)

*Prave et al. (2009) argued that the post-Port Askaig negative d13C anomaly is missing, in which case the Bonahaven anomaly, which is stratigraphically above the glacigenic Port Askaig Formation, might be equivalent to the Tayshir anomaly in Mongolia.

Formation (Hill et al. 2000). The most salient pattern in the d34S data is a coupled increase in d34Ssulphate and decrease in d34Spyrite coincident with the Bitter Springs anomaly (Fig. 4.1), resulting in a transient increase in D34S. Other constraints on the d34Ssulphate of early Neoproterozoic seawater come from studies on successions from Arctic Canada. d34Ssulphate values there range from 14 to 18‰ in the pre-Bitter Springs anomaly (.810 Ma) Gypsum Formation of the Little Dal Group (Turner 2009), while similar values (16 – 19‰) have been reported from the post-Bitter Springs anomaly (,780 Ma) Redstone River Formation of the overlying Coates Lake Group (Strauss 1993). The sulphur-isotope record throughout the remainder of the early –middle Neoproterozoic is thin, but suggests a return to reduced D34S resulting from higher average d34Spyrite sometime prior to the middle Cryogenian glaciation.

Spanning the Cryogenian glaciations Significant discussion remains as to the number, timing and duration of the Cryogenian glaciations, and the base of the Cryogenian period has yet to be formally defined. Two particular persistent sources of controversy and confusion are the possible occurrence of an early (c. 745–750 Ma; Borg et al. 2003) ‘Kaigas’ glaciation (Frimmel et al. 1996) that may not have been global in extent and the seemingly contradictory age constraints on the so-called ‘Sturtian’ or early Cryogenian glaciation. There is no incontrovertible evidence that the Kaigas diamictite in the Gariep belt of southern Namibia (see Macdonald et al. (2011) for a stratigraphic and sedimentological reappraisal of the Kaigas diamictites) and other older diamictites in the Katanga Supergroup in Zambia (Key et al. 2001) and NW China (Xu et al. 2009) are glacial in origin (Hoffman & Li 2009). Chemostratigraphic data spanning these isolated diamictites are also sparse. Therefore, although the putative Kaigas glaciation is plotted in Figure 4.1, we do not review the chemostratigraphic framework for this possible middleNeoproterozoic glaciation. As for the sticky question of the age of early Cryogenian glaciation (i.e. glaciation between the older Kaigas event and the end-Cryogenian ‘Marinoan’ event), reviewed recently by Hoffman & Li (2009), we acknowledge that it is

impossible to discriminate at this point between the model that there were multiple middle Cryogenian glaciations or a single, long-lived glacial epoch that spanned from c. 716 (Macdonald et al. 2010b) to 660 Ma (Fanning & Link 2008). These deficiencies in the Cryogenian chronology are exacerbated by presumably larges hiatuses in the records on continental shelves (from whence most of the chemostratigraphic records are derived) resulting from glacioeustatic fall in sea level and glacial erosion. These complications in the Cryogenian record make it impossible to produce an accurate chemostratigraphic record for the Cryogenian period and the compilation in Figure 4.1 can only be regarded as a work in progress. Nevertheless, it is notable that independent compilations emerging from widely separate sedimentary basins do seem to be converging on a similar overall chemostratigraphic structure for the interval spanning the Cryogenian glaciations (e.g. Halverson et al. 2005; Prave et al. 2009; Macdonald et al. 2010b). The negative d13Ccarb anomalies that coincide with deglaciation and are preserved in transgressive –regressive sequences overlying glacial deposits and their equivalent glacial unconformities (i.e. cap carbonate sequences; Hoffman & Schrag 2002) are well documented (Table 4.1; Williams 1979; Kaufman et al. 1997; Kennedy et al. 1998; Halverson et al. 2005; Hoffman et al. 2007). Therefore, only the most important features are summarized here. Importantly, no more than two distinct, post-glacial negative d13Ccarb anomalies linked to Cryogenian glaciation are known to occur in a single succession (Kennedy et al. 1998). Where there are two such post-glacial anomalies within one succession or, where there is just one of known age, consistent distinctions can be drawn between the two anomalies (Table 4.1; Kennedy et al. 1998; Hoffman & Schrag 2002). As a general rule, the older of the two post-glacial sequences lacks the basal transgressive portion, meaning the maximum flooding surface is the base of the sequence (Hoffman & Schrag 2002; Hoffman et al. 2011). d13Ccarb typically begins negative and trends positive, returning to values .0‰ within a few metres (Kennedy et al. 1998; Halverson et al. 2005). For example, in the Rasthof Formation of NW Namibia, d13Ccarb begins at around –4.3‰ (herein referred to as the Rasthof anomaly) and crosses over abruptly to positive values c. 10 m up-section, at the contact

58

G. P. HALVERSON & G. SHIELDS-ZHOU

between a basal rhythmite member and an overlying chaotic microbialite member (Hoffman et al. 1998; Yoshioka et al. 2003). The thickness of this anomaly in Namibia is in large part a function of the amount of allodapic dolomite within the lower member (Halverson et al. 2005). Only in rare instances is a negative trend preserved in the lowermost part of the sequence (e.g. in the presumed Rasthof-equivalent basal Tayshir Member of the Tsagaan Oloom Formation, Mongolia; Macdonald et al. 2009a). In contrast, the basal transgressive portion of the sequence overlying the end-Cryogenian glacials and defining the base of the Ediacaran Period (Knoll et al. 2006) commonly preserves an upward negative trend, with values typically in the range of –2 to –4‰ (Hoffman & Schrag 2002; Hoffman et al. 2007; Rose & Maloof 2010). In the carbonate-dominated post-glacial platformal sequence in NW Namibia (Maieberg Formation) where the Maieberg anomaly has been most extensively studied, d13Ccarb continues to decline to nearly –6‰ at the maximum flooding surface before gradually returning to higher values over hundreds of metres of the section (Hoffman et al. 1998). However, this apparent protracted recovery is not typical of the basal Ediacaran negative d13Ccarb anomaly, and in many other carbonatedominated locations, including the slope equivalents of the Maieberg Formation in Namibia (Halverson et al. 2005), the Doushantuo Formation of South China (e.g. McFadden et al. 2008), and the Barakunskaya Formation in southern Siberia (Pokrovskii et al. 2006), the anomaly is relatively condensed, spanning only several metres to a few tens of metres. Thus, relative sedimentation rates must be considered when correlating chemostratigraphic signals. Large negative d13Ccarb anomalies also precede two Cryogenian glaciations (Table 4.1; Fig. 4.2). The so-called Trezona anomaly, named after the Trezona Formation in South Australia where the anomaly was first fully documented (McKirdy et al. 2001), occurs prior to the end-Cryogenian (i.e. the ‘Marinoan’ or ‘Elatina’ glaciation in South Australia) glaciation and following the so-called Keele Peak (Kaufman et al. 1997), where d13Ccarb reaches values þ9‰ (Figs 4.1 & 4.2). This negative anomaly is best documented in the Ombaatjie Formation in the Otavi Group, NW Namibia, where it is defined by a gradual, faciesindependent decline from c. þ7 to –5‰ through up to 50 m of dominantly shallow-water carbonates, followed by a positive shift of a few ‰ (Halverson et al. 2002). The extent of truncation of the anomaly can be used as a general indicator of the depth of erosion on the glacial surface. A negative anomaly of similar magnitude, referred to as the Islay anomaly (Prave et al. 2009) after the formation in Scotland (Figs 4.1 & 4.2) where it was first documented (Brasier & Shields 2000; McCay et al. 2006; Sawaki et al. 2010a), occurs prior to the older Cryogenian glaciation in several successions. In NE Svalbard, this anomaly is virtually identical in shape, magnitude and stratigraphic context to the Trezona anomaly in Namibia (Fig. 4.2) and is similarly variably erosionally truncated (Halverson et al. 2004). However, in Scotland, d13C returns to positive values prior to glaciation (Prave et al. 2009). In NW Canada and East Greenland, the presumably (but not definitively) equivalent anomalies in the Coates Lake Group (Halverson 2006) and Bed Groups 19–20 (Fairchild et al. 2000), respectively, feature an even more prominent positive shift, possibly providing an important distinction from the Trezona anomaly. A potentially different late Cryogenian negative d13Ccarb anomaly of similar magnitude has been found in the Tayshir Member of the lower Tsagaan Oloom Formation in Mongolia (Macdonald et al. 2009a). This anomaly has been tentatively correlated with the Bonahaven anomaly in Scotland (Prave et al. 2009) and an anomaly of much smaller magnitude in the Gruis Formation of NW Namibia, between the Rasthof and Trezona anomalies (Fig. 4.2). A similar large negative anomaly in this stratigraphic position has not been documented elsewhere. Therefore, it remains plausible that the anomaly in the Tayshir Member is

equivalent to the Trezona anomaly and the return to positive d13Ccarb values above the anomaly has been removed by erosion as it was in Namibia (Macdonald et al. 2009a). Although the homologous pattern of the Islay and Trezona anomalies suggests a similar origin, it diminishes the utility of these anomalies for correlation purposes. Fortunately, the anomalies are clearly distinguished from one another chemostratigraphically by their strontium-isotope signatures. Whereas 87 Sr/86Sr ratios of 0.7072–0.7073 coincide with the Trezona anomaly (Halverson et al. 2007a), 87Sr/86Sr through the older Islay anomaly distinctly declines from c. 0.7068 to 0.7064 (Fig. 4.2; Table 4.1), with the lowest values in the Islay Formation corresponding to the return towards positive d13Ccarb values (Sawaki et al. 2010a). Similar 87Sr/86Sr values have been reported from the likely equivalent interval in the Coates Lake Group (0.7066; Halverson et al. 2007a) and Bed Group 20 (0.7063; Fairchild et al. 2000). Thus, it appears that the onset of the middle-Cryogenian glaciation is preceded by a 0.0005 decline in the 87Sr/86Sr of seawater following the c. 765 Ma peak of c. 0.7070 (Fig. 4.1) documented in the Backlundtoppen Formation. In contrast to this negative slope leading up to the middle Cryogenian glaciation, the 87Sr/86Sr record in the following interglacial interval has a positive slope (Fig. 4.1), with a smooth and seemingly rapid increase in 87Sr/86Sr from 0.7067 to 0.7073 recorded in the Tsagaan Oloom Formation in Mongolia (Shields et al. 1997). Strontium-rich aragonite and barite cements of the Hayhook Formation in the upper part of the transgressive portion of the post-Ice Brook (base Ediacaran) sequence in the Mackenzie Mountains, NW Canada, have 87Sr/86Sr ratios of c. 0.70715 (Halverson et al. 2007a; Hoffman & Halverson 2011). These values are effectively indistinguishable from a ratio of 0.70718 measured within the pre-glacial Keele Peak in the same section (Halverson et al. 2007a) and therefore indicate no net change in 87Sr/87Sr spanning the end-Cryogenian glaciation. The sulphur-isotope record spanning the Cryogenian glaciations is almost entirely based on pyrite and CAS (there being few evaporites during this time span) and is intensely variable (Fig. 4.1). This variability reflects a combination of low seawater sulphate concentrations and strongly fluctuating environmental conditions tied to global glaciation (Hurtgen et al. 2002), as well as possible post-depositional effects. Sulphur-isotope data (Fig. 4.1) from the Chuar Group in southwestern USA show a return to d34Spyrite values clustering around 0‰ and low d34S values around 750 Ma, coinciding with at least local development of euxinic deep waters as determined from iron-speciation data on mid-shelf black shales (Johnston et al. 2010). Sulphur-isotope data are sparse for the subsequent interval leading up to the middle Cryogenian glaciation (i.e. the c. 717 Ma Rapitan glaciation; Macdonald et al. 2010b). However, abundant pyrite and CAS data have been produced from the interglacial interval, in particular from Namibia (Hurtgen et al. 2002; Gorjan et al. 2003) and Australia (Gorjan et al. 2000; McKirdy et al. 2001). Whereas d34Spyrite data from within the Sturt and equivalent glacial units in Australia range from 0 to –20‰, they are strongly positive, with values commonly 20– 60‰ in post-glacial rocks, such as the Tapley Hill, Aralka, Datangpo, Rasthof and lower Court formations (Fig. 4.1). This extreme, immediately post-glacial trend in pyrite values is matched by highly 34S-enriched d34Ssulphate values from CAS, such that D34S  0‰. d34Spyrite values then decline precipitously in the middle interglacial period, paralleling a smaller decline in d34Ssulphate, resulting in D34S values approaching 30‰. D34S reduces again just before the onset of the endCryogenian glaciation, coinciding with the Trezona negative d13C anomaly (Fig. 4.1; Table 4.1). d34S values are again erratic immediately following the end-Cryogenian glaciation: a trend of rapidly increasing d34Ssulphate that coincides with a wide range in d34Spyrite (– 30 to þ20‰) may be sufficient to distinguish this post-glacial (i.e.

CHEMOSTRATIGRAPHY AND THE NEOPROTEROZOIC GLACIATIONS

59

Fig. 4.2. Summary stratigraphic plots and accompanying d13Ccarb profiles and 87 Sr/86Sr data for key sections spanning the Trezona and Islay pre-glacial (Cryogenian) negative d13C anomalies. Note the variable scales across the non-scaled glaciogenic intervals in each column. Nucc, Nuccaleena Formation; Brach, Brachina Formation; Macdon., Macdonaldryggen Member. Australia data from McKirdy et al. (2001); Namibia data from Halverson et al. (2002, 2007a), which includes data from both the inner and outer continental shelf of the Ombaatjie Formation; Scotland data from Brasier & Shields (2000), McCay et al. (2006) and Prave et al. (2009); Svalbard data from Halverson et al. (2004, 2007a).

basal Ediacaran) sulphur-isotope record from the middle Cryogenian post-glacial record. However, coeval d34S profiles that vary from more proximal to more open ocean post-glacial sections in Namibia (Hurtgen et al. 2006) and South China (Li et al. 2010) suggest significant heterogeneity in the concentration and sulphur-isotope composition of seawater sulphate and generally low sulphate concentrations in the earliest Ediacaran Period. Therefore, sulphur-isotope chemostratigraphy may be useful in establishing an early Ediacaran age, but would not be reliable for detailed correlations for sedimentary rocks of this age.

The Ediacaran Period The beginning of the Ediacaran Period coincides with a major negative d13Ccarb anomaly (the Maieberg anomaly; Fig. 4.1) that punctuates the final global glaciation of the Neoproterozoic (Knoll et al. 2006), as discussed above. Although d13Ccarb values return to near 0‰ within about 3 million years of the end of glaciation (Condon et al. 2005), they then remain relatively low for much of the early Ediacaran Period, with the exception of a positive spike to þ6 –10‰ seen in some sections, such as NW Namibia and NE

60

G. P. HALVERSON & G. SHIELDS-ZHOU

Svalbard (Halverson et al. 2005). This pattern contrasts with the sustained high d13Ccarb values typical of much of the earlier Neoproterozoic. Superimposed on the lower average d13Ccarb values in the Ediacaran Period, the dominant feature in the record is the so-called Shuram (or Shuram-Wonoka) anomaly, named after the Shuram Formation in the Huqf Supergroup of Oman (Burns & Matter 1993). The Shuram anomaly features a precipitous drop to d13Ccarb values , – 10‰, followed by a protracted recovery back to values near 0‰ (Fig. 4.1). This extraordinary anomaly has been the subject of considerable recent research and controversy. However, despite arguments against a primary origin for the anomaly (Bristow & Kennedy 2008; Knauth & Kennedy 2009; Derry 2010), anomalies of strikingly similar magnitude, structure and approximate age occur in multiple sedimentary basins from across the globe (Table 4.1), including Sr-rich limestones in southcentral Siberia (Pokrovskii et al. 2006; Melezhik et al. 2009). Thus, the most parsimonious interpretation is that the Shuram anomaly reflects a primary oceanographic phenomenon. Despite the large number of chemostratigraphic data that have been obtained globally from rocks preserving the Shuram anomaly, the timing and duration of the anomaly and, indeed, the number of large Ediacaran negative d13Ccarb anomalies remain controversial. For example, whereas d13Ccarb records from thick, seemingly continuous carbonate-rich sections in Oman (Fike et al. 2006; Le Guerroue´ et al. 2006) and south-central Siberia (Pokrovskii et al. 2006) clearly show a single, long-lived negative d13Ccarb anomaly, certain sections from the Doushantuo Formation, South China, show a pair of large-magnitude anomalies separated by d13Ccarb values as high as þ6‰ (Jiang et al. 2007; McFadden et al. 2008). Because the Doushantuo pattern is preserved in highly condensed sections (relative to sections from other successions) from a basin that was not demonstrably fully open to the global ocean and has not been reproduced globally, the more consistent and higher-resolution patterns derived from Oman and Siberia are favoured in the compilation in Figure 4.1. However, we acknowledge that the inconsistent patterns between the Doushantuo Formation and correlative units worldwide pose a problem for carbon-isotope chemostratigraphy. Separate data sets from the Wonoka Formation, South Australia (Calver 2000), Oman (Fike et al. 2006) and South China (McFadden et al. 2008) also show a decoupling between the d13Ccarb record spanning the Shuram anomaly and coeval organic carbon (d13Corg) data (Calver 2000; Fike et al. 2006; McFadden et al. 2008), a pattern that lends support to the popular hypothesis that the Ediacaran ocean contained a large, reactive organic carbon pool that strongly modulated d13Ccarb and the oxidation of the deep oceans (Rothman et al. 2003; Condon et al. 2005). However, the relationship between d13Ccarb and d13Corg is not the same in these three successions. For example, d13Corg values are much more negative in South China than in Oman or South Australia, and d13Corg values are more variable through the main part of the anomaly in Oman than elsewhere. Therefore, the significance and reliability of the middle Ediacaran d13Corg record is unclear. Of critical importance in reconstructing the connections between perturbations to the exogenic carbon cycle, global climate and biospheric evolution in the Ediacaran Period is the connection, if any, between the Shuram anomaly and the shortlived c. 580 Ma Gaskiers glaciation (Bowring et al. 2003). Is the Gaskiers glaciation, like the global Cryogenian glaciations, preceded by a large negative d13C anomaly? Unfortunately, no unambiguous evidence for both the glaciation and the full anomaly occurs in any single stratigraphic succession. Although Halverson et al. (2005) interpreted major erosional unconformities in successions hosting the Shuram anomaly, such as the Wonoka Formation (South Australia) and the Johnnie Formation (Death Valley), to be related to Gaskiers glacioeustasy, this argument remains speculative. The problem of resolving the relative timing between d13C

anomalies and glaciation is exacerbated by the possibility that whereas the anomaly should be globally synchronous, the glaciation may be diachronous, or there may be multiple discrete glaciations (e.g. Chumakov 2010). Indeed, d13Ccarb values are highly variable in relation to other reported Ediacaran glacial deposits, such as in the Kimberley of northern Australia, where positive d13C values characterize carbonates overlying the Egan Formation (Corkeron 2007, 2011). Other contrasting d13Ccarb values are found in the Hankalchough Formation in the Tarim Basin, NW China (Xiao et al. 2004), and the Hongtiegou Formation in the Chaidam Basin, NW China (Shen et al. 2010). Some authors have even argued that some of the purported Ediacaran glacial deposits, such as the Hankalchough (Chumakov 2010) and the Hongtiegou (Shen et al. 2010) formations significantly post-date the Gaskiers glaciation (Chumakov 2009). Germs et al. (2009) argue for a latest Ediacaran (c. 547 Ma) ‘Schwarzrand’ glaciation based on evidence from the upper Nama Group, Namibia. Notwithstanding the ambiguity in the age, timing, number and distribution of Ediacaran-aged glacial deposits, carbon-isotope data from the basal Ediacaran Nyborg Formation and the overlying glaciogenic Mortensnes Formation of the Varanger Peninsula, northern Norway, do appear to resolve the relative timing between at least one middle Ediacaran glaciation and the Shuram d13Ccarb anomaly. Dolomite beds within Member E in the uppermost Nyborg Formation, in the least-truncated section beneath the major erosional unconformity at the base of the glaciogenic Mortensnes Formation (Edwards 1984), have d13Ccarb compositions of –7.6 to –9.9‰, while the matrix of a carbonate-dominated diamictite within the Mortensnes Formation has a d13Ccarb composition of –10.4‰ (see Rice et al. 2011). Because the Shuram anomaly is the only interval in the Ediacaran Period where d13Ccarb values are known to drop below –7‰, these data are interpreted to indicate that the Mortensnes glaciation post-dated the nadir of the negative anomaly (Halverson et al. 2005). In this regard, an Ediacaran glaciation, although seemingly not global, may have been triggered by a similar mechanism as the Cryogenian glaciations. The return to positive d13Ccarb values following the Shuram anomaly occurs late in the Ediacaran, by c. 551 Ma, implying that the Ediacaran fauna first appeared during the anomaly (Condon et al. 2005). A subsequent, short-lived negative d13C anomaly coincides with the Precambrian –Cambrian boundary (Magaritz et al. 1986; Knoll et al. 1995; Saylor et al. 1998; Amthor et al. 2003), but there is no evidence to link this anomaly with a glaciation. Strontium-isotope ratios from the Mackenzie Mountains indicate little net change in seawater 87Sr/86Sr spanning the endCryogenian glaciation (Fig. 4.2). However, data from strontiumrich limestones in the lower Maieberg Formation, NW Namibia, show a rapid rise in 87Sr/86Sr to values .0.7078 following post-glacial maximum flooding and prior to recovery in d13Ccarb values to near 0‰ (Halverson et al. 2007; cf. Shields 2007). This positive spike in 87Sr/86Sr is a predictable result of extremely high CO2 levels and corresponding elevated silicate weathering rates following deglaciation (Higgins & Schrag 2003). Simple geochemical modelling also implies that the spike to more radiogenic values should be short-lived, with 87Sr/86Sr approaching background values after CO2 declined to more typical levels within a few million years of deglaciation (Le Hir et al. 2009). This decline may be recorded in data from the Una Group (Ireceˆ basin) on the Sa˜o Francisco Craton, Brazil (Misi & Veizer 1998), where the strontium- and carbonate-isotope compositions suggest an early Ediacaran age, consistent with recent detrital zircon data from the equivalent Bambui Group (Sial et al. 2010). A much better established pattern in the Ediacaran strontiumisotope record is the subsequent rise to 87Sr/86Sr  0.7080 that characterizes the latter part of the Ediacaran Period (e.g. Kaufman et al. 1997; Jacobsen & Kaufman 1999; Shields 2007; Melezhik et al. 2009). Sawaki et al. (2010b) suggested additional

CHEMOSTRATIGRAPHY AND THE NEOPROTEROZOIC GLACIATIONS

structure in the late Ediacaran record based on data they obtained from a drill core through the Doushantuo Formation, including a drop to ,0.7080 prior to the Shuram anomaly and a peak in 87 Sr/86Sr of 0.7090 coinciding with the protracted low d13Ccarb of the anomaly. Regarding the high values, as for Oman, where similarly high values have been reported for the late Ediacaran (Burns et al. 1994), none of these highly enriched values is preserved in samples retaining high strontium concentrations. Conversely, strontium-rich limestones in the upper Doushantuo Formation in the core have 87Sr/86Sr values of c. 0.7085, similar to strontium-isotope data straddling the Precambrian –Cambrian boundary in other well-studied successions (Brasier et al. 1996; Kaufman et al. 1997; Saylor et al. 1998). Gaucher et al. (2009) report 87Sr/86Sr values of ,0.7080 in the late Ediacaran Arroyo del Soldado Group on the Rio de la Plata craton, lending support to the possibility that 87Sr/86Sr may have indeed fluctuated strongly in the late Ediacaran Period, much as it did in the subsequent Early Cambrian (Halverson et al. 2010). Evidently, much work remains in elucidating the fine structure in the late Ediacaran 87Sr/86Sr record. The relationship between middle Ediacaran 87Sr/86Sr and the Gaskiers (or other Ediacaran) glaciation is virtually unconstrained. However, strontium-isotope data from highly strontium-enriched limestones in southern Siberia show 87Sr/86Sr ¼ 0.7081 at about the level of the nadir in the Shuram anomaly, followed by an increase towards 0.7086 up-section, but still within the range of very low d13Ccarb (Pokrovskii et al. 2006; Melezhik et al. 2009). Anomalously 34S-enriched seawater sulphate compositions bookend the Ediacaran Period (Fig. 4.1) in many successions. As discussed above, variable, but predominantly high d34Ssuphate and low D34S values near 0‰ prevail in the basal Ediacaran Period (Fig. 4.1). By the early Cambrian, d34Ssulphate, as measured in a variety of minerals, including evaporites, francolite and carbonate, was again high (35 –40‰) in most basins (e.g. Holser & Kaplan 1966; Strauss 1993; Shields et al. 1999; Kampschulte & Strauss 2004). However, d34Ssulphate from the late Ediacaran in South China (McFadden et al. 2008) and southern Namibia (Ries et al. 2009) are generally lower, with concomitant decreases in D34S. The source of this substantial spatial variability in the late Ediacaran sulphur-isotope record, whether oceanographic, related to post-depositional influences or analytical artefacts, is not resolved. The Ediacaran sulphur-isotope record as compiled in Figure 4.1 comprises a detailed data set of d34Ssulphate (CAS) and d34Spyrite data from the Nafun Group, Oman (Fike et al. 2006) and pyrite data from the Pertatataka Formation, central Australia (Gorjan et al. 2000). The Oman record, together with other fragmentary data (Fig. 4.1), indicates that more typical d34Ssulphate values close to 20‰ are the norm for the Ediacaran Period, although evaporite and CAS data appear to be contradictory at times (cf. Fike et al. 2006 with Schro¨der et al. 2004). At the same time, data from both Oman and central Australia display a spike to positive d34Spyrite values in the early Ediacaran Period, followed by a gradual decrease through the middle Ediacaran Period, resulting in a steady increase in d34S, a predictable result of increasing seawater sulphate concentrations (Halverson & Hurtgen 2007). Average d34Spyrite values in the deepwater Conception Group (Avalon Peninsula, SE Newfoundland; Canfield et al. 2007) show a decrease from about þ20 to 0‰ across the Gaskiers glaciation, but the d34Ssulphate record spanning the glaciation is unknown. Iron-speciation data indicate ferruginous (and presumably low sulphate) conditions during the Gaskiers glaciation (Canfield et al. 2008). Detailed d34Ssulphate (CAS) records have been produced across the onset of the Shuram –Wonoka anomaly. In Oman, a spike in d34Ssulphate to þ29‰ coincides with the anomaly (Fike et al. 2006). Across the presumably equivalent interval in Death Valley (Kaufman et al. 2007) and South China (McFadden et al. 2008), d34Ssulphate decreases to values ,20‰. Thus, although d34S records show tantalizing connections with

61

other major Ediacaran events, either d34S compositions of late Neoproterozoic seawater sulphate were highly heterogeneous (Ries et al. 2009), inferred correlations are inaccurate, or the measured sulphur-isotope compositions in some or all of the sedimentary successions do not preserve reliable signatures of primary seawater composition.

Discussion and conclusions Precise radiometric ages are the gold standard for drawing precise interbasinal correlations, particularly in the absence of a robust biostratigraphy. Direct ages are also required for calibrating the sedimentary record, but they are frustratingly rare in most of the successions that archive important information about the evolution of Earth’s surface environment in the Neoproterozoic. Thus, other tools and many assumptions are required to merge the fragmentary records from across the globe into one coherent Neoproterozoic chronology. Isotope chemostratigraphy has long been heralded as a possible solution (Knoll & Walter 1992). Insofar as the popular proxies for ocean chemistry (d13C, d34S, 87Sr/86Sr) are reliable, geologists can characterize the Neoproterozoic record one fragment at a time. Accordingly, isotope geochemistry has been widely applied to stratigraphic studies of the Neoproterozoic, with a strong focus on the glaciogenic record. An intimate relationship between extreme perturbations to the global carbon cycle, as recorded in the marine d13Ccarb record, and episodes of widespread glaciation (Knoll et al. 1986) has long been the starting point for these studies. However, even as this fundamental observation about the behaviour of the Neoproterozoic Earth has withstood the test of time, diverse studies from across the world have added important new details to this broad pattern. Long-lived positive carbon-isotope trends had been ascribed to pre-glacial oceanographic conditions, while negative carbon-isotope anomalies were attributed to post-glacial ‘cap carbonates’, providing for elegantly simple models for the interrelation between d13Ccarb, carbon burial, atmospheric CO2 and glaciation (e.g. Kaufman et al. 1997). The recognition that a large negative d13Ccarb anomaly actually preceded the endCryogenian glaciation shattered this model of a simple causeand-effect relationship between the two and opened the door for the compelling but controversial Snowball Earth hypothesis (Hoffman et al. 1998), which continues to dominate the debate about the Neoproterozoic glaciations. It is now recognized that negative d13C anomalies also precede middle Cryogenian (McCay et al. 2006; Prave et al. 2009; Macdonald et al. 2010b) and middle Ediacaran glaciations (Halverson et al. 2005). Furthermore, in all cases, it appears that the initiation of glaciation followed a return towards positive d13Ccarb values after the d13Ccarb minima, thus further complicating an already complex relationship between the anomalies and global cooling. This motif in the coupled d13C-climate record renders using the carbon-isotope record alone for global correlations insufficient, but fortunately it projects upon a marine strontium-isotope record with a strong unidirectional trend (Fig. 4.1). 87Sr/86Sr increases steadily during the Neoproterozoic from 0.7055 around the Mesoproterozoic– Neoproterozoic boundary to 0.7085 at the Neoproterozoic –Palaeozoic boundary, interrupted by only a few significant but short-lived declines (Fig. 4.1). Consequently, seawater 87Sr/86Sr ratios are different for each of the large negative anomalies associated with the Neoproterozoic glaciations (Table 4.1). The potential of strontium isotopes as a tracer of global environmental change remains to be fully realized. However, abrupt rises, most notably following the Cryogenian glaciations, are consistent with enhanced chemical weathering after glaciation. Conversely, changes in seawater 87Sr/86Sr appear to have been relatively muted during glaciations. At this stage, the Neoproterozoic record is now sufficiently well understood that the combination of carbon- and strontium-isotope

62

G. P. HALVERSON & G. SHIELDS-ZHOU

stratigraphy, particularly if integrated with stratigraphic and sedimentological data, enables confident correlations and assignment of robust, if only relative, ages on carbonate-rich successions that otherwise lack firm age control. High-quality data sets spanning variable portions of the Neoproterozoic record and tied to an increasing number of precise U –Pb zircon ages have now been produced from across the world (e.g. Macdonald et al. 2010b). The result is a convergence in carbon- and strontiumisotope compilations towards a consistent pattern, even if many of the finer details remain obscure. The integrated stratigraphic, d13C and 87Sr/86Sr records serve as a chronological template to the Neoproterozoic onto which other secular data sets can be added, such as sulphur-isotope compositions, redox specific data, biomarkers and biostratigraphy. The sulphur-isotope record is patchy, particularly for the first half of the Neoproterozoic, and highly variable, with clear instances where data sets spanning equivalent-aged rocks from different parts of the world, and even single basins, are not consistent with one another. Thus, much work remains to be done to establish the reliability of the sulphur-isotope proxy and the causes, either related to oceanographic conditions or diagenetic process, for the high degree of scatter in the record. Nevertheless, several of the salient patterns, mostly notably extremely high d34Ssulphate and d34Spyrite values, are temporally linked to glaciation, indicating that the record can be used to help reconstruct the palaeoenvironmental evolution of Neoproterozoic ocean chemistry. Specifically, large variations in d34S of sulphate and pyrite and the difference between these two values in coeval sediments (D34S) is an indirect measure of fluctuations in the redox state of global seawater (Hurtgen et al. 2005). The sustained and very high d34Spyrite (20 –60‰) values recorded in the post-Sturtian Tapley Hill – Aralka formations in Australia (Gorjan et al. 2000) and the peak in d34Ssulphate straddling the Precambrian – Cambrian boundary stand out as unique chronostratigraphic markers in the otherwise noisy sulphur-isotope record (Fig. 4.1). Iron-speciation data has become a popular new tool for reconstructing water column redox conditions (Lyons & Severmann 2006) that has been successfully applied to the Neoproterozoic (e.g. Canfield et al. 2007, 2008). Although it remains to be shown that distinct trends can be correlated beyond a single basin (Johnston et al. 2010), or are even consistent within a single basin, sharp fluctuations in normalizsed highly reactive iron content in shales (FeHR/FeT), for example, are closely linked to glaciation and other proxy evidence for oscillations in the redox chemistry of the oceans (Canfield et al. 2007, 2008). Although not discussed here, many other chemostratigraphic proxies show great promise for tracing evolving redox conditions through the Neoproterozoic and evaluating the highly unusual environment in the aftermath of global glaciations (Halverson et al. 2010, and references therein). Furthermore, biostratigraphy is increasingly being tightly integrated with chemostratigraphic data in order to evaluate more closely the connections between biospheric and biogeochemical change (e.g. McFadden et al. 2008; Nagy et al. 2009; Macdonald et al. 2010a). Thus, there is little doubt that chemostratigraphy will continue to elucidate the finer details in the number, timing and correlation of Neoproterozoic glaciations and the environmental conditions leading up to and following them. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Alene, M., Jenkin, G. R. T., Leng, M. J. & Darbyshire, D. P. F. 2006. The Tambien Group, Ethiopia: An early Cryogenian (ca. 800– 735 Ma) Neoproterozoic sequence in the Arabian– Nubian Shield. Precambrian Research, 147, 79 –99.

Allen, P. A. & Etienne, J. L. 2008. Sedimentary challenge to Snowball Earth. Nature Geoscience, 1, 817– 825. Amthor, J. E., Grotzinger, J. P., Schro¨der, S., Bowring, S. A., Ramezani, J., Martin, M. W. & Matter, A. 2003. Extinction of Cloudina and Namacalathus at the Precambrian – Cambrian boundary in Oman. Geology, 31, 431–434. Bahlburg, H. & Dobrzinski, N. 2011. A review of the Chemical Index of Alteration (CIA) and its application to the study of Neoproterozoic glacial deposits and climate transitions. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 81 – 92. Bailey, T. R., McArthur, J. M., Prince, H. & Thirlwall, M. F. 2000. Dissolution methods for strontium isotope stratigraphy: whole rock analysis. Chemical Geology, 167, 313–319. Banner, J. L. & Hanson, G. N. 1990. Calculation of simultaneous isotopic and trace element variations during water– rock interaction with applications to carbonate diagenesis. Geochimica et Cosmochimica Acta, 54, 3123– 3137. Bao, H., Lyons, J. R. & Zhou, C. 2008. Triple oxygen isotope evidence for elevated CO2 levels after a Neoproterozoic glaciation. Nature, 453, 504– 506. Bartley, J. K., Semikhatov, M. A., Kaufman, A. J., Knoll, A. H., Pope, M. C. & Jacobsen, S. B. 2001. Global events across the Mesoproterozoic – Neoproterozoic boundary: C and Sr isotopic evidence from Siberia. Precambrian Research, 111, 165– 202. Borg, G., Ka¨rner, K., Buxton, M., Armstrong, R. & van der Merwe, S. W. 2003. Geology of the Skorpion supergene zinc deposit, southern Namibia. Economic Geology, 98, 749–771. Bowring, S., Myrow, P., Landing, E., Ramezani, J. & Grotzinger, J. 2003. Geochronological constraints on terminal Proterozoic events and the rise of the Metazoans. Geophysical Research Abstracts (EGS, Nice), 50, 13219. Brand, U. & Veizer, J. 1981. Chemical diagenesis of a multicomponent carbonate system – 2: stable isotopes. Journal of Sedimentary Petrology, 51, 987– 997. Brasier, M. D. & Lindsay, J. F. 1995. A billion years of environmental stability and the emergence of eukaryotes: New data from northern Australia. Geology, 26, 555–558. Brasier, M. D. & Shields, G. 2000. Neoproterozoic chemostratigraphy and correlation of the Port Askaig glaciation, Dalradian Supergroup of Scotland. Journal of the Geological Society, London, 157, 909–914. Brasier, M. D., Shields, G. A., Kuleshov, V. N. & Zhegallo, E. A. 1996. Integrated chemo- and biostratigraphic calibration of early animal evolution: Neoproterozoic– early Cambrian of southwest Mongolia. Geological Magazine, 133, 445–485. Bristow, T. F. & Kennedy, M. J. 2008. Carbon isotope excursions and the oxidant budget of the Ediacaran atmosphere and ocean. Geology, 36, 863–866. Burns, S. J. & Matter, A. 1993. Carbon isotopic record of the latest Proterozoic from Oman. Eclogae Geologicae Helvetiae, 86, 595– 607. Burns, S. J., Haudenschild, U. & Matter, A. 1994. The strontium isotopic composition of carbonates from the late Precambrian (,560–540 Ma) Huqf Group of Oman. Chemical Geology, 111, 269– 282. Calver, C. 2000. Isotope stratigraphy of the Ediacaran (Neoproterozoic III) of the Adelaide rift complex, Australia, and the overprint of water column stratification. Precambrian Research, 100, 121– 150. Canfield, D. E. 1989. Reactive iron in marine sediments. Geochimica et Cosmochimica Acta, 53, 619– 632. Canfield, D. E. & Teske, A. 1996. Late Proterozoic rise in atmospheric oxygen concentration inferred from phylogenetic and sulfur-isotope studies. Nature, 382, 127–132. Canfield, D. E., Poulton, S. W., Knoll, A. H., Narbonne, G. M., Ross, G., Goldberg, T. & Strauss, H. 2008. Ferruginous conditions dominated later Neoproterozoic deep-water chemistry. Science, 321, 949– 952. Canfield, D. E., Poulton, S. W. & Narbonne, G. M. 2007. LateNeoproterozoic deep-ocean oxygenation and the rise of animal life. Science, 315, 92 – 95. Chumakov, N. M. 2009. The Baykonurian Glaciohorizon of the Late Vendian. Stratigraphy and Geological Correlation, 17, 373–381.

CHEMOSTRATIGRAPHY AND THE NEOPROTEROZOIC GLACIATIONS

Chumakov, N. M. 2010. Neoproterozoic glacial events in Eurasia. In: Gaucher, C., Sial, A. N., Halverson, G. P. & Frimmel, H. E. (eds) Neoproterozoic-Cambrian Tectonics, Global Change and Evolution: A Focus on Southwestern Gondwana. Developments in Precambrian Geology, 16. Elsevier, Dordrecht, 389– 403. Condon, D., Zhu, M., Bowring, S., Jin, Y., Wang, W. & Yang, A. 2005. From the Marinoan glaciation to the oldest bilaterians: U –Pb ages from the Doushantou Formation, China. Science, 308, 95 – 98. Corkeron, M. 2007. ‘Cap carbonates’ and Neoproterozoic glacigenic successions from the Kimberly region, north-west Australia. Sedimentology, 54, 871– 903. Corkeron, M. 2011. Neoproterozoic glacial deposits of the Kimberly Region and northwestern Northern Territory, Australia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 659–672. Corsetti, F. A. & Kaufman, A. J. 2003. Statigraphic investigations of carbon-isotope anomalies and Neoproterozoic ice ages in Death Valley, California. Geological Society of America Bulletin, 115, 916– 932. Cozzi, A., Allen, P. A. & Grotzinger, J. P. 2004. Understanding carbonate ramp dynamics from C profiles: Examples from the Neoproterozoic Buah Formation of Oman. Terra Nova, 16, 62 – 67. Derry, L. A. 2010. A burial diagenesis origin for the Ediacaran ShuramWonoka anomaly. Earth and Planetary Science Letters, 295, 152– 162. Derry, L. A., Kaufman, A. J. & Jacobsen, S. B. 1992. Sedimentary cycling and environmental change in the Late Proterozoic: evidence from stable and radiogenic isotopes. Geochimica et Cosmochimica Acta, 56, 1317–1329. Detmers, J., Bru¨chert, V., Habicht, K. S. & Kuever, J. 2001. Diversity of sulfur isotope fractionations by sulfate-reducing prokaryotes. Applied and Environmental Microbiology, 67, 888– 894. Edmond, J. M. 1992. Himalayan tectonic, weathering processes, and strontium isotope record in marine limestones. Science, 258, 1594– 1597. Edwards, M. B. 1984. Sedimentology of the Upper Proterozoic glacial record, Vestertana Group, Finnmark, North Norway. Norges geologiske Undersøkelse Bulletin, 394, 1– 76. Evans, D. A. D. 2006. Proterozoic low orbital obliquity and axialdipolar geomagnetic field from evaporite paleolatitudes. Nature, 444, 51– 55. Fairchild, I. J., Marshall, J. D. & Bertrand-Sarfati, J. 1990. Stratigraphic shifts in carbon isotopes from Proterozoic stromatolitic carbonates (Mauritania): influences of primary mineralogy and diagenesis. American Journal of Science, 290A, 46 –79. Fairchild, I. J., Spiro, B., Herrington, P. M. & Song, T. 2000. Controls on Sr and C isotope compositions of Neoproterozoic Sr-rich limestones of East Greenland and North China. In: Grotzinger, J. & James, N. (eds) Carbonate Sedimentation, Diagenesis in an Evolving Precambrian World. SEPM Special Publication, 67, 297– 313. Fanning, C. M. & Link, P. K. 2008. Age constraints for the Sturtian glaciation: data from the Adelaide Geosyncline, South Australia and Pocatello Formation, Idaho, USA. Selwyn Symposium 2008, Geological Society of Australia, Extended Abstracts, 91, 57– 62. Farquhar, J., Bao, H. & Thiemens, M. 2000. Atmospheric influence of Earth’s earliest sulfur cycle. Science, 289, 756– 758. Fike, D. A., Grotzinger, J. P., Pratt, L. M. & Summons, R. E. 2006. Oxidation of the Ediacaran ocean. Nature, 444, 744– 747. Fo¨lling, P. G. & Frimmel, H. W. 2002. Chemostratigraphic correlation of carbonate successions in the Gariep and Saldania belts, Namibia and South Africa. Basin Research, 14, 69 –88. Frank, T. D., Kah, L. C. & Lyons, T. W. 2003. Changes in organic matter production and accumulation as a mechanism for isotopic variation in the Mesoproterozoic ocean. Geological Magazine, 140, 397– 420. Frimmel, H. 2009. Trace element distribution in Neoproterozoic carbonates as palaeoenvironmental indicator. Chemical Geology, 258, 338– 353.

63

Frimmel, H. W., Klo¨tzi, U. S. & Siegfried, P. R. 1996. New Pb/Pb single zircon age constraints on the timing of Neoproterozoic glaciation and continental break-up in Namibia. The Journal of Geology, 104, 459–469. Gaucher, C., Sial, A. N., Poire´, D., Go´mez-Peral, L., Ferreira, V. P. & Pimentel, M. M. 2009. Chemostratigraphy. Neoproterozoic – Cambrian evolution of the Rı´o de la Plata Palaeocontinent. In: Gaucher, C., Sial, A. N., Halverson, G. P. & Frimmel, H. E. (eds) Neoproterozoic –Cambrian Tectonics, Global Change and Evolution: A Focus on Southwestern Gondwana. Developments in Precambrian Geology, 16, 115–122. Germs, G. J. B., Miller, R. McG., Frimmel, H. E. & Gaucher, C. 2009. Syn- to late-orogenic sedimentary basins of southwestern Africa. Neoproterozoic to Early Palaeozoic evolution of Southwestern Africa. In: Gaucher, C., Sial, A. N., Halverson, G. P. & Frimmel, H. E. (eds) Neoproterozoic– Cambrian Tectonics, Global Change and Evolution: A Focus on Southwestern Gondwana. Developments in Precambrian Geology, 16, 183– 203. Gorjan, P., Veevers, J. J. & Walter, M. R. 2000. Neoproterozoic sulfurisotope variation in Australia and global implications. Precambrian Research, 100, 151–179. Gorjan, P., Walter, M. R. & Swart, R. 2003. Global Neoproterozoic (Sturtian) post-glacial sulfide-sulfur isotope anomaly recognised in Namibia. Journal of African Earth Sciences, 36, 89 –98. Gorokhov, I. M., Semikhatov, M. A., Baskakov, A. V., Kutyavin, E. P., Melnikov, N. N., Sochava, A. V. & Turchenko, T. L. 1995. Sr isotopic composition in Riphean, Vendian, and Lower Cambrian carbonates from Siberia. Stratigraphy and Geological Correlation, 3, 1 – 28. Grotzinger, J. P., Bowring, S. A., Saylor, B. Z. & Kaufman, A. J. 1995. Biostratigraphic and geochronologic constraints on early animal evolution. Science, 270, 598–604. Halverson, G. P. 2006. A Neoproterozoic Chronology. In: Xiao, S. & Kaufman, A. J. (eds) Neoproterozoic Geobiology and Paleobiology. Springer, 231– 271. Halverson, G. P. & Hurtgen, M. T. 2007. Ediacaran growth of the marine sulfate reservoir. Earth and Planetary Science Letters, 263, 32– 44. Halverson, G. P., Hoffman, P. F., Schrag, D. P. & Kaufman, A. J. 2002. A major perturbation of the carbon cycle before the Ghaub glaciation (Neoproterozoic) in Namibia: prelude to snowball Earth? Geochemistry, Geophysics, Geosystems, 3, 10.1029/2001GC000244. Halverson, G. P., Maloof, A. C. & Hoffman, P. F. 2004. The Marinoan glaciation (Neoproterozoic) in Svalbard. Basin Research, 16, 297– 324. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. 2005. Towards a Neoproterozoic composite carbon-isotope record. Geological Society of America Bulletin, 117, 1181–1207. Halverson, G. P., Dudas, F. O., Maloof, A. C. & Bowring, S. A. 2007a. Evolution of the 87Sr/86Sr composition of Neoproterozoic seawater. Palaeogeography, Palaeoclimatology, Palaeoecology, 256, 103–129. Halverson, G. P., Maloof, A. C., Schrag, D. P., Dudas, F. O. & Hurtgen, M. T. 2007b. Stratigraphy and geochemistry of a ca 800 Ma negative carbon isotope interval in northeastern Svalbard. Chemical Geology, 237, 5 –27. Halverson, G. P., Wade, B. P., Hurtgen, M. T. & Barovich, K. M. 2010. Neoproterozoic chemostratigraphy. Precambrian Research, 182, 337–350. Hayes, J. M., Kaplan, I. R. & Wedeking, K. W. 1983. Precambrian organic geochemistry, preservation of the record. In: Schopf, J. W. (ed.) Earth’s Earliest Biosphere. Its Origin and Evolution. Princeton University Press, 93 – 134. Hayes, J. M., Strauss, H. & Kaufman, A. J. 1999. The abundance of 13C in marine organic matter and isotopic fractionation in the global biogeochemical cycle of carbon during the past 800 Ma. Chemical Geology, 161, 103– 125. Higgins, J. A. & Schrag, A. P. 2003. Aftermath of a snowball Earth. Geochemistry, Geophysics, Geosystems, 4, 1028. Hill, A. C. 2005. Stable isotope stratigraphy, GSWA Lancer 1, Officer Basin, Western Australia. In: Mory, A. J. & Haines, P. W. (eds) GSWA Lancer 1 Well Completion Report (Interpretive Papers)

64

G. P. HALVERSON & G. SHIELDS-ZHOU

Officer and Gunbarrel Basins, Western Australia. Western Australia Geological Survey Record 2005/4, 1– 11. Hill, A. C., Arouri, K., Gorjan, P. & Walter, M. R. 2000. Geochemistry of marine and non-marine environments of a Neoproterozoic cratonic carbonate/evaporite: the Bitter Springs Formation, Central Australia. In: Grotzinger, J. P. & James, N. P. (eds) Carbonate Sedimentation and Diagenesis in an Evolving Precambrian World. SEPM Tulsa Special Publications, 67, 327– 344. Hoffman, P. F. & Halverson, G. P. 2011. Neoproterozoic glacial record in the Mackenzie Mountains, northern Canadian Cordillera. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 397–411. Hoffman, P. F. & Li, Z.-X. 2009. A palaeogeographic context for Neoproterozoic glaciation. Palaeogeography, Palaeoclimatology, Palaeoecology, 277, 158– 172. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129–155. Hoffman, P. F., Kaufman, A. J., Halverson, G. P. & Schrag, D. P. 1998. A Neoproterozoic snowball Earth. Science, 281, 1342– 1346. Hoffman, P. F., Halverson, G. P., Domack, E. W., Husson, J. M., Higgins, J. A. & Schrag, D. P. 2007. Are basal Ediacaran (635 Ma) post-glacial ‘cap dolostones’ diachronous? Earth and Planetary Science Letters, 258, 114– 131. Hoffman, P. F., Macdonald, F. A. & Halverson, G. P. 2011. Chemical sediments associated with Neoproterozoic glaciations: iron formation, cap carbonate, barite and phosphorite. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 67– 80. Hoffmann, K. H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. A U– Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817– 820. Holser, W. T. 1997. Geochemical events documented in inorganic carbon isotopes. Palaeogeography, Palaeoclimatology, Palaeoecology, 132, 173– 182. Holser, W. T. & Kaplan, I. R. 1966. Isotope geochemistry of sedimentary sulfates. Chemical Geology, 1, 93 –135. Hough, M. L., Shields, G. A., Evins, L. Z., Strauss, H., Henderson, R. A. & Mackenzie, S. 2006. A major sulphur-isotope event at c. 510 Ma: a possible anoxia– extinction –volcanism connection during the Early– Middle Cambrian transition? Terra Nova, 18, 257– 263. Hsu, K. J., Oberha¨nsli, H., Gao, J. Y., Shu, S., Haihong, C. & Kra¨henbu¨hl, U. 1985. ‘Strangelove ocean’ before the Cambrian explosion. Nature, 316, 809–811. Hurtgen, M. T., Arthur, M. A., Suits, N. & Kaufman, A. J. 2002. The sulfur isotopic composition of Neoproterozoic seawater sulfate: implications for snowball Earth? Earth and Planetary Science Letters, 203, 413–429. Hurtgen, M. T., Halverson, G. P., Arthur, M. A. & Hoffman, P. F. 2006. Sulfur cycling in the aftermath of a Neoproterozoic (Marinoan) snowball glaciation: Evidence for a syn-glacial sulfidic deep ocean. Earth and Planetary Science Letters, 245, 551–570. Jacobsen, S. B. & Kaufman, A. J. 1999. The Sr, C, and O isotopic evolution of Neoproterozoic seawater. Chemical Geology, 161, 37 – 57. Jiang, G., Kaufman, A. J., Christie-Blick, N., Zhang, S. & Wu, H. 2007. Carbon isotope variability across the Ediacaran Yangtzee platform in South China: Implications for a large surface-to-deep ocean gradient. Earth and Planetary Science Letters, 261, 303– 320. Johnston, D. T., Schmitz, M. D. et al. 2005. Active microbial sulfur disproportionation in the Mesoproterozoic. Science, 310, 1477– 1479. Johnston, D. T., Poulton, S. W., Dehler, C., Porter, S., Husson, J., Canfield, D. E. & Knoll, A. H. 2010. An emerging picture of Neoproterozoic ocean chemistry: Insights from the Chuar Group, Grand Canyon, USA. Earth and Planetary Science Letters, 290, 64– 73. Jones, D. S., Maloof, A. C., Hurtgen, M. T., Rainbird, R. H. & Schrag, D. P. 2010. Regional and global chemostratigraphic correlation of the early Neoproterozoic Shaler Sueprgroup, Victoria Island, Northwestern Canada. Precambrian Research, 181, 43 –63.

Kampschulte, A. & Strauss, H. 2004. The sulfur isotopic evolution of Phanerozoic seawater based on the analysis of structurally substituted sulfate in carbonates. Chemical Geology, 204, 255–286. Kasemann, S. A., Hawkesworth, C. J., Prave, A. R., Fallick, A. E. & Pearson, P. 2005. Boron and calcium isotope composition in Neoproterozoic carbonate rocks from Namibia: evidence for extreme environmental change. Earth and Planetary Science Letters, 231, 73 – 86. Kaufman, A. J. & Knoll, A. H. 1995. Neoproterozoic variations in the Cisotopic composition of seawater. Precambrian Research, 73, 27–49. Kaufman, A. J., Hayes, J. M., Knoll, A. H. & Germs, G. J. B. 1991. Isotopic composition of carbonates and organic carbon from upper Proterozoic successions in Nambia. Precambrian Research, 49, 301– 327. Kaufman, A. J., Jacobsen, S. B. & Knoll, A. H. 1993. The Vendian record of Sr and C isotopic variations in seawater: Implications for tectonics and paleoclimate. Earth and Planetary Science Letters, 120, 409– 430. Kaufman, A. J., Knoll, A. H. & Narbonne, G. M. 1997. Isotopes, ice ages, and terminal Proterozoic earth history. Proceedings of the National Academy of Science (USA), 94, 6600–6605. Kaufman, A. J., Corsetti, F. A. & Varni, M. A. 2007. The effect of rising atmospheric oxygen on carbon and sulfur isotope anomalies in the Neoproterozoic Johnnie Formation, Death Valley, USA. Chemical Geology, 237, 47 –63. Kennedy, M. J., Runnegar, B., Prave, A. R., Hoffmann, K.-H. & Arthur, M. 1998. Two or four Neoproterozoic glaciations? Geology, 26, 1059–1063. Kennedy, M. J., Christie-Blick, N. & Sohl, L. E. 2001. Are Proterozoic cap carbonates and isotopic excursions a record of gas hydrate destabilization following Earth’s coldest intervals? Geology, 29, 443– 446. Key, R. M., Liyungu, A. K., Njamu, F. M., Somwe, V., Banda, J., Mosley, P. N. & Armstrong, R. A. 2001. The western arm of the Lufilian Arc in NW Zambia and its potential for copper mineralization. Journal of African Earth Sciences, 33, 503– 528. Knauth, L. P. & Kennedy, M. J. 2009. The late Precambrian greening of the Earth. Nature, 460, 728– 732. Knoll, A. H. 2000. Learning to tell Neoproterozoic time. Precambrian Research, 100, 3– 20. Knoll, A. H. & Walter, M. R. 1992. Latest Proterozoic stratigraphy and Earth history. Nature, 356, 673–677. Knoll, A. H., Hayes, J. M., Kaufman, A. J., Swett, K. & Lambert, I. B. 1986. Secular variation in carbon isotope ratios from Upper Proterozoic successions of Svalbard and east Greenland. Nature, 321, 832– 837. Knoll, A. H., Grotzinger, J. P., Kaufman, A. J. & Kolosov, P. 1995. Integrated approaches to terminal Proterozoic stratigraphy: An example from the Olenek Uplift, northeastern Siberia. Precambrian Research, 73, 251– 270. Knoll, A. H., Walter, M. R., Narbonne, G. M. & Christie-Blick, N. 2006. The Ediacaran Period: a new addition to the geologic time scale. Lethaia, 39, 13 – 30. Kulp, J. L., Turekian, K. & Boyd, D. W. 1952. Strontium content of limestones and fossils. Geological Society of America Bulletin, 63, 701– 716. Le Guerroue´, E., Allen, P. A., Cozzi, A., Etienne, J. L. & Fanning, M. 2006. 50 Myr recovery from the largest negative d13C excursion in the Ediacaran ocean. Terra Nova, 18, 147–153. Le Hir, G., Donnadieu, Y. et al. 2009. The snowball Earth aftermath: Exploring the limits of continental weathering processes. Earth and Planetary Science Letters, 277, 453– 463. Li, C., Love, G. D., Lyons, T. W., Fike, D. A., Sessions, A. L. & Chu, X. 2010. A stratified redox model for the Ediacaran Ocean. Science, 328, 80– 83. Lyons, T. W. & Severmann, S. 2006. A critical look at iron paleoredox proxies: New insights from modern euxinic marine environments. Geochimica et Cosmochimica Acta, 70, 5698– 5722. Macdonald, F. A., Jones, D. S. & Schrag, D. P. 2009a. Stratigraphic and tectonic implications of a newly discovered glacial diamictitecap carbonate couplet in southwestern Mongolia. Geology, 37, 123– 126.

CHEMOSTRATIGRAPHY AND THE NEOPROTEROZOIC GLACIATIONS

Macdonald, F. A., McClelland, W. C., Schrag, D. P. & Macdonald, W. P. 2009b. Neoproterozoic glaciation on a carbonate platform margin in Arctic Alaska and the origin of the North Slope subterrane. Geological Society of America Bulletin, 121, 448–473. Macdonald, F. A., Cohen, P. A., Duda´s, F. O. & Schrag, D. P. 2010a. Early Neoproterozoic scale microfossils in the Lower Tindir Group of Alaska and the Yukon Territory. Geology, 38, 143–146. Macdonald, F. A., Schmitz, M. D. et al. 2010b. Calibrating the Cryogenian. Science, 327, 1241– 1243. ¨ . & Schrag, Macdonald, F. A., Strauss, J. V., Rose, C., Duda´s, F. O D. P. 2011. Stratigraphy of the Port Nolloth Group of Namibia and South Africa, and implications for the age of Neoproterozoic iron formations. American Journal of Science, 310, 862– 888. Magaritz, M., Holser, W. T. & Kirschvink, J. L. 1986. Carbon-isotope events across the Precambrian/Cambrian boundary on the Siberian Platform. Nature, 320, 258– 259. Maloof, A. C., Halverson, G. P., Kirschvink, J. L., Schrag, D. P., Weiss, B. P. & Hoffman, P. F. 2006. Combined paleomagnetic, isotopic, and stratigraphic evidence for true polar wander from the Neoproterozoic Akademikerbreen Group, Svalbard, Norway. Geological Society of America Bulletin, 118, 1099– 2014. Marenco, P. J., Corsetti, F. A., Kaufman, A. J. & Bottjer, D. J. 2008a. Environmental and diagenetic variations in carbonate associated sulfate: An investigation of CAS in the Lower Triassic of the western USA. Geochimica et Cosmochimica Acta, 72, 1570– 1582. Marenco, P. J., Corsetti, F. A., Hammond, D. E., Kaufman, A. J. & Bottjer, D. J. 2008b. Oxidation of pyrite during extraction of carbonate associated sulfate. Chemical Geology, 247, 124– 132. McCay, G. A., Prave, A. R., Alsop, G. I. & Fallick, A. E. 2006. Glacial trinity: Neoproterozoic Earth history within the British –Irish Caledonides. Geology, 34, 909–912. McFadden, K. A., Huang, J. et al. 2008. Pulsed oxidation and biological evolution in Ediacaran Doushantuo Formation. Proceedings of the National Academy of Sciences (USA), 105, 3197– 3202. McKirdy, D. M., Burgess, J. M. et al. 2001. A chemostratigraphic overview of the late Cryogenian interglacial sequence in the Adelaide Fold-Thrust Belt, South Australia. Precambrian Research, 106, 149– 186. Melezhik, V. A., Gorokhov, I. M., Kuznetsov, A. B. & Fallick, A. E. 2001. Chemostratigraphy of Neoproterozoic carbonates: implications for ‘blind dating’. Terra Nova, 13, 1 –11. Melezhik, V. A., Pokrovsky, B. G., Fallick, A. E., Kuznetsov, A. B. & Bujakaite, M. I. 2009. Constraints on the 87Sr/86Sr of Late Ediacaran seawater: insights from high-Sr limestones. Journal of the Geological Society, London, 166, 183–191. Miller, N. R., Stern, R. J., Avigad, D., Beyth, M. & Schilman, B. 2009. Cryogenian slate-carbonate sequences of the Tambien Group, Northern Ethiopia (I) ‘Pre-Sturtian’ chemostratigraphy and regional correlations. Precambrian Research, 170, 129– 156. Misi, A. & Veizer, J. 1998. Neoproterozoic carbonate sequences of the Una Group, Irece Basin, Brazil: chemostratigraphy, age and correlations. Precambrian Research, 89, 87 –100. Misi, A., Kaufman, A. J., Azmy, K., Dardenne, M. A., Sial, A. N. & de Oliveira, T. F. 2011. Neoproterozoic successions of the Sa˜o Francisco Craton, Brazil: The Bambuı´, Una, Vazante and Vaza Barris/ Miaba groups and their glacigenic deposits. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 509– 522. Nagy, R. M., Porter, S. M., Dehler, C. M. & Shen, Y. 2009. Biotic turnover driven by eutrophication before the Sturtian low-latitude glaciation. Nature Geoscience, 2, 415– 418. Nogueira, A. C. R., Riccomini, C., Sial, A. N., Moura, C. A. V., Trindade, R. I. F. & Fairchild, T. R. 2007. Carbon and strontium isotope fluctuations and paleoceanographic changes in the late Neoproterozoic Araras carbonate platform, southern Amazon craton, Brazil. Chemical Geology, 237, 191– 210. Ono, S., Wing, B., Johnston, D., Farquhar, J. & Rumble, D. 2006. Mass-dependent fractionation of quadruple stable sulfur isotopes as a new tracer of sulfur biogeochemical cycles. Geochimica et Cosmochimica Acta, 70, 2238–2252.

65

Pavlov, A. A. & Kasting, J. F. 2002. Mass-independent fractionation of sulfur isotopes in Archean sediments: strong evidence for an anoxic Archean atmosphere. Astrobiology, 2, 27 –41. Paytan, A., Kastner, M., Campbell, D. & Thiemens, M. H. 1998. Sulfur isotopic composition of Cenozoic seawater sulfate. Science, 282, 1459–1462. Paytan, A., Kastner, M., Campbell, D. & Thiemens, M. H. 2004. Seawater sulfur isotopic variations in the Cretaceous. Science, 304, 1663–1665. Pell, S. D., McKirdy, D. M., Jansyn, J. & Jenkins, R. J. F. 1993, Ediacaran carbon isotope stratigraphy of South Australia — an initial study. Transactions of the Royal Society of South Australia, 117, 153– 161. Pokrovskii, B. G., Melezhik, V. A. & Bujakaite, M. I. 2006. Carbon, oxygen, strontium, and sulfur isotopic compositions in late Precambrian rocks of the Patom Complex, central Siberia: Communication 1. Results, isotope stratigraphy, and dating problems. Lithology and Mineral Resources, 41, 450– 474. Poulton, S. W., Fralick, P. W. & Canfield, D. E. 2004. The transition to a sulphidic ocean 1.84 billion years ago. Nature, 431, 173– 177. Prave, A. R. 1999. Two diamictites, two cap carbonates, two d13C excursions, two rifts: The Neoproterozoic Kingston Peak Formation, Death Valley, California. Geology, 27, 339– 342. Prave, A. R., Fallick, A. E., Thomas, C. W. & Graham, C. M. 2009. A composite C-isotope profile for the Neoproterozoic of Scotland and Ireland. Journal of the Geological Society, London, 166, 845– 857. Rainbird, R. H., Jefferson, C. W. & Young, G. M. 1996. The early Neoproterozoic sedimentary Succession B of northwestern Laurentia: correlations and paleogeographic significance. Geological Society of America Bulletin, 108, 454–470. Raiswell, R., Buckley, F., Berner, R. A. & Anderson, T. F. 1988. Degree of pyritisation of iron as a paleoenvironmental indicator of bottom-water oxygenation. Journal of Sedimentary Petrology, 58, 812– 819. Rice, A. H. N., Edwards, M. B., Hansen, T. A., Arnaud, E. & Halverson, G. P. 2011. Glaciogenic rocks of the Neoproterozoic Smalfjord and Mortensnes Formations, Vestertana Group, E. Finnmark, Norway. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 593–602. Ries, J. B., Fike, D. A., Pratt, L. M., Lyons, T. W. & Grotzinger, J. P. 2009. Superheavy pyrite (d34Spyr . d34SCAS) in the terminal Proterozoic Nama Group, southern Namibia: a consequence of low seawater sulfate at the dawn of animal life. Geology, 37, 743– 746. Rooney, A. D., Selby, D., Houzait, J.-P. & Renne, P. R. 2010. Re– Os geochronology of a Mesoproterozoic sedimentary succession, Taoudeni basin, Mauritania: Implications for basin-wide correlations and Re– Os organic-rich sediments systematic. Earth and Planetary Science Letters, 289, 486–496. Rose, C. V. & Maloof, A. C. 2010. Testing models for post-glacial “cap dolostone” deposition: Nuccaleena Formation, South Australia. Earth and Planetary Science Letters, 296, 165–180. Rothman, D. H., Hayes, J. M. & Summons, R. E. 2003. Dynamics of the Neoproterozoic carbon cycle. Proceedings of the National Academy of Sciences (USA), 100, 124– 129. Sawaki, Y., Kawai, T. et al. 2010a. 87Sr/86Sr chemostratigraphy of Neoproterozoic Dalradian carbonates below the Port Askaig glaciogenic Formation, Scotland. Precambrian Research, 179, 150–164. Sawaki, Y., Ohno, T. et al. 2010b. The Ediacaran radiogenic Sr isotope excursion in the Doushantuo Formation in the Three Gorges area, South China. Precambrian Research, 176, 46 –64. Saylor, B. Z., Kaufman, A. J., Grotzinger, J. P. & Urban, F. 1998. A composite reference section for Terminal Proterozoic strata of southern Namibia. Journal of Sedimentary Research, 68, 1223–1235. Schrag, D. P., Berner, R. A., Hoffman, P. F. & Halverson, G. P. 2002. On the initiation of a snowball Earth. Geochemistry, Geophysics, Geosystems, 31, doi: 10.1029/2001GC000219. Schidlowski, M., Eichmann, R. & Junge, C. E. 1975. Precambrian sedimentary carbonates: carbon and oxygen isotope geochemistry

66

G. P. HALVERSON & G. SHIELDS-ZHOU

and implications for the terrestrial oxygen budget. Precambrian Research, 2, 1 –69. Schro¨der, S., Schreiber, C., Amthor, J. E. & Matter, A. 2004. Stratigraphy and environmental conditions of the terminal Neoproterozoic – Cambrian Period in Oman: evidence from sulphur isotopes. Journal of the Geological Society, London, 161, 489– 499. Shen, B., Xiao, S., Zhou, C., Kaufman, A. J. & Yuan, X. 2010. Carbon and sulfur isotope chemostratigraphy of the Neoproterozoic Quanji Group of the Chaidam Basin, NW China: Basin stratification in the aftermath of an Ediacaran glaciation post-dating the Shuram event? Precambrian Research, 177, 241– 252. Shields, G. A. 1999. Working towards a new stratigraphic calibration scheme for the Neoproterozoic –Cambrian. Eclogae Geologicae Helvetiae, 92, 221– 233. Shields, G. A. 2005. Neoproterozoic cap carbonates: a critical appraisal of existing models and the plumeworld hypothesis. Terra Nova, 17, 299– 310. Shields, G. A. 2007. A normalised seawater strontium isotope curve: possible implications for Neoproterozoic –Cambrian weathering rates and further oxygenation of the Earth. Earth, 2, 35 – 42. Shields, G. & Veizer, J. 2002. Precambrian marine carbonate isotope database: Version 1.1. Geochemistry, Geophysics, Geosystems, 3, doi: 10.1029/2001GC000266. Shields, G., Stille, P., Brasier, M. D. & Atudorei, N.-V. 1997. Stratified oceans and oxygenation of the late Precambrian environments: a post glacial geochemical record from the Neoproterozoic. Terra Nova, 9, 218–222. Shields, G. A., Strauss, H., Howe, S. S. & Siegmund, H. 1999. Sulphur isotope compositions of sedimentary phosphorites from the basal Cambrian of China — implications for Neoproterozoic – Cambrian biogeochemical cycling. Journal of the Geological Society, London, 156, 943 –957. Shields, G., Kimura, H., Yang, J. & Gammon, P. 2004. Sulphur isotopic evolution of Neoproterozoic – Cambrian seawater: new francolitebound sulphate S data and critical appraisal of the existing record. Chemical Geology, 204, 163–182. Sial, A. N., Dardenne, M. A. et al. 2010. The Sa˜o Francisco Paleocontinent. In: Gaucher, C., Sial, A. N., Halverson, G. P. & Frimmel, H. (eds) Neoproterozoic– Cambrian Tectonics, Global Change and

Evolution: A focus on Southwestern Gondwana. Developments in Precambrian Geology, Elsevier, Dordrecht, 16, 31– 69. Strauss, H. 1993. The sulfur isotopic record of Precambrian sulfates: new data and a critical evaluation of the existing record. Precambrian Research, 63, 225– 246. Swanson-Hysell, N. L., Rose, C. V., Calmet, C., Halverson, G. P., Hurtgen, M. T. & Maloof, A. C. 2010. Cryogenian glaciation and onset of carbon-isotope decoupling. Science, 328, 608–611. Tucker, M. E. 1986. Carbon isotope excursions in Precambrian/ Cambrian boundary beds. Nature, 319, 48 –50. Turner, E. C. 2009. Lithostratigraphy and stable isotope values of the early Neoproterozoic Gypsum Formation (Little Dal Group), Mackenzie Mountains Supergroup, NWT. NWT Open Report, 2009-002. Veizer, J., Compston, W., Clauer, N. & Schidlowski, M. 1983. 87 Sr/86Sr in late Proterozoic carbonates: evidence for a ‘mantle’ event at 900 Ma. Geochimica et Cosmochimica Acta, 47, 295– 302. Walter, M. R., Veevers, J. J., Calver, C. R., Gorjan, P. & Hill, A. C. 2000. Dating the 840– 544 Ma Neoproterozoic interval by isotopes of strontium, carbon, and sulfur in seawater and some interpretive models. Precambrian Research, 100, 371–433. Williams, G. E. 1979. Sedimentology, stable-isotope geochemistry and palaeoenvironment of dolostones capping late Precambrian glacial sequences in Australia. Journal of the Geological Society of Australia, 26, 377–386. Workman, R. K., Grotzinger, J. P. & Hart, S. R. 2002. Constraints on Neoproterozoic ocean chemistry from C and B analyses of carbonates from the Witvlei and Nama groups, Namibia. In: Goldschmidt Conference Proceedings (Davos, Switzerland). Xiao, S., Bao, H. et al. 2004. The Neoproterozoic Quruqtagh Group in eastern Chinese Tianshan: evidence for a post-Marinoan glaciation. Precambrian Research, 130, 1 –26. Xu, B., Xiao, S. et al. 2009. SHRIMP zircon U– Pb age constraints on Neoproterozoic Quruqtagh diamictites in NW China. Precambrian Research, 168, 247– 258. Yoshioka, H., Asahara, Y., Tojo, B. & Kawakami, S. 2003. Systematic variations in C, O, and Sr isotopes and elemental concentrations in Neoproterozoic carbonates in Namibia: implications for glacial to interglacial transition. Precambrian Research, 124, 69 –85.

Chapter 5 Chemical sediments associated with Neoproterozoic glaciation: iron formation, cap carbonate, barite and phosphorite PAUL F. HOFFMAN1,2 *, FRANCIS A. MACDONALD1 & GALEN P. HALVERSON3,4 1

Department of Earth and Planetary Sciences, Harvard University, 20 Oxford Street, Cambridge, MA, 02138, USA 2

School of Earth and Ocean Sciences, University of Victoria, Box 1700, Victoria, BC V6W 2Y2, Canada

3

School of Earth and Environmental Sciences, The University of Adelaide, North Terrace, Adelaide, SA 5005, Australia 4

Present address: Department of Earth and Planetary Sciences, McGill University, 3450 University Street, Montre´al, PQ H3A 2K6, Canada *Corresponding author (e-mail: [email protected]) Abstract: Orthochemical sediments associated with Neoproterozoic glaciation have prominence beyond their volumetric proportions because of the insights they provide on the nature of glaciation and the records they hold of the environment in which they were precipitated. Synglacial Fe formations are mineralogically simple (haematite jaspilite), and their trace element spectra resemble modern seawater, with a weaker hydrothermal signature than Archaean– Palaeoproterozoic Fe formations. Lithofacies associations implicate subglacial meltwater plumes as the agents of Fe(II) oxidation, and temporal oscillations in the plume flux as the cause of alternating Fe- and Mn-oxide deposits. Most if not all Neoproterozoic examples belong to the older Cryogenian (Sturtian) glaciation. Older and younger Cryogenian (Marinoan) cap carbonates are distinct. Only the younger have well-developed transgressive cap dolostones, which were laid down during the rise in global mean sea level resulting from ice-sheet meltdown. Marinoan cap dolostones have a suite of unusual sedimentary structures, indicating abnormal palaeoenvironmental conditions during their deposition. Assuming the meltdown of ice-sheets was rapid, cap dolostones were deposited from surface waters dominated by buoyant glacial meltwater, within and beneath which microbial activity probably catalysed dolomite nucleation. Former aragonite seafloor cement (crystal fans) found in deeper water limestone above Marinoan cap dolostones indicates carbonate oversaturation at depth, implying extreme concentrations of dissolved inorganic carbon. Barite is associated with a number of Marinoan cap dolostones, either as digitate seafloor cement associated with Fe-dolomite at the top of the cap dolostone, or as early diagenetic void-filling cement associated with tepee or tepee-like breccias. Seafloor barite marks a redoxcline in the water column across which euxinic Ba-rich waters upwelled, causing simultaneous barite titration and Fe(III) reduction. Phosphatic stromatolites, shrub-like structures and coated grains are associated with a glacioisostatically induced exposure surface on a cap dolostone in the NE of the West African craton, but this appears to be a singular occurrence of phosphorite formed during a Neoproterozoic deglaciation.

The association of chemical sediments with Neoproterozoic glacial deposits has long been known. In fact, Neoproterozoic glaciation was discovered in a number of regions as a result of the search for economic Fe and Fe–Mn deposits (e.g. SW Brazil, NW Canada, Namibia). Glacial associated chemical sediments provide critical evidence concerning the nature of Neoproterozoic glaciations and their aftermaths, in the form of geochemical records of the waters from which they were precipitated, and indirectly the atmosphere with which those waters interacted. Here, we briefly review the distribution, lithological association, sedimentology and palaeoenvironmental significance of orthochemical sediments deposited during Neoproterozoic glaciations and deglaciations. They include sedimentary Fe and Fe –Mn deposits and cap-carbonate sequences, of which ‘cap dolostones’ sensu stricto form the basal transgressive systems tracts. Barite (BaSO4) and phosphorite mineralization occurs locally within Marinoan cap dolostones. More information on geologic setting, stratigraphic relations, geochemistry and isotopic characteristics is given in the appropriate regional chapters. Palaeogeographies of the deposits (Fig. 5.1) are based on global model maps for 715 Ma (Sturtian) and 635 Ma (Marinoan), created by the Tectonics Special Research Centre in Perth, Western Australia (Li et al. 2008; Hoffman & Li 2009). The maps were constructed on the basis of palaeomagnetic constraints, the mantle plume record and palaeocontinental tectonic genealogy, factors that are largely independent of palaeoclimate. The Sturtian glaciation persisted until 659 + 6 Ma in its type area (Fanning & Link 2008), by which time the model palaeogeography (Li et al. 2008) was intermediate between 715 and 635 Ma (Fig. 5.1).

Fe and Fe – Mn deposits Distribution in time and space The palaeogeographic distribution of synglacial Fe and Fe –Mn sedimentary deposits is shown in Figure 5.1 (Table 5.1). The economic Fe – Mn ores of the Jacadigo Group in the Urucum District of southwestern Brazil and eastern Bolivia (see below) were tentatively assigned to the terminal Cryogenian (Marinoan) glaciation (Hoffman & Li 2009) through correlation with the Puga diamictite in the adjacent Paraguay fold belt (Alvarenga & Trompette 1992; Trompette 1994; Trompette et al. 1998). The Puga diamictite is capped by a diagnostic Marinoan-type cap-carbonate sequence (Nogueira et al. 2003, 2007; Trindade et al. 2003; Font et al. 2005, 2006; Alvarenga et al. 2008). However, no cap carbonate is preserved in the Jacadigo Group, which occupies a transverse rift-basin on the cratonic foreland of the Paraguay fold belt (Trompette et al. 1998). The true age of the Jacadigo Group, Sturtian or Marinoan, is unknown. Similarly, Fe formation in the glaciogenic Rizu Formation of central Iran was tentatively assigned to the Marinoan glaciation (Hoffman & Li 2009) on the basis of a reported cap dolostone (Kianian & Khakzad 2008), but details are lacking. The reassignment of the Jacadigo and Rizu Fe formations to the Sturtian glaciation (Fig. 5.1) is therefore permissible, but arbitrary. In either palaeogeographic reconstruction (Fig. 5.1), the Fe- and Fe –Mn deposits formed disproportionately within 308 of the palaeoequator, and at the margins of ocean basins that were both internal (Jiangkou, Sturt, Rapitan, Surprise, Numees and Chuos) and external (Rizu, Tany) to the fragmented Rodinia supercontinent.

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 67– 80. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.5

68

P. F. HOFFMAN ET AL.

(a)

“ Blasskranz Nantuo Ghaub Petite Tereeken Landrigan Ta Olympic Ir Blaini

In

Au

635 Ma

Ka

SC

Ma

Jiangkou Blaubekker Areyonga

Si

TM

Laur

CA

RP Elatina Co Cottons SF Fiq Ar Shareef Bondo Am WA Superieure Kodjari Ba Bakoye Jbeliat Palestina Puga

(b)

Namuskluft Wildrose Vreeland Stelfox Khongoryn Pod’em Marnya Dzemkukan NC

Wilsonbreen Storeelv Stralinchy-Reelan Smalfjord

715 Ma

Numees Surprise Pocatello Toby Rapitan Tindir Maikhan Ul Chivida Kharlukhtakh NC

Ta

Si In

SC

TM

Au Rizu CA Sturt Ma Julius River Laur Konnarock Ka Sv Gubrah RP Tambien Ar Sc Co Ayn Grand Am SF Chuos Ba Akwokwo WA Inferieure Jequitai Chiquerio Jacadigo

Hula Hula Petrovbreen Ulveso Port Askaig Tany Baykonur

Lithological associations Impure to pure haematite –jaspilite occurs within intervals of parallel-laminated argillite, ferruginous argillite and Mn-oxides. Those identified in Figure 5.1 host outsize clasts (lonestones) of intrabasinal and extrabasinal (crystalline basement) derivation, generally interpreted as ice-rafted debris. Many occur in contact with massive to poorly stratified diamictite, inferred to be

Fig. 5.1. Palaeogeographic maps for 715 Ma and 635 Ma (Hoffman & Li 2009) showing distribution of (a) Sturtian and (b) Marinoan glaciogenic deposits (open stars), and synglacial Fe or Fe– Mn deposits (black stars). Palaeocontinents: Ar, Arabia; Am, Amazonia; Au, Australia; Ba, Baltica; CA, Chukotka–Arctic Alaska; Co, Congo; In, India; Ir; central Iran; Ka, Kalahari; Laur, Laurentia (including Sc, Scotland and Sv, Svalbard); Ma, Mawson; NC, North China; SF; Sa˜o Francisco; RP, Rio de la Plata; SC, South China; Si, Siberia; Ta, Tarim; TM, Tuva-Mongolia; WA, West Africa. Palaeocontinents constrained by penecontemporaneous palaeomagnetic data have heavy lines. The Sturtian glaciation persisted to c. 660 Ma in some areas (South Australia and western USA), by which time the model palaeogeography was intermediate between 715 and 635 Ma. Assignment of Fe formations in the Jacadigo Group (Am) and Rizu Formation (Ir) to the Sturtian glaciation is uncertain. Correlation of Fe formation in the Gariep Belt (Ka) with the glaciogenic Numees Formation follows Macdonald et al. (2011).

peri- or subglacial in origin based on the presence of faceted, striated and/or preferentially oriented clasts. Sizeable Fe and Fe –Mn deposits occur in intervals that are sandwiched between composite diamictite horizons, locally of great thickness (e.g. Braemar, Holowilena, Jakkalsberg, Sayunei, Tany). Smaller deposits are intimately interfingered with diamictite (e.g. Braemar, Chuos). Still other deposits lack diamictite but the presence of dropstones of intra- and extrabasinal origin suggest a

Table 5.1. Cryogenian glacigenic Fe and Fe– Mn deposits Palaeocontinent

Location

Host strata

References

Amazonia Australia Baltica Congo Kalahari Laurentia Laurentia* Lut (central Iran) South China Tuva-Mongolia

Urucum Braemar Middle Urals Damara Belt Gariep Belt NWT-Yukon Death Valley Kerman Yangtze platform Erzin

Jacadigo Yudnamutana Vil’va Chuos Numees Rapitan Surprise Rizu Jiangkou Ongoluk

Dorr (1945), Urban et al. (1992), Klein & Ladeira (2004) Whitten (1970), Lottermoser & Ashley (2000) Chumakov (1992, 2007) Martin (1965a), Badenhorst (1988), Clifford (2008) Macdonald et al. (2011) Young (1976), Yeo (1981, 1986), Klein & Beukes (1993) Corsetti & Kaufman (2003) Kianian & Khakzad (2008) Jiafu et al. (1987) Ilyin (2009)

*In the collisional model for Laramide orogeny (Hildebrand 2009), the Death Valley area as well as most other Cryogenian glacial formations in the western USA and British Columbia are considered to be exotic with respect to Laurentia before the Late Cretaceous.

CHEMICAL SEDIMENTS ASSOCIATED WITH NEOPROTEROZOIC GLACIATION

subaqueous proglacial setting (e.g. Urucum). Parallel-laminated facies hosting haematite –jaspilite are commonly interpreted as interglacial or interstadial, but they might alternatively represent maximum glacial stages, when outlet glaciers were blocked by thick multi-annual sea ice (Dowdeswell et al. 2000).

Geochemical characteristics Compared with Archaean –Palaeoproterozoic banded Fe formations, Cryogenian Fe deposits are mineralogically simple, consisting of haematite, chert (jasper) and minor carbonate. Fe contents range up to 50% Fe2O3. Organic contents are negligible and carbonates are moderately depleted in 13C, suggesting that organic matter originally present was respired with Fe(III) serving as electron acceptor. Normalized rare earth element profiles are distinct from Archaean–Palaeoproterozoic Fe formations, being more strongly depleted in light rare earth elements, and having much weaker positive Eu anomalies or none at all (Klein & Beukes 1993; Graf et al. 1994; Klein & Ladeira 2004; Klein 2005). The Cryogenian profiles are more similar to those of modern seawater, and hydrothermal contributions are more diluted than in more ancient Fe formations.

Notable examples Jacadigo Group, Urucum District, Mato Grosso do Sul, Brazil. Eco-

nomic Fe–Mn oxide ores of the Jacadigo Group occur in stratigraphically isolated fault blocks in the Urucum District of southwestern Brazil and adjacent Bolivia (Dorr 1945; Walde et al. 1981; Urban et al. 1992; Trompette et al. 1998). A lower Urucum Formation, comprising conglomerate, sandstone and black shale, is overlain by the Santa Cruz Formation, a sequence of Mn ore horizons, chiefly composed of cryptomelane (K2Mn8O16), interspersed with ferruginous sandstone and up to 270 m of haematite –jaspilite. All units of the Santa Cruz Formation carry dropstones of basement and subordinate carbonate lithologies. They are interpreted to be ice-rafted and the Jacadigo Group to have been deposited in a partly ice-covered fjord-like basin (Urban et al. 1992). The alternation of Mn and Fe ores is attributed to cycles of glacial advance and retreat. Enhanced fluxes of oxygenated subglacial meltwater during glacial retreats favoured Mn precipitation in ice-covered anoxic parts of the basin. Reduced fluxes of subglacial meltwater during glacial advances favoured Fe deposition (Urban et al. 1992). The Jacadigo Group lies in the cratonic foreland of southern Amazonia, close to the Paraguay fold belt of Early Cambrian (Brasiliano) age within which glaciogenic diamictite of the Puga Formation (Alvarenga & Trompette 1992) is sharply overlain by a typical Marinoan-type cap dolostone (Nogueira et al. 2003, 2007; Trindade et al. 2003; Font et al. 2005, 2006; Alvarenga et al. 2008). No cap carbonate is preserved on the Jacadigo Group, however, so the correlation of the Jacadigo and Puga groups is uncertain, as is, therefore, the age of the Urucum Fe– Mn deposits. Rapitan Group, northern Canadian Cordillera. Haematite– jaspilite and ferruginous argillite occur within glaciogenic diamictite of the Rapitan Group and correlatives discontinuously for c. 800 km along the strike of the Ogilvie and Mackenzie mountains (Young 1976; Yeo 1981, 1986; Eisbacher 1985; Klein & Beukes 1993; Macdonald et al. 2011; Hoffman & Halverson 2011; Macdonald & Cohen 2011). The glacial onset coincided with a major flood basalt episode across Arctic Laurentia, bimodal representatives of which underlie and intercalate the basal Rapitan diamictite in the Ogilvie Mountains, constraining its age at 717 Ma (Macdonald et al. 2010). Haematite –jaspilite is up to 120 m thick in an iceproximal section (Iron Creek, NW Mackenzie Mtns), where it is sandwiched between composite diamictite complexes hundreds

69

of metres thick. Jaspilite first appears as 2-m-thick septa between individual diamictite bodies in the upper part of the lower complex (Mount Berg Formation). In a more distal section (Hayhook Lake, SE Mackenzie Mountains), 14 m of ferruginous argillite and haematite –jaspilite conformably overlie 650 m of maroon-coloured siltstones with graded sandy beds and a sprinkling of small dropstones, some of which were redeposited as debrites. The jaspilite itself contains rounded boulders of quartzmonzonite and is overlain disconformably by the upper diamictite complex (Shezal Formation). The basal part of the otherwise olive-coloured Shezal diamictite acquired its maroon colour and numerous jaspilite clasts from the underlying Sayunei Formation. Fe-isotope and Ce anomaly profiles suggest that the haematite – jaspilite records subsidence of the basin floor across a redoxcline in the water column (Klein & Beukes 1993; Halverson et al. 2011). Jakkalsberg Member (Numees Formation, Port Nolloth Group), Gariep Belt, Namibia and South Africa. Haematite- and magnetite –jaspilite

with basement-derived lonestones make up the Jakkalsberg Member of the Numees Formation in thrust sheets intersected by the Orange River (Frimmel & von Veh 2003). The Numees Formation has long been assumed to be the younger of two glaciogenic horizons in the Gariep Belt, the older being the Kaigas Formation (Frimmel 2008, 2011). Recently, a mid-Ediacaran age for the Numees glaciation has been advocated from carbonate Pb/Pb dating (Fo¨lling et al. 2000), 87Sr/86Sr ratios of 0.7082– 0.7085 (Fo¨lling & Frimmel 2002), correlations with dated South American strata (Frimmel 2004, 2008) and micropalaeontological findings (Gaucher et al. 2005). The microfossils do not appear to be diagnostic, however, and the Sr isotope ratios (0.001 higher than typical Marinoan values; Halverson et al. 2007) are potentially attributable to diagenesis, as are the Pb/Pb dates. Subsequent remapping of the least-deformed and least-allochthonous sections prompted a reassessment of stratigraphic correlations between the autochthon and thrust sheets within the belt (Macdonald et al. 2011). The younger autochthonous diamictite, previously correlated with the Numees Formation, is now proposed as an independent glaciogenic horizon, the Namuskluft diamictite (Macdonald et al. 2011). It is capped by a pale dolostone (Dreigratberg Member, previously conflated with the Bloeddrif Member, of the Holgat Formation) containing hallmark features of Marinoan cap dolostones –sheet-crack cements, tubestone stromatolites and giant wave ripples (Hoffman & Macdonald 2010). In contrast, the jaspilite-bearing Numees diamictite is overlain by a dark microbial limestone with roll-up structures, characteristic of Sturtian cap carbonates (see below). Accordingly, the Jakkalsberg jaspilite is probably older Cryogenian (Sturtian) in age, and the Kaigas Formation may represent a preSturtian (c. 0.74 Ga) diamictite of uncertain origin (Macdonald et al. 2011). Central Flinders Ranges (Holowilena) and Nackara arc (Braemar), South Australia. Like the more ice-proximal (Iron Creek) section

in NW Canada, haematite – jaspilite and associated ferruginous argillite in South Australia are sandwiched between a conformably underlying diamictite complex (Pualco ‘Tillite’) and a disconformably overlying complex (Wilyerpa Formation) dominated by diamictite (Whitten 1970; Preiss 1987; Lottermoser & Ashley 2000). In the Flinders Ranges, ferruginous argillite hosts sporadic lenses of boulders, some of which are grooved and striated, ferruginous diamictite, and rare lenses of haematite – jaspilite with rafted dropstones. In the more basinward Nackara arc, haematite –jaspilite is interrupted by thin units of ferruginous diamictite. Sorted sandstones occur within the diamictite complexes in both areas. Correlation of the Holowilena and Braemar jaspilites is uncertain, but they and the bounding Pualco and Wilyerpa diamictites are referred to the Sturtian glaciation (Preiss 1987).

70

P. F. HOFFMAN ET AL.

Cap-carbonate sequences The continuous layers of carbonate that blanket Cryogenian glaciogenic sequences or their equivalent disconformities are called ‘cap’ carbonates. Clearly associated with syndeglacial flooding (Kennedy 1996; Bertrand-Sarfati et al. 1997; Hoffman et al. 2007), they occur on virtually every palaeocontinent and even in siliciclastic-dominated successions (Table 5.2). Lithologically and isotopically, Sturtian and Marinoan cap carbonates are distinct, both from each other (Kennedy et al. 1998; Halverson & Shields 2011) and from most other Neoproterozoic carbonates (Hoffman 2011). Hoffman & Schrag (2002) proposed that depositional sequences related to Neoproterozoic syndeglacial flooding be called ‘cap-carbonate sequences’. ‘Cap dolostones’, sensu stricto, are the transgressive tracts of cap-carbonate sequences and feature a suite of idiosyncratic sedimentary features (e.g. sizegraded peloids, sheet-crack cements, tubestone stromatolites, giant wave ripples, primary and early diagenetic barite). Transgressive cap dolostones have rarely been documented in Sturtian cap-carbonate sequences and are thin (,1.0 m) where present (Smith et al. 1994). Cap-carbonate sequences have aroused intense interest because they are unique to Proterozoic glaciations and record physical, chemical and biological conditions during and immediately after global deglaciation (Aitken 1991; Grotzinger &

Knoll 1995; Kennedy, 1996; Hoffman et al. 1998, 2007; James et al. 2001; Kennedy et al. 2001; Higgins & Schrag 2003; Ridgwell et al. 2003; Trindade et al. 2003; Shields 2005; Allen & Hoffman 2005; Font et al. 2005, 2006; Hurtgen et al. 2006; Jiang et al. 2006; Bao et al. 2008; Le Hir et al. 2009; Hoffman 2011).

Marinoan-type (basal Ediacaran) cap-carbonate sequences The extent and uniqueness of Marinoan cap dolostones was first appreciated in Australia (Dunn et al. 1971; Rankama 1973) and the presence of barite in cap dolostones was first recognized in West Africa (Deynoux & Trompette 1976). The smooth and abrupt, yet conformable, contact between glaciogenic and related detritus and cap dolostones is so distinctive and widespread that it was selected as the basis of the Global Stratotype Section and Point (GSSP) for the Ediacaran Period (Knoll et al. 2006), the first Period boundary to be defined strictly on lithologic grounds. A tuff at the top of the presumed correlative cap dolostone in South China yields a 238U – 206Pb (IDTIMS) zircon date of 635.2 + 0.4 Ma (Condon et al. 2005). A statistically indistinguishable 238U – 206Pb (SHRIMP) zircon date of 636.3 + 4.9 Ma was obtained from a tuff near the base of the underlying Nantuo glaciogenic diamictite (Zhang et al. 2008). The presumed

Table 5.2. Average thickness and idiosyncratic features of Marinoan cap dolostones in numerical order of abundance Palaeocontinent Amazonia Arabia Arctic Alaska Australia Australia Australia Australia Baltica Congo Congo Congo India Kalahari Kalahari Kalahari Laurentia Laurentia Laurentia Laurentia Laurentia Laurentia* Siberia South China Tarim Tuva-Mongolia Tuva-Mongolia West Africa West Africa West Africa

Glaciation Puga Gadir Manqil (Fiq) No deposits Olympic Elatina Cottons Landrigan Smalfjord Ghaub Petit Conglome´rat Upper Tilloid Blaini Namuskluft Bla¨sskranz No deposits Stelfox (Ice Brook) Storeelv Wilsonbreen Stralinchy-Reelan Stelfox (Ice Brook) Wildrose Ulyakha Nantuo Tereeken Khongoryn Khesen Fersiga Jbe´liat Banboli

Cap dolostone (ref.) Mirassol d’Oeste (1)† Hadash (2) Nularvik (3)† Mount Doreen (4) Nuccaleena (5) Cumberland Creek (6) Lower Stein (7) Lower Nyborg (8) Keilberg (9)† Calcaire rose´ (10) C1 Dolomie rose´ (11) Upper Blaini (12) Dreigratberg (13) Tsabesis (14) Bildah (15)† Ravensthroat (16)† Lower Canyon (17) Lower Dracoisen (18) Cranford (19) Hard Luck (20) Noonday (21) Lower Ozerki (22) Lower Doushantuo (23) Lower Zhamoketi (24) Ol (25)† Baxha (26) Oued Djouf (27) Amogjar (28) Mid Sud-Banboli (29)

Metres 24 4.5 35 4 5 6 8.5 5 38 10 10 10 25 21 80 12 10 10 4 4 175 35 4 6 15 4 6 5 1.5

LAC

PEL

GWR

p

p

p

– p p

– p p

– p p

– p

– p p p p p

SCC – – – p p

TBS p – p

– p

– – p p

– – – – – p

p

p

p

p

– – p

– – p

– p p

– p p

– p

– p

– – p

– – p

– – p p

– – – – p

– – p

– p p p

– – – p

– – – p

– – p p

– – – –

– p

p p p p p p p – p p

– – p

– – p – p p p – p

– – – p – – – –

TPB

DGB

SFB

– – – – – – – – –

– – – – – – – – –

– – – p

– – – – – – p

– – – – – – – – – – p

– – – p – – – p p p

– p – – p p

– – – p –

– – – p – – – – – – – – p p – – –

*In the collisional model for Laramide orogeny (Hildebrand 2009), Wildrose-Noonday strata and most other Cryogenian glacial formations in the western USA and British Columbia are considered to be exotic with respect to Laurentia before the Late Cretaceous. † Former-aragonite seafloor cement above cap dolostone. LAC, low-angle cross-laminae; PEL, peloids; GWR, giant wave ripples; SCC, sheet-crack cements; TBS, tubestone stromatolite; TPB, tepee breccia; DGB, diagenetic barite; SFB, seafloor barite. References: (1) Nogueira et al. (2003); (2) Allen et al. (2004); (3) Macdonald et al. (2009); (4) Kennedy (1996); (5) Plummer (1978); (6) Calver & Walter (2000); (7) Corkoron (2007); (8) Edwards (1984); (9) Hoffman et al. (2007); (10) Cahen & Lepersonne (1981); (11) Cahen (1950); (12) Kaufman et al. (2006); (13) Hoffman & Macdonald (2010); (14) PFH observations; (15) Hegenberger (1993); Prave et al. (2011); (16) James et al. (2001); (17) Hambrey & Spencer (1987); (18) Halverson et al. (2004); (19) McCay et al. (2006); (20) Macdonald & Cohen (2011); (21) Corsetti & Grotzinger (2005); (22) Sovetov & Komlev (2005); (23) Jiang et al. (2006); (24) Xiao et al. (2004); (25) Macdonald (2011); (26) Macdonald & Jones (2011); (27) Bertrand-Sarfati et al. (1997); (28) Shields et al. (2007); (29) Ne´de´lec et al. (2007).

CHEMICAL SEDIMENTS ASSOCIATED WITH NEOPROTEROZOIC GLACIATION

correlative Ghaub diamictite in Namibia has an indistinguishable 238 U – 206Pb (IDTIMS) zircon date of 635.6 + 0.5 Ma (Hoffmann et al. 2004). Marinoan cap dolostones are therefore assumed to have been deposited at 635 Ma.

created during Cryogenian glaciations by net erosion, compaction and tectonic subsidence. Glacial palaeofjords, for example, contain expanded cap-carbonate sequences compared with adjacent uplands (Fig. 5.3). Similarly, areas undergoing rapid tectonic subsidence like the Otavi platform, on which the Sturtian and Marinoan cap-carbonate sequences are each 250–400 m thick, contrast with areas like the Taoudeni Basin of the West African craton (Deynoux 1985; Shields et al. 2007), where subsidence rates were low and cap-carbonate sequences are highly condensed, outside of incised palaeovalleys (Fig. 5.3). Condensation may also occur in fully marine settings where sedimentation rates were inadequate to fill the existing accommodation space. On the distal foreslope of the Otavi platform, for example, drowning at the terminal deglaciations was effectively permanent (Halverson et al. 2005). The cause of condensation of the lower Doushantuo Formation cap-carbonate sequence of South China (low sedimentation rate or lack of accommodation?) depends on whether tepeelike breccias and associated early diagenetic barite mineralization in the cap dolostone are submarine (Jiang et al. 2006) or subaerial (Zhou et al. 2010) in origin, respectively.

Distribution and thickness. The median and average thicknesses of 29 Marinoan cap dolostones on 15 palaeocontinents (Table 5.2) are 9 m and 18 m, respectively (Hoffman et al. 2007). As there are large facies-controlled variations in thickness in some areas, these estimates are based on area-weighted average thicknesses, where data permit. On the Otavi platform (Congo palaeocontinent) of Namibia, for example, the Keilberg cap dolostone is 6– 10 m thick on the distal foreslope, up to 100 m on the upper foreslope, up to 75 m on the raised outer platform, and 20– 25 m on the deepened inner platform (Fig. 5.2). In general, cap dolostones are thickest on palaeotopographic highs and thinnest in lows, converse to underlying glacial deposits. Observed inner-shelf deepening is characteristic of glaciated margins (Anderson 1999). If cap dolostones averaged 18 m in thickness over 20% of the present continental surface area (including shelves), they would contain c. 2.6  105 Pg (Gt) of C. This dwarfs the 3.8  104 Pg of dissolved inorganic C in the present ocean. According to the palaeogeography in Figure 5.1a, the maximum thickness of cap dolostones increased from 5 m above 408 palaeolatitude, to 10 m above 308, 25 m above 208, 38 m above 108 and 175 m at the palaeoequator (Hoffman & Li 2009).

Transgressive cap dolostones in expanded sequences. Cap dolostones

are typically pale pinkish (yellowish-grey weathering) dolomicrites, with palimpsest micro- to macro-peloidal textures indicating deposition as silt and sand- and granule-sized aggregates. Those in expanded sequences lack the diagenetic overprint experienced by those in condensed sequences, which we treat separately. The former (e.g. Cumberland Creek, lower Canyon, lower Dracoisen, Dreigratberg, Keilberg, Mirassol d’Oeste, Mount Doreen, Nuccaleena, Ol, lower Pertatataka, Ravensthroat, Sentinel Peak, lower Stein, Tsabesis, and lower Zhamoketi; see Table 5.2) are invariably laminated, with each lamina representing a sedimentation unit composed of well-sorted, graded or reverse-graded peloids, up to 3 mm in diameter, terminating with a micropeloidal drape (Aitken 1991; Kennedy 1996; Calver & Walter 2000; James

Expanded and condensed sequences. The decompacted thickness of

a cap-carbonate sequence (as distinct from a cap dolostone), minus the effect of sediment loading, is a measure of the accommodation created before and during its deposition. The lowering and raising of global mean sea level attending the growth and decay of ice sheets creates no permanent accommodation if global ice-sheet volume was the same before and after a glaciation. Permanent accommodation (i.e. outlasting isostatic adjustments) was

S

Distal Slope

Upper Slope

Outer Platform

5 km

5 km

71

N

Inner Platform 100 km

80

LEGEND Maieberg Formation

(e) δ C

(g)

(f)

13

-5.0

pink marly limestone rhythmite

20

20 0.0

60 0

100

(d)

0 -5.0

with sea-floor crystal fans

40 -5.0

Keilberg Member cap dolostone

δ13C

0.0

δ13C

0.0

P4017

G2008

80

marly micropeloidal dololutite turbidites peloidal dolarenite, giant wave ripples

20

as above, swaley low-angle crossbedding

60

LEGEND

tubestone stromatolitic dolostone peloidal dolarenite, sheet-crack cements

0 -5.0

marly micropeloidal dolostone turbidites

40

δ13C

0.0

P6005

δ 13 C carb sediment δ 13 C carb sheet-crack cement

10 -5.0 -5.0

0.0

10

(a)

0.0

10

(c)

(b)

20

stratified proglacial detrital carbonate with ice-rafted debris

5

Ombaatjie Formation cycle b7-b8

5 5

0

0 0 -5.0 0 δ13C

δ13C

Bethanis Member of the Ghaub Formation

δ13C

P7016

peritial cycles of dolostone ribbonite, stromatolite, grainstone and microbialaminite)

0.0 P7009

P7002

P1607

Fig. 5.2. Representative lithologic columns and d13C profiles of the Keilberg cap dolostone (Marinoan) in contiguous palaeobathymetric zones of the Otavi platform, northern Namibia (Hoffman et al. 2007). The datum for platform columns (e– g) is the inflection point (arrows) in isotopic profiles. Each column records part of an overall sigmoidal profile, implying diachronous deposition during progressive marine inundation. The systematic nature of profiles rules out a detrital origin for dolomite and a diagenetic origin for d13C values. Note that values for isopachous dolomite sheet-crack cement (crosses) in (b) and (c) are indistinguishable from host dolopelmicrite (dots).

72

P. F. HOFFMAN ET AL.

measured sections ( )

(m) SW

NE

0

–50

–100 0

5

10

15 km

Teniagouri Group Atar West African craton

6

green and red shale, siltstone aquamarine chert (silexite), shale violet and green limestone

Jbeliat Group

5 4 3 2 1

dolomitic quartz sandstone, granulestone reddish siltstone grading to fine sandstone cap dolostone, tepee breccia, barite crusts pebbly regolith, sand-wedge polygons terrestrial lodgement tillite and outwash

CCS

Mesoproterozoic Atar, Tifounke, etc. groups microbial limestone, dolostone; fine clastics

Fig. 5.3. Stratigraphic cross-section of the glaciogenic Jbe´liat Group (Marinoan?) based on sections measured along Atar Cliff, Mauritania, by present authors PFH and GPH and Adam C. Maloof (Hoffman & Schrag 2002). The cap-carbonate sequence (CCS) includes cap dolostone (unit 3) and regressive highstand deposits (units 4 and 5) preserved only in palaeovalleys. The datum is the top of the cap-carbonate sequence. Silicified shale (silexite) belongs to the succeeding Te´niagouri depositional sequence and need not be genetically related to glaciation.

et al. 2001; Halverson et al. 2004; Xiao et al. 2004; Font et al. 2006; Ne´de´lec et al. 2007). Locally, cap dolostones begin with a coarsening-upward interval of discrete, parallel-sided, normally graded, sedimentation units interpreted as turbidites, separated by argillaceous partings (Lithofacies I of Kennedy 1996; Hoffman & Macdonald 2010). The turbidites, where present, grade upward into laminated but less discretely bedded dolostone with ubiquitous low-angle crossbedding, including toplaps, onlaps and downlaps (Lithofacies II of Kennedy 1996). Kennedy (1996) attributes the low-angle crossbedding to asymmetric synsedimentary dissolution and/or semi-plastic slumping of originally parallel-sided strata deposited below the storm wave base. A simpler and more conventional interpretation is that low-angle crossbedding records wave action above the storm wave base. This interpretation is consistent with the sorted macropeloids and implies initial shoaling subsequent to the underlying turbidites (Hoffman & Macdonald 2010). Intertidal– supratidal indicators (e.g. fenestral texture or other desiccation features, abraded intraclasts, beach ‘rosettes’, channels and levees, polygonal tepees and related breccias) are absent (Kennedy 1996). The lowangle cross-bedded lithofacies is dominant and within it occurs a triad of highly idiosyncratic structures in a broadly consistent vertical sequence, although all three are rarely present in a single section (Fig. 5.2). The structures are described in more detail in Hoffman (2011, and references therein) and are, in order of appearance: † sheet-crack cements (Kennedy 1996; Corkeron 2007; Hoffman & Macdonald 2010) – bedding-parallel, variably buckled, fibrous isopachous, void-filling dolospar, confined to a continuous metre-thick zone of variable intensity near the base of the cap dolostone in distal slope sections, typically just above basal turbidites;

† tubestone (geoplumb) stromatolites (Cloud et al. 1974; Hegenberger 1993; Corsetti & Grotzinger 2005) – metre- to decametre-scale mounds within which arched microbial growth laminae are interrupted by tube-like structures, invariably oriented palaeovertically, defined by parallel-laminated dolomicrite with meniscus-like curvature, variably replaced by late-stage void-filling cement; † giant wave ripples (Allen & Hoffman 2005; Hoffman & Li 2009) – trochoidal megaripples with high aspect ratios, aggradational development, and crestward-coarsening bidirectional laminae that interdigitate in the crestal region. Interpretations of these structures differ. Kennedy et al. (2001) relate sheet-crack cements to permafrost clathrate destabilization by marine flooding, but d13C values of the isopachous dolospar cement are not extraordinarily depleted (Fig. 5.2), nor is d34S of cap dolostone enriched as predicted (Shields 2005). Corkeron (2007) attributes sheet-crack cements to pore-fluid overpressure associated with differential shale-carbonate compaction. Hoffman & Macdonald (2010) suggest that pore-fluid overpressures signal rapid falls in regional sea level associated with the disappearance of ice sheets and the loss of their gravitational ‘pull’ on adjacent ocean waters. Cloud et al. (1974) interpret the tubular structure in stromatolites as fluid- or gas-escape channels on account of their geoplumb orientation. Corsetti & Grotzinger (2005) interpret them as microbial growth structures. Hoffman (2011) attempts to estimate the inclination of the upper foreslope of the Otavi platform from the mean dip of stromatolitic layering after restoring the contained tubes to vertical. Giant wave ripples were first interpreted as supratidal tepee structures (Eisbacher 1985), but their crestlines are straight and parallel, not polygonal like true tepees, and they lack the breccias and void-filling cements diagnostic of supratidal tepees (Kendall & Warren 1987). Gammon et al. (2005) describe a tepee-like structure in the Nuccaleena cap dolostone and relate it to a growth fault, but giant wave ripples in the same formation lack faults and the one described by Gammon et al. (2005), which is oriented perpendicular to bedding, could not have had a slip vector that was parallel or subparallel to the outcrop surface (i.e. the plane of their twodimensional kinematic analysis), compromising the growth fault interpretation. Allen & Hoffman (2005) interpret them as unusually large, steep and aggradational wave ripples, related to the action of long-period waves in the lower part of the ocean mixed layer, driven by strong sustained winds (not hurricanes). Hoffman & Li (2009) find that crestal orientations globally had meridional (north –south) mean orientations and that zonal (east – west) orientations were absent, consistent with zonal and not with cyclonic winds. Sorted peloids, low-angle crossbedding, stromatolites and giant wave ripples signify that cap dolostones were deposited above the storm wave base and at least partly in the euphotic zone. The occurrence of shallow-water structures over a large range of palaeodepths implies diachronous deposition attending a large rise in sea level, consistent with the fragmented d13C records of cap dolostones from different palaeodepths (Fig. 5.3). The estimated magnitude of the rise of more than a kilometre implicates global ice-sheet melting (so-called ‘glacioeustasy’) and links the timescales of deglaciation and cap-dolostone sedimentation (Kennedy 1996; Bertrand-Sarfati et al. 1997; James et al. 2001; Hoffman et al. 2007). Cap dolostones are locally postglacial, but globally syndeglacial. Seafloor carbonate cement. In expanded sequences, cap dolostones are overlain conformably by deeper water limestone, limestone with dolostone turbidites, marlstone or shale. Seafloor cement, in the form of crystal fans of pseudomorphosed prismatic aragonite, built masses up to 100 m thick, localized by sea-bottom topography (Hoffman 2011). In magnitude, they dwarf other Neoproterozoic seafloor cements and differ from volumetrically comparable

CHEMICAL SEDIMENTS ASSOCIATED WITH NEOPROTEROZOIC GLACIATION

Archaean –Palaeoproterozoic examples in having formed in the presence of micritic sedimentation (Sumner 2002). Highstand deposits of cap-carbonate sequences. Regressive highstand deposits in expanded cap-carbonate sequences may be carbonate or clastic-dominated. The former begin with limestone rhythmite and shoal to dolostone grainstone beneath a welldeveloped subaerial sequence boundary. Grain-sized equivalent terrigenous deposits (shales, siltstones and sandstones) conformably overlie cap dolostones in many areas (e.g. in the Brachina Formation and the ABC Range Quartzite in South Australia). In the cratonic Taoudeni Basin of West Africa, highstand deposits are found in palaeovalleys but are missing on palaeotopographic highs (Fig. 5.3). This has led to misinterpretation. Silicified shale (silexite) of the Te´niagouri Group (Fig. 5.3) and correlatives rest directly on the altered surface of the cap dolostone over vast areas of the craton, leading to the concept of a ‘triad’: tillite –carbonate–silexite. We concur with Bertrand-Sarfati et al. (1997), in that this concept should be abandoned. Deposition of the Te´niagouri Group (and correlative Azlaf Group, Deynoux et al. 2006) cannot be related to the glacioeustatic transgression and could be much younger (BertrandSarfati et al. 1997).

73

ne Cille, Luoquan, Hankalchough and Croles Hill diamictites). Up to 40 cm of argillaceous limestone overlies glacial marine diamictite in two sections of the short-lived, 582 Ma Gaskiers Formation on the Avalon Peninsula of eastern Newfoundland, Canada (Myrow & Kaufman 1999). This cap carbonate is significant as the sole carbonate bed in the nearly 15-km-thick host succession. It lithologically and isotopically resembles the base of many Sturtian cap-carbonate sequences, although the amount of carbonate deposited overall is far smaller. Finally, the Egan diamictite represents a short-lived glacial incursion onto a mid-Ediacaran carbonate platform in the Kimberley region of Western Australia (Corkeron & George 2001). Diamictite and overlying conglomerate are overlain by c.15 m of dolostone with herringbone and trough crossbedding, and an additional c.15 m of interbedded quartz-arenite and silty dolostone with a distinctive stromatolite (Tungussia julia) horizon near the top. The carbonate sequence following the Egan glaciation resembles no other cap-carbonate sequence, including the one on the Landrigan diamictite in the same succession, stratigraphically well below the Egan, which has a recognizable Marinoan-type cap dolostone (Corkeron 2007).

Barite in Marinoan-type (basal Ediacaran) cap dolostones Transgressive cap dolostones in condensed sequences. Where accom-

modation was lacking, cap dolostones aggraded or prograded to the supratidal zone, where they were subjected to intense meteoric and vadose diagenesis (Fig. 5.3). In such cases, primary marine dolopelmicrite was repeatedly fractured and re-cemented in tepee structures and related breccias (Kendall & Warren 1987). Barite is a major void-filling cement in tepee brecciated cap dolostone on the Yangtze platform of South China (Jiang et al. 2006) and the West African craton (Shields et al. 2007). Where cap dolostones were subaerially exposed (Fig. 5.3), highstand deposits are typically absent.

Sturtian-type (Cryogenian) cap-carbonate sequences The contacts between Sturtian glaciogenic deposits and cap carbonates, like Marinoan ones, are characteristically singular, smooth, abrupt and conformable. Sturtian cap carbonates are lithologically distinct: where Marinoan cap dolostones are pale in colour (,0.1 wt% total organic carbon, TOC), arenaceous in texture and display wave-generated structures. Sturtian capcarbonate sequences typically begin with fetid, dark grey, parallel laminated, micritic limestone (rarely rhodochrosite, MnCO3), with or without parallel-sided graded beds (turbidites) and debrites (Tojo et al. 2007). In Namibia, the Sturtian cap-carbonate sequence continues with up to 200 m of continuous sublittoral dolostone microbialaminite, featuring microbial rollup structures, neptunian dykes and inclined zones of high primary porosity overlain by humped microbialaminite (Pruss et al. 2010). The sublittoral microbialaminite extravaganza eventually grades up into crossbedded grainstone, ending at a tepee brecciated sequence boundary. Compared with Marinoan cap-carbonate sequences, where transgressive tracts (i.e. cap dolostones) are well developed, shallow-water transgressive tracts are thin (Smith et al. 1994) or absent in Sturtian cap-carbonate sequences. Typically, deepwater deposits directly overlie beveled supraglacial surfaces. The simplest explanation for the difference is a below-critical saturation state with respect to carbonate in the surface ocean during Sturtian deglaciation (Hoffman & Schrag 2002).

Mid-Ediacaran cap carbonate sequences Most mid-Ediacaran glaciogenic horizons lack cap carbonates (e.g. Serra Azul, Squantum, Mortensnes, Moelv, Vil’chitsi, Loch

Barite (BaSO4) is a major constituent of certain Marinoan cap dolostones (Table 5.2). The mineral is easily recognizable in the field from its vitreous pearly lustre, bladed crystal habit and high specific gravity (4.3 – 5.0). Even where pseudomorphosed by calcite, its bladed crystals are distinct from aragonite, which is acicular (needlelike) in habit, forming pseudohexagonal prisms due to polysynthetic twinning. The palaeoenvironmental significance of barite rests with its highly redox-sensitive solubility in S-bearing aqueous solutions: the concentration of Ba in modern (oxic) seawater is only 1.4 ppb by atom, compared with 51 ppm in crustal rocks from which Ba in seawater is derived. Consequently, barite leached by anoxic pore waters from deep-sea carbonate ooze reprecipitates cumulatively at the sub-seafloor redox front as it migrates through the sediment column in response to sediment accumulation. However, it is doubtful that this process alone could provide a satisfactory explanation for either barite type in Marinoan cap dolostones, given their limited thickness. Two types of barite in cap dolostones: primary and early diagenetic. Two types of barite have been described from Marinoan cap

dolostones (Table 5.2): primary barite in the form of seafloor cement (Kennedy 1996; Hoffman & Halverson 2011) and early diagenetic barite associated with tepee and tepee-like breccias (Jiang et al. 2006; Shields et al. 2007). Seafloor barite forms at a laterally continuous horizon in the top few centimetres of cap dolostones. Tiny barite crystals self-assemble into macroscopic crystal fans or digitate aggregates. Interaction between barite growth and particulate sedimentation forms the basis for determining a seafloor origin, denoting precipitation of barite from the ambient water column. Early diagenetic barite forms void-filling isopachous crustose cements within breccias associated with tepee-like structures. The cause of tepee formation and brecciation is attributed to submarine methane venting (Jiang et al. 2006), or alternatively to evaporative pumping due to subaerial (supratidal) exposure in zones of marine – meteoric groundwater mixing (Shields et al. 2007). The second alternative is the conventional interpretation of tepee structures and associated breccias (Assereto & Kendall 1977; Kendall & Warren 1987). Seafloor barite in central Australia. In the Amadeus and Ngalia basins of central Australia, seafloor barite is associated with ferruginous domal stromatolite at the top of cap dolostones overlying Marinoan glaciogenic deposits of the Olympic and Mount Doreen formations, respectively (Kennedy 1996). Barite occurs

74

P. F. HOFFMAN ET AL.

either as isolated bladed crystals (1 –2 mm) or as large upwardoriented rosettes of bladed crystal (up to 4 cm high). The crystals ‘grew at the sediment –water interface, as indicated by the upward-oriented growth habit, sediment drape defining crystal terminations, and presence of crystal fragments as detrital material within overlying sediment’ (Kennedy 1996). In the Amadeus Basin, the barite-rich horizon is conformably overlain by reddishgrey laminated siltstone with limestone turbidites (Pertatataka Formation); in the Ngalia Basin it is overlain by red laminated siltstone (Red Shale Member of the Mount Doreen Formation). In both cases, in our view, the barite horizon marks a conformable transition from deposition under the influence of storm waves (i.e. size-sorted peloids, low-angle crossbedding) to sedimentation below the storm wave base. Walter & Bauld (1983) suggested that the barite was secondary after primary anhydrite (CaSO4), contingent on their interpretation that the carbonate and sulphate were products of intense evaporation in glacial lakes such as those of the Antarctic Dry Valleys. This is not a credible scenario in an open marine setting in the absence of a local hydrothermal source of Ca. Seafloor barite in northwestern Canada. In the Mackenzie Moun-

tains of northwestern Canada, the top 4– 10 cm of the Ravensthroat cap dolostone contains digitate barite cement (variably calcitized) over a strike length of nearly 200 km (Hoffman & Halverson 2011). The barite coincides with a change in colour of the peloidal dolostone from pale cream to chocolate brown. Locally, the barite-rich layer is developed on a train of giant wave ripples. The barite first appears as randomly oriented, millimetre-scale, bladed crystals, which appear to have formed just below the sediment –water interface. Upwards, the crystals self-organize to form centimetre-scale digitate structures, commonly oversteepened toward the SW (seaward). The digitate structures display coral-like growth laminae, defined by films of opaque minerals. Where calcitized, they resemble microdigitate stromatolites except for their diagnostic bladed crystal habit (so-called barite ‘rosettes’). The flanks of the digitate structures are ragged due to outward growth of barite after each lamina of peloidal Fe-dolomite was deposited between the digits. The interplay between crystal growth and burial by peloids demonstrates that the barite precipitated from the water column, into which the digitate structures projected up to 1.0 cm above the seafloor. The Fe-dolostone layer with barite cement makes a conformable contact with overlying micritic limestone (Hayhook Formation) – epigenetically dolomitized locally – containing acicular crystal fans of pseudomorphosed aragonite cement and reworked detrital dolomite basally (James et al. 2001).

Cryogenian (Marinoan) glaciogenic deposits of the Nantuo Formation on the Yangtze platform (Jiang et al. 2006). Void-filling barite fans occur in the lower part (unit C1) of the cap dolostone on the inner platform and on the slope and basin to the SE. Unit 1 is brecciated everywhere and hosts a variety of sheet-crack cements, cement-filled stromatactis-like cavities, and tepee-like structures (Jiang et al. 2006). Importantly, the base of unit C1 is a sharp, smooth, undulating surface; neither it nor the upper Nantuo Formation experienced the intense brecciation typically observed within unit C1 (Jiang et al. 2006). In addition, possible seafloor barite is described as ‘layer-parallel barite fans growing out of the dolomicrite substrate’ in the lower part of unit C2 in the basin (Jiang et al. 2006). Tepee-like structures occur locally in the lower part of unit C2 on the platform. Unit 3 and the upper part of unit C2 are not brecciated, consisting of parallel- and small-scale cross-laminated peloidal dolopackstone with graded layers. Unit 3 has increased silt and limestone (Jiang et al. 2006). Early diagenetic barite on the Dzabkhan Platform, western Mongolia. A characteristic Marinoan cap dolostone (Table 5.2) overlies

the younger of two Cryogenian glaciogenic horizons (Khongoryn diamictite) on the Dzabkhan platform of western Mongolia (Macdonald et al. 2009, 2011). Barite occurs in the mountain pass (Hoh Davaa) between Bayan-Uul and Jargalan in northern Govi-Altay. There the cap dolostone includes two units, separated by a sharp smooth disconformity veneered by dark brown Fe-dolostone. The first unit (11.1 m) is a pale tan peloidal dolostone with lowangle crossbedding. Isopachous sheet-crack cements occur near its base, above the basal 0.7 m of marly brown ribbon beds. The second unit (8.4 m) consists of variably brecciated medium-grey dolostone with thick crusts of isopachous void-filling barite cement. Toward the top, sheet-crack and stratiform barites form domal structures with up to 1.0 m of relief. The barite domes were ultimately onlapped and buried by a third unit consisting of unbrecciated medium-dark grey dolostone ribbons with subaqueous microbial textured intervals, including rollup structures, that lead into the scree-covered maximum flooding horizon of the Ol cap-carbonate sequence. Although onlap relations clearly show that the domal barite was exposed to seawater at the top of the second unit, proof that it formed on the seafloor in the form of detailed interaction between sedimentation and precipitation has yet to be observed.

Phosphorite in cap-carbonate sequences Taoudeni Basin, West Africa

Early diagenetic barite in the Taoudeni Basin, West Africa. The

common occurrence of barite in cap dolostones terminating the Jbe´liat glaciation of the West African craton has been known for half a century (Deynoux & Trompette 1981). The type area is the Atar Cliff, Mauritania, on the northern edge of the Taoudeni Basin (Fig. 5.3). There, the cap dolostone directly overlies a periglacial permafrost regolith with deep polygonal sand wedges, developed above a terrestrial sub- and proglacial succession of northerly derivation and vast extent (Deynoux 1982, 1985; Deynoux et al. 2006). The cap dolostone is variably brecciated and spectacular metre-scale tepee structures (Kendall & Warren 1987) are developed toward the top near Amogjar, east of Atar. Isopachous barite forms thick crusts around blocks of brecciated dolostone, as well as secondary (remobilized?) vein-fillings (Shields et al. 2007). An early diagenetic origin for the cavityfilling, crustose barite is indicated by detrital barite clasts in coarse-grained quartz sand- and granulestone of the cap-carbonate sequence highstand tract (Fig. 5.3). Early diagenetic barite on the Yangtze Platform, South China. A

relatively thin (2.5 –4.0 m) cap dolostone (lower Doushantuo Formation units C1 –3 of Jiang et al. 2006) blankets terminal

On the northeastern margin of the Taoudeni Basin, diamictite of the Jbe´liat glaciation (Fersiga Formation) is overlain by a capcarbonate sequence correlative with that found in palaeovalleys on the Atar Cliffs (Fig. 5.3), 1250 km to the SW (Bertrand-Sarfati et al. 1997; Deynoux et al. 2006). A thin but continuous cap dolostone (Oued Djouf Formation) is overlain by a regressive highstand sequence (Grizim Formation) up to 80 m thick, composed of glauconitic green shale, siltstone and sandstone, ultimately aeolian (Bertrand-Sarfati et al. 1997). Apatitic phosphorite is concentrated at the base of the Grizim Formation, forming stromatolitic domes, microdigitate clusters and shrub-like colonies over moraines of Fersiga diamictite, where the cap dolostone is brecciated due to subaerial exposure, and crossbedded phospharenite composed of phosphate-coated grains and oncolites in the inter-moraine depressions (Bertrand-Sarfati et al. 1997). Authigenic glauconite occurs in both phosphorite facies and void-filling barite is a minor constituent of the shrub-like phosphate colonies. The phosphorites formed rapidly, during the initial stages of the glacioeustatic transgression (Bertrand-Sarfati et al. 1997). Phosphorites of early Ediacaran age occur in other areas, but their connection to glaciation is tenuous. Khodjari-type

CHEMICAL SEDIMENTS ASSOCIATED WITH NEOPROTEROZOIC GLACIATION

phosphorite in the northern Volta basin was tentatively linked to the terminal Cryogenian deglaciation of West Africa (Trompette et al. 1980), but the phosphorite is younger than regionally extensive silicified green argillite (‘silexite’), which itself disconformably overlies the basal Ediacaran (Marinoan) cap-carbonate sequence in the Taoudeni basin (Fig. 5.3; Bertrand-Sarfati et al. 1997; Shields et al. 2007). Similarly, economic phosphorite in the Khubsugul area of northern Mongolia (Ilyin et al. 1986) was genetically related to the underlying terminal Cryogenian Khesen diamictite (Sheldon 1984). However, Ilyin (2009) and recent mapping has shown that the phosphorite is separated from post-glacial Baxha cap-carbonate sequence by a major hiatus (Macdonald & Jones 2011).

Genesis and significance of glacial-associated chemical sediments Fe and Fe – Mn oxide deposits Ferrous v. euxinic anoxia. Anoxia due to ice cover has long been invoked as a means of mobilizing reduced Fe and Mn in solution as a source for oxide ores within Cryogenian glaciogenic sequences (Martin 1965b; Urban et al. 1992; Klein & Beukes 1993; Lottermoser & Ashley 2000; Klein & Ladeira 2004). Fe(II) concentration in the presence of H2S is limited by pyrite (FeS2) saturation. High degrees of continental ice cover and consequent large sea-level falls favour Fe over S in ocean waters because of reduced riverine sulphate supply and higher Fe/S in deep-sea hydrothermal vent fluids, respectively (Canfield & Raiswell 1999; Kump & Seyfried 2005). Subglacial sulphate-rich ferrous waters. The discovery of modern Fe- and sulphate-rich (3.45 and 50 mM, respectively) waters trapped beneath the Taylor Glacier in the McMurdo Dry Valleys of East Antarctica gives a new perspective on Cryogenian jaspilites (Mikucki et al. 2009). Fe is not titrated as pyrite because there is insufficient organic primary production, due to the glacier cover, to fuel bacterial sulphate reduction. The source of Fe in the trapped former seawater is subaqueous bedrock ‘weathering’, enhanced by the low pH (6.2) of the waters. Fe-oxy-hydroxide forms instantaneously at Blood Falls, where the anoxic waters discharge at the glacier terminus (Mikucki et al. 2009). This demonstrates that waters rich in both dissolved Fe and sulphate can exist if ice cover is perennial and sufficiently thick to preclude phototrophic primary production. Localization of oxidative titration. In anoxic Fe-rich waters, the localization of Fe (and Mn) deposition could relate to the availability of oxygenated subglacial meltwater at the grounding-lines of wet-base ice streams within cold-base ice sheets. This would be consistent with the observed occurrence of Cryogenian jaspilites as lenticular bodies in contact with diamictite complexes whose lithofacies and internal organization is most consistent with deposition in ice-sheet grounding zones. Alternation of Fe and Mn ores could relate to variations of the meltwater flux associated with glacial retreat and advance, respectively, as proposed by Urban et al. (1992).

Cap-carbonate sequences Accumulation rates for syndeglacial cap dolostones. Marinoan cap dolostones were deposited during marine transgressions resulting from the meltdown of grounded ice sheets (Preiss et al., 1978; Kennedy 1996; Bertrand-Sarfati 1997; James et al. 2001; Shields 2005; Fairchild & Kennedy 2007; Hoffman et al. 2007). The timescales for Cryogenian deglaciations have not been specifically targeted in climate modelling. Nevertheless, deglaciation histories of

75

the temperate Quaternary ice sheets, simplified Neoproterozoic deglacial simulations, and the apparent absence of polar continents in Cryogenian palaeogeographic reconstructions (Fig. 5.1) all suggest that Sturtian and Marinoan ice sheets disappeared rapidly. How rapidly? If those middle and low-latitude ice sheets disappeared at the rate most temperate mountain glaciers have been lowered in recent decades (1.1 m/a), their demise would have taken 2 ka assuming their average thickness was comparable with the present East Antarctic Ice Sheet (Lythe et al. 2001). This is close to the pan-deglacial timescale of ,2 ka in a climate model (Hyde et al. 2000). Positive feedbacks – ice-albedo, ice-elevation and various greenhouse-gas feedbacks notably water vapour feedback – contribute to rapid deglaciation. The energy demands for such a deglaciation, amounting to a globally averaged energy flux of c. 11 W m2 (Wallace & Hobbs 1977, p. 320), are small compared to the Neoproterozoic solar irradiance of c. 362 W m2 (c. 94% of present), a high surface albedo notwithstanding. Pending detailed modelling of a global deglaciation, we conservatively take 8 ka, the timescale for Quaternary deglaciations, as an upper limit for the Marinoan deglaciation. Given median (9 m) and maximum (175 m) thicknesses of Marinoan cap dolostones globally (Hoffman et al. 2007), average accumulation rates were then c. 1 and 22 mm/a for cap dolostones of typical and extreme thickness, respectively. If deglaciation occurred in 2 ka, the respective accumulation rates were four times larger. These must be considered minimum rates, as cap dolostones at any given location represent only a fraction of the total deglaciation (Fig. 5.2). We consider it likely that unusual structures in cap dolostones (e.g. sheet-crack cements, tubestone stromatolites and highly aggradational wave ripples) may be fundamentally related to extreme rates of accumulation. The record of geomagnetic reversals and excursions in the Mirassol d’Oeste and Nuccaleena cap dolostones creates a dilemma (Trindade et al. 2003; Raub 2008). Conventional estimates of their duration and frequency (Johnson et al. 1995; Gubbins 1999; Roberts 2008) imply a timescale on the order of 0.1– 1.0 Ma for those cap dolostones, implying that accumulation rates were at least 10 –100 times slower than the ‘conservative’ estimates given in the preceding paragraph. This makes for an irreconcilable conflict between conventional interpretations of geomagnetic polarity reversals and excursions on the one hand, and estimated deglaciation rates on the other. If one accepts a conventional magnetostratigraphic interpretation, cap dolostones are highly condensed (Fairchild & Kennedy 2007). However, compared with other depositional sequences in the same (or other) successions, Marinoan cap dolostones are relatively thick and expanded among transgressive tracts (e.g. Hoffman & Halverson 2008). The resolution of geomagnetic field reversals and excursions is limited by the accumulation rates of the sediments in which the records are kept (Channell & Lehman 1997; Roberts & Lewin-Harris 2000; Roberts 2008). If Marinoan cap dolostones were actually deposited at rates far exceeding the best Quaternary deep-sea records (Channell & Lehman 1997; Laj et al. 2006), might they not display hitherto unsuspected phenomena? This would most likely be the case if the absolute strength of the geomagnetic field was collapsed in 635 Ma, just as it does during field reversals and excursions. The slowness with which the solid inner core reverses, by diffusion, does not limit the speed with which the outer core alone reverses in a weak field (Gubbins 1999). In our view, the issue is unresolved, but neither the sedimentological evidence for high sedimentation rates nor the physical arguments for rapid deglaciation should be dismissed lightly.

Are syndeglacial cap dolostones non-marine? If Marinoan ice sheets had an average thickness of c. 2.2 km over all continents and shelves (Fig. 5.1), their disappearance in 10 ka

76

P. F. HOFFMAN ET AL.

would produce meltwater at a global average rate of c. 0.6  106 m3 s21 (2.01  1017 m3/3.15  1010 s), or c. 6.4 Sv (Sverdrups, where 1.0 Sv ¼ 106 m3 s21). Deglaciation in 2 ka would result in a global meltwater flux of c. 32 Sv. These values correspond to approximately 6 and 32 times present total runoff, respectively. Such an increase in freshwater input, followed by strong surface warming after the ice disappeared, must have impeded ocean mixing by creating a more stable density stratification (Shields 2005). Cold saline deepwater that evolved during the Marinoan glaciation would have been capped by a thickening meltwater-dominated lid. The melting of sea ice, or a sea glacier in the case of a Snowball Earth (Warren et al. 2002; Goodman & Pierrehumbert 2003), followed by the melting of ice sheets, would have contributed up to 0.7 and c. 1.0 km of freshwater respectively. Ocean mixing time would have been many times the present 103 years, ensuring that the surface ocean was brackish when cap dolostones were deposited (Shields 2005). This meltwater-dominated lid has been dubbed ‘plumeworld’ (Shields 2005), and we refer to the Marinoan example as ‘Glacial Lake Harland’ (GLH). (W. Brian Harland of Cambridge University was a pioneer advocate of Neoproterozoic glaciations as chronostratigraphic markers.) Shields (2005, p. 305) gives many reasons why GLH was more favourable than seawater for rapid carbonate production.

Abiotic or biogenic dolomite? The dolomicrospar of cap dolostones preserves primary structures, makes conformable contact with overlying and (rare) underlying limestones, and near the base of the former is reworked as intraclastic dolomite grains. These are good reasons to conclude that the dolomite was either primary or a product of very early diagenesis near the sediment – water interface. It has recently been demonstrated that dolomite-forming reactions are catalysed at low temperatures by microbial activity utilizing different metabolic pathways: sulphate reduction (van Lith et al. 2003), anaerobic methane oxidation (Moore et al. 2004) and methanogenesis by Archaea (Kenward et al. 2009). The first of these has been applied to Marinoan cap dolostones (Shields 2005; Font et al. 2006; Ne´de´lec et al. 2007). However, methanogenesis by Archaea is particularly appealing because the formation of ordered dolomite at low temperature (48C) has been demonstrated experimentally in a low sulphate, low Mg:Ca (,1), acidic (pH ¼ 6.7), freshwater environment, not unlike GLH (see above). As the methanogenic consortium (Kenward et al. 2009) is anaerobic, formation of cap dolostone by this means must have occurred below the sediment –water interface, assuming deglacial surface waters were well oxygenated. Sources of alkalinity for cap dolostones. Syndeglacial carbonate pro-

duction in surface waters was driven by warming (DT  508C; Pierrehumbert 2002), inundation of shallow shelves and platforms where carbonate burial is favoured (Ridgwell et al. 2003), alkalinity provided by anaerobic respiration (Kennedy et al. 2001; but see Shields 2005 for critical discussion), and carbonate weathering (Higgins & Schrag 2003; Anderson 2007; see Le Hir et al. 2009 for critical discussion, but note that they consider only silicate weathering as a source of alkalinity, not carbonate weathering which is orders of magnitude more rapid). Significance of seafloor aragonite cements. Seafloor cements com-

posed of former aragonite occurring in limestones above Marinoan cap dolostones formed below the storm wave base, at or near the maximum flooding horizon (Grotzinger & Knoll 1995; James et al. 2001; Nogueira et al. 2003; Hoffman 2011). They represent a brief revival of more ancient conditions following a long decline in the importance of seafloor cement in the latter half of the Proterozoic (Grotzinger & James 2000; Sumner 2002). What was the

source of alkalinity driving carbonate production in deep water? One possibility is anaerobic respiration, which operates as an alkalinity pump in modern super-anoxic fjords (Anderson et al. 1987). Another is that extreme concentrations of dissolved inorganic carbon (DIC), expected for an ocean equilibrated with carbonate sediments and a c. 0.1 bar CO2 atmosphere, flattened the negative gradient in carbonate saturation (V) with increasing depth, favouring the precipitation of cement on the seafloor, irrespective of the oxidant or intensity of organic carbon cycling (Higgins et al. 2009).

Barite in cap dolostones Seafloor barite. Seafloor barites in basal Ediacaran (Marinoan) cap dolostones of central Australia and northwestern Canada are strikingly similar in form, dimensions and precise stratigraphic position within their respective cap-carbonate sequences. We infer that they formed where the seabed intersected the interface between Ba-rich euxinic deeper water and an oxic Fe(III)laden mixed layer. Localized upwelling would then have caused sulphate production and barite titration. Simultaneously, anaerobic respiration, where Fe was the electron acceptor, caused Fe(III) reduction, leading to Fe-carbonate (Fe-dolostone) production coincident with the barite cement horizon. It is well established that mixed-layer waters from which cap dolostones were deposited were Fe(III)-laden (Embleton & Williams 1986; Li 2000; Font et al. 2005), accounting for the pinkish tint of many cap dolostones. Ba in euxinic glacial deep waters would derive from seawater – basalt exchange reactions and seafloor weathering of detrital feldspar. We see no need for the involvement of methane venting, for which there is neither physical nor isotopic evidence at seafloor barite horizons in Marinoan cap dolostones (Kennedy 1996; Hoffman & Halverson 2011). Euxinic deep water that evolved during the Marinoan glaciation (Hurtgen et al. 2006) may have contrasted with Fe(II) sulphate deep water (Mikucki et al. 2009), which may have developed during the Sturtian glaciation, accounting for the deposition of haematite –jaspilite at that time. Early diagenetic (void-filling) barite. Tepee breccias in cap dolostones and associated void-filling isopachous barite cement have been ascribed to vadose diagenesis in West Africa (BertrandSarfati et al. 1997; Shields et al. 2007). In South China, similar features have been attributed to submarine cold seeps (Jiang et al. 2006) or, alternatively, to vadose diagenesis (Zhou et al. 2010). In both interpretations, barite precipitation is thought to have resulted from mixing of an oxic sulphate-rich fluid (seawater) and a highly reducing, Ba- and possibly methane-rich fluid, derived from the destabilization of permafrost gas hydrate in underlying glaciogenic sediments from which the Ba was scavenged (Jiang et al. 2006; Shields et al. 2007). It remains to be seen if this attractive scenario is compatible with a basic field observation in both areas, that the contact surface between the glaciogenic diamictite and the cap dolostone is smooth and unbroken (authors’ observations). Is it conceivable that methane and other gases, generated within the till (permafrost melting), could thoroughly brecciate the cap dolostone on the seafloor, while leaving their mutual contact surface undisturbed? Methane involvement is supported by extreme d13C depletion (below –10‰ Vienna Pee Dee Belemnite standard) in certain void-filling cement phases in brecciated cap dolostone on the Yangtze platform (Jiang et al. 2003; Wang et al. 2008). However, such strong depletions are not observed elsewhere, despite thousands of published analyses, and those in South China come mostly from late-stage calcite void-fillings that demonstrably post-date coexisting barite and void-filling dolomite cements (Bristow et al. 2011; Zhou et al. 2011). Moreover, d34S values in cap dolostones are not strongly enriched, as predicted if anaerobic methane oxidation (an alkalinity pump) was fueling bacterial sulphate reduction (Shields 2005).

CHEMICAL SEDIMENTS ASSOCIATED WITH NEOPROTEROZOIC GLACIATION

Phosphorite in cap dolostones It is not possible to generalize about Neoproterozoic glacial-related phosphorite from a single example, no matter how impressive and well documented it is (Bertrand-Sarfati et al. 1997). Its existence supports assumed high productivity during deglaciation (Shields 2005; Font et al. 2006; Ne´de´lec et al. 2007).

Conclusions Synglacial Fe and Fe–Mn deposits are mostly, if not all, associated with the older Cryogenian (Sturtian) glaciation. They indicate anoxic (not euxinic) deep water. Lithofacies associations suggest that subglacial meltwater plumes provided oxidant. Alternating Fe and Mn deposits may reflect oscillating plume fluxes associated with glacial cycles. Sturtian and Marinoan cap carbonates are distinct. Transgressive cap dolostones are almost entirely limited to the Marinoan deglaciation, and were deposited diachronously, mainly above the storm wave base, during coastal inundation by meltwaterdominated surface waters. Biogenic dolomite nucleation in waters of low ionic strength and low Mg:Ca ratio was likely microbially mediated. Barite occurs in Marinoan cap dolostones both as seafloor cements in terminal Fe-dolomite beds and as early diagenetic voidfilling crusts within tepee and tepee-like breccias. Seafloor barite cement formed where euxinic Ba-rich deep waters upwelled into oxic Fe(III)-rich surface waters. Early diagenetic void-filling barite cement in cap dolostones has been attributed to subaerial and submarine methane seepage, driven by destabilization of permafrost hydrates in underlying sediments following marine inundation. Although attractive in principle, undisturbed contact surfaces between glaciogenic diamictites and internally brecciated cap dolostones may be difficult to reconcile with seepage from below, unless brecciation was strictly limited to the horizon of methane oxidation. Fieldwork was supported by research grants from the Earth System History (ESH), Arctic Natural Science (ANS), and Geobiology & Environmental Geochemistry (GEG) programmes of the US National Science Foundation (NSF). PFH receives additional support from the Canadian Institute for Advanced Research (CIFAR), Harvard University Center for the Environment (HUCE), US Social Security, and the Canada Pension Plan. FAM and GPH thank the Yukon Geological Survey for additional support. We gratefully acknowledge constructive reviews by G. Jiang and P. K. Link, which prompted several improvements. This represents a contribution of the IUGS- and UNESCOfunded IGCP (International Geoscience Programme) Project #512.

References Aitken, J. D. 1991. The Ice Brook Formation and post-Rapitan, late Proterozoic glaciation, Mackenzie Mountains, Northwest Territories. Geological Survey of Canada Bulletin, 404, 1 – 43. Allen, P. A. & Hoffman, P. F. 2005. Extreme winds and waves in the aftermath of a Neoproterozoic glaciation. Nature, 433, 123–127. Allen, P., Leather, J. & Brasier, M. D. 2004. The Neoproterozoic Fiq glaciation and its aftermath, Huqf Supergroup of Oman. Basin Research, 16, 507– 534. Alvarenga, C. J. S. de & Trompette, R. 1992. Glacially influenced sedimentation in the Later Proterozoic of the Paraguay belt (Mato Grosso, Brazil). Palaeogeography, Palaeoclimatology, Palaeoecology, 92, 85 – 105. Alvarenga, C. J. S. de, Dardenne, M. A. et al. 2008. Isotope stratigraphy of Neoproterozoic cap carbonates in the Araras Group, Brazil. Gondwana Research, 13, 469–479. Anderson, J. B. 1999. Antarctic Marine Geology. Cambridge University Press, Cambridge. Anderson, S. P. 2007. Biogeochemistry of glacial landscape systems. Annual Review of Earth and Planetary Sciences, 35, 375– 399.

77

Anderson, L. G., Dyrssen, D. & Skei, J. 1987. Formation of chemogenic calcite in super-anoxic seawater –Framvaren, Southern-Norway. Marine Chemistry, 20, 361–376. Assereto, R. L. A. M. & Kendall, C. G. St. C. 1977. Nature, origin and classification of peritidal tepee structures and related breccas. Sedimentology, 24, 153–210. Badenhorst, F. P. 1988. The lithostratigraphy of the Chuos mixtite in part of the southern central zone of the Damara Orogen, South West Africa. Communications of the Geological Survey of South West Africa/Namibia, 4, 103– 110. Bao, H., Lyons, J. R. & Zhou, C. 2008. Triple oxygen isotope evidence for elevated CO2 levels after a Neoproterozoic glaciation. Nature, 452, 504–506. Bertrand-Sarfati, J., Flicoteaux, R., Moussine-Pouchkine, A. & Aı¨t Kaci Ahmed, A. 1997. Lower Cambrian apatitic stromatolites and phospharenites related to the glacio-eustatic cratonic rebound (Sahara, Algeria). Journal of Sedimentary Research, 67, 957–974. Bristow, T. F., Boniface, M., Derkowski, A., Eller, J. M. & Grotzinger, J. P. 2011. A hydrothermal origin for isotopically anomalous cap dolostone cements from south China. Nature, 474, 68 –72. Cahen, L. & Lepersonne, J. 1981. Proterozoic diamictites of Lower Zaire. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s PrePleistocene Glacial Record. Cambridge University Press, Cambridge, 153– 157. Cahen, L. 1950. Le Calcaire de Sekelolo, le Complexe tillitique et la Dolomie rose C1 dans l’Anticlinal de Congo dia Kati (Bas-Congo). Annales du Muse´e du Congo Belge, Sciences Ge´ologiques, 7, 13– 54. Calver, C. R. & Walter, M. R. 2000. The late Neoproterozoic Grassy Group of King Island, Tasmania: correlation and palaeogeographic significance. Precambrian Research, 100, 299– 312. Canfield, D. E. & Raiswell, R. 1999. The evolution of the sulfur cycle. American Journal of Science, 299, 697–723. Channell, J. E. T. & Lehman, B. 1997. The last two geomagnetic polarity reversals recorded in high-deposition-rate sediment drifts. Nature, 389, 712– 715. Chumakov, N. M. 1992. The problems of old glaciations: Pre-Pleistocene glaciogeology in the USSR. Soviet Science Reviews, Section G Geology, 1, 1 –208. Chumakov, N. M. 2007. Climates and climate zonality of the Vendian: geological evidence. In: Vickers-Rich, P. & Komarower, P. (eds) The Rise and Fall of the Ediacaran Biota. Geological Society, London, Special Publication, 286, 15– 26. Clifford, T. N. 2008. The geology of the Neoproterozoic Swakop– Otavi transition zone in the Outjo Distroct, northern Damara Orogen, Namibia. South African Journal of Geology, 111, 117– 140. Cloud, P., Wright, L. A., Williams, E. G., Diehl, P. & Walter, M. R. 1974. Giant stromatolites and associated vertical tubes from the upper Proterozoic Noonday Dolomite, Death Valley region, eastern California. Geological Society of America Bulletin, 85, 1869– 1882. Condon, D., Zhu, M., Bowring, S. A., Wang, W., Yang, A. & Jin, Y. 2005. U– Pb ages from the Neoproterozoic Doushantuo Formation, China. Science, 308, 95 –98. Corkeron, M. 2007. ‘Cap carbonates’ and Neoproterozoic glacigenic successions from the Kimberley region, north-west Australia. Sedimentology, 54, 871–903. Corkeron, M. L. & George, A. D. 2001. Glacial incursion on a Neoproterozoic carbonate platform in the Kimberley region, Australia. Geological Society of America Bulletin, 113, 1121– 1132. Corsetti, F. A. & Grotzinger, J. P. 2005. Origin and significance of tube structures in Neoproterozoic post-glacial cap carbonates: example from Noonday Dolomite, Death Valley, United States. Palaios, 20, 348– 363. Corsetti, F. A. & Kaufman, A. J. 2003. Stratigraphic investigations of carbon isotope anomalies and Neoproterozoic ice ages in Death Valley, California. Geological Society of America Bulletin, 115, 916– 932. Deynoux, M. 1982. Periglacial polygonal structures and sand wedges in the late Precambrian glacial formations of the Taoudeni Basin in Adrar of Mauretania (West Africa). Palaeogeography, Palaeoclimatology, Palaeoecology, 39, 55 –70. Deynoux, M. 1985. Terrestrial or waterlain glacial diamictites? Three case studies from the late Proterozoic and late Ordovician glacial

78

P. F. HOFFMAN ET AL.

drifts in West Africa. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 97– 141. Deynoux, M. & Trompette, R. 1976. Discussion: Late Precambrian mixtites: glacial and/or nonglacial? Dealing especially with the mixtites of West Africa. American Journal of Science, 276, 1302– 1315. Deynoux, M. & Trompette, R. 1981. Late Precambrian tillites of the Taoudeni Basin, West Africa. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 123– 131. Deynoux, M., Affaton, P., Trompette, R. & Villeneuve, M. 2006. Pan-African tectonic evolution and glacial events registered in Neoproterozoic to Cambrian cratonic and foreland basins of West Africa. Journal of African Earth Sciences, 46, 397–426. Dorr, J. V. N. II. 1945. Manganese and iron deposits of Morro do Urucum, Mato Grosso, Brazil. United States Geological Survey Bulletin, 946-A, 1– 47. Dowdeswell, J. A., Whittington, J. A., Jennings, A. E., Andrews, J. T., Mackensen, A. & Marienfield, P. 2000. An origin for laminated glacimarine sediments through sea-ice build-up and suppressed iceberg rafting. Sedimentology, 47, 557– 576. Dunn, P. R., Thompson, B. P. & Rankama, K. 1971. Late Pre-Cambrian glaciation in Australia as a stratigraphic boundary. Nature, 231, 498– 502. Edwards, M. B. 1984. Sedimentology of the Upper Proterozoic glacial record, Vestertana Group, Finnmark, North Norway. Norges Geologiske Undersøkelse Bulletin, 394, 76. Eisbacher, G. H. 1985. Late Proterozoic rifting, glacial sedimentation and sedimentary cycles in the light of Windermere deposition, western Canada. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 231– 254. Embleton, B. J. J. & Williams, G. E. 1986. Low latitude of deposition for late Precambrian periglacial varvites in South Australia: implications for palaeoclimatology. Earth and Planetary Science Letters, 79, 419– 430. Fairchild, I. J. & Kennedy, M. J. 2007. Neoproterozoic glaciation in the Earth System. Journal of the Geological Society, London, 164, 895– 921. Fanning, C. M. & Link, P. K. 2008. Age constraints for the Sturtian glaciation: data from the Adelaide Geosyncline, South Australia and Pocatello Formation, Idaho, USA. Selwyn Symposium 2008, Geological Society of Australia, Extended Abstracts, 91, 57 –62. Fo¨lling, P. G. & Frimmel, H. E. 2002. Chemostratigraphy correlation of carbonate successions in the Gariep and Saldania Belts, Namibia and South Africa. Basin Research, 14, 69 – 88. Fo¨lling, P. G., Zartman, R. E. & Frimmel, H. E. 2000. A novel approach to double-spike Pb/Pb dating of carbonate rocks: examples from Neoproterozoic sequences in southern Africa. Chemical Geology, 171, 97 –122. Font, E., Ne´de´lec, A., Trindade, R. I. F., Macouin, M. & Charrie`re, A. 2006. Chemostratigraphy of the Neoproterozoic Mirassol d’Oeste cap dolostones (Mato Grosso, Brazil): an alternative model for Marinoan cap dolostone formation. Earth and Planetary Science Letters, 250, 89 –103. Font, E., Trindade, R. I. F. & Ne´de´lec, A. 2005. Detrital remanent magnetization in haematite-bearing Neoproterozoic Puga cap dolostone, Amazon craton: a rock magnetic and SEM study. Geophysical Journal International, 163, 491–500. Frimmel, H. E. 2004. Neoproterozoic sedimentation rates and timing of glaciations – a southern African perspective. In: Eriksson, P. G., Altermann, W., Nelson, D. R., Mueller, W. U. & Catuneanu, O. (eds) The Precambrian Earth: Tempos and Events. Elsevier, Amsterdam, 459– 473. Frimmel, H. E. 2008. Neoproterozoic Gariep Orogen. In: Miller, R. McG. (ed.) The Geology of Namibia (In Three Volumes), Volume 2 (Neoproterozoic to Lower Palaeozoic). Geological Survey of Namibia, Windhoek, 14-1–14-39. Frimmel, H. E. & Von Veh, M. W. 2003. Numees formation (including the Jakkalsberg Member). In: Johnson, M. R. (ed.) Catalogue of South African Lithostratigraphic Units. South African Committee for Stratigraphy, Pretoria, 7-25– 7-28.

Frimmel, H. E. 2011. The Kaigas and Numees Formations, Port Nolloth Group, in South Africa and Namibia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 223– 231. Gammon, P. R., McKirdy, D. M. & Smith, H. D. 2005. The timing and environment of tepee formation in a Marinoan cap carbonate. Sedimentary Geology, 177, 195– 208. Gaucher, C., Frimmel, H. E. & Germs, G. J. B. 2005. Organic-walled microfossils and biostratigraphy of the upper Port Nolloth Group (Namibia): implications for latest Neoproterozoic glaciations. Geological Magazine, 142, 539–559. Goodman, J. & Pierrehumbert, R. T. 2003. Glacial flow of floating marine ice in ‘Snowball Earth’. Journal of Geophysical Research, 108, 3308, doi: 10.1029/2002JC001471. Graf, J. L., Jr, O’Connor, E. A. & Van Leeuwin, P. 1994. Rare earth element evidence of origin and depositional environment of Late Proterozoic ironstone beds and manganese-oxide deposits, SW Brazil and SE Bolivia. Journal of South American Earth Sciences, 7, 115– 133. Grotzinger, J. P. & James, N. P. 2000. Precambrian carbonates: evolution of understanding. In: Grotzinger, J. P. & James, N. P. (eds) Carbonate Sedimentation and Diagenesis in the Evolving Precambrian World. SEPM (Society for Sedimentary Geology) Special Publication, 67, 3– 20. Grotzinger, J. P. & Knoll, A. H. 1995. Anomalous carbonate precipitates: is the Precambrian the key to the Permian? Palaios, 10, 578– 596. Gubbins, D. 1999. The distinction between geomagnetic excursions and reversals. Geophysical Journal International, 137, F1– F3. Halverson, G. P. & Shields, G. 2011. Chemostratigraphy and the Neoproterozoic glaciations. In: Arnaud, E., Halverson, G. P. & Shields, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 51 –66. Halverson, G. P., Maloof, A. C. & Hoffman, P. F. 2004. The Marinoan glaciation (Neoproterozoic) in northeast Svalbard. Basin Research, 16, 297– 324. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Toward a Neoproterozoic composite carbon-isotope record. Geological Society of America Bulletin, 117, 1181– 1207. ¨ ., Maloof, A. C. & Bowring, S. A. Halverson, G. P., Duda´s, F. O 2007. Evolution of the 87Sr/86Sr composition of Neoproterozoic seawater. Palaeogeography, Palaeoclimatology, Palaeoecology, 256, 103– 129. Halverson, G. P., Poitrasson, F., Hoffman, P. F., Ne´de´lec, A., Montel, J.-M. & Kirby, J. 2011. Fe isotope and trace element geochemistry of the Neoproterozoic syn-glacial Raptian iron formation. Earth and Planetary Science Letters, 309, 100–112. Hegenberger, W. 1993. Stratigraphy and sedimentology of the Late Precambrian Witvlei and Nama Groups, East of Windhoek. Geological Survey of Namibia Memoir, 17, 82. Higgins, J. A. & Schrag, D. P. 2003. Aftermath of a snowball Earth. Geophysics, Geochemistry, Geosystems, 4, doi:10.1029/2002GC000403. Higgins, J. A., Fischer, W. W. & Schrag, D. P. 2009. Oxygenation of the ocean and sediments: consequences for the seafloor carbonate factory. Earth and Planetary Science Letters, 284, 25 – 33. Hildebrand, R. S. 2009. Did westward subduction cause Cretaceous– Tertiary orogeny in the North American Cordillera? Geological Society of America, Special Paper, 457, 71. Hoffman, P. F. 2011. Strange bedfellows: glacial diamictite and cap carbonate from the Marinoan (635 Ma) glaciation in Namibia. Sedimentology, 58 – 119. Hoffman, P. F. & Halverson, G. P. 2008. Otavi group of the western northern platform, the eastern Kaoko Zone and the western northern Margin Zone. In: Miller, R. McG. (ed.) The Geology of Namibia, Vol. 2: Neoproterozoic to Lower Palaeozoic. Handbook of the Geological Survey of Namibia, Windhoek, 13.69– 13.136. Hoffman, P. F. & Halverson, G. P. 2011. Neoproterozoic glacial record in the Mackenzie Mountains, northern Canadian Cordillera. In: Arnaud, E., Halverson, G. P. & Shields, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 397 –411.

CHEMICAL SEDIMENTS ASSOCIATED WITH NEOPROTEROZOIC GLACIATION

Hoffman, P. F. & Li, Z. X. 2009. A palaeogeographic context for Neoproterozoic glaciation. Palaeogeography, Palaeoclimatology, Palaeoecology, 277, 158–172. Hoffman, P. F. & Macdonald, F. A. 2010. Sheet-crack cements and early regression in Marinoan (635 Ma) cap dolostones: regional benchmarks of vanishing ice-sheets? Earth and Planetary Science Letters, 300, 374– 384. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129– 155. Hoffman, P. F., Kaufman, A. J., Halverson, G. P. & Schrag, D. P. 1998. A Neoproterozoic Snowball Earth. Science, 281, 1342– 1346. Hoffman, P. F., Halverson, G. P., Domack, E. W., Husson, J. M., Higgins, J. A. & Schrag, D. P. 2007. Are basal Ediacaran (635 Ma) post-glacial ‘cap dolostones’ diachronous? Earth and Planetary Science Letters, 258, 114– 131. Hoffmann, K.-H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. U– Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817– 820. Hurtgen, M. T., Halverson, G. P., Arthur, M. A. & Hoffman, P. F. 2006. Sulfur cycling in the aftermath of a 635-Ma snowball glaciation: evidence for a syn-glacial sulfidic deep ocean. Earth and Planetary Science Letters, 245, 551–570. Hyde, W. T., Crowley, T. J., Baum, S. K. & Peltier, W. R. 2000. Neoproterozoic ‘snowball Earth’ simulations with a coupled climate/ice-sheet model. Nature, 405, 425–429. Ilyin, A. V. 2009. Neoproterozoic banded iron formations. Lithology and Mineral Resources, 44, 78 –86. Ilyin, A. V., Zaitsev, N. S. & Bjamba, Z. 1986. Khubsugul, Mongolia people’s republic. In: Cook, P. J. & Shergold, J. H. (eds) Phosphate Deposits of the World, Vol. 1: Proterozoic and Cambrian Phosphorites. Cambridge University Press, Cambridge. James, N. P., Narbonne, G. M. & Kyser, T. K. 2001. Late Neoproterozoic cap carbonates: Mackenzie Mountains, northwestern Canada: precipitation and global glacial meltdown. Canadian Journal of Earth Sciences, 38, 1229–1262. Jiafu, T., Heqin, F. & Zhioiu, G. 1987. Stratigraphy, type and formation conditions of the Late Precambrian banded iron ores in South China. Chinese Journal of Geochemistry, 6, 332–351. Jiang, G., Kennedy, M. J. & Christie-Blick, N. 2003. Stable isotopic evidence for methane seeps in Neoproterozoic postglacial cap carbonates. Nature, 426, 822– 826. Jiang, G., Kennedy, M. J., Christie-Blick, N., Wu, H. & Zhang, S. 2006. Stratigraphy, sedimentary structures and textures of the late Neoproterozoic Doushantuo cap carbonate in South China. Journal of Sedimentary Research, 76, 978– 995. Johnson, H. P., Van Patten, D., Tivey, M. & Sager, W. W. 1995. Geomagnetic polarity reversal rate for the Phanerozoic. Geophysical Research Letters, 22, 231– 234. Kaufman, A. J., Jiang, G., Christie-Blick, N., Banerjee, D. M. & Rai, V. 2006. Stable isotope record of the terminal Neoproterozoic Krol platform in the Lesser Himalayas of northern India. Precambrian Research, 147, 156– 185. Kendall, C. G. St. C. & Warren, J. 1987. A review of the origin and setting of tepees and their associated fabrics. Sedimentology, 34, 1007– 1027. Kennedy, M. J. 1996. Stratigraphy, sedimentology and isotopic geochemistry of Australian Neoproterozoic postglacial cap dolostones: deglaciation, d13C excursions and carbonate precipitation. Journal of Sedimentary Research, 66, 1050– 1064. Kennedy, M. J., Runnegar, B., Prave, A. R., Hoffmann, K.-H. & Arthur, M. A. 1998. Two of four Neoproterozoic glaciations? Geology, 26, 1059– 1063. Kennedy, M. J., Christie-Blick, N. & Sohl, L. E. 2001. Are Proterozoic cap carbonates and isotopic excursions a record of gas hydrate destabilization following Earth’s coldest intervals? Geology, 29, 443– 446. Kenward, P. A., Goldstein, R. H., Gonza´lez, L. A. & Roberts, J. A. 2009. Precipitation of low-temperature dolomite from an aerobic microbial consortium: the role of methanogenic Archaea. Geobiology, 7, 1 – 10. Kianian, M. & Khakzad, A. 2008. Geochemistry of glaciogenic Neoproterozoic banded iron-formations from Kerman District (Iran). 33rd International Geological Congress, Oslo, Abstracts, Session CGC-04.

79

Klein, C. 2005. Some Precambrian banded iron-formations (BIFs) from around the world: their age, geologic setting, mineralogy, metamorphism, geochemistry and origin. American Mineralogist, 90, 1473–1499. Klein, C. & Beukes, N. J. 1993. Sedimentology and geochemistry of the glacigenic Late Proterozoic Rapitan iron-formation in Canada. Economic Geology, 88, 542–565. Klein, C. & Ladeira, E. A. 2004. Geochemistry and mineralogy of Neoproterozoic banded iron-formations and some selected, siliceous manganese formations from the Urucum District, Mato Grosso do Sul, Brazil. Economic Geology, 99, 1233– 1244. Knoll, A. H., Walter, M. R., Narbonne, G. M. & Christie-Blick, N. 2006. The Ediacaran Period: a new addition to the geologic time scale. Lethaia, 39, 13– 30. Kump, L. R. & Seyfried, W. E. Jr. 2005. Hydrothermal Fe fluxes during the Precambrian: effect of low oceanic sulfate concentrations and low hydrostatic pressure on the composition of black smokers. Earth and Planetary Science Letters, 235, 654– 662. Laj, C., Kissel, C. & Roberts, A. P. 2006. Geomagnetic field behavior during the Iceland Basin and Laschamp geomagnetic excursions: a simple transitional field geometry? Geochemistry, Geophysics, Geosystems, 7, Q03004, doi:10.1029/2005GC001122. Le Hir, G. & Donnadieu, Y. et al. 2009. The snowball Earth aftermath: exploring the limits of continental weathering processes. Earth and Planetary Science Letters, 277, 453– 463. Li, Z. X. 2000. New palaeomagnetic results from the ‘cap dolomite’ of the Neoproterozoic Walsh Tillite, northwestern Australia. Precambrian Research, 100, 359–370. Li, Z. X. & Bogdanova, S. V. et al. 2008. Assembly, configuration and break-up history of Rodinia: a synthesis. Precambrian Research, 160, 179– 210. Lottermoser, B. G. & Ashley, P. M. 2000. Geochemistry, petrology and origin of Neoproterozoic ironstones in the eastern part of the Adelaide Geosyncline. Precambrian Research, 101, 49 –67. Lythe, M. B., Vaughan, D. G. & BEDMAP CONSORTIUM 2001. BEDMAP: a new thickness and subglacial topographic model of Antarctica. Journal of Geophysical Research, 106, 11335–11351. Macdonald, F. A. 2011. The Tsagaan Oloom Formation, southwestern Mongolia. In: Arnaud, E., Halverson, G. P. & Shields, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 331– 337. Macdonald, F. A. & Cohen, P. A. 2011. The Tatonduk inlier, Alaska– Yukon border. In: Arnaud, E., Halverson, G. P. & Shields, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 389–396. Macdonald, F. A. & Jones, D. S. 2011. The Khubsugul Group, northern Mongolia. In: Arnaud, E., Halverson, G. P. & Shields, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 339– 345. Macdonald, F. A., Mcclelland, W. C., Schrag, D. P. & Macdonald, W. P. 2009. Neoproterozoic glaciation on a carbonate platform margin in Arctic Alaska and the origin of the North Slope subterrane. Geological Society of America Bulletin, 121, 448– 473. Macdonald, F. A., Schmidtz, M. D. et al. 2010. Calibrating the Cryogenian. Science, 327, 1241– 1243. ¨ . & Schrag, Macdonald, F. A., Strauss, J. V., Rose, C., Duda´s, F. O D. P. 2011. Stratigraphy of the Port Nolloth Group of Namibia and South Africa and implications for the age of Neoproterozoic iron formations. American Journal of Science, 310, 862–888. Martin, H. 1965a. The Precambrian Geology of South West Africa and Namaqualand. University of Cape Town, Precambrian Research Unit Bulletin, 1, 1– 159. Martin, H. 1965b. Beobachtungen zum Problem der jung-pra¨kambrischen Glazialen Ablagerungen in Su¨dwestafrika (Observations concerning the problem of the late Precambrian glacial deposits in South West Africa). Geologische Rundschau, 54, 115– 127. Mikucki, J. A., Pearson, A. et al. 2009. A contemporary microbially maintained subglacial ferrous ‘ocean’. Science, 324, 397–400. Moore, T. S., Murray, R. W., Kurtz, A. C. & Schrag, D. P. 2004. Anaerobic methane oxidation and the formation of dolomite. Earth and Planetary Science Letters, 229, 141–154.

80

P. F. HOFFMAN ET AL.

Myrow, P. M. & Kaufman, A. J. 1999. A newly discovered cap carbonate above Varanger-age glacial deposits in Newfoundland, Canada. Journal of Sedimentary Research, 69, 784– 793. Ne´de´lec, A., Affaton, P., France-Lanord, C., Charrie`re, A. & Alvaro, J. 2007. Sedimentology and chemostratigraphy of the Bwipe Neoproterozoic cap dolostones (Ghana, Volta Basin): a record of microbial activity in a peritidal environment. C. R. Geoscience, 339, 223– 239. Nogueira, A. C. R., Riccomini, C., Sial, A. N., Moura, C. A. V. & Fairchild, T. R. 2003. Soft-sediment deformation at the base of the Neoproterozoic Puga cap carbonate (southwestern Amazon craton, Brazil): confirmation of rapid icehouse to greenhouse transition in snowball Earth. Geology, 31, 613–616. Nogueira, A. C. R., Riccomini, C., Sial, A. N., Moura, C. A. V., Trindade, R. I. F. & Fairchild, T. R. 2007. Carbon and strontium isotope fluctuations and paleoceanographic changes in late Neoproterozoic Araras carbonate platform, southern Amazon craton, Brazil. Chemical Geology, 237, 168–190. Pierrehumbert, R. T. 2002. The hydrologic cycle in deep-time climate problems. Nature, 419, 191–198. Plummer, P. S. 1978. Note on the palaeoenvironmental significance of the Nuccaleena Formation (upper Precambrian), central Flinders Ranges, South Australia. Journal of the Geological Society of Australia, 25, 395– 402. Prave, A. R., Hoffmann, K.-H., Hegenberger, W. & Fallick, A. E. 2011. The Witvlei Group of east-central Namibia. In: Arnaud, E., Halverson, G. P. & Shields, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 211– 216. Preiss, W. V. 1987. The Adelaide Geosyncline: Late Proterozoic stratigraphy, sedimentation, palaeontology and tectonics. Geological Survey of South Australia Bulletin, 53, 438. Preiss, W. V., Walter, M. R., Coates, R. P. & Wills, A. T. 1978. Lithological correlation of the Adelaidean glaciogenic rocks in parts of the Amadeus, Ngalia & Georgina basins. Bureau of Mineral Resources, Journal of Australian Geology and Geophysics, 3, 43 –53. Pruss, S. B., Bosak, T., Macdonald, F. A., Mclane, M. & Hoffman, P. F. 2010. Microbial facies in a Sturtian cap carbonate, the Rasthof Formation, Otavi Group, northern Namibia. Precambrian Research, 181, 187– 108. Rankama, K. 1973. The Late Precambrian glaciation, with particular reference to the Southern Hemisphere. Journal and Proceedings, Royal Society of New South Wales, 106, 89– 97. Raub, T. D. 2008. Prolonged deglaciation of ‘Snowball Earth’. PhD thesis, Yale University, CT. Ridgwell, A. J., Kennedy, M. J. & Caldeira, K. 2003. Carbonate deposition, climate stability and Neoproterozoic ice ages. Science, 302, 859– 862. Roberts, A. P. 2008. Geomagnetic excursions: knowns and unknowns. Geophysical Research Letters, 35, L17307, doi: 10.1029/ 2008GL034719. Roberts, A. P. & Lewin-Harris, J. C. 2000. Marine magnetic anomalies: evidence that ‘tiny wiggles’ represent short-period geomagnetic polarity intervals. Earth and Planetary Science Letters, 183, 375– 388. Sheldon, R. P. 1984. Ice-ring origin of the Earth’s atmosphere and hydrosphere and Late Proterozoic –Cambrian hypothesis. Geological Survey of India Special Publication, 17, 17 –21. Shields, G. A. 2005. Neoproterozoic cap carbonates: a critical appraisal of existing models and the plumeworld hypothesis. Terra Nova, 17, 299– 310. Shields, G. A., Deynoux, M., Strauss, H., Paquet, H. & Nahon, D. 2007. Barite-bearing cap dolostone of the Taoude´ni Basin, northwest Africa: sedimentary and isotopic evidence for methane seepage after a Neoproterozoic glaciation. Precambrian Research, 154, 209– 235. Smith, L. H., Kaufman, A. J., Knoll, A. H. & Link, P. K. 1994. Chemostratigraphy of predominantly siliciclastic Neoproterozoic successions: a case study of the Pocatello Formation and lower Brigham Group, Idaho. Geological Magazine, 131, 301–314.

Sovetov, Yu. K. & Komlev, D. A. 2005. Tillites at the base of the Oselok Group, foothills of the Sayan Mountains, and the Vendian lower boundary in the southwestern Siberian Platform. Stratigraphy and Geological Correlations, 13, 337–366. Sumner, D. Y. 2002. Decimetre-thick encrustations of calcite and aragonite on the sea-floor and implications for Neoarchaean and Neoproterozoic ocean chemistry. Special Publications of the International Association of Sedimentologists, 33, 107–120. Tojo, B., Katsuta, N., Takano, M., Kawakami, S. & Ohno, T. 2007. Calcite – dolomite cycles in the Neoproterozoic Cap carbonates, Otavi Group, Namibia. In: Vickers-Rich, P. & Komarower, P. (eds) The Rise and Fall of the Ediacaran Biota. Geological Society, London, Special Publication, 286, 103– 113. Trindade, R. I. F., Font, E., D’agrella-Filho, M. S., Nogueira, A. C. R. & Riccimini, C. 2003. Low-latitude and multiple geomagnetic reversals in the Neoproterozoic Puga cap carbonate, Amazon craton. Terra Nova, 15, 441–446. Trompette, R. 1994. Geology of Western Gondwana (2000– 500 Ma): Pan-African –Brasiliano Aggregation of South America and Africa. Balkema, Amsterdam. Trompette, R., Affaton, P., Joulia, F. & Marchand, J. 1980. Stratigraphic and structural controls of late Precambrian phosphate deposits of the Northern Volta Basin in Upper Volta, Niger and Benin, West Africa. Economic Geology, 75, 62– 70. Trompette, R., de Alvarenga, C. J. A. & Wade, D. 1998. Geological evolution of the Neoproterozoicv Corumba´ graben system (Brazil). Depositional context of the stratified Fe and Mn ores of the Jacadigo Group. Journal of South American Earth Sciences, 11, 587–597. Urban, H., Stribny, B. & Lippolt, H. J. 1992. Iron and manganese deposits of the Urucum District, Mato Grosso do Sul, Brazil. Economic Geology, 87, 1375–1892. van Lith, Y., Warthmann, R., Vasconcelos, C. & McKenzie, J. A. 2003. Sulphate-reducing bacteria induce low-temperature Cadolomite and high Mg-calcite formation. Geobiology, 1, 71 – 79. Walde, D. H. G., Gierth, E. & Leonardos, O. H. 1981. Stratigraphy and mineralogy of the manganese ores of Urucum, Mato Grosso, Brazil. Geologische Rundschau, 70, 1077– 1085. Wallace, J. M. & Hobbs, P. V. 1977. Atmospheric Science: An Introductory Survey. Academic Press, San Diego. Walter, M. R. & Bauld, J. 1983. The association of sulphate evaporites, stromatolitic carbonates and glacial sediments: examples from the Proterozoic of Australia and the Cainozoic of Antarctic. Precambrian Research, 21, 129– 148. Wang, J., Jiang, G., Xiao, S., Li, Q. & Wei, Q. 2008. Carbon isotope evidence for widespread methane seeps in the ca 635 Ma Doushantuo cap carbonate in South China. Geology, 36, 347– 350. Warren, S. G., Brandt, R. E., Grenfell, T. C. & Mckay, C. P. 2002. Snowball Earth: ice thickness on the tropical ocean. Journal of Geophysical Research, 107, 3167, doi: 10.1029/2001JC001123. Whitten, G. F. 1970. The investigation and exploitation of the Razorback Ridge iron deposit. Geological Survey of South Australia Reports of Investigations, 33, 165. Xiao, S., Bao, H. et al. 2004. The Neoproterozoic Quruqtagh Group in eastern Chinese Tianshan: evidence for a post-Marinoan glaciation. Precambrian Research, 130, 1 –26. Yeo, G. M. 1981. The Late Proterozoic Rapitan glaciation in the northern Cordillera. In: Campbell, F. H. A. (ed.) Proterozoic Basins of Canada. Geological Survey of Canada Paper, 81– 10, 25 –46. Yeo, G. M. 1986. Iron-formation in the late Proterozoic Rapitan Group, Yukon and Northwest Territories. In: Morin, J. A. (ed.) Mineral Deposits of Northern Cordillera. Canadian Institute of Mining and Metallurgy Special Volume, 37, 142– 153. Young, G. M. 1976. Iron-formation and glaciogenic rocks of the Rapitan Group, Northwest Territories, Canada. Precambrian Research, 3, 137– 158. Zhang, S., Jiang, G. & Han, Y. 2008. The age of the Nantuo Formation and Nantuo glaciation in South China. Terra Nova, 20, 289–294. Zhou, C., Bao, H., Peng, Y. & Yuan, X. 2010. Timing of deposition of 17 O-depleted barite at the Nantuo glacial meltdown in South China. Geology, 38, 903–906.

Chapter 6 A review of the Chemical Index of Alteration (CIA) and its application to the study of Neoproterozoic glacial deposits and climate transitions HEINRICH BAHLBURG1* & NICOLE DOBRZINSKI2 1

Institut fu¨r Geologie und Pala¨ontologie, Westfa¨lische Wilhelms-Universita¨t, 48149 Mu¨nster, Germany 2

RWE Dea AG, Wietze Laboratory, 29323 Wietze, Germany

*Corresponding author (e-mail: [email protected]) Abstract: The Chemical Index of Alteration (CIA) is the most accepted of available weathering indices. Past conditions of physical and chemical weathering can be reliably inferred if application of the CIA is combined with a comprehensive facies analysis. When applied to the reconstruction of climate conditions during Neoproterozoic times, CIA data provide crucial insights into the changes in the relative contributions of chemical and physical weathering in the production of sedimentary detritus. CIA data are thus instrumental not only in documenting changes between icehouse and greenhouse climates, but also in recognizing shorter-term climate oscillations between glacial and warm– humid conditions. Concerning the Neoproterozoic glacial periods, sedimentological and CIA data sets give strong evidence of a functioning hydrological cycle, operative sediment routing systems, and variable climate conditions oscillating between dry–cool and glacial, and warm–humid and interglacial. These findings are incompatible with the hypothesis of a totally ice-covered Snowball Earth.

Climate exerts the major control on weathering processes affecting the upper continental crust. Direct evidence of past weathering conditions and thus climate can be obtained from palaeosols through a combination of, among other things, field observation, petrography (particularly diagenetic phases), X-ray diffractometry and whole rock and isotope geochemistry. However, palaeosols are restricted to continental environments, which commonly have a relatively low preservation potential. Even though palaeosoils as old as Palaeoproterozoic are reported in the literature (e.g. Gutzmer & Beukes 1998; Nedachi et al. 2005), most information on weathering conditions and thus climate has to be gleaned from reworked siliciclastic material deposited in marine or lacustrine environments. In the Precambrian eon, which represents nearly 90% of Earth history, marine siliciclastic sedimentary rocks are the most widely available source of information on past climates. In Phanerozoic rocks, information derived from siliciclastic sedimentary rocks may be supplemented by ecological indicators derived from fossils. However, the bearing of palaeofaunal data on the interpretations of palaeoclimate conditions needs to be considered very carefully, because differences between continental regimes of weathering and marine environments of deposition may be very large. This is illustrated for example by the pronounced difference at the coast of northern Chile between conditions of extreme subtropical aridity in the Atacama desert and the cold waters of the Humboldt current originating in the periAntarctic westwind drift. Extreme climate states are of particular interest to palaeoclimate research. Among these, those that figure most prominently are the icehouse climates that happened in at least five cycles of highly variable duration and extent during Earth history, that is, the Palaeoproterozoic and Neoproterozoic, Ordovician, Carboniferous –Permian and Neogene glaciations (Crowley & North 1991; Crowell 1999). Of these, the Neoproterozoic glaciations are some of the most discussed and contentious (e.g. Hoffman et al. 1998; Evans 2000; Hoffman & Schrag 2002; Eyles & Januszczak 2007; Fairchild & Kennedy 2007; Etienne et al. 2008; Eyles 2008). The timing and causes of the inception and termination of Neoproterozoic glacial states as well as variations in the severity of the icehouse conditions themselves remain unresolved issues (e.g.

Schrag et al. 2002; Baum & Crowley 2003; Fairchild & Kennedy 2007; Shields 2008). Unfortunately, in many cases, facies analysis of sedimentary rocks does not supply unambiguous information on palaeoclimate conditions. This conundrum is well illustrated by the mutually exclusive interpretations of the Neoproterozoic, that is, Marinoan, Ghaub Formation diamictites (Otavi Group) in northern Namibia. These have been interpreted either as glaciogenic sediments (Gevers 1931; Martin 1964; Hoffmann & Prave 1996; Hoffman & Schrag 2002) or as tectonically triggered mass flow deposits (Schermerhorn 1974; Martin et al. 1985; Eyles & Januszczak 2007). However, these issues may be moved closer to resolution through a combination of facies analysis with considerations of weathering as a function of climate (e.g. Young & Nesbitt 1999; Young 2001; Scheffler et al. 2003). In this contribution we will review the most important weathering proxies applied to siliciclastic sedimentary rocks. We will concentrate on the CIA (Nesbitt & Young 1982) and its potential and limitations for deriving palaeoclimate information from glacial and pre- and post-glacial siliciclastic successions. We will preferentially but not exclusively address Neoproterozoic glacial successions based on a combination of published information from the Nanhuan –Sinian glacial succession in South China (Dobrzinski et al. 2004; Dobrzinski & Bahlburg 2007) and new data on the Port Askaig Formation (Scotland), the Ghaub Formation (northern Namibia) and the Mortensnes and Smalfjord formations (northern Norway).

Weathering and weathering indices Exposed rocks are affected to variable degrees by a combination of chemical and physical weathering (Bland & Rolls 1998). Progressive chemical weathering of labile minerals like feldspar leads to the loss of Ca2þ, Kþ and Naþ and the transformation to minerals more stable under surface conditions (Fedo et al. 1995). Ultimately, it results in the formation of shales rich in clay minerals such as illite and kaolinite, and Fe-oxyhydrates such as goethite. Physical weathering, in turn, leads to the degradation of rocks to smaller grain sizes, ideally without causing geochemical and

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 81– 92. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.6

82

H. BAHLBERG & N. DOBRZINSKI

mineralogical changes. If physical weathering, and the grinding action of moving ice-lodged debris in particular, degrades a source rock into a clay-sized deposit, it should essentially preserve the mineralogical and geochemical composition of the original rock (Nesbitt & Young 1982, 1996). Consequently, the character of the climate framework and the way it governs weathering conditions and reactions is reflected by the mineralogical and mobile element geochemical composition of the resulting deposits, and the shales and clay-sized materials in particular. Ultimately, chemical weathering of silicate rocks leads through hydrolysis to an exchange of the cations Naþ, Kþ, Ca2þ and Mg2þ for Hþ, and maybe a loss of Si4þ (Kramer 1968). Naþ, Kþ and Ca2þ are commonly supplied by the weathering of feldspar and volcanic glass, together accounting for c. 58% of the exposed crust. Mg2þ is derived from glasses, sheet silicates and mafic minerals and resides in chloritic and smectitic clays (Nesbitt & Young 1984; Pettijohn et al. 1987). In general, hydrolytic weathering causes a progressive transformation of affected components into clay minerals, ultimately kaolinite. In a qualitative way, mineralogical changes can be detected in shales and sandstones using X-ray diffractometry. Optical analysis is a quantitative option in sandstones only. Here, the Mineralogical Index of Alteration (MIA), given by the ratio of quartz to the sum of quartz þ K-feldspar þ plagioclase, permits an assessment of weathering effects (Johnsson 1993). Quantitative estimations of weathering in fine-grained as well as coarse-grained rocks are relatively easily achieved by using whole-rock geochemical data to calculate geochemical weathering proxies. A comprehensive review of a great number of geochemical weathering indices has been presented by Duzgoren-Aydin et al. (2002). Recently, von Eynatten et al. (2003) and von Eynatten (2004) presented a quantitative approach, the t-index, to statistically model linear compositional and weathering trends. If corroborated by further studies the t-index will likely lead to more quantitative definitions of weathering trends. A related statistical approach has been taken by Ohta & Arai (2007), whose W index is based on principal component analysis of eight major oxides. As yet, it has been applied only to igneous rocks. Geochemical estimations of weathering effects need to be considered carefully, because the major cations Naþ, Kþ and Ca2þ may also be mobile under diagenetic conditions (Wintsch & Kvale 1994). Geochemical weathering proxies make use of the

changes of bulk-rock geochemical composition caused by chemical alteration. A very simple proxy is the Ruxton Ratio R (Ruxton 1968) given by the SiO2/Al2O3 ratio. This assumes that Al2O3 remains immobile during weathering, and changes in R therefore reflect silica loss as a proxy for total element loss. The Ruxton Ratio may be useful when weathering profiles on rocks of felsic and intermediate composition are considered, but was found to be poorly correlated to the actual weathering grade of a silicate rock (Duzgoren-Aydin et al. 2002). As the transformation of feldspar to clay minerals and the coincident mobility of the main cations is a major process of chemical (i.e. hydrolytical) weathering, Parker (1970) considered it more useful to mirror changes in Naþ, Kþ, Ca2þ and Mg2þ and created a weathering index (WIP, Weathering Index of Parker) given by

WIP ¼

   Na Mg K Ca :35 þ :9 þ :25 þ :7  100 0 0 0 0

(1)

where the cations* represent the atomic percentage of an element divided by the atomic weight. Parker (1970) also considered the susceptibility of these elements to weathering by including in the denominator Nicholls’ values of bond strength as a measure of the energy necessary to break the cation-to-oxygen bonds of the respective oxides. These different values are considered to reflect the probability of an element being mobilized during the weathering process. Values of WIP are commonly between 100 and 0, with the least weathered rocks having the highest values (Fig. 6.1). The WIP implicitly assumes that all Ca2þ in a silicate rock is contained in silicate minerals. This simplification is a source of imprecision of the index, particularly if larger amounts of carbonate detritus or cements are present in the rock. More problematic still is the lack of consideration of a relatively immobile reference phase like Al2O3 in the formula, which would help to monitor relative changes of composition of the relevant mineral components. The disadvantages of the WIP are overcome in the Chemical Index of Alteration (CIA) using whole-rock geochemical data of major element oxides (Nesbitt & Young 1982). The index is essentially based on the same considerations that led Kramer (1968) to

Fig. 6.1. Relationship between two weathering proxies, WIP (Parker 1970) and CIA (Nesbitt & Young 1982). Shown are data obtained from the matrix of Neoproterozoic glacial diamictites from South China (Dobrzinski et al. 2004), the Port Askaig Formation, Scotland (Panahi & Young 1997, and this study, Table 6.3), the Ghaub Formation of northern Namibia (Table 6.4), and the Smalfjord and Mortensnes formations, northern Norway (Table 6.4). Also shown are data from non-glacial shales and siltstones including those underlying, intercalated with and overlying the South China glacial diamictites, and well-weathered Palaeozoic shales from the Cantabrian mountain belt of Hercynian age in northern Spain (Table 6.1).

A REVIEW OF THE CHEMICAL INDEX OF ALTERATION

83

Table 6.1. Sampled formations and compositional groupings discussed in this contribution Lithostratigraphic units, terminology, this paper

Formation

Region

Rock types

Devonian and Carboniferous formations, Cantabria

Vergan˜o, Van˜es, Carmen, Potes Group, Murcia

Northern Spain

Mortensnes and Smalfjord formations Port Askaig Formation

Mortensnes, Smalfjord

Northern Norway

Quartz sandstone, sandstone, siltstone, shale Diamictite, shale

Port Askaig

Scotland

Ghaub diamictites, Namibia

Ghaub

Non-glacial, South China (post-glacial)

Doushantuo, Jinjiadong

Glacial, South China, (upper diamictites)

Nantuo, Hongjiang, Leigongwu

Non-glacial, South China (inter-glacial)

Datanpo, Xianmeng, Lantian

Glacial, South China, (lower diamictites)

Dongshanfeng, Jiangkou, Tie-si-ao

Non-glacial, South China (pre-glacial)

Xieshuihe, Wuqiangxi, Zitang

Otavi Mountains, northern Namibia Yangtze platform, South China Yangtze platform, South China Yangtze platform, South China Yangtze platform, South China Yangtze platform, South China



 Al2 O3  100 Al2 O3 þ Na2 O þ K2 O þ CaO

Depositional age

Ref.

Shale, siltstone

Late Devonian to Carboniferous

Wagner & Wagner-Gentis (1963), IGME (1984), ITGE (1994), Keller et al. (2008)

Siltstone, shale

Ediacaran, Neoproterozoic

Diamictite

Diamictite matrix: shale, siltstone

Cryogenian, Neoproterozoic

Diamictite, siltstone

Silicate shale matrix, siltstone

Cryogenian, Neoproterozoic

Rice & Hoffmann (2001), Arnaud & Eyles (2002), Rice et al. (2011) Panahi & Young (1997), Arnaud & Eyles (2006), Benn & Prave (2006), Arnaud & Fairchild (2011) Hoffman & Schrag (2002), Eyles (2007), Hoffman (2011)

Carbonate, shale

Shale

Ediacaran, Neoproterozoic

Diamictite

Diamictite matrix: shale, siltstone Shale

Cryogenian, Neoproterozoic

Diamictite matrix: shale, siltstone Shale

Cryogenian, Neoproterozoic

Carbonate, shale Diamictite

Sandstone, shale

conclude that monitoring the hydrolysis of feldspar and volcanic glass and the respective changes in the content of the major cations offers the best quantitative measure of chemical weathering. It represents a ratio of predominantly immobile Al2O3 to the mobile cations Naþ, Kþ and Ca2þ given as oxides. The CIA is defined as

CIA ¼

Sampled lithology

(2)

where the major element oxides are given in molecular proportions. CaO* represents the CaO content of silicate minerals only (Fedo et al. 1995) and thus eliminates one of the disadvantages of the WIP. Kaolinite has a CIA value of 100 and represents the highest degree of weathering. Illite is between 75 and c. 90, muscovite is at 75, and the feldspars at 50. Fresh basalts have values between 30 and 45, fresh granites and granodiorites between 45 and 55 (Nesbitt & Young 1982; Fedo et al. 1995; Fig. 6.1). Careful attention has to be paid to the potential presence of clastic carbonate grains or of carbonate cement in the sedimentary rock. If undetected but abundant, both would lead to very low and unrealistic CIA values. The presence of carbonate grains or cements (calcite or dolomite) and the ratio of calcite to dolomite have to be determined petrographically. Both can be effectively accounted for geochemically by determining total MgO and CaO and the total inorganic carbon (TIC) content of a sample, all of which can then be used to calculate relative contributions to CO2, and finally CaO* (Fedo et al. 1995; Tables 6.2, 6.3, 6.4).

Cryogenian, Neoproterozoic

Tonian and Cryogenian, Neoproterozoic

Dobrzinski et al. (2004), Dobrzinski & Bahlburg (2007), Zhang et al. (2011) Dobrzinski et al. (2004), Dobrzinski & Bahlburg (2007), Zhang et al. (2011) Dobrzinski et al. (2004), Dobrzinski & Bahlburg (2007), Zhang et al. (2011) Dobrzinski et al. (2004), Dobrzinski & Bahlburg (2007), Zhang et al. (2011) Dobrzinski et al. (2004), Dobrzinski & Bahlburg (2007), Zhang et al. (2011)

Ca in phosphates may not be considered in calculation of the CIA, because CIA values will increase by not more than c. 1 unit if all P2O5 present in common siliciclastic rocks is assigned to apatite. The CIA has been successfully applied in a large number of studies involving glacial deposits (e.g. Young & Nesbitt 1999; Young 2001; Condie et al. 2001; Scheffler et al. 2003; Dobrzinski et al. 2004; Young et al. 2004; Rieu et al. 2007) and other depositional environments (e.g. Gallet et al. 1998; Aristiza´bal et al. 2005; Kahmann et al. 2008).

A brief discussion of WIP and CIA Results of a comparison of WIP and CIA are displayed in Fig. 6.1. It combines data on Neoproterozoic glaciomarine deposits from South China (Dobrzinski et al. 2004) and the Port Askaig Formation (Scotland; Panahi & Young 1997; and our own analyses) with new data on the Ghaub Formation (Namibia) and the Mortensnes and Smalfjord formations of northern Norway (Tables 6.1 to 6.4). For comparison we also show hitherto unpublished results on a number of late Devonian and early Carboniferous siliciclastic shales and siltstones from the Palentinian foreland basin (Keller et al. 2007, 2008) of the Hercynian Cantabrian Mountain belt in northern Spain (Table 6.1). In the Late Palaeozoic, the Palentinian Basin was located at near-equatorial latitudes (Weil et al. 2001). We use the Cantabrian data as an example of well-weathered detritus developed under non-glacial conditions. The WIP values of the glacial Ghaub, Port Askaig, Mortensnes and Smalfjord formations fall between 29 and 107, with a majority between 40 and 80 (averages: Ghaub, 76; Mortensnes and

84

H. BAHLBERG & N. DOBRZINSKI

Fig. 6.2. The Yangjiaping section in South China and its Neoproterozoic climate record indicated by CIA values. PG, preglacial units; LD, lower diamictite; DF, Datangpo Formation; UD, upper diamictite; DSF, Doushantuo Formation (Table 6.1). Note the initially high CIA values in the lower part of the upper diamictite unit followed by a decrease interpreted as the initial and then waning incorporation of chemically weathered detritus into the glacial deposit. Source: modified from Dobrzinski et al. (2004).

Smalfjord, 56; Port Askaig, 66; Fig. 6.1). In contrast, the Cantabrian formations demonstrate their well-weathered character with values between 0 and 75, with a majority below 40 (average, 19). Considering these formations, the WIP appears to differentiate well between glacial and non-glacial weathering types. The CIA values of the glacial units fall within 55 and 77 with a majority below 70 (averages: Ghaub, 63; Mortensnes and Smalfjord, 66; Port Askaig, 70; Fig. 6.1). These are typical values of unweathered to slightly weathered detritus and conform well to a glacial weathering regime. The Cantabrian shales and siltstones, in turn, reflect the dominance of warm –humid weathering conditions with a CIA between 56 and 100 and an average of 79. The glacial and non-glacial deposits from South China seem to complicate the picture because they both have relatively low as well as high WIP values typical of intense and weak chemical weathering, respectively (Fig. 6.1). The WIP range of the glacial deposits is 31–121 with an average of 57. The respective values of the non-glacial formations are 2 –118 and 52. CIA values are distributed between 50 and 85 in the glacial deposits (average, 68) and between 55 and 85 with the rare sample having a value of close to 100 (average, 71; Fig. 6.1; Dobrzinski et al. 2004). The values of both indices seem to demonstrate that both the glacial and non-glacial Neoproterozoic sedimentary rocks in

South China consist of a mixture of detritus of glacial and nonglacial weathering provenance, albeit in different proportions. The initial incorporation of weathered older detritus into the Neoproterozoic glacial deposits was demonstrated by Dobrzinski et al. (2004) by upward trends to lower CIA values in several sections in South China (Fig. 6.2). Similar observations were made by Panahi & Young (1997) in the Port Askaig Formation, Scotland. K-feldspar and plagioclase are relatively common in the Chinese glacial and non-glacial deposits. The presence of labile feldspar indicates that the non-glacial rocks did not form predominantly under humid and warm –humid conditions. Regarding the weathering sensitive cations Kþ and Naþ, average values for Na2O in both the glacial and non-glacial deposits from South China are 1.3 wt%, whereas the average value of detritus derived from the upper continental crust is 1.2 wt% (PAAS, Post-Archaean average Australian shale, Taylor & McLennan 1985). Condie (1993) gives an estimate of Na2O in the average Proterozoic shale (APS) derived from the Proterozoic upper crust of 1.1 wt%. Average K2O values of 3.3 wt% coincide with the Proterozoic upper crust, upper continental crust and Average Proterozoic shale values of 3.3, 3.4 and 3.6 wt%, respectively (Condie 1993; McLennan 2001). Considering the calculations of the respective indices, an increase in, for example, Na2O of 0.5 wt% causes a decrease in the CIA of c. 2.5 units, and an increase of WIP of 5 units. A decrease by 0.5 wt% results in similar, but opposing, changes in magnitude. In the case of K2O the effect is similar; however, the change in CIA units is c.1.5 per 0.5 wt% change. The respective changes in WIP are c. + 4 units. Our data demonstrate that both the WIP and CIA appear to be equally sensitive to small changes in the concentrations of the major cations, and thus, by inference, of dominant warm –humid or dry –cold weathering conditions. Disadvantages of the WIP include the lack of consideration of mobile relative to immobile phases, and the fact that it is based on whole-rock CaO and not on CaO* of the silicate fraction only. Another advantage of the CIA lies in the way it permits the definition and prediction of weathering trends of silicate rocks (Nesbitt & Young 1982, 1984). Data and trends can then be displayed well in A – CN– K (Al2O3 – CaO* þ Na2O – K2O) ternary diagrams (Fig. 6.3; Nesbitt & Young 1984). The combination of these features makes the CIA the presently preferred weathering index.

Provenance and reconstruction of original compositions Provenance pathways and the distribution of lithologies in source regions exert a first-order control on the composition of siliciclastic deposits (Johnsson 1993). Differences in provenance are therefore reflected also in the weathering indices (Fedo et al. 1995). In the ternary A –CN –K diagram of Figure 6.3, the provenance compositions and weathering trends can be depicted and predicted. Hydrolytic weathering leads to a loss of Naþ and Ca2þ and changes rock compositions towards the A apex and ever higher CIA values. Transport sorting exerts another major influence on the composition of clastic sediments. Along a river and transport path, labile mineral grains will be comminuted preferentially. In the presence of chemical weathering, labile grains like feldspar will be prone to progressive decay along cleavages, cracks and other zones of lattice weakness. Feldspars will eventually be transformed into clay minerals and a marked compositional difference, in addition to the comminution of mineral grains, between source and sediment will become evident (Johnsson et al. 1988). Sorting will thus lead to a higher proportion of clay minerals downriver and consequently higher CIA values (Nesbitt et al. 1997; Fig. 6.3). Thus, when CIA values of different successions are compared, similar grain sizes need to be considered. Only exceptional circumstances permit the transfer of labile mineral grains from source to sink without any mineralogical

A REVIEW OF THE CHEMICAL INDEX OF ALTERATION

85

Fig. 6.3. Major oxides in molecular ratios with compositions of typical magmatic source rock types (Fedo et al. 1995) and average compositions of the UCC through time (UCC, McLennan 2001; PAAS, Taylor & McLennan 1985; all others, Condie 1993). Note that the lower part of the diagram with A , 40 is not shown. 1, gabbro; 2, tonalite; 3, granodiorite; 4, granite; 5, A-type granite; 6, charnokite; 7, potassic granite. Ideal weathering trends of UCC-type source lithologies would be parallel to the predicted weathering trend (Nesbitt & Young 1984). The statistically modelled t-index weathering trend (von Eynatten 2004) is based on data obtained from the world’s major rivers and erosional products of major denudation areas (McLennan 1993). Trends of K-metasomatism and its graphical correction to pre-metasomatic values on predicted weathering trends is indicated by grey arrows originating or pointing at the K apex of the diagram (Fedo et al. 1995). Alteration trend of Neoproterozoic glacial successions estimated from the distribution of samples in Fig. 6.4. The dotted double-headed arrow outlines the effect of grain sorting on the chemical composition of sediments and consequently on CIA values; c, coarse, sand; f, fine, clay (Nesbitt et al. 1997).

and thus compositional changes, that is, without the influence of any chemical weathering. In such an ideal case transport will lead only to comminution of mineral grains; the final fine-grained deposit will have the same composition as the coarser source (Nesbitt & Young 1996), and CIA values will be unaffected. Diagenetic addition of potassium to a deposit during illitization of kaolinite and the replacement of plagioclase or illite by K-feldspar, the so called K-metasomatic effect of Fedo et al. (1995), causes weathered compositions to plot closer to the K apex (Fig. 6.3), thus resulting in diagenetically lowered CIA values. This can be checked by comparisons with average upper crustal compositions (see above) and by comparing CIA values with K/Cs ratios (McLennan et al. 1993); a K-metasomatic

effect would disturb the pronounced negative correlation between the K/Cs ratio and weathering intensity. The distribution of data in the A –CN –K diagram (Fig. 6.4) indicates that in particular some of the Chinese sedimentary rocks discussed in this paper show a weak K-metasomatic effect (Dobrzinski et al. 2004). If projected back to the predicted weathering trend originating from average granodiorite or the upper continental crust (UCC) (Fig. 6.3), the originally higher CIA values of these samples can be reconstructed (Tables 6.3 and 6.4), thus permitting the subsequent inference of original provenance compositions (Figs 6.3 and 6.4; Fedo et al. 1995). As the CIA is founded essentially in considerations of the mobility of the major cations, geochemical estimations of provenance

Fig. 6.4. A–CN– K (Al2O3 –CaO*þ Na2O– K2O) diagram with data of Neoproterozoic glacial units and reference units and their composition (PAAS, Taylor & McLennan 1985; UCC, McLennan 2001; all others, Condie 1993). Note that the lower part of the diagram with A , 40 is not shown. The left side of the figure shows the range and averages (black dots) of CIA values of the Neoproterozoic units corrected for K-metasomatism according to Fedo et al. (1995). For comparison, the figure also shows data obtained from shales of several late Devonian and Carboniferous formations from the Palentinian foreland basin of the Hercynian Cantabrian mountain belt in northern Spain (Keller et al. 2008), which conform to the predicted weathering trend (Tables 6.1, 6.2; Nesbitt & Young 1984).

86

H. BAHLBERG & N. DOBRZINSKI

Fig. 6.5. Zr/Sc v. Th/Sc diagram of McLennan et al. (1993) demonstrating that all studied sedimentary rocks have an upper crustal composition and were affected to minor degrees by recycling, which is indicated by high Zr/Sc values. This effect is most prominently shown in some of the Late Palaeozoic sedimentary rocks from Cantabria. It is noteworthy that all the Neoproterozoic samples cluster around the composition of the average Proterozoic shale (APS, Condie 1993). Reference units and their composition: PAAS, Taylor & McLennan (1985); UCC, McLennan (2001); all others, Condie (1993).

should be combined preferably with ratios of immobile elements including the high field strength elements La, Th, Sc and Zr (Bhatia & Crook 1986; McLennan et al. 1993; Bahlburg 1998). All the Neoproterozoic sedimentary rocks considered in this paper have Zr/Sc and Th/Sc ratios similar to or higher than the Upper Continental Crust (UCC) and PAAS (Fig. 6.5; Taylor & McLennan 1985; McLennan 2001). The rock compositions all cluster around the composition of the APS (Fig. 6.5; Condie 1993) as a proxy for material derived from the average Proterozoic upper crust. This provides another indication that processes beyond original weathering and recycling did not significantly alter the composition of the studied Neoproterozoic sedimentary rocks.

Weathering trends Thermodynamic, kinetic, experimental and observational evidence define weathering trends for silicate rocks that are (sub)parallel to the CN– A join of the A – CN– K diagram (Nesbitt & Young 1982, 1984). Weathering of average granodiorite or UCC along the predicted weathering trend (Fig. 6.3) will result in the transformation of labile components including the feldspars first to illite, thus causing the sample’s composition to plot ever closer to the A –K join and the illite composition in A –CN –K space (Nesbitt & Young 1984; Fig. 6.3). Increasing formation of kaolinite during progressive weathering will curve the weathering trend towards the A apex when approaching the A – K join (Nesbitt & Young 1984) in a way similar to the calculated weathering trend of von Eynatten et al. (2004; Fig. 6.3). The validity of this general weathering behaviour and trend is substantiated by the fact that the compositions of shale composites like PAAS and APS as averaged samples of the weathered and recycled upper crust plot very near the predicted weathering trend (Fig. 6.3). Also, the Late Palaeozoic reference samples from Cantabria, Spain, which were well weathered under non-glacial conditions, plot along this trend as predicted by CIA systematics (Figs 6.3 & 6.4; Nesbitt & Young 1982, 1984). In many cases, sedimentary rocks do not plot on the predicted weathering trend but below it, thus reflecting an increase in Kþ in the samples (e.g. Panahi & Young 1997; Rieu et al. 2007). A correction of this K-metasomatic effect according to Fedo et al. (1995) of those CIA data points below the predicted weathering trend line results in an increase of CIA values (Figs 6.3, 6.4; Tables 6.3, 6.4). This increase depends on the compositional difference between the sample and the predicted original value and reflects the degree of K-metasomatism having affected a sample’s detritus if the original composition was granodioritic and close to that of the UCC (Fig. 6.3; Fedo et al. 1995). If the original composition was richer in Kþ than granodiorite or UCC, the predicted weathering trend would shift to the right but remain

(sub)parallel to the predicted weathering trend (Nesbitt & Young 1984). Corrections for K-metasomatism and resulting increases in CIA values would be correspondingly smaller. A correction to the predicted weathering trend thus represents a maximum correction if average granodiorite or UCC is the starting point. The average compositions of post Archaean shale and Proterozoic shale (PAAS, Taylor & McLennan 1985; APS, Condie 1993) plot below the predicted weathering trend towards the K apex of the A –CN –K diagram. This may be due either to starting compositions richer in Kþ than UCC and the average Proterozoic upper crust (Fig. 6.3), or a mild K-metasomatic effect including the conversion under increased temperature of illite to secondary K-feldspar (Nesbitt & Young 1989; Fedo et al. 1995). The composition of almost all considered Neoproterozoic units and samples appears to systematically deviate from the predicted weathering trend, in a way here preliminarily called the observed Neoproterozoic alteration trend (Fig. 6.3). In discussing this trend, two things have to be borne in mind: the similarity of the estimates for the UCC and the average Proterozoic crust in A – CN –K space, and the fact that glaciogenic deposits commonly represent averaged samples of the available upper crust exposed to denudation. It is a reasonable assumption that the original average geochemical composition of the silicate sources feeding the Neoproterozoic glaciogenic deposits was similar to the average upper crustal compositions. Repeated weathering and recycling processes may lead to a stepwise increase of Kþ in the resulting deposits, as indicated by the offsets of the average shale compositions from the predicted weathering trend (Fig. 6.3). The observed Neoproterozoic alteration trend may consequently reflect the degree to which unweathered and weathered crust and recycled sedimentary rocks have been mixed and incorporated into the glaciogenic deposits. The correction of CIA values for a K-metasomatic effect according to Fedo et al. (1995) will thus reconstitute CIA values to the predicted weathering trend originating in fresh upper crust, which may, however, be unrepresentative of the actually eroded continental surface. If CIA values were referred back only to unweathered compositions, irrespective of the regional context and the recycling history, estimates of weathering effects during the last sedimentary cycle will be exaggerated and may consequently be misleading.

The evidence of glaciomarine deposits Successions of modern and ancient glacial deposits and of glacially derived detritus are ideally characterized by low CIA values reflecting the dominance of physical over chemical weathering processes (Nesbitt & Young 1982; Young 2001). A typical example is the Palaeoproterozoic Gowganda Formation (Young

A REVIEW OF THE CHEMICAL INDEX OF ALTERATION

87

Table 6.2. Major element whole-rock compositions and weathering indices of shales Sample*

SiO2

Al2O3

Fe2O3

MgO

CaO‡

Na2O

K2O

TiO2

P2O5

MnO

LOI

Total

BT Nr.64 MB 87-K MB P-05 MB P-25 MB P-16 BT PO 28 BT PO 29 PO 1 PO 2-1 PO 3 PO 4-2 PO 4-4 PO 5 PO 6 PO 9 PO 11 BT PO 4 BT PO 5 BT PO 16 BT PO 30/2 FC 1 FC 2 FC 3 FC 5 CRM 1A CRM 2 CRM 3 CM 1A CM 3A CM 3B CM 4A CM 4A_2 CM 5A VN 1 VN 2 VN 3A VN 3B VN 4 VN 5

96.83 95.33 87.40 83.00 60.87 84.47 92.10 75.34 79.79 88.09 74.77 67.92 37.88 45.19 54.00 84.53 71.88 68.46 72.33 68.44 73.68 77.89 87.51 74.16 85.36 84.48 84.32 79.37 84.89 90.78 89.18 89.09 76.72 79.31 79.70 81.63 82.34 79.00 81.63

1.85 2.40 3.54 3.47 19.13 6.02 3.78 8.04 9.72 7.16 9.96 6.79 3.92 2.17 3.36 6.15 6.79 6.44 7.15 12.29 2.87 3.83 4.35 6.86 6.00 4.90 6.80 3.11 3.25 3.82 4.10 4.16 5.54 10.69 10.79 9.43 8.93 11.45 9.40

0.60 0.61 0.83 2.19 5.87 2.79 0.82 7.37 3.36 1.74 6.38 9.36 4.89 1.41 2.81 3.20 4.90 5.28 8.37 5.23 7.31 10.22 4.40 11.27 3.72 3.58 3.88 9.94 6.37 2.13 3.27 3.43 9.19 3.19 2.82 2.84 2.78 2.83 2.62

0.04 0.02 0.12 0.29 0.60 0.66 0.14 1.04 0.62 0.15 0.42 0.87 0.45 0.34 0.39 0.18 1.22 1.15 1.27 1.17 0.50 0.80 0.43 0.57 0.28 0.30 0.17 0.86 0.60 0.32 0.44 0.45 1.02 0.82 0.81 0.79 0.75 0.82 0.75

– – 2.49 0.61 3.14 1.40 1.02 0.55 0.05 – 0.92 4.40 27.10 26.81 19.65 0.34 4.29 6.87 2.11 2.68 6.40 0.40 0.07 0.13 0.38 1.78 0.11 0.64 0.22 0.04 0.04 0.05 0.34 0.06 0.06 0.07 0.06 0.04 0.11

– – 0.07 0.03 0.60 0.15 0.21 0.03 0.10 – 0.40 0.34 0.03 0.02 0.12 0.05 0.25 0.08 0.21 0.31 0.15 0.15 0.20 0.40 0.04 0.05 0.40 0.11 0.12 0.22 0.23 0.23 0.36 0.51 0.52 0.55 0.52 0.48 0.42

– 0.04 0.41 0.73 2.68 1.08 0.76 1.06 1.45 0.75 1.23 1.43 0.49 0.16 0.34 0.93 1.19 0.94 1.04 1.72 0.55 0.53 0.54 1.20 1.21 1.03 1.05 0.36 0.36 0.47 0.51 0.51 0.81 1.65 1.76 1.50 1.50 2.00 1.44

0.23 0.23 0.20 0.09 1.02 0.62 0.11 0.52 0.56 0.53 0.58 0.40 0.19 0.11 0.16 0.46 0.37 0.35 0.46 0.70 0.21 0.24 0.33 0.50 0.41 0.34 0.41 0.21 0.25 0.25 0.30 0.31 0.61 0.65 0.66 0.59 0.55 0.66 0.81

0.08 0.04 0.07 0.03 0.09 0.08 0.02 0.07 0.04 0.02 0.09 0.07 0.07 0.14 0.16 0.09 0.08 0.08 0.20 0.09 0.14 0.09 0.07 0.14 0.10 0.08 0.11 0.06 0.04 0.03 0.05 0.06 0.08 0.08 0.10 0.09 0.10 0.07 0.14

0.01 0.01 0.05 0.03 0.07 0.03 0.03 0.18 0.01 0.00 0.15 0.28 0.09 0.12 0.17 0.03 0.08 0.08 0.13 0.06 0.24 0.21 0.06 0.23 0.05 0.06 0.06 0.20 0.16 0.01 0.01 0.01 0.33 0.02 0.01 0.02 0.01 0.01 0.01

0.42 0.72 3.02 2.20 6.79 1.57 1.57 4.30 2.55 1.62 3.67 6.74 23.38 21.91 16.83 2.04 6.91 9.26 7.16 5.60 7.90 5.60 2.20 4.60 2.10 3.10 2.40 5.00 3.60 2.00 1.40 1.40 4.80 2.50 2.50 2.10 2.30 2.60 2.40

100.06 99.41 98.21 93.28 100.87 100.28 101.58 98.50 98.30 100.06 98.56 98.59 98.49 98.38 97.99 98.35 97.97 98.99 100.44 98.30 99.98 99.97 100.18 100.10 99.67 99.72 99.74 99.87 99.87 100.08 99.55 99.71 99.82 99.52 99.77 99.64 99.87 100.00 99.77

TC

TIC

0.72

0.54

0.94

0.45

0.74

0.58

0.30 1.31 6.04 5.86 4.42

0.16 1.21 5.92 5.76 4.33

1.80 2.40 1.83 1.21 1.92

1.59 2.24 1.22 0.75 1.81

WIP

CIA†

0 0 11 9 38 16 11 14 15 7 18 29 75 71 55 10 27 29 20 28 24 9 8 16 12 15 13 8 6 7 8 8 14 21 22 20 20 24 18

100 98 86 64 83 60 56 87 84 90 83 76 87 91 86 78 80 85 83 84 77 71 81 76 74 52 77 64 77 80 80 80 73 80 79 78 78 79 79

*Formations, late Devonian– Carboniferous, of the Cantabrian Mountains, northern Spain: Vergan˜o, VN; Van˜es, CM; Carmen, FC, CRM; Potes Group, PO, BT PO; Murcia, BT, MB. Major elements were determined by ICP-ES after a LiBO2 fusion at ACME Analytical Labs, Vancouver, Canada. † CIA has been calculated according to Nesbitt and Young (1982) and Fedo et al. (1995). Ca in phosphates has not been considered in the calculation of the CIA, because CIA values increase by only c.1 unit if all P2O5 is assigned to apatite. ‡ Calculation of CaO* is based on values of TC (total carbon) and TIC (total inorganic carbon) obtained by analysis using CS-Mat 5500, in the Department of Geology and Paleontology, Mu¨nster University. Only samples with increased values of CaO were selected. Those with low values did not show a reaction with HCl (10%); dolomite was not observed in thin section. CaO was consequently considered to be equal to CaO*. –, below detection limit; blank spaces, not measured.

& Nesbitt 1999). Within this unit, rocks of diamictite facies commonly have the lowest CIA values (between 50 and 70). Diamictites in general represent either lodgement tills or moraine material that was rapidly redeposited by mass wasting in glaciomarine environments (e.g. Scheffler et al. 2003; Dobrzinski et al. 2004; Young et al. 2004; Rieu et al. 2007; Fig. 6.4). Such deposits with low CIA values are frequently associated with finer-grained and laminated glaciomarine sandstones, siltstones and shales, which may include ice-rafted debris, with higher CIA values between 70 and 85 (Fig. 6.4; Panahi & Young 1997; Dobrzinski et al. 2004; Young et al. 2004; Rieu et al. 2007). Lower CIA values in glaciomarine deposits can be caused by several factors acting alone or in concert. K-metasomatism in connection with the conversion of illite to K-feldspar, or the illitization of kaolinite, is the most commonly cited cause (Nesbitt & Young 1989; Fedo et al. 1995). However, the grain-size of the analysed rock is also an important factor. Sands formed from chemically

pre-weathered rocks and soils generally have lower CIA values and plot lower in the A –CN –K triangle than associated muds, because feldspars and other labile minerals tend to be concentrated in the coarser-grained fraction (Nesbitt et al. 1996; Fig. 6.3). Furthermore, sorting or mixing of known and cryptic sources may result in an unexpected change of the amount of feldspar in a deposit. A sedimentary enrichment of K-feldspar relative to the assumed source rock causes the distribution of data in the A –CN –K diagram to mimic a K-metasomatic effect. In such a case the predicted weathering trend would shift to the right and point more closely to a muscovite rather than an illite composition on the A –K join of the diagram, or a CIA of 80 instead of 90 (Fig. 6.4; Nesbitt et al. 1997). A correction of the CIA data for an alleged K-metasomatism according to Fedo et al. (1995; Figs 6.3 and 6.4) would in this case lead to an overadjustment. A cause of increased or high CIA values in glaciomarine rocks may also be

88

H. BAHLBERG & N. DOBRZINSKI

Table 6.3. CIA and WIP values of the Neoproterozoic Port Askaig Formation, Scotland, based on the data of Panahi & Young (1997) and our own analyses Sample

TC

TIC

WIP

CIA*

CIAcorr†

94-102 94-106 94-108 94-78 94-79 94-87 94-97 95-62 95-66 95-67 95-69 95-70 95-71 95-73 95-76 95-79 95-80 95-83 95-84

0.58 0.68 0.90 4.14 4.52 1.95 0.56 3.39 5.80 5.11 3.37 3.92 6.19 4.93 5.21 3.57 5.03 1.69 1.56

0.45 0.44 0.62 3.39 4.17 1.23 0.45 2.85 5.00 4.75 2.91 3.37 4.90 4.04 4.25 2.93 4.66 1.32 0.51

33 56 61 75 72 57 36 69 84 75 75 70 85 79 78 64 80 53 56

68 65 64 74 74 63 64 77 74 72 70 68 71 71 75 73 70 66 61

84 73 72 83 86 72 71 87 87 87 86 78 87 84 83 87 81 74 66

*Ca in phosphates has not been considered in the calculation of the CIA, because CIA values increase by only c.1 unit if all P2O5 is assigned to apatite. CIA has been calculated according to Nesbitt & Young (1982) and Fedo et al. (1995). † CIAcorr: CIA value corrected to the predicted weathering trend according to Fedo et al. (1995). Calculation of CaO* is based on values of TC (total carbon) and TIC (total inorganic carbon) obtained by analysis using CS-Mat 5500, in the Department of Geology and Paleontology, Mu¨nster University.

the incorporation of older, strongly weathered material into the glacial deposits (Nesbitt & Young 1997; Dobrzinski et al. 2004; Fig. 6.2). Many Neoproterozoic glaciomarine successions are characterized by a deviation of the observed weathering trend from the predicted one and towards the K apex (Figs 6.3 and 6.4), including the lower and upper diamictites of the Yangtze platform in South China (Table 6.1; Dobrzinski et al. 2004), the Port Askaig Formation (Nesbitt & Young 1997; Tables 6.1 and 6.3), the Ghaub Formation of northern Namibia, the Norwegian Smalfjord and Mortensnes formations (Tables 6.1 and 6.4), and the Fiq Formation in Oman (Rieu et al. 2007). The same effect is also evident in the glacial Ordovician Table Mountain Group in South Africa (Young et al. 2004). Regarding the Port Askaig and Fiq formations, the Chinese diamictites and the Table Mountain Group, K-metasomatism has in fact been cited as the cause of the deviation from the commonly used predicted weathering trend (Nesbitt & Young 1997; Dobrzinski et al. 2004; Young et al. 2004; Rieu et al. 2007). However, almost all the mentioned cases include analyses of matrix material of glaciomarine diamictites and associated coarsegrained strata. Some of the Neoproterozoic diamictites, including the ones from South China (Dobrzinski & Bahlburg 2007), have an impure shaley matrix rich in silt or fine sand. The presence of this coarser material in the matrix of the diamictites may have contributed consequently to a deviation of the observed weathering trends from the predicted one. A combination of grain-size effects and the variable incorporation of older weathered material into the Neoproterozoic glaciomarine deposits may account for the measured CIA values (Fig. 6.4; Tables 6.2 to 6.4). In view of the corrected CIA values

Table 6.4. Major element whole-rock compositions and weathering indices Sample*†

SiO2

Al2O3

Fe2O3

MgO

CaO}

Na2O

K2O

TiO2

P2O5

MnO

LOI

Total

TC

TIC

WIP

CIA§

CIAcorrk

GF 1 GF 2 GF 3 GF 4 GF 5 GF 7 GF 8 GF 9 GF 11 GF 12 GF 13

70.77 81.78 72.25 28.42 38.51 27.84 59.81 33.92 71.33 73.82 7.93

6.66 8.45 7.50 6.46 7.33 4.28 16.53 7.96 13.72 12.63 0.84

15.52 2.31 10.25 2.82 3.44 3.20 7.32 4.97 3.68 3.39 0.48

0.4 0.42 0.95 10.93 9.29 12.98 3.15 10.47 1.11 0.96 19.65

0.05 0.01 0.47 19.89 15.52 19.08 0.70 15.03 0.56 0.31 27.73

0.05 0.1 0.09 0.86 0.83 0.22 0.74 0.91 2.20 2.06 0.02

3.17 5.22 4.79 2.14 2.55 1.79 5.36 2.82 4.93 4.49 0.29

1.24 0.43 0.34 0.40 0.43 0.24 0.87 0.47 0.44 0.42 0.06

0.12 0.06 0.35 0.12 0.14 0.10 0.18 0.14 0.14 0.12 0.07

0.01 0.01 0.04 0.06 0.06 0.08 0.11 0.14 0.03 0.05 0.60

1.50 1.20 2.70 27.80 21.70 30.00 5.00 23.00 1.70 1.60 42.30

99.50 99.98 99.73 99.91 99.80 99.82 99.80 99.83 99.84 99.86 99.98

0.03 0.01 0.06 7.48 5.66 7.91 0.23 6.09 0.07 0.03 11.13

–‡ – 0.01 7.26 5.34 7.65 0.09 5.62 0.06 0.02 10.92

29 46 45 107 94 102 63 99 67 61 127

65 59 55 63 64 65 69 64 59 59 71

87 87 79 72 74 81 80 74 68 68 85

S1 S2 S3 S4 S5 S6 M1 M2 M3 M4 M5 M6

73.68 73.22 67.36 59.38 69.49 65.02 64.35 64.14 69.19 67.96 65.56 63.64

10.67 11.95 13.36 11.26 12.75 14.83 15.52 15.93 10.5 11.91 11.05 14.2

3.59 4.10 5.94 4.65 3.91 5.73 6.23 5.49 3.88 4.44 5.19 5.79

1.89 1.72 2.94 4.85 2.33 2.6 3.13 2.67 3.05 2.88 3.01 3.39

1.42 0.41 0.27 4.74 1.48 1.05 0.33 0.89 2.98 2.35 3.43 2.00

1.45 1.03 1.41 1.37 1.91 1.86 2.05 2.36 1.45 1.96 1.85 2.45

2.82 3.66 3.91 3.48 3.39 4.00 3.58 3.76 2.45 2.42 2.13 2.61

0.5 0.65 0.83 0.79 0.61 0.70 0.78 0.82 0.51 0.58 0.54 0.69

0.12 0.11 0.17 0.16 0.11 0.13 0.14 0.15 0.11 0.12 0.10 0.13

0.03 0.02 0.10 0.06 0.03 0.04 0.05 0.04 0.05 0.06 0.10 0.10

3.20 2.50 3.10 8.80 3.40 3.50 3.30 3.20 5.20 4.70 6.20 4.40

99.38 99.38 99.40 99.55 99.42 99.48 99.48 99.46 99.38 99.40 99.17 99.43

0.36 0.08 0.09 1.88 0.46 0.24 0.06 0.16 0.95 0.78 1.11 0.46

0.27 – 0.04 1.56 0.32 0.17 0.06 0.14 0.73 0.56 0.79 0.29

46 46 55 68 57 61 59 63 50 53 52 59

65 65 67 65 65 65 68 65 68 67 67 64

70 73 74 73 70 69 72 69 73 69 69 64

*Neoproterozoic Ghaub Formation (GF), northern Namibia, and the Smalfjord (S) and Mortensnes (M) formations, northern Norway. Samples of the Ghaub Formation diamictites were taken on the Ghaub and Jakkalumuramba farms NE of the town of Otavi. Analyses represent matrix compositions. † Major elements were determined by ICP-ES after a LiBO2 fusion at ACME Analytical Labs, Vancouver, Canada. –below detection limit. § CIA has been calculated according to Nesbitt & Young (1982) and Fedo et al. (1995). Ca in phosphates has not been considered in the calculation of the CIA, because CIA values increase by only c.1 unit if all P2O5 is assigned to apatite. k CIAcorr: CIA value corrected to the predicted weathering trend according to Fedo et al. (1995). } Calculation of CaO* is based on values of TC (total carbon) and TIC (total inorganic carbon) obtained by analysis using CS-Mat 5500, in the Department of Geology and Paleontology, Mu¨nster University.

A REVIEW OF THE CHEMICAL INDEX OF ALTERATION

Fig. 6.6. Variations in the CIA with stratigraphic height in the Fiq Formation, Oman. MB Fm., Masirah Bay Formation. Modified from Rieu et al. (2007).

shown in Figure 6.4 and the discussion in the preceding paragraphs, we conclude that in particular the high values are overcorrected maximum values very likely exaggerating the magnitude of changes due to chemical weathering.

Implications for the study of Neoproterozoic climate change The origin of Neoproterozoic diamictite successions was debated long before the ‘Snowball Earth’ hypothesis was proposed (Kirschvink 1992; Hoffman et al. 1998; Hoffman & Schrag 2002). Classic examples like the Ghaub diamictites in northern Namibia were repeatedly interpreted either as tillites and glaciomarine sediments (e.g. Gevers 1931; Martin 1964; Hoffmann & Prave 1996; Hoffman & Schrag 2002) or as tectonically triggered mass flow deposits (e.g. Schermerhorn 1974; Martin et al. 1985; Eyles & Januszczak 2007). Furthermore, there are also cases where alleged glacial deposits were proven to be in fact of depositional origin unrelated to ice action and cold climates. Examples include purported tillites of Tremadoc age in the Andes of northwestern Argentina (Keidel 1943), which were demonstrated to be coastal conglomerates (Bahlburg 1990), and conglomeratic deposits in northern Chile allegedly linked to the Late Palaeozoic Gondwana glaciation (Cecioni 1979, 1981), which were shown to be of mass-flow origin in turbidite complexes (Charrier 1986; Bahlburg & Breitkreuz 1993). It is usually necessary to support one line of evidence with at least a second, independent one in order to have confidence that an interpretation is correct. This applies not only to controversial sedimentological cases. The analysis of the CIA is a powerful tool in studies of the palaeoclimatic record preserved in siliciclastic sedimentary successions. However, as discussed previously, numerous factors can influence the major element geochemical composition of a siliciclastic rock. Thus, CIA values alone are

89

almost meaningless as climate indicators when not considered in the context of the stratigraphic framework and facies of the analysed sedimentary rock. This is demonstrated very well in a study of the Fiq Formation in Oman (Rieu et al. 2007; Fig. 6.6), an association of alternating glacial diamictites, debris-flow deposits, turbidite sandstones, hemipelagic shales, and wave-rippled shoreface deposits. The facies changes are connected to coincident changes between low CIA values of 62 –70 for the glacial diamictite matrix, and high ones between 80 and 85 (corrected after Fedo et al. 1995) for shales, siltstones and fine-grained sandstones. Facies changes and CIA data give evidence of variations in the intensity of chemical weathering as a function of climate oscillations between glacial and warm –humid interglacial conditions (Rieu et al. 2007). Scheffler et al. (2003) presented similar data in a study of the waning phase of the Late Palaeozoic glaciation in South Africa recorded in the Dwyka Group. In the Neoproterozoic as well as in the Late Palaeozoic, these climate oscillations together with the facies of the respective deposits give evidence of a functioning hydrological cycle and sediment dispersal systems. In the Neoproterozoic case these data are in direct opposition to assumptions of a totally ice-covered Snowball Earth as proposed by Hoffman & Schrag (2002). These CIA values are compatible with a prominent influence of physical weathering on the production of the diamictite detrital silicate matrix. Together with sedimentological features including the presence of dropstones in laminated facies, these values strengthen interpretations of a glacially connected origin of the Ghaub diamictites. The interpretation of the Neoproterozoic Smalfjord and Mortensnes formations of northern Norway is less controversial and centres around the question whether the strata represent redeposited detritus originally produced in a glacial environment (Arnaud & Eyles 2002) or deposits of direct glacial action (Edwards 2004; Rice 2004). Application of the CIA cannot solve these differences in a fundamentally sedimentological controversy. The uncorrected CIA values of the Smalfjord and Mortensnes formations (Table 6.4) are uniformly low between 65 and 68, and between 64 and 74 (averages of 72 and 69, respectively) if corrected according to Fedo et al. (1995; Fig. 6.4). It is thus legitimate to infer at least a glacial origin for the detritus. The Neoproterozoic lower and upper glacial diamictites of the Yangtze platform represent associations of lodgement till, subglacial melt-out deposits of grounded glaciers, sediments of concentrated density flows and glacial outwash, and turbidites and laminated deposits containing ice-rafted debris (Dobrzinski & Bahlburg, 2007). CIA values of both diamictite units are between 50 and 90 (Fig. 6.4; Table 6.1; Dobrzinski et al. 2004). Correction of values following Fedo et al. (1995) does not change their range but leads to a shuffling of values within this range. The corrected average value is 72 (Fig. 6.4). The combination of low values in the matrix of glacial diamictites and higher values in associated glacial facies is interpreted as a result of reworking of older, weathered material into the glacial deposits (Dobrzinski et al. 2004). This is evident in some sections showing a decrease in CIA values upsection (Fig. 6.2). The non-glacial units of the China platform consisting of preglacial siliciclastics, the shales and carbonates of the Datangpo Formation intercalated between the glacial units, and the postglacial carbonates and shales of the Doushantuo Formation (Table 6.1) have CIA values ranging between 59 and 97 with abundant higher values. The corrected ranges are rather similar, with a corrected average of 76 (Fig. 6.4). The higher values of the non-glacial deposits relative to the diamictite successions are taken to reflect more humid and warmer weathering and climate conditions. Particularly noteworthy is the presence of K-feldspar and plagioclase in the silicate fraction contained in the Marinoan cap carbonates in South China indicating the minor influence of chemical weathering on the silicate detritus in these strata (Dobrzinski 2005).

90

H. BAHLBERG & N. DOBRZINSKI

Conclusions The CIA (Nesbitt & Young 1982) is the most widely applied and most indicative of the available weathering indices. When the intricacies of the weathering systems and of applying the index are appropriately considered (Fedo et al. 1995; Nesbitt & Young 1996; Nesbitt et al. 1997; Young 2001; and discussion of Fig. 6.4), the index is a very valuable tool in the assessment of past climate change as recorded by siliciclastic sedimentary rocks, with one critical caveat: it needs to be applied in conjunction with a comprehensive facies analysis. Concerning the Neoproterozoic glacial periods, the combination of both data sets gives strong evidence of (i) a functioning hydrological cycle, (ii) operative sediment routing systems, and (iii) variable climate conditions oscillating between dry –cool and glacial, and warm –humid and interglacial. These findings are incompatible with the hypothesis of a totally ice-covered Snowball Earth. H. Rice (Vienna, Austria) kindly supplied samples of the Smalfjord and Mortensnes formations of northern Norway. We thank G. Young (London, Ontario, Canada) for sending us sample powders of the Panahi & Young (1997) samples of the Port Askaig Formation. This study was supported by the German Research Foundation DFG (grants Ba 1011/23-1,2,3). This paper is a contribution to IGCP project 512 ‘Neoproterozoic Ice Ages’. We thank A.R. Prave (St. Andrews, Scotland), C. Augustsson and C. Reimann (Mu¨nster, Germany), for commenting on earlier versions of the manuscript. We appreciate the very constructive reviews by Hilmar v. Eynatten (Go¨ttingen, Germany) and an anonymous reviewer. B. Fister (Mu¨nster) kindly re-drafted Figure 6.6.

References Aristiza´bal, E., Roser, B. & Yokota, S. 2005. Tropical chemical weathering of hillslope deposits and bedrock source in the Aburra´ Valley, northern Colombian Andes. Engineering Geology, 81, 389– 406. Arnaud, E. & Eyles, C. 2002. Glacial influence on Neoproterozoic sedimentation: the Smalfjord Formation, northern Norway. Sedimentology, 49, 765– 788. Arnaud, E. & Eyles, C. 2006. Neoproterozoic environmental change recorded in the Port Askaig Formation, Scotland: climatic vs tectonic controls. Sedimentary Geology, 183, 99 –124. Arnaud, E. & Fairchild, I. J. 2011. The Port Askaig Formation, Dalradian Supergroup, Scotland. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 635–642. Bahlburg, H. 1990. The Ordovician basin in the Puna of NW Argentina and N Chile: geodynamic evolution from back-arc to foreland basin. Geotektonische Forschungen, 75, 1 –107. Bahlburg, H. 1998. The geochemistry and provenance of Ordovician turbidites in the Argentinian Puna. In: Pankhurst, R. J. & Rapela, C. W. (eds) The Proto-Andean Margin of Gondwana. Geological Society, London, Special Publication, 142, 127– 142. Bahlburg, H. & Breitkreuz, C. 1993. Differential response of a Devonian– Carboniferous platform-deeper basin system to sea-level change and tectonics, N. Chilean Andes. Basin Research, 5, 21– 40. Baum, S. K. & Crowley, T. J. 2003. The snow/ice instability as a mechanism for rapid climate change: a Neoproterozoic snowball Earth model example. Geophysical Research Letters, 30, 2030, doi: 10.1029/2003GL017333. Benn, D. I. & Prave, A. R. 2006. Subglacial and proglacial glacitectonic deformation in the Neoproterozoic Port Askaig Formation, Scotland. Geomorphology, 75, 266– 280. Bhatia, M. R. & Crook, K. A. W. 1986. Trace element characteristics of graywackes and tectonic setting discrimination of sedimentary basins. Contributions to Mineralogy and Petrology, 92, 181–193. Bland, W. & Rolls, D. 1998. Weathering. An Introduction to the Scientific Principles. Arnold Publishers, London. Cecioni, G. 1979. Grupo El Toco, desierto de Atacama, Chile. Revista de la Asociacio´n Geolo´gica Argentina, 34, 211–223. Cecioni, G. , 1981. Triassic El Toco Group, Atacama Desert, Chile. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge.

Charrier, R. 1986. The Gondwana glaciation in Chile: description of alleged glacial deposits and paleogeographic conditions bearing on the extension of the ice cover in Southern South America. Palaeogeography, Palaeoclimatology, Palaeoecology, 56, 151–175. Condie, K. C. 1993. Chemical composition and evolution of the upper continental crust: contrasting results from surface samples and shales. Chemical Geology, 104, 1– 37. Condie, K. C., Des Marais, D. J. & Abbott, D. 2001. Precambrian superplumes and supercontinents: a record in black shales, carbon isotopes, and paleoclimates? Precambrian Research 106, 239–260. Crowell, J. C. 1999. Pre-Mesozoic ice ages: their bearing on understanding the climate system. Geological Society of America Memoir, 192, 1 –106. Crowley, T. J. & North, G. R. 1991. Paleoclimatology. Oxford Monographs on Geology and Geophysics, 18, 349. Dobrzinski, N. 2005. Das Paradoxon a¨quatornah abgelagerter glazialer Sedimentfolgen: Sedimentologische und geochemische Klimaindizien von der neoproterozoischen Yangtze Plattform (Su¨dchina). Dissertation, Westfa¨lische Wilhelms-Universita¨t, Mu¨nster, Germany. Dobrzinski, N. & Bahlburg, H. 2007. Sedimentology and environmental significance of the Cryogenian successions of the Yangtze platform, South China block. Palaeogeography, Palaeoclimatology, Palaeoecology, 254, 100–122. Dobrzinski, N., Bahlburg, H., Strauss, H. & Zhang, Q. R. 2004. Geochemical climate proxies applied to the Neoproterozoic glacial succession on the Yangtze Platform, South China. In: Jenkins, G., McMenamin, M., McKay, C. P. & Sohl, L. (eds) The Extreme Proterozoic: Geology, Geochemistry and Climate. American Geophysical Union Monograph Series, 146, 13– 32. Duzgoren-Aydin, N. S., Aydin, A. & Malpas, J. 2002. Re-assessment of chemical weathering indices: case study of pyroclastic rocks of Hong Kong. Engineering Geology, 63, 99 –119. Edwards, M. B. 2004. Glacial influence on Neoproterozoic sedimentation: the Smalfjord Formation, northern Norway — discussion. Sedimentology, 51, 1409– 1417. Etienne, J. L., Allen, P. A., Rieu, R. & Le Gerroue´, E. 2008. Neoproterozoic glaciated basins: a critical review of the Snowball Earth hypothesis by comparison with Phanerozoic glaciations. Special Publication of the International Association of Sedimentologists Special Publication 39. Evans, D. A. D. 2000. Stratigraphic, geochronological, and paleomagnetic constraints upon the Neoproterozoic climatic paradox. American Journal of Science, 300, 347– 433. Eyles, N. 2008. Glacio-epochs and the supercontinent cycle after 3.0 Ga: Tectonic boundary conditions for glaciation. Palaeogeography, Palaeoclimatology, Palaeoecology, 258, 89 – 125. Eyles, N. & Januszczak, N. 2007. Syntectonic subaqueousmass flows of the Neoproterozoic Otavi Group, Namibia: where is the evidence of global glaciation? Basin Research, 19, 179– 198. Fairchild, I. J. & Kennedy, M. J. 2007. Neoproterozoic glaciation in the Earth system. Journal of the Geological Society, London, 164, 895– 921. Fedo, C. M., Nesbitt, H. W. & Young, G. M. 1995. Unraveling the effects of potassium metasomatism in sedimentary rocks and paleosols, with implications for paleoweathering conditions and provenance. Geology, 23, 921– 924. Gallet, S., Jahn, B., Van Vliet Lano, B., Dia, A. & Rosello, E. 1998. Loess geochemistry and its implications for particle origin and composition of the upper continental crust. Earth and Planetary Science Letters, 156, 157– 177. Gevers, T. W. 1931. An ancient tillite in South West Africa. Transactions of the Geological Society of South Africa, 34, 1 –17. Gutzmer, J. & Beukes, N. J. 1998. Earliest laterites and possible evidence for terrestrial vegetation in the Early Proterozoic. Geology, 26, 263– 266. Hoffman, P. F. 2011. Glaciogenic and associated strata of the Otavi carbonate platform and foreslope, northern Namibia: evidence for large base-level and glacioeustatic changes. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of the Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 195– 209.

A REVIEW OF THE CHEMICAL INDEX OF ALTERATION

Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129– 155. Hoffman, P. F., Kaufman, A. J., Halverson, G. P. & Schrag, D. P. 1998. A Neoproterozoic snowball Earth. Science, 281, 1342–1346. Hoffmann, K. H. & Prave, A. R. 1996. A preliminary note on a revised subdivision and regional correlation of the Otavi group based on glacigenic diamictites and associated cap dolostones. Communications of the Geological Survey of Namibia, 11, 81 – 86. IGME 1984. Mapa Geolo´gica de Espan˜a a escala 1:50.000, Hoja no. 107 (Barruelo de Santullan), 113. ITGE 1994. Mapa Geolo´gica de Espan˜a a escala 1:50.000, Hoja no. 81 (Potes), 128. Johnsson, M. J. 1993. The system controlling the composition of clastic sediments. In: Johnsson, M. J. & Basu, A. (eds) Processes Controlling the Composition of Clastic Sediments. Geological Society of America Special Paper, 285, 1 – 19. Johnsson, M. J., Stallard, R. F. & Meade, R. H. 1988. First-cycle quartz arenites in the Orinoco River Basin, Venezuela and Colombia. Journal of Geology, 96, 263– 277. Kahmann, J. A., Seaman, J. III & Driese, S. G. 2008. Evaluating trace elements as paleoclimate indicators: multivariate statistical analysis of Late Mississippian Pennington Formation paleosols, Kentucky, U.S.A. Journal of Geology, 116, 254– 268. Keidel, J. 1943. El Ordovı´cico inferior en los Andes del norte Argentino y sus depo´sitos marino glaciales. Boletin de la Academı´a Nacional de Ciencias de Co´rdoba, 36, 140– 229. Keller, M., Bahlburg, H. & Reuther, C.-D. 2008. The transition from passive to active margin sedimentation in the Cantabrian Mountains, Northern Spain: Devonian or Carboniferous? Tectonophysics, doi: 10.1016/j.tecto.2008.06.022 Keller, M., Bahlburg, H., Reuther, C.-D. & Weh, A. 2007. Flexural to broken foreland basin evolution as a result of Variscan collisional events in northwestern Spain. In: Hatcher, R. D. Jr, Carlson, M. P., McBride, J. H. & Martı´nez Catala´n, J. R. (eds) The 4D Framework of Continental Crust. Geological Society of America Memoir, 200, 489– 510. Kirschvink, J. L. 1992. Late Proterozoic low-latitude global glaciation: the snowball earth. In: Schopf, J. W & Klein, C. (eds) The Proterozoic Biosphere. Cambridge University Press, Cambridge, 51 – 52. Kramer, J. R. 1968. Mineral-water equilibria in silicate weathering. International Geological Congress, 23rd session, Section 6, 149–160. Martin, H. 1964. Beobachtungen zum Problem der jung-pra¨kambrischen glazialen Ablagerungen in Su¨dwestafrika. Geologische Rundschau, 54, 115– 127. Martin, H., Porada, H. & Walliser, O. H. 1985. Mixtite deposits of the Damara sequence, Namibia: problem of interpretation. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 159– 196. McLennan, S. M. 1993. Weathering and global denudation. Journal of Geology, 101, 295– 303. McLennan, S. M. 2001. Relationships between the trace element composition of sedimentary rocks and upper continental crust. Geochemistry, Geophysics, Geosystems (G3), 2, doi:10.1029/2000GC000109. McLennan, S. M., Hemming, S., McDaniel, D. K. & Hanson, G. N. 1993. Geochemical approaches to sedimentation, provenance and tectonics. In: Johnsson, M. J. & Basu, A. (eds) Processes Controlling the Composition of Clastic Sediments. Geological Society of America Special Paper, 285, 21 – 40. Nedachi, Y., Nedachi, M., Bennett, G. & Ohmoto, H. 2005. Geochemistry and mineralogy of the 2.45 Ga Pronto paleosols, Ontario, Canada. Chemical Geology, 214, 21 – 44. Nesbitt, H. W. & Young, G. M. 1982. Early Proterozoic climates and plate motions inferred from major element chemistry of lutites. Nature, 199, 715– 717. Nesbitt, H. W. & Young, G. M. 1984. Prediction of some weathering trends of plutonic and volcanic rocks based on thermodynamic and kinetic considerations. Geochimica et Cosmochimica Acta, 48, 1523– 1534. Nesbitt, H. W. & Young, G. M. 1989. Formation and diagenesis of weathering profiles. Journal of Geology, 97, 129–147. Nesbitt, H. W. & Young, G. M. 1996. Petrogenesis of sediments in the absence of chemical weathering: effects of abrasion and

91

sorting on bulk composition and mineralogy. Sedimentology, 42, 341– 358. Nesbitt, H. W., Fedo, C. M. & Young, G. M. 1997. Quartz and feldspar stability, steady and non-steady-state weathering, and petrogenesis of siliciclastic sands and muds. Journal of Geology, 105, 173– 191. Nesbitt, H. W., Young, G. M., McLennan, S. M. & Keays, R. R. 1996. Effects of chemical weathering and sorting on the petrogenesis of siliciclastic sediments, with implications for provenance studies. Journal of Geology, 104, 525–542. Ohta, T. & Arai, H. 2007. Statistical empirical index of chemical weathering in igneous rocks: a new tool for evaluating the degree of weathering. Chemical Geology, 240, 280– 297. Panahi, A. & Young, G. M. 1997. A geochemical investigation into the provenance of the Neoproterozoic Port Askaig Tillite, Dalradian Supergroup, western Scotland. Precambrian Research, 85, 81 –96. Parker, A. 1970. An index of weathering for silicate rocks. Geological Magazine, 107, 501– 504. Pettijohn, F. J., Potter, P. E. & Siever, R. 1987. Sand and Sandstone. Springer, New York. Rice, A. H. N. 2004. Glacial influence on Neoproterozoic sedimentation: the Smalfjord Formation, northern Norway – discussion. Sedimentology, 51, 1419–1422. Rice, A. H. N. & Hofmann, C. C. 2001. The transition from Neoproterozoic glacial to interglacial sedimentation near Hammarnes, East Finnmark, North Norway. Norsk Geologisk Tidsskrift, 81, 257– 262. Rice, A. H. N., Edwards, M. B., Hansen, T. A., Arnaud, E. & Halverson, G. P. 2011. Glaciogenic rocks of the Neoproterozoic Smalfjord and Mortensnes Formations, Vestertana Group, E. Finnmark, Norway. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 593– 602. Rieu, R., Allen, P. A., Plotze, M. & Pettke, T. 2007. Compositional and mineralogical variations in a Neoproterozoic glacially influenced succession, Mirbat area, south Oman: Implications for paleoweathering conditions. Precambrian Research, 154, 248– 265. Ruxton, B. P. 1968. Measures of the degree of chemical weathering of rocks. Journal of Geology 76, 518–527. Scheffler, K., Hoernes, S. & Schwark, L. 2003. Global changes during Carboniferous–Permian glaciation of Gondwana: Linking polar and equatorial climate evolution by geochemical proxies. Geology, 31, 605– 608. Schermerhorn, L. J. G. 1974. Late Precambrian mixtites: glacial and/or non-glacial? American Journal of Science, 274, 673– 824. Schrag, D. P., Berner, R. A., Hoffman, P. F. & Halverson, G. P. 2002. On the initiation of Snowball Earth. Geochemistry, Geophysics, Geosystems, 3, doi:10.1029/2001GC000219. Shields, G. A. 2008. Palaeoclimate – Marinoan meltdown. Nature Geoscience, 1, 351– 353. Taylor, S. R. & McLennan, S. M. 1985. The Continental Crust: Its Composition and Evolution. Blackwell, Oxford. von Eynatten, H. 2004. Statistical modelling of compositional trends in sediments. Sedimentary Geology, 171, 79– 89. von Eynatten, H., Barcelo´-Vidal, C. & Pawlowsky-Glahn, V. 2003. Modelling compositional change: the example of chemical weathering of granitoid rocks. Mathematical Geology, 35, 231– 251. Wagner, R. H. & Wagner-Gentis, C. H. T. 1963. Summary of the stratigraphy of Upper Paleozoic rocks in NE. Palencia, Spain. Proceedings Kongelige Nederlandse Akademie Wetenschappen (B) LXVI, 3, 149– 163. Weil, A. B., Van der Voo, R. & Van der Pluijm, B. A. 2001. Oroclinal bending and evidence against the Pangea megashear; the Cantabria – Asturias Arc (northern Spain). Geology, 29, 991– 994. Wintsch, R. P. & Kvale, C. M. 1994. Differential mobility of elements in burial diagenesis of siliciclastic rocks. Journal of Sedimentary Research, A64, 349– 361. Young, G. M. 2001. Comparative geochemistry of Pleistocene and Paleoproterozoic (Huronian) glaciogenic laminated deposits: relevance to crustal and atmospheric composition in the last 2.3 Ga. Journal of Geology, 109, 463– 477.

92

H. BAHLBERG & N. DOBRZINSKI

Young, G. M. & Nesbitt, H. W. 1999. Paleoclimatology and provenance of the glaciogenic Gowganda Formation (Paleoproterozoic), Ontario, Canada: a chemostratigraphic approach. Geological Society of America Bulletin, 111, 264–274. Young, G. M., Minter, W. E. L. & Theron, J. N. 2004. Geochemistry and palaeogeography of upper Ordovician glaciogenic sedimentary

rocks in the Table Mountain Group, South Africa. Palaeogeography, Palaeoclimatology, Palaeoecology, 214, 323– 345. Zhang, Q.-R., Chu, X.-L. & Feng, L.-J. 2011. Neoproterozoic glacial records in the Yangtze Region, China. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 357–366.

Chapter 7 Neoproterozoic glacial palaeolatitudes: a global update D. A. D. EVANS1* & T. D. RAUB2 1

Department of Geology & Geophysics, Yale University, 210 Whitney Avenue, New Haven, CT 06520-8109, USA 2

Division of Geological and Planetary Sciences, 100-23 Caltech, Pasadena CA 91125, USA *Corresponding author (e-mail: [email protected])

Abstract: New stratigraphic, geochronological and palaeomagnetic constraints allow updates to be made to a synthesis of Neoproterozoic glacial palaeolatitudes, including modifications to some reliability estimates. The overall pattern of a Neoproterozoic climatic paradox persists: there is an abundance of tropical palaeolatitudes and near to complete absence of glaciogenic deposits demonstrably laid down between latitudes of 608 and 908. In addition to 12 units with palaeolatitude estimates that are somewhat reliable, estimates with moderate to high reliability now include Konnarock (less than 108 from the palaeo-equator), Elatina, Rapitan, Mechum River, Grand Conglomerat (10– 208), Upper Tindir, Puga (20– 308), Nantuo, Gaskiers (30–408) and Walsh (40– 508). Among these, Elatina, Upper Tindir and Nantuo are considered to have the highest reliability, all with estimates of low to moderate palaeolatitude. The Elatina result stems from sedimentary rocks with quantitative correction of inclination-shallowing effects, and the Upper Tindir result stems from data collected from igneous rocks that are precisely coeval with the glacial deposits. Despite continuing debate on the global character of Neoproterozoic ice ages, their pan-glacial extent (ice extending to low latitude in a low-obliquity world) is well demonstrated.

Palaeomagnetism of glaciogenic deposits has provided a quantitative basis for hypotheses of extreme climatic shifts in the late Neoproterozoic Era. The most recent global compilations of palaeomagnetic depositional latitudes for Proterozoic ice ages indicate a dominant mode near the palaeo-equator (Evans 2000, 2003), with no robust palaeo-polar deposits yet discovered. Such results could therefore support either the Snowball Earth (Kirschvink 1992) or the high-obliquity (Williams 1993) hypotheses for Precambrian ice ages, but would appear to reject the uniformitarian comparison to polar/temperate-restricted Phanerozoic glaciogenic deposits (Evans 2000). Hoffman (2009) has suggested that Neoproterozoic ice ages represent a globally all-encompassing ‘pan-glacial’ state of the Earth’s climate system, fundamentally distinct from either partially ice-covered ‘glacial –interglacial’ or ice-free ‘nonglacial’ palaeoclimates experienced during the Phanerozoic Eon. Several reviews of stratigraphic (Halverson et al. 2005, 2007), sedimentological (Hoffman & Schrag 2002; Eyles & Januszczak 2004; Fairchild & Kennedy 2007; Hoffman et al. 2007; Allen & Etienne 2008; Hoffman 2009) and palaeomagnetic (Trindade & Macouin 2007; Hoffman & Li 2009) data sets pertaining to Neoproterozoic ice ages have appeared recently. These discussions of glacial deposits owe much to the pioneering synthesis of Hambrey & Harland (1981), but the recent reviews arrive at differing conclusions regarding the extent and severity of Neoproterozoic ice ages. In particular, the study by Eyles & Januszczak (2004) is commonly cited as a palaeogeographic alternative that avoids the need for nonuniformitarian processes to account for the advance of widespread continental ice sheets into tropical palaeolatitudes. It becomes useful, then, to review the global evidence for or against low-latitude glaciation. Herein, we reassess Neoproterozoic glacial palaeolatitudes in light of new stratigraphic, geochronological and palaeomagnetic data obtained within the last decade, providing the first comprehensive update and revision of the palaeographic analyses of Evans (2000).

Methods Table 7.1 lists the known or alleged Neoproterozoic glaciogenic deposits, revising the unit-identifying numbering scheme introduced by Evans (2000), but with cross-references to that study,

as well as those of Hambrey & Harland (1981), Eyles & Januszczak (2004, table 1), Trindade & Macouin (2007) and Hoffman & Li (2009). Deposits are numbered by present geographical location, differentiating units that are separated from each other on the scale of 100 km or more, or in some cases distinguishing units of uncertain relative correlation that are presently adjacent via tectonic stacking in orogenic belts. Within each numbered location, the units are denoted (a, b, c, etc.) in ascending stratigraphic order. Many global chronostratigraphic schemes have been proposed, that assign ages to undated deposits via correlation, but herein, we adopt the approach used by Evans (2000) to consider each unit’s age constraints in isolation. This splitting, rather than lumping, approach serves to illustrate how few of the deposits are precisely dated or studied palaeomagnetically. Many of the deposits have been classified using the loose terms ‘Sturtian’, ‘Marinoan’, or ‘Ediacaran’ based on the lithographic character of either diamictites or their overlying cap carbonates (e.g. Hoffman & Li 2009). Recent U – Pb geochronology has thus far permitted many so-called Marinoan deposits to record the synchronous end of a widespread ice age of unknown duration, which ended at 635 Ma (Hoffmann et al. 2004; Condon et al. 2005; Zhang et al. 2008), and a mid-Ediacaran ice age at c. 580 Ma (Bowring et al. 2002, abstract only). U –Pb ages of so-called Sturtian deposits range from c. 765 Ma (Key et al. 2001) to possibly as young as c. 660 Ma (Fanning & Link 2008; from the type area in South Australia). Use of the lithological characteristics of cap carbonates as global correlation tools appears successful in some instances, but may be problematic in others (Corsetti & Lorentz 2006; Kendall et al. 2009). Finally, several of the units discussed by Evans (2000) are now considered irrelevant to discussions on Neoproterozoic ice ages, as they are likely nonglacial, or demonstrably Cambrian or younger in age. Those units are included in Table 7.1, but are now stripped of their numerical codes. In a few cases, questions remain about these issues, and the deposits retain their numerical status until documented otherwise. Table 7.1 also lists our preferred interpretation on the reliability of palaeolatitude determination for the deposits, using the threelevel qualitative scale after Evans (2000). Although quantitative measures of palaeomagnetic pole reliability exist (e.g. Van der Voo 1990), they have generally been tailored towards plate reconstructions rather than palaeoclimatic problems. Herein, we regard

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 93– 112. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.7

94

Table 7.1 Published assessments of glacial influence and depositional palaeolatitude for Neoproterozoic strata #

Deposit name

Lat (88 N)

HH81

E00

EJ04

– –

– – – –

– Uncertain (F12) Yes (F13) – Yes (F14) Yes (F15) Yes (F16) – – – – Yes (F17) Yes (F18) – – Yes (E18 –E20) – – Yes (F23,F24) Yes (F25) Yes (F25) Yes (E10) Yes (E10) Yes (E10) Yes (F26) –

– – No pmag data (#3) – ** 06 þ8/27 (#1) Unreliable pmag (#2) No pmag data (#4) No pmag data (#4) No pmag data (#5) * 08 +4 (#6) No pmag data (#7) No pmag data (#7) No pmag data (#7) No pmag data (#8) No pmag data (#8) – * 01 +4 (#9) – * 03 +8 (#10) See Mechum R (#11) ** 20–21 +4 (#11) – Unreliable pmag (#12) – No pmag data (#12) No pmag data (#14) Unreliable pmag (#15) Unreliable pmag (#15) No pmag data (#16) No pmag data (#16) No pmag data (#17) No pmag data (#18) No pmag data (#19)

Yes (E12) Yes (E12) Uncertain (E13) Yes (E14/E15) Yes (E16) Yes (E24 –E28) Yes (E30) Yes (E31)

** 33 þ14/212 (#20) ** 33 þ14/212 (#20) Too young (#21) No pmag data (#22) Unreliable pmag (#23) Unreliable pmag (#24) No pmag data (#25) No pmag data (#26)

Laurentia and environs 1a Hula Hula 1b Katakturuk unit 2 2a Upper Tindir: unit 2 2b Upper Tindir: unit 3b 3a Rapitan 3b Ice Brook (Stelfox) 4 Mt Lloyd George 5 Deserters Range 6 Mount Vreeland 7 Toby 8 Edwardsburg 9 Pocatello: Scout Mtn 10 Mineral Fork/Dutch Pk 11a Kingston Peak: Surprise 11b Kingston Peak: Wildrose 11c Ibex 11d Johnnie Rainstorm 12 Cerro Las Bolas – Florida Mtns 13 Konnarock/Grandf. Mtn 14a Mechum River 14b Fauquier 15a Port Askaig 15b Stralinchy (Reelan) 15c Loch na Cille 16 Ga˚seland/Charcot Land 17a Tillite Gp: Ulvesø 17b Tillite Gp: Storeelv 18a Elbobreen (Petrovbreen) 18b Wilsonbreen 19 West Spitsbergen 20 Moraenesø 21 Pearya

69.5 69.5 65 65 64 63.5 58 57 54.5 49.5 45 42.5 40.5 36 36 36 36 29 32 36.5 38.5 39 55.5 55.5 55.5 71 73 73 79.5 79.5 78 82 82.5

215.5 214.5 222 219 230 231.5 235 235.5 239 243 244.5 248 247.5 243.5 243.5 243.5 244 250.5 252.5 278.5 281.5 282.5 353.5 353.5 353.5 330.5 336 336 18 18 13 326 281.5

Baltica 22a 22b – 23 24 25 26 27a

70 70 68 64 61.5 53 60 58.5

28 28 19 14.5 11 32 57.5 59

Vestertana: Smalfjord Vestertana: Mortensnes Sito/Vakkejokk La˚ngmarkberg/Lillfja¨llet Moelv Vilchitsy/Blon N Urals: Churochnaya C Urals: Tany

Yes (F10) – Yes (F11) – Yes (F12)

Yes ?? – – – – – – – No No – – – – No – – Yes – – – Yes Yes – – – – – Yes (a or b) Yes (a or b) – – Yes – – –

TM07

HL09

This study

– – – – Q ¼ 4; 06 +4 – – – – – No pmag data Q ¼ 6; 08 +3 – – – – – – – – – – No pmag data – – – – – – – – – –

Hu(S) 0 –15 – Ti(S) 0 –15 – Ra(S) 0 –15 IB(M) 0 –15 – – Vr(M) 0–15 To(S) 0 –15 – Po(S) 0– 15 – Su(S) 0– 15 Wr(M) 0– 15 – – – – Kn(S) 15–30 – – Pt(S) 30– 45 Re(M) 30– 45 Lo(E) 75–90 – Ul(S) 15– 30 St(M) 30– 45 Pb(S) 15– 30 Wb(M) 30–45 – – –

No pmag data No pmag data *** 21 +3 No pmag data ** 18 +3 Unreliable pmag No pmag data No pmag data No pmag data No pmag data No pmag data No pmag data No pmag data No pmag data No pmag data No pmag data * 01 +4 No pmag data Too young ** 06 +5 ** 16 +3 * 78 +12 Unreliable pmag No pmag data Unreliable pmag No pmag data Unreliable pmag Unreliable pmag No pmag data No pmag data No pmag data No pmag data No pmag data

Q ¼ 3; 33 +9 Q ¼ 3; 33 +9 – – No pmag data – – –

Sm(M) 45–60 Mt(E) 45– 60 – – Mo(E) 45– 60 Vi(E) 45–60 Cn(E) 30– 45 Ty(S) 45– 60

* 33 þ14/212 No pmag data Too young No pmag data Unreliable pmag Unreliable pmag No pmag data No pmag data

D. A. D. EVANS & T. D. RAUB

Long (88 E)

27b 27c 28

C Urals: Koyva C Urals: Staryye Pechi S Urals: Kurgashlya

58.5 58.5 53.5

59 59 57.5

Yes (E31) Yes (E31) Yes (E32)

No pmag data (#26) No pmag data (#26) Unreliable pmag (#27)

– – –

– – –

– –

– – – – – – –

44 44 47 47 51 61 61 54.5 59 59

72 72 96 96 100 92 92 99 115 115

Yes (C27) Variable (C22– C28) – – – Nonglacial (C30) – – – Uncertain (C31)

No pmag data (#31) No pmag data (#31) No pmag data (#32) – No pmag data (#30) Unreliable pmag (#28) – – – No pmag data (#29)

China cratons 35a Aksu: Qiaoenbrak 35b Aksu: Yuermeinbrak 36a Quruqtagh: Bayisi 36b Quruqtagh: Tereeken 36c Quruqtagh: Hankalchough 37 Qaidam: Hongtiegou 38 Luoquan 39a Chang’an/Tiesiao/Jiangkou 39b Nantuo

41 41 41.5 41.5 41.5 37.5 34 27 27

79.5 79.5 87.5 87.5 87.5 96 115 111 111

Yes (C33) Yes (C33) Yes (C33) Yes (C33) Yes (C33) Yes (C33) Yes (C33,C34) Yes (C33,C35) Yes (C33,C35)

* 08 +8 (#33) * 08 +8 (#33) * 08 +8 (#33) * 08 +8 (#33) * 08 +8 (#33) – Unreliable pmag (#34) *** 30–40 +12 (#35) *** 30–40 +12 (#35)

India to Nubia 40 Blaini – Penganga 41 Rizu 42a Ghubrah 42b Fiq 43a Ayn 43b Shareef 44 Tambien 45 Atud

30 19.5 31 23 23 17.5 17.5 14 26

78.5 75 56 58 58 55 55 39 35

Yes (C14)

217 218 218.5 217 217 221.5 223 223

126 126.5 126.5 129 129 140 134 135

Australia and Mawsonland 46 Walsh 47a Landrigan 47b Egan 48a Fargoo 48b Moonlight Valley 49 Little Burke 50a Areyonga/Naburula/Yardida 50b Olympic/Mount Doreen

– – – – – – – – Yes (D16) Yes (D16) Yes (D16) Yes (D16) Yes (D16) Yes (D19) Yes (D17,D18) Yes (D17)

No pmag data No pmag data Unreliable pmag

– Br(S) 45– 60 Mk(S) 0–15 Kg(M) 0– 15 – Cv(S) 15–30 Pd(M) 0–15 Ma(M) 0 –15 Kh(S) 0 –15 Dz(M) 0– 15

No pmag data No pmag data Unreliable pmag * 03 + 11 Unreliable pmag Unreliable pmag No pmag data No pmag data No pmag data No pmag data

Yes Yes (a or b) Yes (a or b)

– – Q ¼ 4; 01 +3 – – – – – Q ¼ 6; 04 +4

– – By(S*) 30–45 Te(M) 15– 30 Ha(E) 15– 30 – Lq(E) 0 –15 Ji(S) 15–30 Na(M) 15– 30

No pmag data * 27 + 9 * 01 þ4/22 No pmag data No pmag data No pmag data Unreliable pmag No pmag data *** 37 +9

Unreliable pmag (#37) No pmag data (#38) – Unreliable pmag (#36) No pmag data (#36) Unreliable pmag (#36) No pmag data (#36) – –

– – – – Yes (’Shuram’) – – – –

– – – No pmag data Q ¼ 6; 15 +4 – – – –

Bl(M) 0– 15 – Ri(M) 0– 15 Gu(S) 15–30 Fi(M) 30– 45 Ay(S) 15–30 Sh(M) 30– 45 Ta(S) 15–30 –

Unreliable pmag Nonglacial Too young? No pmag data * 13 +7 Unreliable pmag * 18 +7 No pmag data No pmag data

** 45 þ14/212 (#42) No pmag data (#44) * 21 +8 (#43) No pmag data (#44) No pmag data (#44) Unreliable pmag (#40) No pmag data (#39) No pmag data (#40)

– – – –

Q ¼ 5; 45 +5 – – – – – – –

– La(M) 0– 15 Eg(E) 15–30 – – – Ar(S) 15– 30 Ol(M) 0 –15

** 45 þ14/212 No pmag data * 21 +8 No pmag data No pmag data Unreliable pmag No pmag data No pmag data

Blank – – – – – – – – – – – – –

Yes – – –

– –

GLOBAL NEOPROTEROZOIC GLACIAL PALAEOLATITUDES

Altaids and Siberia 29a West Altaids (Satan/Dzhetym) 29b West Altaids (Baykonur) 30a Tsagaan Oloom (Maikhan Ul) 30b Tsagaan Oloom (Khongoryn) 31 East Sayan/Khubsugul: Zabit 32a Chivida 32b Pod’em 33 Marnya 34a Patom: Kharluktakh 34b Patom: Dzhemkukan

– – –

(Continued)

95

96

Table 7.1 Continued Deposit name

Lat (88 N)

Long (88 E)

HH81

E00

EJ04

TM07

Boondawari/Wahlgu/Turkey Hill Chambers Bluff Sturtian (type area) Elatina Yancowinna Teamsters Creek Cottons Julius River Croles Hill Goldie

225 227 232 232 231.5 231.5 240 241 240.5 283

124 134 139 139 141.5 141.5 144 145 145 160

– –

– –

Yes (D21) Yes (D21) Yes (D20) Yes (D20) Uncertain (D22) Uncertain (D22) – –

No pmag data (#40) No pmag data (#39) Unreliable pmag *** 03–09 +4 (#40) No pmag data (#39) No pmag data (#40) Unreliable pmag (#41) No pmag data (#41) – No pmag data (#45)

– – – – – –

– – No pmag data Q ¼ 4,6; 05 +6 – – No pmag data No pmag data No pmag data –

– – St(S) 0– 15 El(M) 0–15 – – Co(M) 0– 15 Ju(S) 0 –15 Cr(E) 0 –15 –

Unreliable pmag No pmag data Unreliable pmag *** 10– 14 No pmag data No pmag data Unreliable pmag No pmag data No pmag data No pmag data

Kalahari and environs 58a Blaubeker/Court 58b Blasskrans/Naos 59a Kaigas 59b Numees 59c Namaskluft 59d Schwarzrand 60a Karoetjes Kop 60b Aties 61 Dernburg

223.5 223.5 228.5 228.5 228.5 227 231.5 232 228

17.5 17.5 16.5 16.5 16.5 18 18.5 18.5 15.5

Yes (A29) Map only (A29) Nonglacial (A29) Yes (A29) Yes (A29) Yes (A29) – Nonglacial (A29) –

Unreliable pmag (#46) No pmag data (#47) No pmag data (#48) No pmag data (#49) No pmag data (#49) * 38 +3 (#51) – No pmag data (#50) –

– – – – – – – – –

– – No pmag data – – – – – –

Bb(S) 0– 15 Bk(M) 0– 15 Ka(S*) 0 –15 Nu(M) 0– 15 0–15 – – – –

Unreliable pmag No pmag data No pmag data No pmag data No pmag data * 38 +3 No pmag data No pmag data No pmag data

Congo-Sa˜o Francisco 62a Grand Conglome´rat 62b Petit Conglome´rat 63 Geci 64 Tshibangu 65 Bunyoro 66a Akwokwo/Bandja 66b Bondo 67 Mintom 68a Sergipe: Jueteˆ/Ribeiro´polis 68b Sergipe: Palestina 69 Bebedouro 70 Rio Preto: Canabravinha 71 Brasilia int: Ibia´/Cristalina 72a Vazante: St. Antoˆnio do Bonito 72b Vazante: Lapa 73 Carandaı´ 74a Bambui: Jequitaı´/Macau´bas 74b Bambui: Inhau´ma 74c Bambui: Lagoa Formosa

211 211 213 22 1.5 3 5 3 211 211 211 211 218 217.5 217.5 221 217 219.5 218.5

27 27 35 29 31.5 24 20 13 322 322 319 314 313 313 313 316 316 315.5 313.5

Yes (A28) Uncertain (A28) – – Yes (A25) – – – – – Uncertain (G28) – Uncertain (G27) – – Uncertain (G30) Yes (G31) – –

No pmag data (#52) No pmag data (#53) – No pmag data (#54) No pmag data (#55) No pmag data (#55) No pmag data (#56) – No pmag data (#56) No pmag data (#56) No pmag data (#58) – No pmag data (#59) – – No pmag data (#60) Unreliable pmag (#57) – –

– – – – – – – – – – – – – – – –

Q ¼ 5; 10 +6 – – – – – – – – – – – – – – – – – –

Gr(S*) 15–30 Pe(M) 0 –15 – – – Ak(S) 30–45 Bo(M) 15– 30 – – Pa(M) 30–45 – – – – – – Je(S) 30–45 – –

** 10 +5 No pmag data No pmag data No pmag data No pmag data No pmag data No pmag data No pmag data No pmag data No pmag data Unreliable pmag No pmag data No pmag data No pmag data No pmag data No pmag data Unreliable pmag No pmag data No pmag data

#

Yes Blank

No – –

This study

D. A. D. EVANS & T. D. RAUB

51 52 53a 53b 54a 54b 55 56a 56b 57

HL09

320.5 14 14 16 16

West Africa and Hoggar 78a Basal Atar Gp 78b Jbe´liat/Bthaat Ergil (triad) 79 Mali Gp/Bakoye 80 Rokel River 81a Kodjari 81b Tamale/Obosum 82a Se´rie Verte (Tafeliant) 82b Se´rie Pourpre´e 83 Se´rie Tirririne 84a Sarhro 84b Tiddiline

22 22 15 8.5 10 8 21 24 21 31 30

352 352 351 348 0 0 2 2 8.5 352.5 352

Avalonia and Cadomia 85 Squantum 86 Gaskiers 87 Gwna/Mona 88 Brioverian (Granville) 89 Clanzschwitz/Weesenstein

42.5 47.5 53 48.5 51

Rio Pardo: Salobro West Congo Infe´rieure West Congo Supe´rieure Chuos/Varianto Ghaub

Uncertain (G29) Nonglacial (A26,A27) Nonglacial (A26,A27) Variable (A29) Uncertain (A29)

Uncertain glacial (#57) No pmag data (#61) No pmag data (#62) * 10 +5 (#63) Unreliable pmag (#64)

–

– Nonglacial (A18)

Unreliable pmag (#66) * 30– 70 (#65) Same as #65 (#67) No pmag data (#68) No pmag data (#69) No pmag data (#70) Unreliable pmag (#71) Same as #65 (#72) Unreliable pmag (#71) No pmag data (#73) No pmag data (#73)

289 307 355.5 358 13.5

Uncertain (F19) Yes (F20) – Nonglacial (E21) Correlative to (E22)?

* 55 þ8/27 (#74) * 31 þ10/28 (#75) – No pmag data (#76) –

Yes (G26) – – – – – – –

Amazonia and environs 90a Puga 90b Serra Azul 91 Chiquerı´o

215.5 215.5 215.5

303 303 285

Rio Plata and environs 92a Zanja del Tigre 92b Las Ventanas/Playa Hermosa 93 Picada das Grac¸as 94 Iporanga 95 Sierra del Volca´n

234.5 234.5 231 224.5 238

305 305 305.5 311.5 302

Yes (A19) Yes (A19) Yes (A20) Yes (A21) Yes (A21) Yes (A23) Yes (A22) Yes (A23)

– – –

– In(S) 30–45 Sp(M) 15– 30 Ch(S) 15–30 Gh(M) 0– 15

Nonglacial No pmag data No pmag data * 09 +5 Unreliable pmag

– – – – – –

– – – – – – – –

– Jb(M) 45–60 Ba(M) 30– 45 – Ko(M) 30–45 – – –

– –

– –

– –

Nonglacial? No pmag data No pmag data No pmag data No pmag data No pmag data Unreliable pmag No pmag data Unreliable pmag No pmag data Unreliable pmag

No No – Yes

– – – – No pmag data

–

Q ¼ 3; 55 +3 No pmag data – – –

Sq(E) 75– 90 Ga(E) 75– 90 – Gr(E) 45– 60 –

* 55 þ8/27 ** 34 þ8/27 No pmag data No pmag data No pmag data

No pmag data (#77) – No pmag data (13)

– – –

Q ¼ 5; 22 +6 – –

Pu(M) 30– 45 Az(E) 60– 75 Cq(S) 15–30

** 22 þ6/25 No pmag data No pmag data

– – Unreliable pmag (#78) – –

– – – – –

– – – – Q ¼ 5; 48 +7

– – – – –

No Yes – No

Nonglacial? Unreliable pmag Unreliable pmag No pmag data Too young?

Deposits are numbered by geographical area; within each region, they are lettered by ascending stratigraphic order. Glaciogenic influence assessments: HH81, Hambrey & Harland (1981); EJ04, Eyles & Januszczak (2004, table 1). The terms ‘yes’, ‘no’ and ‘uncertain’ refer to the question of whether a glaciogenic influence has been demonstrated according to each compilation. ‘Blank’ refers to blank entries in Eyles & Januszczak (2004, table 1). Palaeomagnetic constraint assessments: E00, Evans (2000); TM07, Trindade & Macouin (2007); HL09, Hoffman & Li (2009). Correlations by Hoffman & Li (2009) presented in parentheses, follow the deposit abbreviations used in their work: E, Ediacaran; M, Marinoan; S, Sturtian; S*, nominally Sturtian but with possibly older ages than the deposits labelled ‘S’. Asterisks in the columns for E00 and this study indicate the relative reliability (*, low; **, moderate; ***, high) of glacial palaeolatitudes assessed by Evans (2000) or herein. Q values use the reliability scheme of Van der Voo (1990). Throughout the table, dashed entries indicate that the deposit was not mentioned by the particular compilation.

GLOBAL NEOPROTEROZOIC GLACIAL PALAEOLATITUDES

215.5 25 25 221 221

75 76a 76b 77a 77b

97

98

D. A. D. EVANS & T. D. RAUB

reliability of glacial palaeolatitudes to depend mainly on the quality of palaeomagnetic data, but also on the confidence in chrono/lithostratigraphic correlations, if necessary for palaeolatitude estimation, and to a lesser extent the amount of consensus on the glaciogenicity of the deposits. In addition, some of our assessments in Table 7.1 are labelled ‘unreliable pmag [palaeomagnetic data]’, which merely indicates that the data are not applicable for reliable estimates of glacial palaeolatitude, although in some cases they could be useful for other purposes such as tectonic reconstructions. We do not provide references to all the stratigraphic, geochronological, and palaeomagnetic constraints on the deposits, but rather only discuss those instances in which our current assessment differs from previously published global compilations. Our global palaeolatitude estimates for Neoproterozoic ice ages, using this quality filtering method, are summarized graphically in Fig. 7.1. We begin with the assumption that Earth’s time-averaged (.103 years) magnetic field has the geometry of a geocentric-axial dipole (GAD). This assumption is justified to first order by the comparison of evaporite palaeolatitudes in the palaeomagnetic reference frame, relative to their expected palaeolatitudes according to intertropical convergence of zonal atmospheric circulation. Agreement between these two reference frames for most Proterozoic evaporite deposits implies that the GAD hypothesis is tenable for most of the last two billion years (Evans 2006). Departures from this model are possible, due to data gaps, over intervals as long as 100– 200 million years; notably, the Neoproterozoic glacial interval does indeed lack voluminous evaporite deposits. Ephemeral, dominantly non-GAD fields (on the order of 10 million years duration or shorter) are also permissible, as will be discussed further below, but in light of the broader record of a uniformitarian geodynamo (see also Swanson-Hysell et al. 2009), a burden of proof must lie with these alternatives for any specific age.

Palaeomagnetic constraints on Neoproterozoic glacial palaeolatitudes Laurentia and environs Macdonald et al. (2010a) have produced direct, high-precision, U –Pb age constraints on the Upper Mount Harper Group diamictite-bearing succession, at 716.5 Ma, which is precisely coeval with the immediately underlying Mount Harper Volcanic Complex and the transcontinental Franklin large igneous province. The latter has yielded a high-quality palaeomagnetic pole representing data measured in several laboratories and spanning the width of Arctic Canada plus northwestern Greenland (Denyszyn et al. 2009). The Mount Harper and underyling Fifteenmile groups are well correlated, lithostratigraphically, to the Tindir Group across the length of the Oglivie Mountains, and the Franklin palaeomagnetic pole implies a depositional palaeolatitude of 21+38 for the Tindir/Mount Harper glaciogenic succession. We assign this result to the highest reliability in the present analysis, and it is noteworthy by the fact that it derives from igneous rocks that are immune from any possible effects of systematic inclination shallowing. We assign only slightly less reliability to the application of the Franklin mean palaeomagnetic pole to the Rapitan glaciogenic deposits in the adjacent Mackenzie Mountains (implied depositional palaeolatitude of 18 + 38) because that succession lacks the volcanic manifestation of the Franklin event. Correlation among the Upper Tindir, Upper Mount Harper and Rapitan diamictite successions is moderately strong, because of their close associations with Fe formation, similarity of overlying shale units, and reasonable matches of chemostratigraphic profiles from underlying units (Macdonald et al. 2010a). The palaeolatitudes determined herein for these deposits are slightly higher (c. 208) than the previous estimate (06 þ8/–78; Evans 2000),

(a)

Somewhat reliable

Shareef

Moderately reliable Chuos

Fiq

Bayisi

Grand Congl.

Very reliable

Khongoryn

Egan

Uniform distribution (n=22)

YuerSmalMechum meinbrak fjord

Johnnie Rapitan

Konnarock 0

Schwarzrand

Elatina 10

Puga

Gaskiers

U.Tindir Nantuo unit 2 20

30

Squantum

Walsh 40

50

Fauquier

60

70

80

90

Palaeolatitude (°) Shareef Area-normalized distribution Fiq Chuos Schwarz. Bayisi Grand C. Egan Smalfj. Khong. Mechum Yuerm. Fauquier Johnnie Rapitan Puga Gaskiers Konnar. Elatina U.Tindir Nantuo Walsh Squant. 0 10 20 30 40 50 60 70 80

(b)

90

Fig. 7.1. Reliability of palaeomagnetic depostional latitudes for Neoproterzoic glacial deposits. (a) Unit-weight given to each estimate, following Evans (2000). Thick lines indicate the normalized probability at each latitude band, for a uniform (random) distribution over the surface of the Earth. (b) Latitude band-area normalization of the palaeolatitude estimates.

which was determined by direct measurements of the Rapitan succession that apparently were substantially affected by sedimentary inclination shallowing (implied flattening factor, f  0.3 calculated here; see Tauxe et al. 2008). Nonetheless, general concordance of 750– 700 Ma palaeomagnetic results from Laurentia (Fig. 7.2) indicates that the Earth’s geomagnetic field was stable at that time. Distinctly higher in the Tindir stratigraphy, a second diamictite level (unit 3b) and overlying cap carbonate bear striking resemblance to the Stelfox/Ravensthroat diamictite/cap carbonate couplet in the nearby Mackenzie Mountains (Macdonald et al. 2010b). Neither of the younger diamictite/cap carbonate pairs at the two localities have palaeomagnetic constraints on depositional palaeolatitude. Inliers on Alaska’s North Slope belong to an Arctic Alaska – Chukotka microplate of questionable Laurentian or Siberian affinity. Regionally, diamictite (informally named Hula Hula) lies above post-760 Ma volcanic rocks, and locally, it is interbedded with basaltic flows (Macdonald et al. 2009a). The overlying Katakturuk dolomite contains a diagnostic unit (Katakturuk 2) considered correlative to classic Marinoan cap carbonates elsewhere in the world. Although the Hula Hula-associated flows might correlate to the Franklin event and merit application of that palaeomagnetic pole, the allochthonous nature of the terrane and uncertain timing and style of its accretion to Alaska render these deposits palaeomagnetically unconstrained. New U –Pb ages on zircon from Neoproterozoic successions in Idaho support several distinct pulses of glaciation in that region. In the Pocatello region, Fanning & Link (2004) produced U –Pb SHRIMP ages constraining the Scout Mountain Member (mainly diamictite) of the Pocatello Formation, to slightly younger than 709 + 5 Ma and substantially older than 667 + 5 Ma. The younger age was determined using a zircon subpopulation from a channelized tuff horizon that lies closely beneath a crystal fanbearing marble. The fans resemble those in post-glacial cap carbonates. It was suggested that the Scout Mountain Member represents a distinct glacial event from the Edwardsburg metadiamictites dated by Lund et al. (2003) at c. 685 Ma. More recently, however, the older age has been ‘corrected’ to 686 + 4 Ma by

GLOBAL NEOPROTEROZOIC GLACIAL PALAEOLATITUDES

1a

17a/ 18a

2a 15a

3a 7 9/10 11a

14a 13

Fig. 7.2. Palaeoequators plotted across Laurentia, according to the named palaeomagnetic poles from 750 to 700 Ma, as discussed in the text and in Evans (2000). Numbered stars correspond to glaciogenic deposits keyed to Table 7.1. Star colours are keyed to the reliability scheme of Figure 7.1, or if uncoloured, stars represent deposits typically correlated with the Rapitan glaciation. References to palaeoequators given in Evans (2000) except for the recently updated Franklin LIP mean (Denyszyn et al. 2009).

the same authors (Fanning & Link 2008). These dates strengthen the evidence for glaciogenic deposits in the Cordilleran rift basins that are younger than 700 Ma. Correlations across the Cordilleran margin of Laurentia have emphasized the close association of allegedly glaciogenic diamictites and volcanic rocks in that time frame. Evans (2000) suggested that all of these deposits could be correlative, with an age of c. 740–720 Ma. In that case, palaeomagnetic poles from the c. 720 Ma Franklin event could be used to infer palaeolatitudes of the Cordillera (applying the GAD field assumption and a rigid Laurentian plate). However, new data suggest some of the deposits of this diamictite –volcanic association are demonstrably younger, and therefore applications of the Franklin palaeomagnetic pole to undated Cordilleran deposits are tenuous. For these reasons, the Toby Formation’s deep-tropical palaeolatitude estimate (Evans 2000) is no longer considered to be valid. Elsewhere in western Laurentia, the Johnnie Rainstorm Member’s incised canyons and features reminiscent of postglacial cap carbonates (Corsetti & Kaufman 2003) still retain the nearequatorial palaeolatitude of 01 + 48 as assessed by Evans (2000). Stratigraphically below that unit, but above the Kingston Peak glaciogenic unit, the Ibex Formation contains a polymictic conglomerate and overlying pink dolostone with negative carbon isotope values similar to many postglacial cap carbonates (Corsetti & Kaufman 2005). However, it is impossible to apply palaeomagnetic constraints from better dated units elsewhere in Laurentia without precise age constraints. A newly recognized glacial/capcarbonate succession has been described in Sonora, Mexico (Corsetti et al. 2007), lacking both a precise age and a palaeomagnetic estimate of depositional palaeolatitude. Finally, although Evans (2000) included the Florida Mountains (New Mexico) diamictite

99

in his summary figure on glacial palaeolatitudes, that unit is demonstrably early Palaeozoic in age and is therefore excluded from the present analysis. In the central Appalachians of eastern Laurentia, pebbly mudstone diamictites of the Grandfather Mountain Formation lie immediately above rhyolitic flows dated at 742 + 2 Ma; this succession has been correlated with the Konnarock Formation diamictites and rhythmites with dropstones. In the Blue Ridge Mountains of Virginia, the c. 730–700 Ma Mechum River succession also contains glaciogenic features (Bailey et al. 2007). Evans (2000) correlated all three successions for the sake of simplicity, and estimated depositional palaeolatitudes based on a palaeomagnetic pole from the Franklin LIP at c. 720 Ma. In the present compilation, however, we distinguish the Grandfather Mountain and Konnarock formations from the Mechum River succession, based on slightly different ages as reviewed above. For the former units, we apply a new palaeomagnetic pole from the Kwagunt Formation in the Grand Canyon (Weil et al. 2004), which is precisely coeval at 742 + 6 Ma (Karlstrom et al. 2000). Assuming negligible rotation of the Colorado Plateau relative to cratonic North America, as considered by Weil et al. (2004), this pole implies a palaeolatitude of 06 + 58 for the Konnarock/Grandfather Mountain region. Alternatively, a near-maximum estimate of Colorado Plateau rotation of 128 anticlockwise would imply a palaeolatitude for Konnarock/Grandfather Mountain at 04 + 58, within the error of the no-rotation model. This estimate is rated as only moderately reliable, because the dated Grandfather Mountain diamictites are not yet conclusively demonstrated as glacially influenced, and the presumed correlative Konnarock glaciogenic strata have more lax age constraints. The Mechum River succession remains within the age range of the c. 720 Ma Franklin LIP, and the new grand-mean pole for that event (Denyszyn et al. 2009), implies a depositional palaeolatitude of 16 + 38 (assuming an internally rigid Laurentian plate and a geocentric axial dipole geomagnetic field). Stratigraphically higher in the succession, the Fauquier Formation contains a basal diamictite member, followed by a thick sandstone interval that is capped by a thin carbonate unit with a negative d13C excursion; in turn, this unit is followed conformably by basalts of the c. 570 Ma Catoctin Formation (Hebert et al. 2010). Palaeomagnetic results from the latter volcanic succession include two remanence components of greatly differing inclination, hence, implied palaeolatitude (Meert et al. 1994). The coexistence of both highand low-inclination directions in these rocks is symptomatic of many broadly coeval igneous formations in eastern North America, leading to a fundamental debate on their general interpretation (e.g. Meert & Van der Voo 2001; Pisarevsky et al. 2001). The highpalaeolatitude Catoctin component is corroborated by other data of generally higher quality than the low-palaeolatitude data from the Iapetus margin of Laurentia (McCausland et al. 2007), so we prefer its implied palaeolatitude of 78 + 128 for the conformably underlying Fauquier Formation. This is the first near-polar estimate for a Neoproterozoic glacial deposit among recent palaeogeographic syntheses, but it is only somewhat reliable for the following reasons: (i) the Catoctin palaeomagnetic data remain ambiguous with respect to the two remanence components, (ii) the corroborating data from elsewhere in eastern Laurentia are also not of the highest quality, (iii) although the contact between the Fauquier Formation and the overlying basalts appears to be conformable, the basal Fauquier diamictites lie as much as hundreds of metres lower in the stratigraphy, and (iv) the 580– 570 Ma interval appears to harbour very rapid apparent polar wander for Laurentia (McCausland et al. 2007), so absolute age uncertainties of only a few million years can permit a wide range of possible palaeolatitudes (tens of degrees) for sedimentary basins on that continent. In Scotland and stratigraphically correlative areas of northern Ireland, the Port Askaig glaciogenic level is now generally correlated to the Sturtian ice ages (Prave 1999; Condon & Prave 2000;

100

D. A. D. EVANS & T. D. RAUB

McCay et al. 2006), but with the uncertainty surrounding the numerical age of Sturtian deposits, it is not possible to apply extrabasinal palaeomagnetic data to address the palaeolatitude of Port Askaig deposition. Older palaeomagnetic data obtained directly from the Port Askaig succession were summarized by Evans (2000) as unreliable, a conclusion followed here. McCay et al. (2006) also identified a so-called Marinoan equivalent in northern Ireland, comprising the Stralinchy and Reelan Formations, based on the presence of diamictite, exotic clasts interpreted as dropstones, and a cap carbonate with a negative d13C anomaly. As with other putative Laurentian glaciogenic deposits of this age (assumed to be c. 635 Ma by correlation; see above), there are no available palaeomagnetic data from Laurentia that can be applied either directly from the relevant sedimentary basin or from elsewhere across the palaeocontinent. The youngest purported Neoproterozoic glaciogenic unit in the Scottish and northern Irish Dalradian stratigraphic succession is the Loch na Cille boulder bed (and its correlative units). It closely overlies the Tayvallich Volcanics that have been recently re-dated at 601 + 4 Ma (Dempster et al. 2002). The two relevant palaeomagnetic poles from Laurentia that approximate this age are from the Long Range dykes of Newfoundland and Labrador, dated by U – Pb at 615+2 and 614 þ 6/–4 Ma (Kamo et al. 1989; Kamo & Gower 1994) and the Grenville Dykes of southeastern Ontario, dated at 590 þ 2/ –1 Ma (Kamo et al. 1995). Both data sets contain an unusually large spread of inclinations and hence implied palaeolatitudes of emplacement, confounding simple interpretations (reviewed by Hodych & Cox 2007; McCausland et al. 2007). Low to moderate palaeolatitudes for the Long Range dykes (Murthy et al. 1992) are supported by subsequent work (McCausland et al. 2009, abstract only). For the younger, Grenville Dykes, a positive baked-contact test demonstrates that the steep palaeomagnetic B remanence found in some of the dykes is primary (Murthy 1971; Hyodo & Dunlop 1993). If these results withstand further scrutiny, then the Loch na Cille boulder bed would have been deposited during a migration of Laurentia from low to high latitudes. For the East Greenland Neoproterozoic succession, Mac Niocaill et al. (2008) report in abstract some preliminary palaeomagnetic data that would appear to indicate rapid motion of Laurentia across palaeolatitudes. The upper Tillite Group glaciogenic unit bears a high-palaeolatitude magnetic remanence direction, but further sample analysis is needed to confirm that result.

Baltica Several new geochronological and palaeomagnetic constraints have forced revision of Evans’s (2000) estimate of depositional palaeolatitude of Neoproterozoic glaciogenic deposits on Baltica. First, the Vestertana Group, in which the palaeomagnetically studied Nyborg Formation lies between diamictite units of the underlying Smalfjord Formation and overlying Mortensnes Formation, has an imprecise palaeolatitude of 33 þ 14/– 128 (Torsvik et al. 1995). The reliability of this estimate has been downgraded because recent syntheses of Baltica’s apparent polar wander path (Iglesia-Llanos et al. 2005) show the Nyborg pole to coincide with a c. 510 Ma overprint pole from elsewhere in Baltica. Because Middle Cambrian Finnmarkian deformation is prevalent in the northernmost Caledonides (Roberts 2003), there is a distinct possibility that the Nyborg palaeomagnetic remanence is an overprint acquired shortly before Finnmarkian folding, and thus not indicative of depositional palaeolatitudes. In addition, given the possibility of substantial age difference between the Smalfjord and Mortensnes Formations (e.g. correlations by Hoffman & Li 2009), the Nyborg pole is now considered as only applicable (if primary) to the conformably underlying Smalfjord Formation. Elsewhere in Baltica, depositional palaeolatitudes on Ediacaran glacial deposits are similarly or even more poorly constrained. New ages exist for the Moelv diamictites of the southern

Norwegian sparagmite basins: younger than 620 + 15 Ma detrital zircons in a distally underlying quartzite (Bingen et al. 2005) or perhaps even younger than 561 + 4 Ma based on Re – Os dating of the immediately underlying Biri black shale (Hannah et al. 2007, abstract only). Nonetheless, the mid-Ediacaran apparent polar wander path for Baltica is highly uncertain (Iglesia-Llanos et al. 2005; Meert et al. 2007). Where Vilchitsy and Blon diamictite units are in close stratigraphic proximity to c. 551 Ma Volhyn volcanic rocks of the southwestern Baltic craton, palaeomagnetic results from the latter units imply a depositional palaeolatitude of 50 + 38 (A component of Nawrocki et al. 2004); however, nearly coeval sedimentary rocks of the White Sea region (555.3 + 0.3 Ma; Martin et al. 2000) yield a more convincingly primary palaeomagnetic pole (Popov et al. 2002, 2005; Iglesia-Llanos et al. 2005) that implies a palaeolatitude of 28 + 38 for the Vilchitsy/Blon region. Any of these younger poles could also apply to the Moelv diamictite and correlative units in the Caledonides, now dated as late Ediacaran as noted above. Ediacaran palaeogeography of Baltica is usually paired with eastern Laurentia, but both cratons’ palaeomagnetic data appear to indicate rapid motions that challenge uniformitarian platetectonic interpretations (Abrajevitch & Van der Voo 2010; McCausland et al. 2011). Amid these uncertainties, the mid –late Ediacaran glacial palaeolatitudes of Baltica must be considered unresolved at present.

Altaids and Siberia Chumakov (2009) has summarized the terminal Proterozoic to Cambrian stratigraphy across Kazakhstan and Kyrgyzstan, distinguishing between two prominent glacial levels (Satan, Rang, or Dzhetym Formations below, and Baykonur above). Both levels are younger than the c. 700 Ma volcanic complexes lying unconformably below. He also has described the Zabit Formation, in the East Sayan region, as glaciogenic and correlative with the Baykonur level, which is in turn correlated with the uppermost glacial level in Tarim (Hankalchough Formation; see below). A complex palaeomagnetic data set was obtained from the succession c. 1000 m or more above the Zabit diamictites into Cambrian strata (Kravchinsky et al. 2010), which may be useful for tectonic reconstructions but shed no direct light on glacial palaeolatitudes. Two glaciogenic levels are also now recognized in the Dzhabkhan region of western Mongolia (Macdonald et al. 2009b): the basal, Maikhan Ul Member of the Tsagaan Oloom Formation, and the Khongoryn Member in the middle of the same formation. Chemostratigraphy of carbon isotopes in this succession supports correlation with the mid-Cryogenian (so-called Sturtian) and late Cryogenian (Marinoan) ice ages worldwide. Kravchinsky et al. (2001) found several components of magnetization through mainly the middle and upper parts of the Tsagaan Oloom Formation, and preferred a high-unblocking-temperature, lowpalaeolatitude component as tentatively primary. That component, although of dual polarity, is not supported by field stability or reversals tests on the magnetization age, so is only somewhat reliable. Levashova et al. (2010) found a high-unblockingtemperature component of magnetization in the underlying Dzabkhan volcanic rocks, with U –Pb laser ablation ages of c. 800– 770 Ma, with moderate palaeolatitudes of igneous emplacement. This estimate, however, would precede Maikhan Ul glaciogenic deposition by nearly 100 million years if the latter is correlated to the Rapitan ice age as discussed above (Macdonald et al. 2010a). New recognition of late Neoproterozoic glaciation in southern Siberia has focused on the Yenisei ridge and the pre-Sayan region NW of Irkutsk. Sovetov (2002) and Sovetov & Komlev (2005) describe regional stratigraphic correlations around two distinct levels of diamictites with various glaciogenic features. The older Chivida unit of the Chingasan Group is more limited in geographic distribution in the Teya-Chapa trough (Yenisei region)

GLOBAL NEOPROTEROZOIC GLACIAL PALAEOLATITUDES

relative to the younger Pod’em Formation that is correlated to the Marnya Formation best exposed along the Biryusa and Uda Rivers (pre-Sayan). The Chivida diamictites are closely associated with volcanic rocks and are traditionally associated with an age of c. 700– 750 Ma (Khomentovsky 1986), but they are not directly dated with modern methods. The Marnya Formation contains clasts that are speculated to derive from the Nersa Complex, a nearby mafic intrusive suite, according to Sovetov & Komlev (2005). Those authors quoted an Ar/Ar age of 611 + 3 Ma for the Nersa Complex, but that age (now revised to 612 + 6 Ma), in fact, derives from other mafic intrusive suites in southern Siberia (Gladkochub et al. 2006); thus the Marnya Formation diamictites lack any precise age constraints. In the Patom region NE of Lake Baikal, at least two widespread glaciogenic levels have been reported: the older Kharlaktakh unit of the Ballaganakh Group and the younger Dzhemkukan or Bol’shoy Patom Formation of the Dal’nyaya Taiga Group (Pokrovskii et al. 2006; Sovetov 2008). No results from these successions are available in the global palaeomagnetic database. The late Neoproterozoic to Early Cambrian palaeomagnetic apparent polar wander path for Siberia is highly contentious with a wide scatter of poles that universally imply low palaeolatitude for the craton; and if one includes the full range of data (Kirschvink & Rozanov 1984; Kravchinsky et al. 2001; Gallet et al. 2003; Metelkin et al. 2005) rather than a simplified selection (Cocks & Torsvik 2007), the problem is compounded further.

China cratons Recent work has helped refine the chronology and palaeomagnetism of Neoproterozoic volcanic-sedimentary successions on the margin of the Tarim block in NW China. Two principal regions contain diamictites interpreted as glaciogenic deposits: Aksu in the west, and Quruqtagh in the east. The Aksu succession contains two diamictite levels, and a coarse magnetostratigraphic study by Li et al. (1991) yielded near-equatorial palaeolatitudes throughout the succession. More recently, Zhan et al. (2007) conducted a more detailed palaeomagnetic investigation of only the Sugetbrak Formation, which rests with disconformity upon the upper diamictite level (Yuermeinbrak Formation). Although Zhan et al. (2007) suggested that their high-stability, dual-polarity direction for the upper Aksu succession, which was acquired prior to folding of unspecified age, implies a palaeolatitude of c. 278 for the Yuermeinbrak ice age, there is lingering concern over why the palaeomagnetic results by Li et al. (1991) and Zhan et al. (2007), in part from the same sampling region on the same stratigraphic units, are so different. We tentatively accept the new constraint as more reliable, because the pole by Li et al. (1991) is similar to Cretaceous poles from Tarim and could represent an overprint of that age. Farther east in the Tarim block, the Quruqtagh region contains three Neoproterozoic diamictite levels: Bayisi, Altungol-Tereeken and Hankalchough. The oldest level lies immediately above bimodal volcanic rocks dated by U –Pb SHRIMP at 740 + 7 Ma (previously 755 + 15 Ma from the same sample; Xu et al. 2005), and below volcanic rocks dated at 725 + 10 Ma (Xu et al. 2009). The proximity of ages between volcanism and diamictite deposition allows the use of palaeomagnetic data from the former rocks (Huang et al. 2005) to constrain the latitude of glaciation, assuming the Bayisi diamictites are glaciogenic. That palaeomagnetic work found a characteristic remanence component of high consistency but no field tests were performed on its stability; it is therefore unclear whether its near-equatorial palaeolatitude (01 þ 4/ –28) is primary or secondary. It is rated here as only somewhat reliable. The Altungol-Tereeken diamictites, overlain by a notable pink cap carbonate, lie above the 725 Ma volcanic rocks and below another volcanic horizon dated at 615 + 6 Ma (Xu et al. 2009). The Altungol-Tereeken interval has not been studied palaeomagnetically. In summary, both of the two recent

101

palaeomagnetic studies from Tarim demand further work to verify their applicability to Neoproterozoic glacial palaeolatitudes. In the Qaidam terrane near the boundary between the Tarim and North China cratons, the Hongtiegou diamictite and overlying cap carbonate have been correlated, via stable isotope stratigraphy, to the uppermost Ediacaran (Sinian) successions in South China (Shen et al. 2010). Palaeomagnetic data are not available from the Neoproterozoic of Qaidam. Chronology of the Sinian sedimentary basins in South China has improved vastly during the past few years. Two glacial levels are described with various local names, but generally referred to as the older Chang’an (or Tiesiao, or Jiangkou) episode and the younger Nantuo ice age. The Chang’an level is now bracketed in age between 725 + 10 Ma (youngest U – Pb SHRIMP population from detrital zircons in the underlying Danzhou Group; Zhang et al. 2008) and 663 + 4 Ma (Zhou et al. 2004). The Nantuo diamictite is younger than 654.5 + 3.8 Ma and contains a basal tuff layer dated at 636.3 + 4.9 Ma (Zhang et al. 2008), with a cap carbonate dated at 635.2 + 0.6 Ma (Condon et al. 2005). Prior to the acquisition of these ages, Evans (2000) assigned a moderate depositional palaeolatitude to the Nantuo Formation based on results from the immediately underlying Liantuo redbeds (Evans et al. 2000). However, given that the Liantuo Formation contains an ash bed dated at 748 + 12 Ma (Evans et al. 2000), the new dates from South China now preclude use of the Liantuo pole to constrain the significantly younger ice ages (Zhang et al. 2009). Macouin et al. (2004) conducted a palaeomagnetic study on the Doushantuo Formation, which conformably overlies the Nantuo diamictites across most of South China. They obtained consistent results throughout the entire formation, implying near-equatorial palaeolatitudes, except for the postglacial cap carbonate at its base, which was palaeomagnetically unstable. Given that the Doushantuo Formation is now constrained to span more than 80 million years of time (Condon et al. 2005), and that the direction obtained by Macouin et al. (2004) is of single geomagnetic polarity, we suspect that it is a secondary overprint. The positive fold test only constrains the age of magnetization to pre-Mesozoic, and the direction is similar to both Early Cambrian and Silurian directions for South China. Zhang et al. (2009) correct numerous misconceptions about the stratigraphy and palaeomagnetic database for the Sinian of South China, and the best estimate for depositional palaeolatitude of the Nantuo diamictite reverts to the value obtained by Zhang & Piper (1997) from Yunnan Province, at 37 + 98. An identical result has been obtained by Zhang et al. (2006, abstract only) from Guizhou Province, supporting this conclusion. The soft-sediment fold test reported in the earlier study (Zhang & Piper 1997) imparts a high level of confidence to the moderate-palaeolatitude determination.

India to Nubia Although Evans (2000) speculated that conglomerates within the Penganga Group might be glaciogenic, more thorough study of that region suggests no demonstrable glacial influence, as is evident, for example, in the discussions by Mukhopadhyay & Chaudhuri (2003) and Chakraborty et al. (2010). Other Neoproterozoic glaciogenic deposits in southern Asia are listed in Table 7.1. In Iran, the so-called Infracambrian stratigraphic succession, between Neoproterozoic metamorphic basement and overlying nonmetamorphosed Cambrian sedimentary rocks, is variable in lithology but contains diamictite and lonestone-bearing units in its lower part, named the Rizu Formation (Huckriede et al. 1962; Hamdi 1992). The upper part of the succession is equally lithologically variable, as currently correlated across the country (Alsharhan & Nairn 1997), but contains rhyolites thought to be consanguineous with c. 530 Ma intrusions (Ramezani & Tucker 2003). No reliable palaeomagnetic data have yet been published on the Rizu strata or correlative rocks.

102

D. A. D. EVANS & T. D. RAUB

Despite a new, precise U–Pb TIMS age of 711.5+ 1.1 Ma for the Ghubrah diamictite in the Jebel Akhdar region of northeastern Oman (Bowring et al. 2007; all sources of error included), its palaeolatitude remains poorly constrained. However, unmetamorphosed deposits of the possibly Ghubrah-correlative, lower Mirbat sandstone/diamictite assemblage, now renamed as the Ayn Formation (Rieu et al. 2006, 2007), yield a low-reliability palaeomagnetic direction (Kempf et al. 2000). This under-defined magnetization (n ¼ 10 samples from two sites) would imply a palaeolatitude of 09 + 48 for the Ayn Formation, but we cannot include it in our final compilation due to the severely limited sample size. On the younger and less precisely dated Fiq diamictite and its overlying Hadash cap carbonate, a two-polarity, pre-fold palaeomagnetic remanence implies a palaeolatitude of 13 + 78 for the Jebel Akhdar region, and 18 + 78 for the Mirbat region (Kilner et al. 2005), the latter area hosting the erosional remnants of an upper diamictite unit (Shareef Formation) that is correlated with the Fiq (Rieu et al. 2007). The folding is probably Palaeozoic in age, suggesting that the remanence could be primary. However, Rowan & Tait (2010, abstract only) have undertaken a more detailed magnetostratigraphic study of the Hadash cap carbonate in two sections with greatly differing structural attitudes, and have found the same remanence direction to be definitively postfolding, and possibly associated with regional remagnetization during emplacement of the Semail ophiolite. Given that only about 20% of the samples in the study by Kilner et al. (2005) yielded the preferred remanence direction, and only a few of their sites contributed significantly to their fold test, the lowpalaeolatitude estimates for Fiq and Shareef diamictites can only be considered as somewhat reliable. Allen (2007) points out additional problems with the published data from Oman, in terms of their stratigraphic correlations and tectonic implications. In the northern East African Orogen of Ethiopia, the Tambien Group contains a glaciogenic/cap-carbonate succession that has only recently been described in detail (Beyth et al. 2003; Miller et al. 2003). The Tambien Group is tightly folded into structures that are crosscut by the Mereb granite suite, dated by zircon Pb-evaporation on two plutons at 606.0 + 0.9 and 613.4 + 0.9 Ma (Miller et al. 2003), and by zircon U –Pb SHRIMP at 612.3 + 7.5 Ma (Avigad et al. 2007). Detrital zircons in the Negash diamictite have yielded near-concordant U – Pb SHRIMP analyses as young as c. 750 Ma (Avigad et al. 2007; all younger diamictite detrital ages quoted in that work are more than 10% discordant). No palaeomagnetic data are available for the Tambien Group. Farther to the north in the Arabian –Nubian Shield, widespread banded Fe-formation is associated with diamictites of the Atud and Nuwaybah Formations (Stern et al. 2006; Ali et al. 2009).

Australia and Mawsonland The palaeolatitude assessments by Evans (2000) remain valid for most of the many glaciogenic deposits scattered across the late Neoproterozoic glacial record of Australia. Ediacaran stratigraphic correlations in Australia are now under intense debate, following the recognition of the possibly glaciogenic Croles Hill Diamictite in Tasmania coincident with 582 + 4 Ma rhyolite volcanism (Calver et al. 2004). Whether those diamictites correlate to the Elatina Formation in South Australia (by way of the Cottons Breccia on King Island), or whether they are substantially younger than Elatina, carries implications as profound as a factorof-two uncertainty in the duration of the Ediacaran Period that begins, as formally defined, with deposition of the post-Elatina, Nuccaleena cap carbonate. No published palaeomagnetic data are available for the Tasmanian diamictites, but data are reported in abstract for the Cottons Breccia, with an implied palaeolatitude of 21 þ 9/ –88; and the subsequent igneous activity dated at c. 580 Ma (Meffre et al. 2004), with implied palaeolatitudes

in the range of c. 12 –218 (McWilliams & Schmidt 2003) that could be representative of Croles Hill deposition. As these data are reported in abstract only, they are omitted from our final compilation. As for the Elatina glacial deposits and their continent-wide correlative units, new palaeomagnetic results add to an already long history of palaeomagnetic studies of this unit (reviewed by Evans 2000; Williams et al. 2008). Schmidt et al. (2009) presented a detailed analysis of palaeomagnetic data from both the Elatina and Nuccaleena formations in the Flinders Ranges, combining new results with previously published work. Based on moderate differences in remanent inclination between the two formations, they concluded that deglaciation coincided with rapid drift of the proto-Australian continent during a hiatus in deposition between the two units. The observed mean remanent inclination of 12.98 from the Elatina Formation would correspond to a depositional palaeolatitude of 6.58 if the geomagnetic field was dipolar and sedimentary inclination shallowing was negligible. For the Nuccaleena Formation, the mean inclination of 34.98 would correspond to a palaeolatitude of 19.28. Schmidt et al. (2009) also reported results from four measurements designed to quantify the effects of remanence shallowing in the sedimentary rocks. In the first measurement, a correction for their measured values of remanence anisotropy would increase the Elatina inclination to 14.08; however, Tauxe et al. (2008) showed that this method commonly underestimates the total amount of inclination shallowing. Indeed, two other measurements by Schmidt et al. (2009) appear to justify such an assessment. Their second test, elongation/inclination analysis on the distribution of data around the mean direction (using the assumed TK03 model of geomagnetic secular variation), would correct the Elatina inclination to 198 (bootstrapgenerated uncertainties were not shown but are typically large for such tests). The third analysis, a dip test on crossbedded sandstones in the succession, would correct the Elatina inclination to between 18.4 and 20.28. Because sandstones should be the least compacted of all the Elatina siliciclastic facies, this correction of the aggregate palaeomagnetic results include data from both sandstones and siltstones. The fourth test considered differences in inclination between carbonates and mudstones in the Nuccaleena Formation, the latter of which are expected to be more compacted than the former. The carbonates’ mean inclination was 33.68 (a95 ¼ 7.78), substantially greater than the mudstones’ mean inclination of 21.18 (a95 ¼ 7.88). Schmidt et al. (2009) claimed that these two means (taking into account both declination and inclination) were indistinguishable, using a test from McFadden & Lowes (1981). However, the test they used from that paper (equation 25) assumes an equivalency in concentration parameters between the two sample sets, which for the Nuccaleena data can be rejected with .99% confidence (Fisher et al. 1987, p. 219). If the correct test is used (the more general equation 23 from the same paper), the Nuccaleena data-derived statistic increases to 4.40, exceeding the p ¼ 0.05 F[2,76] value of 3.12; thus, the carbonates’ and mudstones’ inclinations are significantly distinct at more than 95% confidence, and the Nuccaleena data therefore do support inclination shallowing having affected any fine-grained sedimentary rocks contributing to the mean directions in the sampling area (contra Schmidt et al. 2009). In summary, the various inclination corrections to the Elatina palaeomagnetic data, as reported by Schmidt et al. (2009), imply a ‘true’ inclination value around 208, with an implied palaeolatitude of approximately 108 for the sampled areas in the Flinders Ranges. Nuccaleena inclinations are steeper, due to the lesscompacted carbonate lithology and perhaps also partly due to Australian plate motion between the ice age and the deglacial interval. Raub (2008) found a mean inclination of 278 (a95 ¼ 3.58) for the Nuccaleena Formation, with several correlatable magnetozones spanning geodynamo reversals during deposition. That inclination value indicates a primary palaeolatitude of 14 + 28, between the estimates by Schmidt et al. (2009) for their compaction-corrected

GLOBAL NEOPROTEROZOIC GLACIAL PALAEOLATITUDES

Elatina and Nuccaleena data. Rare stratigraphic sections preserve sedimentological conformity (mixed transition) between siliciclastic rocks indistinguishable from Elatina Formation sandstones and Nuccaleena Formation cap-carbonate beds (Raub et al. 2007a), and the onlap relationships between the two formations described by Schmidt et al. (2009) can be accounted for by sediment draping of an irregular post-glacial topography, perhaps with diachronous onset of cap-carbonate deposition, without a substantial lacuna between the formations (Rose & Maloof 2010). For this reason, we propose that 10–148 of palaeolatitude best characterizes the Elatina/Nuccaleena glacial/postglacial palaeogeography. The 10–208 palaeolatitude range (as indicated in Fig. 7.1) also well represents the distribution of likely Elatina-correlative deposits across the Australian –Mawsonland palaeocontinent (Fig. 7.3). The compaction-corrected Elatina/Nuccaleena result is considered here to be one of the two most reliable palaeolatitude determinations among all Neoproterozoic glaciogenic deposits. Grey et al. (2005) produced a landmark correlation of Neoproterozoic strata among deep boreholes throughout the Officer Basin of Western Australia, using diamictites and other indicators. New palaeomagnetic data from the Wahlgu diamictite level (correlated with the Elatina Formation) in the Lancer-1 borehole (Pisarevsky et al. 2007) generate a palaeolatitude of 7 + 178, but the data are highly scattered and merely consistent with the range of previous results from the Elatina Formation. Haines et al. (2008) revise the correlation of one diamictite and dropstone-bearing unit (therein named the Pirrilyungka Formation) to an age substantially older than the Wahlgu and Turkey Hill Formations, perhaps correlating with the Chambers Bluff diamictite in South Australia. In the Kimberley region of northern Australia, the cap carbonate of the Walsh Tillite (western Kimberley) has yielded a moderatepalaeolatitude magnetic remanence supported by a tectonic fold test and two polarities of magnetization (Li 2000). Although the Walsh Tillite is variously correlated with either the Sturtian or Marinoan glacial level across Australia (see Corkeron 2007 for a summary), that issue is irrelevant for the reliability of the palaeomagnetic pole and its implications for deglaciation of the Walsh ice age. We accept the correlation by Corkeron & George (2001) of the Egan glaciogenic deposit (central Kimberley) with the Boonall dolomite in eastern Kimberley, and by further correlation using stromatolites, the mid-Ediacaran Julie Formation in central Australia. A palaeomagnetic pole from the latter region (Kirschvink 1978) implies a depositional palaeolatitude of 21 + 88 for the Egan, assuming that the pole is primary and the correlations are correct.

103

Figure 7.3 shows two possible palaeogeographies of Australia during the time of terminal Cryogenian glaciation. The first representation (Fig. 7.3a) retains coherence of cratonic Australia in present-day coordinates, with palaeoequators according to the various palaeomagnetic studies of the Elatina or Nuccaleena formations. The second (Fig. 7.3b) incorporates a new model of Proterozoic Australia, with the northern part of the continent displaced from the western and southern parts by a local Euler rotation angle of 408 (Li & Evans 2011). In this model, the palaeoequator determined from the Walsh Tillite cap carbonate (Li 2000) returns to a position that is much closer to the palaeoequator attributed to the Nuccaleena cap carbonate in South Australia. Southern Australia may also have been glaciated in the midEdiacaran interval. Gostin et al. (2011) describe possible icerafting features in the Bunyeroo Formation and equivalent units, which are demonstrated to have a pre-folding palaeomagnetic remanence that was acquired about 158 from the equator. In addition, Jenkins (2011) proposes a glacial influence on strata of the Billy Springs Formation higher in the stratigraphy. Most tectonic syntheses of the region formerly known as East Gondwanaland now subdivide the cratonic areas of Antarctica into (at least) three fragments that collided during Cambrian time (e.g. Fitzsimons 2000). The Mawson Continent, or Mawsonland, comprises South Australia and adjacent regions of Antarctica in a pre-breakup Gondwanaland fit; those regions extend much of the length of the Transantarctic Mountains, from Victoria Land toward the South Pole. The Goldie Formation (redefined by Myrow et al. 2002) contains diamictites of uncertain depositional setting. Although no estimates of palaeolatitude are available from the Goldie Formation in the Transantarctic Mountains, its age is now tightly constrained by U –Pb on zircon, 668 + 1 Ma, from interleaved mafic (meta)volcanic rocks (Goodge et al. 2002).

Kalahari and environs Neoproterozoic glaciogenic deposits on the Kalahari craton are restricted to its present western margin, where as many as two diamictite levels are preserved in any single succession. In addition, several levels of canyon incision with diamictic infill are present within the Schwarzrand Subgroup of the Nama Group, spanning the Ediacaran –Cambrian boundary (reviewed by Evans 2000). No new palaeomagnetic constraints are available for any of the pre-Schwarzrand glacial levels, and the Schwarzrand palaeolatitude of 38 + 38 (implied by the pole from the 547 + 4 Ma Sinyai metadolerite on the Congo craton; Meert &

Fig. 7.3. (a) Palaeoequators plotted across Australia and Mawsonland, according to various palaeomagnetic studies of the Elatina Formation, or, as indicated, of the Nuccaleena Formation (Nucc). (b) Restoration of Ediacaran tectonic rotation between the northern and western/southern portions of cratonic Australia, following Li & Evans (2011). For star legend, see caption to Figure 7.2 (uncoloured stars representing deposits typically correlated with the Elatina glaciation. Palaeoequators are referenced as follows: EW86, Embleton & Williams (1986); SWE91, Schmidt et al. (1991); SW95, Schmidt & Williams (1995); SCBK, Sohl et al. (1999); R08, Raub (2008); SWM09, Schmidt et al. (2009); Walsh, Li (2000).

104

D. A. D. EVANS & T. D. RAUB

Van der Voo 1996) is assigned only low reliability because, despite excellent age constraints and the consistency of earliest Cambrian Gondwanaland palaeomagnetic poles from a variety of areas within the megacontinent (Mitchell et al. 2010), it remains unclear whether the Schwarzrand succession is truly glaciogenic. Among older glaciogenic strata of the Gariep belt, Macdonald et al. (2010c) have revised correlations to distinguish two separate stratigraphic levels among outcrops previously lumped together as the Numees Formation. In this new correlation, the name ‘Numees’ is retained for an older level associated with an organicrich cap carbonate typical of ’Sturtian’ strata worldwide, whereas a new name, ‘Namaskluft’, is introduced for a younger level associated with a light-coloured peloidal cap carbonate typical of ‘Marinoan’ strata in other regions. Kaufman et al. (2009) review these successions and present an unconventional set of correlations with the global stratigraphic record, but it is not clear how such correlations might change according to the proposed distinction between Numees and Namuskluft glacial horizons. Farther south in the parautochthonous Gariep belt stratigraphy, the Aties Formation as listed by Evans (2000) has been subdivided into several new units, which include two glaciogenic levels recognized by Frimmel (2011): the lower Karoetjes Kop Formation, and the upper Bloupoort Formation. The former is correlated with the Kaigas diamictites in the Gariep belt to the north, whereas the latter is correlated to the Numees diamictites (s.l.), although further work is required to discern whether Bloupoort would correspond to Numees or Namuskluft, as distinguished by Macdonald et al. (2010c). In the allochthonous Marmora terrane of the Gariep Belt, Namibia, diamictites and dropstone-bearing stratigraphic units are identified within the Dernburg Formation. These deposits are interpreted as having been deposited on a seamount in the Adamastor Ocean that lay between the Kalahari and Rio de la Plata cratons prior to their collision (Frimmel & Jiang 2001; Frimmel et al. 2002). No palaeomagnetic data are available for the Marmora terrane, although extensive carbonates and evaporites in the succession suggest low to moderate palaeolatitudes of deposition. Gaucher et al. (2009) included the Marmora terrane within a crustal block they named ‘Arachania’, inferred to have transferred from the Rio de la Plata craton to the Kalahari craton between c. 630 and 550 Ma.

Several recent papers provide additional insight on correlations of diamictite units across the present eastern and northern edges of the Congo – Sa˜o Francisco craton. Melezhik et al. (2006) provide stable isotopic data bearing on regional and global stratigraphic correlations of the Geci metacarbonates in northern Mozambique, which are associated with metadiamictites that may be glaciogenic (Pinna et al. 1993). To the NW, Poidevin (2007) reviews stratigraphic data from the Oubanguide belt in central Africa, whereas Caron et al. (2010) report the discovery of a new diamictite cap-carbonate succession in southern Cameroon. Lack of precise ages on all of these units, as well as a general dearth of late Neoproterozoic palaeomagnetic data from the Congo craton, prohibit assignment of depositional palaeolatitudes. Continuing around the Sa˜o Francisco side of the craton, in Brazil, the various lithostratigraphic names for late Neoproterozoic diamictites are well summarized by Misi et al. (2007) and Sial et al. (2009). As many as two or perhaps three glaciogenic levels are known from each area, with few precise age constraints and thus, many possible correlations across the region (see also Kaufman et al. 2009). A comprehensive summary of all these possible relationships is beyond the scope of the present paper, but Table 7.1 splits the deposits into many distinct geographical and tectonostratigraphic subdivisions, which later can be combined as correlations gain reliability. In the cratonic autochthon, the basal Bambuı´ cap carbonate (Sete Lagoas Formation) has yielded a Pb/Pb isochron age of 740 + 22 Ma (Babinski et al. 2007), and the authors of that study infer a depositional palaeolatitude according to c. 750 Ma palaeomagnetic data from Africa; however, we consider the Sete Lagoas age constraints lax enough to render its palaeolatitude assignment too tentative. Another age constraint of particular note is an astounding Re – Os black shale isochron in the 1100–1000 Ma range (Azmy et al. 2008) for the Lapa Formation (uppermost formation in the diamictite-bearing Vazante Group) that, if correct, would represent the first discovery of a Mesoproterozoic ice age, anywhere in the world. However, Rodrigues et al. (2008, extended abstract) find ,1000 Ma detrital zircons in all Vazante units, in direct contrast to the Re –Os ages. As for palaeomagnetic constraints, the Salitre cap carbonate of the Bebedouro Formation has only yielded a Cambrian remagnetization (Trindade et al. 2004). No reliable palaeomagnetic data are available for the other diamictites of the Sa˜o Francisco craton.

Congo – Sa˜o Francisco West Africa and Hoggar The Katangan succession in east-central Africa contains two prominent diamictite horizons with overlying cap carbonates, named the Grand Conglomerat and Petit Conglomerat, in stratigraphic order. U– Pb SHRIMP ages of 765–735 Ma from volcanic strata, interbedded with diamictites that are correlated with the Grand Conglomerat (Key et al. 2001), overlap with a 748 + 6 Ma date on the syenitic phase of the Mbozi complex from the same craton (Mbede et al. 2004, abstract only). This allows the Grand Conglomerat’s depositional latitude to be estimated indirectly at 10 + 58, using the near-coeval Mbozi palaeomagnetic pole (Meert et al. 1995). This estimate is considered more reliable than the same pole applied for the Chuos Formation in northern Namibia, which might be substantially younger than the c. 750 Ma Mbozi pole. Both of these units are included separately in our new compilation because of the c. 1000 km lateral distance between their exposures, as well as the possibility that they are not precisely coeval. Our new estimate of Chuos depositional palaeolatitude differs slightly from that of Evans (2000) due to a refined location for the centre of Chuos-correlated strata in the northern and central Damara belt. As for the younger portion of these successions, there is still no reliable palaeomagnetic constraint on that unit, despite the recent U –Pb TIMS age constraint of 635.5 + 1.2 Ma on a likely correlative of the Ghaub Formation in the internal sector of the Damara orogen (Hoffmann et al. 2004; but see also Kaufman et al. 2009).

Although Evans (2000), following earlier syntheses across the Taoudeni basin, considered the possibility that the glaciogenic ‘triad’ succession is late Ediacaran to earliest Cambrian in age, a more detailed study of that succession in the Volta basin (Porter et al. 2004) provides compelling chemostratigraphic comparisons with basal Ediacaran cap carbonates elsewhere in the world that are dated at 635 Ma (South China and Namibia, see above). Further, the assigned Early Cambrian age of the glaciogenic Hassanah Diallo Formation in Senegal, based on a small shelly fossil assemblage in dolomite resembling cap carbonate, is now retracted with the interpretation that the fossiliferous sample was collected from a loose boulder fallen from overlying strata (Shields et al. 2007). For these reasons, the indirect estimation of triad palaeolatitudes from an Early Cambrian aggregate Gondwana Land apparent polar wander path (Evans, 2000) now appears inappropriate. As discussed in Evans (2000), previous palaeomagnetic data from the Taoudeni sedimentary succession are of questionable reliability. However, a new result from one site of the ‘triad’ cap carbonate in the Gourma region indicates a magnetic remanence that was acquired prior to Pan-African folding in the area (Boudzoumou et al. 2011). If primary, the result would imply a depositional palaeolatitude of 08 + 78, but because of the small sample size, the result warrants further study before it can be considered as reliable as the other determinations shown in Figure 7.1.

GLOBAL NEOPROTEROZOIC GLACIAL PALAEOLATITUDES

Alleged periglacial features from the base of the Atar Group (Trompette 1994), which were discussed by Evans (2000) with additional references, must now be considered as Mesoproterozoic, given multi-sample, black shale Re –Os isochron ages of c. 1105 Ma (Rooney et al. 2010) from somewhat higher levels in the succession. The ages are reminiscent of those from the Vazante Group and its diamictites, marginal to the Sa˜o Francisco craton (see above), but the Atar Re –Os isochrons have lower mean square of weighted deviates values and thus, are more reliable. The alleged periglacial influence on the basal Atar sediments should be tested by further study. Despite new U –Pb ages from the Anti-Atlas Mountains of Morocco, late Neoproterozoic glaciogenic deposits there remain unconstrained in depositional palaeolatitude. The older Sarhro diamictites in the Sirwa region were deformed and metamorphosed c. 660 Ma (Thomas et al. 2002, 2004). Although Thomas et al. (2004) advocate the traditional correlation of the Sarhro Group with the Tiddiline succession in the Bou Azzer –El Graara inlier, the latter diamictite-bearing succession contains boulder-sized clasts of the Bleı¨da granodiorite now dated at 579.4 + 1.2 Ma (Inglis et al. 2004). Therefore, the new stratigraphic constraints appear to distinguish the two successions in age by c. 100 million years. No reliable palaeomagnetic data exist for the older age interval, but the Tiddiline succession depositional palaeolatitude could be estimated by the immediately overlying, and penecontemporaneous (c. 575 –560 Ma; Thomas et al. 2002; Maloof et al. 2005), Ouarzazate volcanic group. Unfortunately, the existing palaeomagnetic results from those rocks are of near-zero reliability, as summarized by Tohver et al. (2006).

Avalonia and Cadomia The Gaskiers glaciation is now dated by U –Pb TIMS to c. 580 Ma (Bowring et al. 2002, abstract only), and serendipitously, the age of the Marystown volcanic rocks (used by Evans 2000 to estimate Gaskiers depositional palaeolatitudes) has also been revised to c. 585–575 Ma (McNamara et al. 2001). The concordance of these ages increases the likelihood that the Marystown palaeolatitude of 34 þ 8/– 78 applies to the Gaskiers deposit. It is considered here as only moderately reliable as an indicator for global palaeoclimate models, because the Gaskiers unit is geographically limited and of short-lived duration (Bowring et al. 2002), and thus, it may represent only local, alpine glaciation in an activemargin tectonic setting. In the Boston basin, the Roxbury Conglomerate has yielded a magnetization in nonglacial red slates bracketing the glaciogenic Squantum Member, which is consistent with that of unconformably underlying volcanic rocks, and partially supported by a positive conglomerate test on those volcanic rocks (Fang et al. 1986). Further palaeomagnetism and geochronology of this same succession (Thompson et al. 2007) has found that the c. 600– 595 Ma, pre-Squantum, Mattapan volcanic rocks yielded a distinct magnetization that had not been recovered previously, and the earlier characteristic magnetization of Fang et al. (1986) was not reproduced. Thompson et al. (2007) reported a possibly primary component from only one site of the Squantum unit, and did not discuss the discrepancy between their results and those of the earlier study. Because the Squantum Member is younger than 593 Ma (U– Pb on detrital zircons; Thompson & Bowring 2000) but otherwise poorly constrained in age, it is possible that the difference is due to plate motions and rotations of the Avalonian arc, between deposition of the volcanic rocks and the overlying glaciogenic succession. We continue to use the directly obtained result of Fang et al. (1986) in our analysis, implying a depositional palaeolatitude of 55 þ 8/– 78. In the Anglesey region of Wales, an East Avalonian accretionary ‘Mona’ complex includes heterolithic, cobble-supporting, volcaniclastic mudstone (Gwna Group) of interpreted ophiolitic affinity, plausibly representing ice-rafted debris of

105

Gaskiers-correlative age in an open ocean setting between 595 and 550 Ma (Kawai et al. 2008). The authors of the study inferred a moderate depositional palaeolatitude from the presumably adjacent Avalonian arc: the 580– 575 Ma Marystown Volcanics (see above) and the c. 603 Ma Caldecote Volcanics plus related intrusions (Vizan et al. 2003). We suspect that this estimate is unreliable, given the combined uncertainties in glacial influence of the Gwna beds, palaeogeographic affinity of the Mona complex relative to the Avalonian arc, and age of the succession. In the Cadomian terrane, Bohemian massif, marine diamictite lenses associated with inferred, glacially influenced, base level changes are found in the Clanzschwitz and Weesenstein Groups. These lenses contain detrital zircons as young as c. 570 Ma, and therefore, their deposition is also approximately the same age (Linnemann et al. 2007). These formations, in the Saxo-Thuringian Zone, were likely proximal to West Africa during Ediacaran to early Palaeozoic time; however, lack of reliable Ediacaran palaeomagnetic data from either the Cadomian Belt itself or the West African craton prohibit assignment of depositional palaeolatitudes for the Clanzschwitz and Weesenstein units.

Amazonia and environs In the Paraguay belt at the southeastern margin of the Amazon craton, the Puga diamictite is overlain by the Mirassol d’Oeste cap carbonate. A characteristic component of magnetization has been found in the cap dolostone (Trindade et al. 2003), carried by both magnetite and haematite with high rock-magnetic stability (Font et al. 2005) and multiple reversals were observed through the stratigraphic section (Font et al. 2010) exposed in a single rock quarry. If primary, that magnetic remanence would indicate a depositional palaeolatitude of 22 þ 6/– 58 for Puga deglaciation. The direction, however, is coincident with the Mesozoic to recent expected directions at the sampling sites. Although the presence of two magnetic polarities is encouraging, remagnetizations of that nature are well known in the palaeomagnetic literature (e.g. Kent & Dupuis 2003), and the Mirassol d’Oeste characteristic remanence is so far observed at only a single sampling locality with no field stability tests on its age. For this reason, it is considered here to be only moderately reliable. Higher in the stratigraphy, the Serra Azul Formation is now recognized as a second glaciogenic level in the Paraguay belt (summarized by Alvarenga et al. 2009). Age constraints are lax, and its depositional palaeolatitude is unknown. Along coastal Peru, the Arequipa massif contains a diamictite unit, the Chiquerı´o Formation, which has been correlated to glaciogenic deposits in eastern Laurentia (see references in Evans 2000). Recent U –Pb dating of its host terrane, the Arequipa-Antofalla block (AAB), however, has identified more compelling tectonic affinities to the Amazon craton (Chew et al. 2007; Casquet et al. 2010). In either case, without a precise age, the Chiquerı´o diamictite’s depositional palaeolatitude remains unconstrained.

Rio de la Plata craton and environs The Las Ventanas and Playa Hermosa Formations, near the eastern edge of the Rio de la Plata craton in southern Uruguay, contain the most convincingly glaciogenic features in the entire region (Pazos et al. 2003, 2008; Pecoits et al. 2008). Their age constraints are not well defined within the Ediacaran Period. Palaeomagnetic results obtained directly from the Playa Hermosa Formation suggest a tropical palaeolatitude of 13 þ 9/ –88 (Sa´nchez-Bettucci & Rapalini 2002), but we exclude this result from our final analysis because of very limited sampling in that preliminary study. Lower in the stratigraphy, within the same tectonic unit (Nico Pe´rez terrane and adjacent portion of the Dom Feliciano belt), the Zanja del Tigre Formation is reported to contain a thin metadiamictite horizon as part of a succession (Lavalleja Group) that

106

D. A. D. EVANS & T. D. RAUB

metamorphosed at c. 630 Ma (Pazos et al. 2008, fig. 4). A glaciogenic influence is disputed for these metamorphic rocks (Gaucher & Poire´ 2009; Pecoits & Aubet, pers. comm. 2009) and, regardless, depositional palaeolatitudes for this unit are unconstrained. Within the southern Brazilian portion of the Dom Feliciano belt, Eerola (2006) identified isolated cobbles within schists of the Brusque Group, which he correlated with the c. 750 Ma Chuos Formation in Namibia. However, a much older age for the Brusque succession is possible, as it contains no detrital zircons younger than 2000 Ma (Hartmann et al. 2003). A glaciogenic origin for this unit is not yet well demonstrated, but interestingly, Pazos et al. (2008) suggest correlation of the Brusque Group with the Lavalleja Group. This correlation suggests that scattered remnants of a pre-630 Ma ice age could be present throughout the region. However, the name Lavalleja Group has been applied to various metamorphic rocks with protolith ages now found to range from Archaean to Ediacaran (Bossi & Cingolani 2009); thus, all correlations should be treated with caution. Eerola (1995) also described diamictites and dropstone-bearing laminated mudstones in the Camaqua˜ basin, referred to as the Passo da Areia Formation (see also Pazos et al. 2008) or, alternatively, the Picada das Grac¸as Formation (Eerola 2006; Janikian et al. 2008). Any proposed glacial influence for that unit was discounted by the latter study (Janikian et al. 2008), which provided a direct U –Pb SHRIMP zircon age of 580.0 + 3.6 Ma from an interbedded tuff. Janikian et al. (2008) cite a palaeolatitude of 23 þ 10/– 78 from the slightly older Hilario Formation volcanic rocks, but as reviewed by Tohver et al. (2006), that study lacks field tests and is only of moderate reliability. Nonetheless, farther north in the Ribeira belt, Campanha et al. (2008) have determined a maximum age limit of c. 590– 580 Ma for diamictites of the highly deformed Iporanga Formation, which contain exotic clasts and thus, could be of glaciogenic origin. Given the more convincing glacial influence on the Las Ventanas and Playa Hermosa formations of the same general age range on the proximal Rio de la Plata craton, the Iporanga strata should be investigated further to confirm or refute the influence of ice during sedimentation. Trindade & Macouin (2007) refer to glaciation at moderate to high palaeolatitudes in the Rio de la Plata craton, constrained by the ‘La Tinta’ palaeomagnetic pole (see below), and depicted as such in a 620 Ma global reconstruction. That citation probably refers to the Sierra del Volca´n Formation, recognized as bearing glaciogenic features with a possible Sturtian age (Pazos et al. 2003). However, the unit lacks precise geochronological data (Pazos et al. 2008) and in fact could be as young as Ordovician (unpublished detrital zircon ages cited by Poire´ & Gaucher 2009). Regardless, the La Tinta pole is now considered to represent a probable Cambrian overprint (Rapalini & Sa´nchez-Bettucci 2008), leaving these deposits completely unconstrained in depositional palaeolatitude.

Discussion Figure 7.1 summarizes the results of our updated synthesis. Although several of the assessments of glacial palaeolatitudes have changed in detail, owing to new geochronologic or palaeomagnetic data, the general result is essentially the same as that determined by Evans (2000): low and moderate palaeolatitudes are abundant, at all reliability levels, whereas few if any near-polar (.608) glacial deposits have thus far been identified in the Neoproterozoic rock record. From a total of 137 stratigraphic units distinguished in the present analysis, only 22 deposits have depositional palaeolatitude constraints that are at least ‘somewhat’ reliable, and among these, merely 10 are constrained moderately well. Only three deposits (Upper Tindir, Nantuo and Elatina) are considered to have the highest degree of reliability, and all were laid down in low to moderate palaeolatitudes. The lowest of these, Elatina, is supported by numerous field stability tests to

verify the primary age of magnetization; reproducibility by four different palaeomagnetic laboratories through 15 years of study; and several independent, quantitative methods to correct for modest amounts of sedimentary inclination shallowing. Several lines of evidence demonstrate the effects of regionally cold climates near sea level during the Elatina ice age (Williams et al. 2008). The conclusion appears inescapable: Neoproterozoic continental ice sheets extended deep within the tropics. Among the other high- to moderate-reliability data shown in Figure 7.1, the only other results obtained from sedimentary rocks are the Puga, Nantuo and Walsh. Puga and Walsh magnetizations are dominated by carbonate-associated haematite, so these results also may be less prone to inclination bias. The Nantuo result represents siliciclastic rocks, so its remanence could be slightly shallowed in inclination and hence, shallowed in palaeolatitude. However, estimates for the Konnarock, Rapitan, Mechum River, Grand Conglomerat, Upper Tindir and Gaskiers deposits are calculated indirectly from coeval igneous rocks on the same cratons. Those results are not affected by inclination shallowing. Concerns that the Neoproterozoic geomagnetic field was either nonaxial or nondipolar are valid in principle, but many palaeomagnetic results, including those from the ElatinaNuccaleena succession, show typical patterns of a sporadically reversing field with circularly symmetric secular variation about the time-averaged mean (e.g. Raub 2008; Schmidt et al. 2009; Font et al. 2010). It should therefore be emphasized that a nonactualistic geodynamo invoked to produce strictly polar or temperate glacial palaeolatitudes from the available palaeomagnetic database is as shockingly nonuniformitarian to geophysicists as near-equatorial glaciation is to palaeoclimatologists. Similarly, hypotheses invoking rapid palaeolatitude shifts of continents to generate erroneous palaeomagnetic latitudes, due to lax age constraints in some instances, would require such motions at rates beyond what is normally considered reasonable for plate tectonics. Such shifts would need to be consistently biased to produce wholly low- to mid-palaeolatitudes at the expense of any polar results. Evans (2006) showed that Proterozoic evaporites produce subtropical palaeomagnetic latitudes, as expected for a uniformitarian geodynamo, as well as low planetary obliquity. Williams (2008) attempted to discredit the latter conclusion by appealing to uncertainties in geomagnetic polarity, to generate an equatorial mean for the evaporites, as predicted by numerical simulations of highobliquity palaeoclimate. However, that issue was already foreseen and refuted by Evans (2006), who wrote (p. 53) ‘A zero mean could be obtained if half of the palaeolatitudes were considered to represent the opposite hemisphere (geomagnetic polarity being generally unknown in the Precambrian), but even then the total distribution would be significantly bimodal, in contrast with the high-obliquity model predictions.’ Our summary analysis (Fig. 7.1) differs from the similar figure in Evans (2000) in several minor ways; most importantly (i) a slight increase in the estimated depositional palaeolatitude of the Rapitan deposits, based on new geochronology that allows precise application of the higher-reliability Franklin large igneous province pole that is immune to sedimentary inclination shallowing, (ii) a slight increase in the estimated depositional palaeolatitude of the Elatina deposits, based on new studies that quantify such inclination shallowing by several independent methods, (iii) exclusion of the Jbeliat palaeolatitude estimate due to its more conclusively pre-Cambrian age and thus inappropriateness of applying Early Cambrian palaeolatitudes from the Gondwana Land apparent polar wander path, and (iv) inclusion of the somewhat reliable, near-polar palaeolatitude for the Fauquier Formation as determined by palaeomagnetism on the overlying Catoctin basalts and related intrusions. This last entry needs to be confirmed, but if so, it has the potential to strike down a pillar of the Precambrian high-obliquity Earth model (Williams 1993, 2008) – that Precambrian glaciogenic deposits are only known from the palaeo-tropics, rather than the palaeo-poles. In general,

GLOBAL NEOPROTEROZOIC GLACIAL PALAEOLATITUDES

however, our update does not significantly affect the basic conclusion by Evans (2000) of a preponderance of low- to midlatitudes for Neoproterozoic glaciation. Our analysis is the most detailed of its kind in the published literature. Although our conclusions are nearly opposite to those of Eyles & Januszczak (2004), we note that their analysis included less than a fifth of the number of deposits considered herein. Among the deposits accepted by them as having ‘demonstrated’ glacial influence, five have palaeolatitudes of at least somewhat reliable quality ranging between 13 and 378. One of the deposits (Rapitan) has moderate palaeomagnetic reliability indicating tropical palaeolatitudes. Although we did not assess the evidence for or against glacial influence on sedimentation in our present analysis, we anticipate that future efforts in that regard will consider the full extent of units listed in Table 7.1. Trindade & Macouin (2007) also considered only a small fraction of the deposits listed in Table 7.1, although their conclusion is similar to ours. Considering only the subset of deposits discussed by Trindade & Macouin (2007), the most striking differences between their conclusions and ours are as follows: (i) our exclusion of the Scout Mountain palaeolatitude based on an updated age for that unit, which differs from the palaeomagnetically studied rock units elsewhere in Laurentia, (ii) our inclusion of a Gaskiers palaeolatitude applied indirectly from a palaeomagnetic study of precisely coeval igneous rocks in Avalon, (iii) our estimate of mid-latitude deposition for the Nantuo ice age, rather than near-equatorial, based on our selection of the direct Nantuo constraint, rather than the (in our opinion) lower-reliability Doushantuo result, and (iv) our estimate of low-latitude deposition of the Playa Hermosa Formation, preferred over the higher latitudes implied by the now-obsolete La Tinta Formation. The most interesting comparison to earlier palaeogeographic syntheses is between our present analysis and that of Hoffman & Li (2009). The essence of our conclusions is similar: only lowand mid-latitudes have been moderately or highly reliably determined for most Neoproterozoic glacial deposits. The principal exception to this rule stems from Ediacaran deposits such as Loch na Cille, Squantum, Gaskiers and Serra Azul, all of which are estimated at .608 palaeolatitude by Hoffman & Li (2009). Among those units, we only consider Squantum and Gaskiers as having direct palaeomagnetic constraints classified as even somewhat reliable, and the Gaskiers palaeolatitude, by our estimation, is remarkably different at 31 þ 10/– 88. These discrepancies can be attributed to the very distinct approaches between the two studies. We took a bottom-up approach that considered ages and palaeolatitudes unconstrained unless demanded by strictly relevant data, whereas Hoffman & Li (2009) adopted a top-down approach that assumed known glacial ages within a specified global platekinematic model. That model was based on a mostly independent palaeomagnetic database obtained primarily from well-dated igneous rocks with little relationship to glacial units. So, although both studies derive from the palaeomagnetic method, they are quite distinct in the assumptions and selection of data. The principal discordance between our work and that of Hoffman & Li (2009) occurs during the mid-Ediacaran interval, when the global palaeogeographic models are most uncertain. This period is associated with the fastest continental motions in Earth history, which may be attributed, in part, to global tectonic reorganization (Hoffman 1999), rapid true polar wander (Raub et al. 2007b) and an isolated episode of nonuniformitarian geodynamo behaviour (Abrajevitch & Van der Voo 2010). If such an isolated episode occurred, and it is far from proven, then it would appear to be limited to the midEdiacaran interval rather than the late Cryogenian, Elatina or Puga ice ages (with uniformitarian symmetry in two-polarity palaeomagnetic directions), or the older Cryogenian, Upper Tindir/ Rapitan ice age (as judged by the highly stable Franklin palaeomagnetic pole). In summary, the palaeomagnetic case for low-latitude Neoproterozoic glaciation is strong, and Snowball Earth remains an

107

attractive model for evaluating the enigmatic geological record of icehouse and hothouse environmental conditions just prior to the dawn of animal life on our planet. We thank the following individuals for constructive criticisms on drafts of this manuscript: N. Aubet, M. Babinski, C. Gaucher, Z.-X. Li, A. Maloof, P. Pazos, E. Pecoits, S. Pisarevsky, P. Schmidt, R. Van der Voo, and G. Williams. We would particularly like to thank E. Arnaud for her remarkable patience throughout the submission and revision process. Our research on Neoproterozoic ice ages has been supported by the David and Lucile Packard Foundation, Agouron Institute for Geobiology, National Science Foundation and Australian Research Council’s Tectonics Special Research Centre. This represents a contribution of the IUGSand UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Abrajevitch, A. & Van der Voo, R. 2010. Incompatible Ediacaran paleomagnetic directions suggest an equatorial geomagnetic dipole hypothesis. Earth and Planetary Science Letters, 293, 164– 170. Ali, K. A., Stern, R. J., Manton, W. I., Johnson, P. R. & Mukherjee, S. K. 2009. Neoproterozoic diamictite in the Eastern Desert of Egypt and Northern Saudi Arabia: evidence of 750 Ma glaciation in the Arabian– Nubian Shield? International Journal of Earth Sciences, doi: 10.1007/s00531-009-0427-3. Allen, P. A. 2007. The Huqf Supergroup of Oman: basin development and context for Neoproterozoic glaciation. Earth-Science Reviews, 84, 139– 185. Allen, P. A. & Etienne, J. L. 2008. Sedimentary challenge to Snowball Earth. Nature Geoscience, 1, 817– 825. Alsharhan, A. S. & Nairn, A. E. M. 1997. Chapter 3: Infracambrian of the Middle East. In: Alsharhan, A. S. & Nairn, A. E. M. (eds) Sedimentary Basins and Petroleum Geology of the Middle East. Elsevier, Amsterdam, 65 – 86. Alvarenga, C. J. S. de, Boggiani, P. C., Babinski, M., Dardenne, M. A., Figueiredo, M., Santos, R. V. & Dantas, E. L. 2009. The Amazonian palaeocontinent. In: Gaucher, C., Sial, A. N., Halverson, G. P. & Frimmel, H. E. (eds) Neoproterozoic– Cambrian Tectonics, Global Change and Evolution. Elsevier, Amsterdam, Developments in Precambrian Geology, 16, 15 –28. Avigad, D., Stern, R. J., Beyth, M., Miller, N. & McWilliams, M. O. 2007. Detrital zircon U– Pb geochronology of Cryogenian diamictites and Lower Paleozoic sandstone in Ethiopia (Tigrai): age constraints on Neoproterozoic glaciation and crustal evolution of the southern Arabian– Nubian Shield. Precambrian Research, 154, 88– 106. Azmy, K., Kendall, B., Creaser, R. A., Heaman, L. & de Oliveira, T. F. 2008. Global correlation of the Vazante Group, Sa˜o Francisco Basin, Brazil: Re– Os and U– Pb radiometric age constraints. Precambrian Research, 164, 160– 172. Babinski, M., Vieira, L. C. & Trindade, R. I. F. 2007. Direct dating of the Sete Lagoas cap carbonate (Bambuı´ Group, Brazil) and implications for the Neoproterozoic glacial events. Terra Nova, 19, 401– 406. Bailey, C. M., Peters, S. E., Morton, J. & Shotwell, N. L. 2007. The Mechum River Formation, Virginia Blue Ridge: a record of Neoproterozoic and Paleozoic tectonics in southeastern Laurentia. American Journal of Science, 307, 1– 22. Beyth, M., Avigod, D., Wetzel, H.-U., Matthews, A. & Berhe, S. M. 2003. Crustal exhumation and indications for Snowball Earth in the East African Orogen: north Ethiopia and east Eritrea. Precambrian Research, 123, 187–201. Bingen, B., Griffin, W. L., Torsvik, T. H. & Saeed, A. 2005. Timing of Late Neoproterozoic glaciation on Baltica constrained by detrital zircon geochronology in the Hedmark Group, south-east Norway. Terra Nova, 17, 250–258. Bossi, J. & Cingolani, C. 2009. Extension and general evolution of the Rı´o de la Plata Craton. In: Gaucher, C., Sial, A. N., Halverson, G. P. & Frimmel, H. E. (eds) Neoproterozoic –Cambrian Tectonics, Global Change and Evolution. Elsevier, Amsterdam, Developments in Precambrian Geology, 16, 73– 85. Boudzoumou, F., Vandamme, D., Affaton, P. & Gattacceca, J. 2011. Neoproterozoic paleomagnetic poles in the Taoudeni basin (West Africa). Comptes Rendus Geoscience, 343, 284– 294.

108

D. A. D. EVANS & T. D. RAUB

Bowring, S. A., Landing, E., Myrow, P. & Ramezani, J. 2002. Abstract #13045-Geochronological constraints on terminal Neoproterozoic events and the rise of metazoans. Astrobiology, 2, 457– 458. Bowring, S. A., Grotzinger, J. P., Condon, D. J., Ramezani, J., Newall, M. J. & Allen, P. A. 2007. Geochronologic constraints on the chronostratigraphic framework of the Neoproterozoic Huqf Supergroup, Sultanate of Oman. American Journal of Science, 307, 1097– 1145. Calver, C. R., Black, L. P., Everard, J. L. & Seymour, D. B. 2004. U –Pb zircon age constraints on late Neoproterozoic glaciation in Tasmania. Geology, 32, 893–896. Campanha, G. A. C., Basei, M. S., Tassinari, C. C. G., Nutman, A. P. & Faleiros, F. M. 2008. Constraining the age of the Iporanga Formation with SHRIMP U– Pb zircon: implications for possible Ediacaran glaciation in the Ribeira Belt, SE Brazil. Gondwana Research, 13, 117– 125. Caron, V., Ekomane, E., Mahieux, G., Moussango, P. & Ndjeng, E. 2010. The Mintom Formation (new): sedimentology and geochemistry of a Neoproterozoic, Paralic succession in south-east Cameroon. Journal of African Earth Sciences, 57, 367– 385. Casquet, C., Fanning, C. M., Galindo, C., Pankhurst, R. J., Rapela, C. W. & Torres, P. 2010. The Arequipa Massif of Peru: new SHRIMP and isotope constraints on a Paleoproterozoic inlier in the Grenvillian orogen. Journal of South American Earth Sciences, 29, 128– 142. Chakraborty, P. P., Dey, S. & Mohanty, S. P. 2010. Proterozoic platform sequences of Peninsular India: implications towards basin evolution and supercontinent assembly. Journal of Asian Earth Sciences, 39, 589– 607. Chew, D., Kirkland, C., Schaltegger, U. & Goodhue, R. 2007. Neoproterozoic glaciation in the Proto-Andes: tectonic implications and global correlation. Geology, 35, 1095–1098. Chumakov, N. M. 2009. The Baykonurian glaciohorizon of the Late Vendian. Stratigraphy and Geological Correlation, 17, 373– 381. Cocks, L. R. M. & Torsvik, T. H. 2007. Siberia, the wandering northern terrane, and its changing geography through the Palaeozoic. EarthScience Reviews, 82, 29 –74. Condon, D. J. & Prave, A. R. 2000. Two from Donegal: Neoproterozoic glacial episodes on the northeast margin of Laurentia. Geology, 28, 951– 954. Condon, D., Zhu, M., Bowring, S., Wang, W., Yang, A. & Jin, Y. 2005. U –Pb ages from the Neoproterozoic Doushantuo Formation, China. Science, 308, 95 –98. Corkeron, M. 2007. ‘Cap carbonates’ and Neoproterozoic glacigenic successions from the Kimberley region, north-west Australia. Sedimentology, 54, 871– 903. Corkeron, M. L. & George, A. D. 2001. Glacial incursion on a Neoproterozoic carbonate platform in the Kimberley region, Australia. Geological Society of America Bulletin, 113, 1121– 1132. Corsetti, F. A. & Kaufman, A. J. 2003. Stratigraphic investigations of carbon isotope anomalies and Neoproterozoic ice ages in Death Valley, California. Geological Society of America Bulletin, 115, 916– 932. Corsetti, F. A. & Kaufman, A. J. 2005. The relationship between the Neoproterozoic Noonday Dolomite and the Ibex Formation: new observations and their bearing on ‘Snowball Earth’. Earth-Science Reviews, 73, 63– 78. Corsetti, F. A. & Lorentz, N. J. 2006. On Neoproterozoic cap carbonates as chronostratigraphic markers. In: Xiao, S. & Kaufman, A. J. (eds) Neoproterozoic Geobiology and Paleobiology. Springer, New York, 273– 294. Corsetti, F. A., Stewart, J. H. & Hagadorn, J. W. 2007. Neoproterozoic diamictite-cap carbonate succession and d13C chemostratigraphy from eastern Sonora, Mexico. Chemical Geology, 237, 129– 142. Dempster, T. J., Rogers, G. et al. 2002. Timing of deposition, orogenesis and glaciation within the Dalradian rocks of Scotland: constraints from U– Pb zircon ages. Journal of the Geological Society of London, 159, 83 – 94. Denyszyn, S. W., Halls, H. C., Davis, D. W. & Evans, D. A. D. 2009. Paleomagnetism and U– Pb geochronology of Franklin dykes in High Arctic Canada and Greenland: a revised age and paleomagnetic pole

constraining block rotations in the Nares Strait region. Canadian Journal of Earth Sciences, 46, 689–705. Eerola, T. 2006. Myo¨ha¨isproterotsooiset ilmastonmuutokset Tutkimuksia Etela¨-Brasiliassa [Neoproterozoic climate changes; research on southern Brazil; Finnish with English abstract and bilingual figure captions]. Geologi, 58, 164– 174. Embleton, B. J. J. & Williams, G. E. 1986. Low palaeolatitude of deposition for late Precambrian periglacial varvites in South Australia: implications for palaeoclimatology. Earth and Planetary Science Letters, 79, 419– 430. Evans, D. A. D. 2000. Stratigraphic, geochronological, and paleomagnetic constraints upon the Neoproterozoic climatic paradox. American Journal of Science, 300, 347– 433. Evans, D. A. D. 2003. A fundamental Precambrian – Phanerozoic shift in Earth’s glacial style? Tectonophysics, 375, 353– 385. Evans, D. A. D. 2006. Proterozoic low orbital obliquity and axial-dipolar geomagnetic field from evaporite palaeolatitudes. Nature, 444, 51 – 55. Evans, D. A. D., Li, Z. X., Kirschvink, J. L. & Wingate, M. T. D. 2000. A high-quality mid-Neoproterozoic paleomagnetic pole from the South China block, with implications for ice ages and the breakup configuration of Rodinia. Precambrian Research, 100, 313– 335. Eyles, N. & Januszczak, N. 2004. ‘Zipper-rift’: a tectonic model for Neoproterozoic glaciations during the breakup of Rodinia after 750 Ma. Earth-Science Reviews, 65, 1– 73. Fairchild, I. J. & Kennedy, M. J. 2007. Neoproterozoic glaciation in the Earth System. Journal of the Geological Society of London, 164, 895–921. Fang, W., Van der Voo, R. & Johnson, R. J. E. 1986. Eocambrian paleomagnetism of the Boston basin: evidence for displaced terrane. Geophysical Research Letters, 13, 1450–1453. Fanning, C. M. & Link, P. K. 2004. U –Pb SHRIMP ages of Neoproterozoic (Sturtian) glaciogenic Pocatello Formation, southeastern Idaho. Geology, 32, 881–884. Fanning, C. M. & Link, P. K. 2008. Age constraints for the Sturtian Glaciation; data from the Adelaide Geosyncline, South Australia and Pocatello Formation, Idaho, USA. Geological Society of Australia Abstracts, 91, 57 – 62. Fisher, N. I., Lewis, T. & Embleton, B. J. J. 1987. Statistical Analysis of Spherical Data. Cambridge, Cambridge University Press. Fitzsimons, I. C. W. 2000. Grenville-age basement provinces in East Antarctica: evidence for three separate collisional orogens. Geology, 28, 879– 882. Font, E., Trindade, R. I. F. & Ne´de´lec, A. 2005. Detrital remanent magnetization in haematite-bearing Neoproterozoic Puga cap dolostone, Amazon craton: a rock magnetic and SEM study. Geophysical Journal International, 163, 491– 500. Font, E., Ne´de´lec, A., Trindade, R. I. F. & Moreau, C. 2010. Fast or slow melting of the Marinoan Snowball Earth? The cap dolostone record. Palaeogeography, Palaeoclimatology, Palaeoecology, 295, 215– 225. Frimmel, H. E. 2011. The Karoetjes Kop and Bloupoort Formations, Gifberg Group, South Africa, In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 233 –237. Frimmel, H. E. & Jiang, S.-Y. 2001. Marine evaporites from an oceanic island in the Neoproterozoic Adamastor ocean. Precambrian Research, 105, 57– 71. Frimmel, H. E., Fo¨lling, P. G. & Eriksson, P. G. 2002. Neoproterozoic tectonic and climatic evolution recorded in the Gariep Belt, Namibia and South Africa. Basin Research, 14, 55 –67. Gallet, Y., Pavlov, V. & Courtillot, V. 2003. Magnetic reversal frequency and apparent polar wander of the Siberian platform in the earliest Palaeozoic, inferred from the Khorbusuonka river section (northeastern Siberia). Geophysical Journal International, 154, 829– 840. Gaucher, C. & Poire´, D. G. 2009. Palaeoclimatic events. Neoproterozoic– Cambrian evolution of the Rı´o de la Plata Palaeocontinent. In: Gaucher, C., Sial, A. N., Halverson, G. P. & Frimmel, H. E. (eds) Neoproterozoic– Cambrian Tectonics, Global Change and Evolution. Elsevier, Amsterdam, Developments in Precambrian Geology, 16, 123–130.

GLOBAL NEOPROTEROZOIC GLACIAL PALAEOLATITUDES

Gaucher, C., Frimmel, H. E. & Germs, G. J. B. 2009. Tectonic events and palaeogeographic evolution of southwestern Gondwana in the Neoproterozoic and Cambrian. In: Gaucher, C., Sial, A. N., Halverson, G. P. & Frimmel, H. E. (eds) Neoproterozoic–Cambrian Tectonics, Global Change and Evolution. Elsevier, Amsterdam, Developments in Precambrian Geology, 16, 295–316. Gladkochub, D. P., Wingate, M. T. D., Pisarevsky, S. A., Donskaya, T. V., Mazukabzov, A. M., Ponomarchuk, V. A. & Stanevich, A. M. 2006. Mafic intrusions in southwestern Siberia and implications for a Neoproterozoic connection with Laurentia. Precambrian Research, 147, 260– 278. Goodge, J. W., Myrow, P., Williams, I. S. & Bowring, S. A. 2002. Age and provenance of the Beardmore Group, Antarctica: constraints on Rodinia supercontinent breakup. Journal of Geology, 110, 393– 406. Gostin, V. A., McKirdy, D. M., Webster, L. J. & Williams, G. E. 2011. Mid-Ediacaran ice-rafting in the Adelaide Geosyncline and Officer Basin, South Australia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 673– 676. Grey, K., Hocking, R. M. et al. 2005. Lithostratigraphic nomenclature of the Officer Basin and correlative parts of the Paterson Orogen, Western Australia. Western Australia Geological Survey, Report, 93, 89. Haines, P. W., Hocking, R. M., Grey, K. & Stevens, M. K. 2008. Vines 1 revisited: are older Neoproterozoic glacial deposits preserved in Western Australia? Australian Journal of Earth Sciences, 55, 397– 406. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Toward a Neoproterozoic composite carbon-isotope record. Geological Society of America Bulletin, 117, 1181–1207. ¨ ., Maloof, A. C. & Bowring, S. A. Halverson, G. P., Duda´s, F. O 2007. Evolution of the 87Sr/86Sr composition of Neoproterozoic seawater. Palaeogeography, Palaeoclimatology, Palaeoecology, 256, 103– 129. Hambrey, M. J. & Harland, W. B. (eds) 1981. Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 1004. Hamdi, B. 1992. Late Precambrian glacial deposits in central Iran. 29th International Geological Congress, Abstracts, 2, 232. Hannah, J., Yang, G., Bingen, B., Stein, H. & Zimmerman, A. 2007. 560 and 300 Ma Re– Os ages constrain Neoproterozoic glaciation and record Variscan hydrocarbon migration on extension of Oslo rift. In: Redfield, T., Buiter, S. J. H. & Smethurst, M. A. (eds) geodynamics, Geomagnetism and Paleogeography: A 50 Year Celebration. NGU Report 2007.057 (Trondheim: Geological Survey of Norway), 50. Hartmann, L. A., Bitencourt, M. F., Santos, J. O. S., McNaughton, N. J., Rivera, C. B. & Betiollo, L. 2003. Prolonged Paleoproterozoic magmatic participation in the Neoproterozoic Dom Feliciano belt, Santa Catarina, Brazil, based on zircon U– Pb SHRIMP geochronology. Journal of South American Earth Sciences, 16, 477–492. Hebert, C. L., Kaufman, A. J., Penniston-Dorland, S. C. & Martin, A. J. 2010. Radiometric and stratigraphic constraints on terminal Ediacaran (post-Gaskiers) glaciation and metazoan evolution. Precambrian Research, 182, 402–412. Hodych, J. P. & Cox, R. A. 2007. Ediacaran U– Pb zircon dates for the Lac Matape´dia and Mt. St.-Anselme basalts of the Quebec Appalachians: support for a long-lived mantle plume during the rifting phase of Iapetus opening. Canadian Journal of Earth Sciences, 44, 565– 581. Hoffman, P. F. 1999. The break-up of Rodinia, birth of Gondwana, true polar wander and the Snowball Earth. Journal of African Earth Sciences, 28, 17– 33. Hoffman, P. F. 2009. Pan-glacial — a third state in the climate system. Geology Today, 25, 107–114. Hoffman, P. F. & Schrag, D. P. 2002. The Snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129– 155. Hoffman, P. F. & Li, Z.-X. 2009. A palaeogeographic context for Neoproterozoic glaciation. Palaeogeography, Palaeoclimatology, Palaeoecology, 277, 158– 172. Hoffman, P. F., Halverson, G. P., Domack, E. W., Husson, J. M., Higgins, J. A. & Schrag, D. P. 2007. Are basal Ediacaran (635

109

Ma) post-glacial ‘cap dolostones’ diachronous? Earth and Planetary Science Letters, 258, 114–131. Hoffmann, K.-H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. U– Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817– 820. Huang, B., Xu, B., Zhang, C., Li, Y. & Zhu, R. 2005. Paleomagnetism of the Baiyisi volcanic rocks (ca. 740 Ma) of Tarim, Northwest China: a continental fragment of Neoproterozoic Western Australia? Precambrian Research, 142, 83 –92. Huckriede, R., Ku¨rsten, M. & Venzlaff, H. 1962. Zur Geologie des Gebiets zwischen Kerman und Saghand (Iran). Beihefte zum Geologischen Jahrbuch, 51, 197. Hyodo, H. & Dunlop, D. J. 1993. Effect of anisotropy on the paleomagnetic contact test for a Grenville dike. Journal of Geophysical Research, 98, 7997–8017. Iglesia-Llanos, M. P., Tait, J. A., Popov, V. & Abalmassova, A. 2005. Palaeomagnetic data from Ediacaran (Vendian) sediments of the Arkhangelsk region, NW Russia: an alternative apparent polar wander path of Baltica for the Late Proterozoic – Early Palaeozoic. Earth and Planetary Science Letters, 240, 732– 747. Inglis, J. D., MacLean, J. S., Samson, S. D., D’Lemos, R. S., Admou, H. & Hefferan, K. 2004. A precise U –Pb zircon age for the Bleı¨da granodiorite, Anti-Atlas, Morocco: implications for the timing of deformation and terrane assembly in the eastern Anti-Atlas. Journal of African Earth Sciences, 39, 277–283. Janikian, L., Almeida, R. P. et al. 2008. The continental record of Ediacaran volcano-sedimentary successions in southern Brazil and their global implications. Terra Nova, 20, 259– 266. Jenkins, R. J. F. 2011. Billy Springs glaciation, South Australia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36 693– 699. Kamo, S. L. & Gower, C. F. 1994. U –Pb baddeleyite dating clarifies age of characteristic paleomagnetic remanence of Long Range dykes, southeastern Labrador. Atlantic Geology, 30, 259–262. Kamo, S. L., Gower, C. F. & Krogh, T. E. 1989. Birthdate for the Iapetus Ocean? A precise U –Pb zircon and baddeleyite age for the Long Range dikes, southeast Labrador. Geology, 17, 602–605. Kamo, S. L., Krogh, T. E. & Kumarapeli, P. S. 1995. Age of the Grenville dyke swarm, Ontario-Quebec: implications for the timing of Iapetus rifting. Canadian Journal of Earth Sciences, 32, 273– 280. Karlstrom, K. E., Bowring, S. A. et al. 2000. Chuar Group of the Grand Canyon: record of breakup of Rodinia, associated change in the global carbon cycle, and ecosystem expansion by 740 Ma. Geology, 28, 619– 622. Kaufman, A. J., Sial, A. N., Frimmel, H. E. & Misi, A. 2009. Neoproterozoic to Cambrian palaeoclimatic events in southwestern Gondwana. In: Gaucher, C., Sial, A. N., Halverson, G. P. & Frimmel, H. E. (eds) Neoproterozoic –Cambrian Tectonics, Global Change and Evolution. Elsevier, Amsterdam, Developments in Precambrian Geology, 16, 369– 388. Kawai, T., Windley, B. F., Terabayashi, M., Yamamoto, H., Isozaki, Y. & Maruyama, S. 2008. Neoproterozoic glaciations in the midoceanic realm: an example from hemi-pelagic mudstones on Llanddwyn Island, Anglesey, U.K. Gondwana Research, 14, 105–114. Kempf, O., Kellerhals, P., Lowrie, W. & Matter, A. 2000. Paleomagnetic directions in Late Precambrian glaciomarine sediments of the Mirbat sandstone formation, Oman. Earth and Planetary Science Letters, 175, 181– 190. Kendall, B., Creaser, R. A., Calver, C. R., Raub, T. D. & Evans, D. A. D. 2009. Correlation of Sturtian diamictite successions in southern Australia and northwestern Tasmania by Re– Os black shale geochronology and the ambiguity of ‘Sturtian’-type diamictite– cap carbonate pairs as chronostratigraphic marker horizons. Precambrian Research, 172, 301– 310. Kent, D. V. & Dupuis, C. 2003. Paleomagnetic study of the Paleocene– Eocene Tarawan chalk and Esna shale: dual polarity remagnetizations of Cenozoic sediments in the Nile valley (Egypt). Micropaleontology, 49, 139– 146. Key, R. M., Liyungu, A. K., Njamu, F. M., Somwe, V. & Banda, J. 2001. The western arm of the Lufilian Arc in NW Zambia and its

110

D. A. D. EVANS & T. D. RAUB

potential for copper mineralization. Journal of African Earth Sciences, 33, 503– 528. Khomentovsky, V. V. 1986. The Vendian System of Siberia and a standard stratigraphic scale. Geological Magazine, 123, 333– 348. Kilner, B., Mac Niocaill, C. & Brasier, M. 2005. Low-latitude glaciation in the Neoproterozoic of Oman. Geology, 33, 413– 416. Kirschvink, J. L. 1978. The Precambrian-Cambrian boundary problem: paleomagnetic directions from the Amadeus Basin, Central Australia. Earth and Planetary Science Letters, 40, 91– 100. Kirschvink, J. L. 1992. Late Proterozoic low-latitude global glaciation: the Snowball Earth. In: Schopf, J. W. & Klein, C. C. (eds) The Proterozoic Biosphere: A Multidisciplinary Study. Cambridge University Press, Cambridge, 51 –52. Kirschvink, J. L. & Rozanov, A. Yu. 1984. Magnetostratigraphy of Lower Cambrian strata from the Siberian platform: a paleomagnetic pole and a preliminary polarity time-scale. Geological Magazine, 121, 189– 203. Kravchinsky, V. A., Konstantinov, K. M. & Cogne´, J.-P. 2001. Palaeomagnetic study of Vendian and Early Cambrian rocks of South Siberia and Central Mongolia: was the Siberian platform assembled at this time? Precambrian Research, 110, 61– 92. Kravchinsky, V. A., Sklyarov, E. V., Gladkochub, D. P. & Harbert, W. P. 2010. Paleomagnetism of the Precambrian Eastern Sayan rocks: implications for the Ediacaran –Early Cambrian paleogeography of the Tuva-Mongolian composite terrane. Tectonophysics, 486, 65 – 80. Levashova, N. M., Kalugin, V. M., Gibsher, A. S., Yff, J., Ryabinin, A. B., Meert, J. G. & Malone, S. J. 2010. The origin of the Baydaric microcontinent, Mongolia: constraints from paleomagnetism and geochronology. Tectonophysics, 485, 306–320. Li, Y., Li, Y., Sharps, R., McWilliams, M. & Gao, Z. 1991. Sinian paleomagnetic results from the Tarim block, western China. Precambrian Research, 49, 61 –71. Li, Z.-X. 2000. New palaeomagnetic results from the ‘cap dolomite’ of the Neoproterozoic Walsh Tillite, northwestern Australia. Precambrian Research, 100, 359– 370. Li, Z.-X. & Evans, D. A. D. 2011. Late Neoproterozoic 408 intraplate rotation within Australia allows for a tighter-fitting and longer-lasting Rodinia. Geology, 39, 39– 42. Linnemann, U., Gerdes, A., Drost, K. & Buschmann, B. 2007. The continuum between Cadomian orogenesis and opening of the Rheic Ocean: constraints from LA– ICP– MS U–Pb zircon dating and analysis of plate-tectonic setting (Saxo-Thuringian zone, northeastern Bohemian Massif, Germany. In: Linnemann, U., Nance, R. D., Kraft, P. & Zulauf, G. (eds) The Evolution of the Rheic Ocean: From Avalonian –Cadomian Active Margin to Alleghenian– Variscan Collision. Geological Society of America, Special Paper, 423, 61 –96. Lund, K., Aleinikoff, J. N. & Evans, K. V. 2003. SHRIMP U –Pb geochronology of Neoproterozoic Windermere Supergroup, central Idaho: implications for rifting of western Laurentia and synchroneity of Sturtian glacial deposits. Geological Society of America Bulletin, 115, 349– 372. Macouin, M., Besse, J., Ader, M., Gilder, S., Yang, Z., Sun, Z. & Agrinier, P. 2004. Combined paleomagnetic and isotopic data from the Doushantuo carbonates, South China: implications for the “Snowball Earth” hypothesis. Earth and Planetary Science Letters, 224, 387– 398. Macdonald, F. A., McClelland, W. C., Schrag, D. P. & Macdonald, W. P. 2009a. Neoproterozoic glaciations on a carbonate platform margin in Arctic Alaska and the origin of the North Slope subterrane. Geological Society of America Bulletin, 121, 448–473. Macdonald, F. A., Jones, D. S. & Schrag, D. P. 2009b. Stratigraphic and tectonic implications of a newly discovered glacial diamictite– cap carbonate couplet in southwestern Mongolia. Geology, 37, 123–126. Macdonald, F. A., Schmitz, M. D. et al. 2010a. Calibrating the Cryogenian. Science, 327, 1241–1243. ¨ . & Schrag, D. P. 2010b. Macdonald, F. A., Cohen, P. A., Duda´s, F. O Early Neoproterozoic scale microfossils in the Lower Tindir Group of Alaska and the Yukon Territory. Geology, 38, 143–146. ˜ . & Schrag, Macdonald, F. A., Strauss, J. V., Rose, C. V., Duda´s, F. O D. P. 2010c. Stratigraphy of the Port Nolloth Group of Namibia and

South Africa and implications for the age of Neoproterozoic iron formations. American Journal of Science, 310, 862– 888. Mac Niocaill, C., Kilner, B., Stouge, S., Knudsen, M. F., Harper, D. A. T. & Christiansen, J. L. 2008. The Neoproterozoic drift history of Laurentia: a critical evaluation and new palaeomagnetic data from Northern and Eastern Greenland. EOS Transactions of the American Geophysical Union, 89, S514–05. Maloof, A. C., Schrag, D. P., Crowley, J. L. & Bowring, S. A. 2005. An expanded record of Early Cambrian carbon cycling from the AntiAtlas Margin, Morocco. Canadian Journal of Earth Sciences, 42, 2195– 2216. Martin, M. W., Grazhdankin, D. V., Bowring, S. A., Evans, D. A. D., Fedonkin, M. A. & Kirschvink, J. L. 2000. Age of Neoproterozoic bilaterian body and trace fossils, White Sea, Russia: implications for metazoan evolution. Science, 288, 841– 845. Mbede, E. I., Kampunzu, A. B. & Armstrong, R. A. 2004. Neoproterozoic inheritance during Cainozoic rifting in the western and southwestern branches of the East African rift system: evidence from carbonatite and alkaline intrusions. Conference abstract, The East African Rift System: Development, Evolution and Resources, Addis Ababa, Ethiopia, June 20– 24, 2004. McCausland, P. J. A., Van der Voo, R. & Hall, C. M. 2007. Circum-Iapetus paleogeography of the Precambrian – Cambrian transition with a new paleomagnetic constraint from Laurentia. Precambrian Research, 156, 125–152. McCausland, P. J. A., Smirnov, A., Evans, D. A. D., Izard, C. & Raub, T. 2009. Low-latitude Laurentia at 615 Ma: Paleomagnetism of the Long Range Dykes and Coeval Lighthouse Cove Formation, Northern Newfoundland and SE Labrador. Abstracts, AGU-GAC/MAC-CGU Joint Assembly, Toronto, May 2009. McCausland, P. J. A., Hankard, F., Van der Voo, R. & Hall, C. M. 2011. Ediacaran paleogeography of Laurentia: paleomagnetism and 40 Ar-39Ar geochronology of the 583 Ma Baie des Moutons syenite, Quebec. Precambrian Research, 187, 58– 78. McCay, G. A., Prave, A. R., Alsop, G. I. & Fallick, A. E. 2006. Glacial trinity: Neoproterzoic Earth history within the British-Irish Caledonides. Geology, 34, 909–912. McFadden, P. L. & Lowes, F. J. 1981. The discrimination of mean directions drawn from Fisher distributions. Geophysical Journal of the Royal Astronomical Society, 67, 19– 33. McNamara, A. K., Mac Niocaill, C., Van der Pluijm, B. A. & Van der Voo, R. 2001. West African proximity of the Avalon terrane in the latest Precambrian. Geological Society of America Bulletin, 113, 1161– 1170. McWilliams, M. & Schmidt, P. W. 2003. Paleomagnetism of Grassy Group rocks, King Island, Tasmania. European Geophysical Society, Geophysical Research Abstracts, 5, no. 12406. Meert, J. G. & Van der Voo, R. 1996. Paleomagnetic and 40Ar/39Ar study of the Sinyai Dolerite, Kenya: implications for Gondwana assembly. Journal of Geology, 104, 131– 142. Meert, J. G. & Van der Voo, R. 2001. Comment on ‘New palaeomagnetic result from Vendian red sediments in Cisbaikalia and the problem of the relationship of Siberia and Laurentia in the Vendian’ by S. A. Pisarevsky, R. A. Komissarova and A. N. Khramov. Geophysical Journal International, 146, 867– 870. Meert, J. G., Van der Voo, R. & Payne, T. W. 1994. Paleomagnetism of the Catoctin volcanic province: a new Vendian –Cambrian apparent polar wander path for North America. Journal of Geophysical Research, 99, 4625– 4641. Meert, J. G., Van der Voo, R. & Ayub, S. 1995. Paleomagnetic investigation of the Neoproterozoic Gagwe lavas and Mbozi complex, Tanzania and the assembly of Gondwana. Precambrian Research, 74, 225– 244. Meert, J. G., Walderhaug, H. J., Torsvik, T. H. & Hendriks, B. W. H. 2007. Age and paleomagnetic signature of the Alnø carbonatite complex (NE Sweden): additional controversy for the Neoproterozoic paleoposition of Baltica. Precambrian Research, 154, 159– 174. Meffre, S., Direen, N. G., Crawford, A. J. & Kamenetsky, V. 2004. Mafic volcanic rocks on King Island, Tasmania: evidence for 579 Ma break-up in east Gondwana. Precambrian Research, 135, 177– 191.

GLOBAL NEOPROTEROZOIC GLACIAL PALAEOLATITUDES

Melezhik, V. A., Kuznetsov, A. B., Fallick, A. F., Smith, R. A., Gorokhov, I. M., Jamal, D. & Catuane, F. 2006. Depositional environments and an apparent age for the Geci meta-limestones: constraints on the geological history of northern Mozambique. Precambrian Research, 148, 19– 31. Metelkin, D. V., Belonosov, I. V., Gladkochub, D. P., Donskaya, T. V., Mazukabzov, A. M. & Stanevich, A. M. 2005. Paleomagnetic directions from Nersa intrusions of the Biryusa terrane, Siberian Craton, as a reflection of tectonic events in the Neoproterozoic. Russian Geology and Geophysics, 46, 398– 413. Miller, N. R., Alene, M., Sacchi, R., Stern, R. J., Conti, A., Kro¨ner, A. & Zuppi, G. 2003. Significance of the Tambien Group (Tigrai, N. Ethiopia) for Snowball Earth events in the Arabian– Nubian Shield. Precambrian Research, 121, 263– 283. Misi, A., Kaufman, A. J. et al. 2007. Chemostratigraphic correlation of Neoproterozoic successions in South America. Chemical Geology, 237, 143– 167. Mitchell, R. N., Evans, D. A. D. & Kilian, T. M. 2010. Rapid Early Cambrian rotation of Gondwana. Geology, 38, 755–758. Mukhopadhyay, J. & Chaudhuri, A. K. 2003. Shallow to deep-water deposition in a Cratonic basin: an example from the Proterozoic Penganga Group, Pranhita– Godavari Valley, India. Journal of Asian Earth Sciences, 21, 613–622. Murthy, G. S. 1971. The paleomagnetism of diabase dikes from the Grenville Province. Canadian Journal of Earth Sciences, 8, 802– 812. Murthy, G., Gower, C., Tubrett, M. & Pa¨tzold, R. 1992. Paleomagnetism of Eocambrian Long Range dykes and Double Mer Formaation from Labrador, Canada. Canadian Journal of Earth Sciences, 29, 1224– 1234. Myrow, P. M., Pope, M. C., Goodge, J. W., Fischer, W. & Palmer, A. R. 2002. Depositional history of pre-Devonian strata and timing of Ross orogenic tectonism in the central Transantarctic Mountains, Antarctica. Geological Society of America Bulletin, 114, 1070– 1088. Nawrocki, J., Boguckij, A. & Katinas, V. 2004. New Late Vendian palaeogeography of Baltica and the TESZ. Geological Quarterly, 48, 309– 316. Pazos, P. J., Bettucci, L. S. & Loureiro, J. 2008. The Neoproterozoic glacial record in the Rı´o de la Plata Craton: a critical reappraisal. In: Pankhurst, R. J., Truow, R. A. J., Brito Neves, B. B. & de Wit, M. J. (eds) West Gondwana: Pre-Cenozoic Correlations Across the South Atlantic Region. Geological Society, London, Special Publication, 294, 343– 364. Pazos, P. J., Sa´nchez-Bettucci, L. & Tofalo, O. R. 2003. The record of the Varanger glaciation at the Rı´o de la Plata craton, VendianCambrian of Uruguay. Gondwana Research, 6, 65 –77. Pecoits, E., Gingras, M., Aubet, N. & Konhauser, K. 2008. Ediacaran in Uruguay: palaeoclimatic and palaeobiological implications. Sedimentology, 55, 689– 719. Pinna, P., Jourde, G., Calvez, J. Y., Mroz, J. P. & Marques, J. M. 1993. The Mozambique Belt in northern Mozambique: Neoproterozoic (1100 –850 Ma) crustal growth and tectogenesis, and superimposed Pan-African (800– 550 Ma) tectonism. Precambrian Research, 62, 1 – 59. Pisarevsky, S. A., Komissarova, R. A. & Khramov, A. N. 2001. Reply to comment by J. G. Meert and R. Van der Voo on ‘New palaeomagnetic result from Vendian red sediments in Cisbaikalia and the problem of the relationship of Siberia and Laurentia in the Vendian’. Geophysical Journal International, 146, 871– 873. Pisarevsky, S. A., Wingate, M. T. D., Stevens, M. K. & Haines, P. W. 2007. Palaeomagnetic results from the Lancer 1 stratigraphic drillhole, Officer Basin, Western Australia, and implications for Rodinia reconstructions. Australian Journal of Earth Sciences, 54, 561– 572. Poidevin, J.-L. 2007. Stratigraphie isotopique du strontium et datation des formations carbonate´es et glacioge´niques ne´oprote´rozoiques du Nord et de l’Ouest du craton du Congo. Comptes Rendus Geoscience, 339, 259– 273. Poire´, D. G. & Gaucher, C. 2009. Lithostratigraphy. Neoproterozoic – Cambrian evolution of the Rı´o de la Plata Palaeocontinent. In: Gaucher, C., Sial, A. N., Halverson, G. P. & Frimmel, H. E. (eds) Neoproterozoic– Cambrian Tectonics, global Change and

111

Evolution. Elsevier, Amsterdam, Developments in Precambrian Geology, 16, 87 –101. Pokrovskii, B. G., Melezhik, V. A. & Bujakaite, M. I. 2006. Carbon, oxygen, strontium, and sulfur isotopic compositions in late Precambrian rocks of the Patom Complex, central Siberia: communication 1. results, isotope stratigraphy, and dating problems. Lithology and Mineral Resources, 41, 450– 474. Popov, V., Iosifidi, A., Khramov, A., Tait, J. & Bachtadse, V. 2002. Paleomagnetism of Upper Vendian sediments from the Winter Coast, White Sea region, Russia: implications for the paleogeography of Baltica during Neoproterozoic times. Journal of Geophysical Research, 107, doi: 10.1029/2001JB001607. Popov, V. V., Khramov, A. N. & Bachtadse, V. 2005. Palaeomagnetism, magnetic stratigraphy, and petromagnetism of the Upper Vendian sedimentary rocks in the sections of the Zolotitsa River and in the Verkhotina Hole, Winter Coast of the White Sea, Russia. Russian Journal of Earth Sciences, 7, 1– 29. Porter, S. M., Knoll, A. H. & Affaton, P. 2004. Chemostratigraphy of Neoproterozoic cap carbonates from the Volta Basin, West Africa. Precambrian Research, 130, 99– 112. Prave, A. R. 1999. The Neoproterozoic Dalradian Supergroup of Scotland: an alternative hypothesis. Geological Magazine, 136, 609– 617. Ramezani, J. & Tucker, R. D. 2003. The Saghand region, central Iran: U–Pb geochronology, petrogenesis and implications for Gondwana tectonics. American Journal of Science, 303, 622–665. Rapalini, A. E. & Sa´nchez-Bettucci, L. 2008. Widespread remagnetization of late Proterozoic sedimentary units of Uruguay and the apparent polar wander path for the Rio de La Plata craton. Geophysical Journal International, 174, 55 –74. Raub, T. D. 2008. Prolonged deglaciation of Snowball Earth. PhD thesis, Yale University. Raub, T. D., Evans, D. A. D. & Smirnov, A. V. 2007a. Siliciclastic prelude to Elatina deglaciation: lithostratigraphy and rock magnetism of the base of the Ediacaran System. In: Vickers-Rich, P. & Komarower, P. (eds) The Rise and Fall of the Ediacaran Biota. Geological Society, London, Special Publication, 286, 53– 76. Raub, T. D., Kirschvink, J. L. & Evans, D. A. D. 2007b. True polar wander: linking deep and shallow geodynamics to hydro- and biospheric hypotheses. In: Kono, M. (ed.) Treatise on Geophysics Volume 5: Geomagnetism. Elsevier, Amsterdam, 565–589. Rieu, R., Allen, P. A., Cozzi, A., Kosler, J. & Bussy, F. 2007. A composite stratigraphy for the Neoproterozoic Huqf Supergroup of Oman: integrating new litho-, chemo- and chronostratigraphic data of the Mirbat area, southern Oman. Journal of the Geological Society of London, 164, 997– 1009. Rieu, R., Allen, P. A., Etienne, J. L., Cozzi, A. & Wiechert, U. 2006. A Neoproterozoic glacially influenced basin margin succession and ‘atypical’ cap carbonate associated with bedrock palaeovalleys, Mirbat area, southern Oman. Basin Research, 18, 471–496. Roberts, D. 2003. The Scandinavian Caledonides: event chronology, palaeogeographic settings and likely modern analogues. Tectonophysics, 365, 283–299. Rodrigues, J. B., Pimentel, M. M., Buhn, B., Dardenne, M. A. & Alvarenga, C. J. S. 2008. Provenance of Vazante Group — Preliminary data. VI South American Symposium on Isotope Geology, San Carlos de Bariloche, Argentina, 4. Rooney, A. D., Selby, D., Houzay, J.-P. & Renne, P. R. 2010. Re– Os geochronology of a Mesoproterozoic sedimentary succession, Taoudeni basin, Mauritania: implications for basin-wide correlations and Re– Os organic-rich sediments systematics. Earth and Planetary Science Letters, 289, 486–496. Rose, C. V. & Maloof, A. C. 2010. Testing models for post-glacial ‘cap dolostone’ deposition: Nuccaleena Formation, South Australia. Earth and Planetary Science Letters, 296, 165–180. Rowan, C. J. & Tait, J. 2010. Oman’s Low Latitude ‘Snowball Earth’ Pole Revisited: Late Cretaceous remagnetisation of Late Neoproterozoic Carbonates in Northern Oman. American Geophysical Union, Fall Meeting Abstract GP33C-0959. Sa´nchez-Bettucci, L. & Rapalini, A. E. 2002. Paleomagnetism of the Sierra de Las Animas Complex, southern Uruguay: its implications in the assembly of western Gondwana. Precambrian Research, 118, 243– 265.

112

D. A. D. EVANS & T. D. RAUB

Schmidt, P. W. & Williams, G. E. 1995. The Neoproterozoic climatic paradox: equatorial palaeolatitude for Marinoan glaciation near sea level in South Australia. Earth and Planetary Science Letters, 134, 107– 124. Schmidt, P. W., Williams, G. E. & Embleton, B. J. J. 1991. Low palaeolatitude of Late Proterozoic glaciation: early timing of remanence in haematite of the Elatina Formation, South Australia. Earth and Planetary Science Letters, 105, 355–367. Schmidt, P. W., Williams, G. E. & McWilliams, M. O. 2009. Palaeomagnetism and magnetic anisotropy of late Neoproterozoic strata, South Australia: implications for the palaeolatitude of late Cryogenian glaciation, cap carbonate and the Ediacaran System. Precambrian Research, 174, 35 – 52. Shen, B., Xiao, S., Zhou, C., Kaufman, A. J. & Yuan, X. 2010. Carbon and sulfur isotope chemostratigraphy of the Neoproterozoic Quanji Group of the Chaidam Basin, NW China: basin stratification in the aftermath of an Ediacaran glaciation postdating the Shuram event? Precambrian Research, 177, 241– 252. Shields, G. A., Deynoux, M., Culver, S. J., Brasier, M. D., Affaton, P. & Vandamme, D. 2007. Neoproterozoic glaciomarine and cap dolostone facies of the southwestern Taoude´ni Basin (Walidiala Valley, Senegal/Guinea, NW Africa). Comptes Rendus Geoscience, 339, 186– 199. Sial, A. N., Dardenne, M. A. et al. 2009. The Sa˜o Francisco palaeocontinent. In: Gaucher, C., Sial, A. N., Halverson, G. P. & Frimmel, H. E. (eds) Neoproterozoic– Cambrian Tectonics, Global Change and Evolution. Elsevier, Amsterdam, Developments in Precambrian Geology, 16, 31 –69. Sohl, L. E., Christie-Blick, N. & Kent, D. V. 1999. Paleomagnetic polarity reversals in Marinoan (ca. 600 Ma) glacial deposits of Australia: implications for the duration of low-latitude glaciation in Neoproterozoic time. Geological Society of America Bulletin, 111, 1120– 1139. Sovetov, J. K. 2002. Vendian foreland basin of the Siberian cratonic margin: Paleopangean accretionary phases. Russian Journal of Earth Sciences, 4, 363– 387. Sovetov, J. 2008. Marinoan Glaciation in the Siberian Craton: Locality, Erosional Forms, Deposits and Constraints to Age. 33rd International Geological Congress, Abstracts, Session CGC-04. Sovetov, Yu. K. & Komlev, D. A. 2005. Tillites at the base of the Oselok Group, foothills of the Sayan Mountains, and the Vendian lower boundary in the southwestern Siberian Platform. Stratigraphy and Geological Correlation, 13, 337– 366. Stern, R. J., Avigad, D., Miller, N. R. & Beyth, M. 2006. Evidence for the Snowball Earth hypothesis in the Arabian – Nubian Shield and the East African Orogen. Journal of African Earth Sciences, 44, 1– 20. Swanson-Hysell, N. L., Maloof, A. C., Weiss, B. P. & & Evans, D. A. D. 2009. No asymmetry in geomagnetic reversals recorded by 1.1-billion-year-old Keweenawan basalts. Nature Geoscience, 2, 713– 717. Tauxe, L., Kodama, K. P. & Kent, D. V. 2008. Testing corrections for paleomagnetic inclination error in sedimentary rocks: a comparative approach. Physics of the Earth and Planetary Interiors, 169, 152– 165. Thomas, R. J., Chevallier, L. C. et al. 2002. Precambrian evolution of the Sirwa Window, Anti-Atlas Orogen, Morocco. Precambrian Research, 118, 1 –57. Thomas, R. J., Fekkak, A. et al. 2004. A new lithostratigraphic framework for the Anti-Atlas Orogen, Morocco. Journal of African Earth Sciences, 39, 217– 226. Thompson, M. D. & Bowring, S. A. 2000. Age of the Squantum ‘Tillite’, Boston Basin, Massachusetts: U– Pb zircon constraints on terminal Neoproterozoic glaciation. American Journal of Science, 300, 630– 655. Thompson, M. D., Grunow, A. M. & Ramezani, J. 2007. Late Neoproterozoic paleogeography of the southeastern New England Avalon zone: insights from U– Pb geochronology and paleomagnetism. Geological Society of America Bulletin, 119, 681–696. Tohver, E., D’Agrella-Filho, M. S. & Trindade, R. I. F. 2006. Paleomagnetic record of Africa and South America for the 1200–500 Ma interval, and evaluation of Rodinia and Gondwana assemblies. Precambrian Research, 147, 193–222.

Torsvik, T. H., Lohmann, K. C. & Sturt, B. A. 1995. Vendian glaciations and their relation to the dispersal of Rodinia: Paleomagnetic constraints. Geology, 23, 727–730. Trindade, R. I. F. & Macouin, M. 2007. Palaeolatitude of glacial deposits and palaeogeography of Neoproterozoic ice ages. Comptes Rendus Geoscience, 339, 200–211. Trindade, R. I. F., D’Agrella-Filho, M. S., Babinski, M., Font, E. & Brito Neves, B. B. 2004. Paleomagnetism and geochronology of the Bebedouro cap carbonate: evidence for continental-scale Cambrian remagnetization in the Sa˜o Francisco craton, Brazil. Precambrian Research, 128, 83– 103. Trindade, R. I. F., Font, E., D’Agrella-Filho, M. S., Nogueira, A. C. R. & Riccomini, C. 2003. Low-latitude and multiple geomagnetic reversals in the Neoproterozoic Puga cap carbonate, Amazon craton. Terra Nova, 15, 441–446. Trompette, R. 1994. Geology of Western Gondwana (2000– 500 Ma): Pan-African –Brasiliano Aggregation of South America and Africa. Balkema, Rotterdam. Van der Voo, R. 1990. The reliability of paleomagnetic data. Tectonophysics, 184, 1– 9. Vizan, H., Carney, J. N., Turner, P., Ixer, R. A., Tomasso, M., Mullen, R. P. & Clarke, P. 2003. Late Neoproterozoic to Early Palaeozoic palaeogeography of Avalonia: some palaeomagnetic constraints from Nuneaton, central England. Geological Magazine, 140, 685– 705. Weil, A. B., Geissman, J. W. & Van der Voo, R. 2004. Paleomagnetism of the Neoproterozoic Chuar Group, Grand Canyon Supergroup, Arizona: implications for Laurentia’s Neoproterozoic APWP and Rodinia break-up. Precambrian Research, 129, 71 –92. Williams, G. E. 1993. History of the Earth’s obliquity. Earth-Science Reviews, 34, 1– 45. Williams, G. E. 2008. Proterozoic (pre-Ediacaran) glaciation and the high obliquity, low-latitude ice, strong seasonality (HOLIST) hypothesis: principles and tests. Earth-Science Reviews, 87, 61 –93. Williams, G. E., Gostin, V. A., McKirdy, D. M. & Preiss, W. V. 2008. The Elatina glaciation, late Cryogenian (Marinoan Epoch), South Australia: sedimentary facies and palaeoenvironments. Precambrian Research, 163, 307– 331. Xu, B., Jian, P., Zheng, H., Zou, H., Zhang, L. & Liu, D. 2005. U–Pb zircon geochronology and geochemistry of Neoproterozoic volcanic rocks in the Tarim Block of northwest China: implications for the breakup of Rodinia supercontinent and Neoproterozoic glaciations. Precambrian Research, 136, 107– 123. Xu, B., Xiao, S. et al. 2009. SHRIMP zircon U– Pb age constraints on Neoproterozoic Quruqtagh diamictites in NW China. Precambrian Research, 168, 247– 258. Zhan, S., Chen, Y., Xu, B., Wang, B. & Faure, M. 2007. Late Neoproterozoic paleomagnetic results from the Sugetbrak Formation of the Aksu area, Tarim basin (NW China) and their implications to paleogeographic reconstructions and the Snowball Earth hypothesis. Precambrian Research, 154, 143–158. Zhang, Q.-R. & Piper, J. D. A. 1997. Palaeomagnetic study of Neoproterozoic glacial rocks of the Yangzi Block: palaeolatitude and configuration of South China in the late Proterozoic Supercontinent. Precambrian Research, 85, 173– 199. Zhang, Q.-R., Li, X.-H., Feng, L.-J., Huang, J. & Song, B. 2008. A new age constraint on the onset of the Neoproterozoic glaciations in the Yangtze Platform, South China. Journal of Geology, 116, 423– 429. Zhang, Q.-R., Chu, X. L. & Feng, L. J. 2009. Discussion on the Neoproterozoic glaciations in the South China Block and their related paleolatitudes. Chinese Science Bulletin, 54, 1797– 1800. Zhang, S., Evans, D. A. D. et al. 2006. New paleomagnetic results from the Nantuo Formation in south China and their paleogeographic implications. EOS, Transactions of the American Geophysical Union, abstract GP34A-04. Zhang, S., Jiang, G. & Han, Y. 2008. The age of the Nantuo Formation and Nantuo glaciation in South China. Terra Nova, 20, 289– 294. Zhou, C., Tucker, R., Xiao, S., Peng, Z., Yuan, X. & Chen, Z. 2004. New constraints on the ages of Neoproterozoic glaciations in south China. Geology, 32, 437– 440.

Chapter 8 Biostratigraphy and stratigraphic subdivision of Cryogenian successions of Australia in a global context KATHLEEN GREY1*, ANDREW C. HILL2 & CLIVE CALVER3 1

Geological Survey of Western Australia, Department of Mines and Petroleum, 100 Plain Street, Western Australia, 6076 2

Centro de Astrobiologı´a (INTA-CSIC), Instituto Nacional de Te´cnica Aeroespacial, Ctra de Ajalvir, km 4, 28850 Torrejo´n de Ardoz, Madrid, Spain 3

Mineral Resources Tasmania, PO Box 56, Rosny Park, Tasmania 7018, Australia *Corresponding author (e-mail: [email protected])

Abstract: Cryogenian correlation in Australia is based on an extensive data set from the Centralian Superbasin and Adelaide Rift Complex and integrates biostratigraphy and isotope chemostratigraphy to provide a three-dimensional interpretation based on outcrop and drill holes. Studies are ongoing, but newer data are consistent with the distributions discussed here. From the chemostratigraphic and biostratigraphic viewpoint, the first appearance of the acritarch Cerebrosphaera buickii, coupled with a large negative isotope excursion at c. 800 Ma, supported by the first appearance of the stromatolite Baicalia burra, seems to have potential for boundary placement. It is widely recognized across Australia and seems to have potential globally.

Australia has extensive Cryogenian successions that can be readily correlated using a variety of techniques. Cryogenian correlation is based on both field sections and continuous core from drill holes. In Western Australia, this includes drill holes Empress 1 and 1A, Lancer 1, Vines 1, other drill holes, and outcrops in the western Officer and Amadeus Basins and the Kimberley area. Results from Western Australia are consistent with data from other drill holes and field sections examined from across Australia, including those from the eastern Officer, Amadeus, Georgina, and Ngalia Basins, the Adelaide Rift Complex and Tasmania (Figs 8.1– 8.4). In general, biostratigraphic correlations (Figs 8.1 –8.3) correspond to those obtained using lithostratigraphy, well-log and other geophysical data and chemostratigraphy (d13C and 87Sr/86Sr; Walter et al. 1995, 2000; Morton & Drexel 1997; Hill & Walter 2000; Hill et al. 2000a, b; Preiss 2000; Haines et al. 2004, 2008; Grey et al. 2005; Mory & Haines 2005; Gorter et al. 2007; Hill, unpublished data). The Cryogenian was originally defined as a Neoproterozoic chronometric period extending from 850 Ma to 650 Ma. Since the name Cryogenian is preoccupied, it is questionable whether it should be used for any proposed chronostratigraphic units below the Ediacaran. However, for practical purposes, the term Cryogenian is here used for the interval from the chronometrically defined base of the Cryogenian System and Period at 850 Ma to the chronometrically defined top of the Cryogenian at 650 Ma (Gradstein et al. 2004), but also incorporates the interval that lies between the 650 Ma boundary and the chronostratigraphically defined base of the succeeding Ediacaran System and Period with its Global Boundary Stratotype Section and Point (GSSP) at the base of the Nuccaleena Formation, Flinders Ranges, South Australia (Knoll et al. 2006). This pragmatic approach facilitates discussion pending decisions of the International Subcommission on Neoproterozoic (Cryogenian and Ediacaran) Stratigraphy. Only Australian Cryogenian successions are discussed here in detail because it is still difficult to correlate between Australia and successions elsewhere with any certainty. However, consistencies observed across Australia auger well for eventual global correlation. There are no unequivocal Tonian-age (.850 Ma) sedimentary rocks in Australia, although several possible sections require further investigation. The base of the recognized Cryogenian

succession is probably no older than c. 830 Ma (Fig. 8.5), based on U –Pb ages for the Gairdner Dyke Swarm, which either underlies or just intrudes the base of the succession in South Australia (see below), and global chemostratigraphic correlation (Macdonald et al. 2010). Early Cryogenian successions are widespread and have been identified in the Adelaide Rift Complex (Callanna and Burra Groups), the Amadeus Basin (Bitter Springs Formation), the Georgina Basin (Yackah beds), the Ngalia Basin (Vaughan Springs Quartzite and Albinia Formation), the Officer Basin (Buldya Group and lateral equivalents in both Western and South Australia, the Kimberley area (Ruby Plains Group), parts of the Birrindudu Basin (Redcliff Pound Group), and Tasmania (lower Togari Group and its equivalents). This part of the succession underlies the Sturt glaciation (Preiss et al. 2011), named after the Sturt Tillite in the Adelaide Rift Complex), and which includes the Areyonga Formation in the Amadeus Basin and lateral equivalents in central Australia and Tasmania, but is represented by a hiatus in Western Australia, except in Vines 1, where evidence of two glaciations, represented by the older Pirrilyungka Formation and the younger Wahlgu Formation, may be present (Haines et al. 2008; Hill et al. 2011). The interval between the Sturt Tillite and Elatina Formation and their equivalents is not well represented across Australia. There is a near-continuous succession in the Adelaide Rift Complex (the Umberatana Group) and a partial succession in the Amadeus Basin (Aralka Formation). In Western Australia, this interval is marked by a heavily karstified surface and disconformity in both Empress 1A and Lancer 1. The succeeding Elatina glaciation (Williams et al. 2008) and its equivalents (previously referred to as the Marinoan glaciation) appear to be represented in all Australian Neoproterozoic basins. In Western Australia, the continuously cored Empress 1A and Lancer 1 successions can be tied to each other and to other Officer Basin drill holes using well-log data, lithostratigraphy, seismic interpretation and chemostratigraphy (Grey et al. 2005). Acritarch and stromatolite ranges are consistent across the basin and proposed biostratigraphic correlations correspond closely to non-biostratigraphic correlations. There is evidence to suggest that Australian successions can be similarly correlated to successions elsewhere in the world.

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 113– 134. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.8

114

K. GREY ET AL.

120°

130°

140°

150° 10°

10°

KIMBERLEY REGION

GEORGINA

20°

T6

BASIN AS1 Arunta Province O1 MW1 Bb1 W1 MC1

LD1 D1 H1 L1 Y2

T9 Western Officer Basin

Lu1

Probable extent of Centralian Superbasin

Y3

K1A

E1A

20°

NGALIA BASIN

AMADEUS BASIN

Musgrave Province

PEAKE & DENISON RANGES

V1 B1

OFFICER BASIN

Eastern Officer Basin

Curnamona Province

G1 Gawler Craton

B4

STUART SHELF

30°

30°

ADELAIDE RIFT COMPLEX

P12

Drillhole with Bitter Springs Formation Assemblage Drillhole with Cerebrosphaera buickii Assemblage Drillhole with both assemblages

KING ISLAND

Outcrop with Bitter Springs Formation Assemblage

1000 km 40°

40°

TASMANIA 110°

120°

130°

120°

140°

130°

150°

140°

160°

Fig. 8.1. Locality map showing distribution of Neoproterozoic basins in Australia and localities sampled for microfossil and palynology studies in the Australian Cryogenian (adapted from Walter et al. 1995; Preiss 2000; Grey 2005b). Distribution data from Grey & Cotter (1996); Cotter 1997, 1999; Grey 1999a, 2005a, b; Grey et al. 2005b. Abbreviations: AS1, BMR Alice Springs 1; Bb1, Bluebush 1; B1, Birksgate 1; B4, Western Mining BLD 04; D1, Eagle Dragoon 1; E1A, GSWA Empress 1/1A; G1, Comalco Giles 1; H1, Eagle Corporation Hussar 1; K1A, Shell Kanpa 1A; L1, GSWA Lancer 1; LD1, Lake Disappointment 1 (LDDH 1); Lu1, Shell Lungkarta 1; MC1, Transoil Mount Charlotte 1; MW1, Mount Winter 1; O1, Magellan Ooraminna 1; PP12; North Broken Hill PP 12; T6, GSWA TWB 6 (Trainor Water Bore 6); T9, GSWA TWB 9 (Trainor Water Bore 9); V1, GSWA Vines 1; W1, Wallara 1; Y2, Shell Yowalga 2; Y3, Shell Yowalga 3.

150° 10°

10°

KIMBERLEY REGION R

GEORGINA

20°

AMADEUS BASIN BS

T

NGALIA BASIN

T T S

LB LB L1

Western Officer Basin

P

BS BS W1

Y3 LB LB

Probable extent of Centralian Superbasin

20°

BASIN

Arunta Province BS

LB Musgrave Province M5 BG

E1A

OFFICER BASIN

Eastern Officer Basin

BG

Curnamona Province

BG Gawler Craton

STUART SHELF

30°

PEAKE & DENISON RANGES

30°

BG

ADELAIDE RIFT COMPLEX

W1 T LB R S

Drillhole

Outcrop

Stratigraphy

Tarcunyah Group Lower Buldya Group Ruby Plains Group Sunbeam Group

BS P BG

Bitter Springs Formation Plenty Group Burra Group

KING ISLAND

1000 km

40°

40°

TASMANIA 110°

120°

130°

140°

150°

160°

Fig. 8.2. Locality map showing distribution of the Acaciella australica stromatolite assemblage in Australia. Only localities with identified specimens are shown (data from Walter 1972; Preiss 1973b, 1976, 1987; Walter et al. 1979; Grey 1995, 1999b, 2005a, unpublished data; Stevens & Grey 1997; Grey & Blake 1999; Hill et al. 2000a; Grey et al. 2005). Abbreviations: T, Tarcunyah Group; LB, Lower Buldya Group; R, Ruby Plains Group; S, Sunbeam Group; BS, Bitter Springs Formation; P, Plenty Group; BG, Burra Group; E1A, GSWA Empress 1A; L1, GSWA Lancer 1; M5, Manya 5; W1, Wallara 1; Y3, Shell Yowalga 3.

AUSTRALIAN CRYOGENIAN CORRELATIONS

120°

130°

140°

115

150° 10°

10°

KIMBERLEY REGION

GEORGINA

20°

NGALIA BASIN

AMADEUS BASIN L1

UB

20°

BASIN

Arunta Province J W1

H1 UB

Western Officer Basin

UB

Probable extent of Centralian Superbasin

UB Musgrave Province

UB UB

E1A

OFFICER BASIN

N1

PEAKE & DENISON RANGES

BG

Curnamona Province

Eastern Officer Basin

Gawler Craton

BG BG

STUART SHELF

30°

BG

BG BG

W1

Drillhole

Fig. 8.3. Locality map showing distribution of the Baicalia burra stromatolite assemblage in Australia. Only localities with identified specimens are shown (data from Preiss 1972, 1974, 1976, 1985, 1987; Griffin & Preiss 1976; Stevens & Grey 1997; Grey 1999b, 2005a, unpublished data; Hill et al. 2000a). Abbreviations: UB, Upper Buldyah Group; J, Johnnys Creek beds; BG, Burra Group; Tg, Togari Group; E1A, GSWA Empress 1A; H1, Eagle Corporation Hussar 1; L1, GSWA Lancer 1; N1, NJD 1; W1, Wallara 1.

30°

BG

ADELAIDE RIFT COMPLEX

Outcrop

Stratigraphy UB J

Upper Buldya Group Johnnys Creek Beds

BG Tg

Burra Group Togari Group

KING ISLAND

1000 km

40°

Tg

40°

TASMANIA 110°

120°

130°

120°

140°

130°

150°

160°

140°

150° 10°

10°

KIMBERLEY REGION

GEORGINA

20°

AMADEUS BASIN

NGALIA BASIN Arunta Province A1

I W1

L Western Officer Basin

E

Probable extent of Centralian Superbasin

LS D A2 A3 B MC

Musgrave Province

M5

OFFICER BASIN

Eastern Officer Basin

C

PEAKE & DENISON RANGES Curnamona Province SR DW B2 SC O LN S3 M ET SL EC B1 S1 DC ADELAIDE SD W2 UBMF RIFT COMPLEX W5

UBPD

Gawler Craton

STUART SHELF

30°

Drillhole

20°

BASIN

30°

Outcrop

KING ISLAND

1000 km

F A

40°

40°

TASMANIA 110°

120°

130°

140°

150°

160°

Fig. 8.4. Locality map showing distribution of Neoproterozoic basins in Australia and localities sampled for chemostratigraphic studies in the Australian Cryogenian (adapted from Walter et al. 1995; Hill & Walter 2000; Preiss 2000; McKirdy et al. 2001; Grey 2005b). Drill hole abbreviations for chemostratigraphy: A1, AS 27; A2, AS 28; A3, AS 3; B1, BTD 1; B2, Blinman 2; E, Empress 1A; ET, ETM 5A-1; F, Forest 1; I, Illogwa Creek 6; L, Lancer 1; LN, LNM 10-1; M5, Manya 5; MC, Mount Charlotte 1; S1, SAS 1; S3, SAR 3; SC, SCYW 1a; SL, SLT 104; SR, SR 6; W1, Wallara 1, W2, Wokurna 2; W5, Wokurna 5. Field section abbreviations for chemostratigraphy: A, Arthur River; B, Bluebush; C, Coominaree Dolomite type section; D, Dump; DC, Depot Creek; DW, Doodney’s Well Hills; EC, Enorama Creek; LS, Limbla Syncline; M, Mallee Water; O, Oraparinna Diapir; SD, Skillogalee Dolomite in Port Germein Gorge; UBMF, Upper Burra Group, mid-Flinders Ranges; UBPD, Upper Burra Group, Peake and Denison Ranges.

K. GREY ET AL.

PER- AGE IOD (Ma) SS 543 550

4

NW

Central

Eastern

Northeastern

NGALIA BASIN

Cambrian

Cambrian

Cambrian

Cambrian

Cambrian

McFadden Lungkarta Fm Fm

ADELAIDE RIFT GEORGINA COMPLEX BASIN Cambrian

12

Julie Formation

Boondawari Fm W

?

WDM Pioneer/Olympic

?

rhyodacite 11

WILPENA GROUP

Newhaven Shale Member

Pertatataka Formation

GROUP

625

volcanics CHD

volcanics MOPUNGA GROUP

UNGOOLYA GROUP

3

NW KIMBERLEY TASMANIA REGION

Cambrian

‘lower’ Arumbera ‘lower’ Yuendumu Sandstone Sandstone

575

600

KING ISLAND

MDDM

Keppel Creek Fm

GRASSY GROUP

EDIACARAN

C

AMADEUS

LOUISA

116

W

Y

Black/Oorabra

UGE

9

10

650 2 Y

S

illite

8

7 5

6 Pirrilyungka Fm

illite

Member

675

T Steptoe Fm

4

725

Kanpa Fm

BURRA GROUP

’

750

BULDYA GROUP

‘CRYOGENIAN’

700

1

775

? Hussar Fm

3 Member

825

WM

CVE

SUNBEAM GROUP

CDE Alinya Fm

850

YCE

Loves

LM

Albinia Fm Units Gillen Member

TQ

1

Key:

1

Yackah beds

CALANNA GROUP

800

?

?

2

Curdimurka Subgroup

ROCKY CAPE GROUP

Arkaroola Subgroup

V

1

BASEMENT Neoproterozoic glacial or partially glacial units Cap dolomite Neoproterozoic sediments

Sequence boundary

1

Geochronological age Volcanic intrusion Volcanics

Unconformity/disconformity

Fig. 8.5. Australian Neoproterozoic stratigraphy and correlations assuming a c. 635 Ma age (Hoffmann et al. 2004; Condon et al. 2005) for the Elatina Formation and equivalents; however, see Williams et al. (2008) for an extended discussion of age constraints (adapted from Preiss 1987, 2000; Walter et al. 1995; Morton 1997; Grey et al. 2005; Dunster et al. 2007). The length of time breaks at major unconformities and sequence boundaries are estimates only. Amadeus Basin successions were described in Wells et al. (1970), Kennard et al. (1986), Korsch & Kennard (1991) and Lindsay (1993). Stratigraphic subdivision of the Adelaide Rift Complex and Stuart Shelf follow Preiss (1987, 2000 and references therein), Haines (1988, 1990) and Reid & Preiss (1999). Officer Basin successions and correlations were detailed in Jackson & van de Graaff (1981), Phillips et al. (1985), Townson (1985), Williams (1992), Preiss (1993), Walter & Gorter (1994), Walter et al. (1994), Bagas et al. (1995, 1999), Lindsay (1995), Lindsay & Reine (1995), Morton (1997), Hill et al. (2000a), Grey et al. (2005) and Haines et al. (2008). Georgina Basin successions were described by Smith (1972), Walter (1980) and Dunster et al. (2007), the Ngalia Basin successions in Wells & Moss (1983), the Kimberley region in Dow & Gemuts (1969), Blake et al. (1998), Corkeron & George (2001) and Corkeron (2007), and in Tasmania by Calver (1998), Calver & Walter (2000) and Calver et al. (2004). Australia-wide correlations were discussed in Preiss et al. (1978), Coats & Preiss (1980), Walter et al. (1995, 2000), Calver & Walter (2000), Hill & Walter (2000), Hill et al. (2000a, b), Calver et al. (2004) and Grey et al. (2005). SS, Supersequence number, Centralian Superbasin (Walter et al. 1995); CDE, Coominaree Dolomite equivalent; CHD, Croles Hill Diamictite; CVE, Cadlareena Volcanics equivalent; LM, Lancer Member; MDDM, Mount Davenport Diamictite Member; TQ, Townsend

AUSTRALIAN CRYOGENIAN CORRELATIONS

Geochronological constraints Tonian (1000 –850 Ma) successions are probably not widespread in Australia, although many units are inadequately dated. Units such as the Pindyin Sandstone and Alinya Formation in the eastern Officer Basin, and their presumed lateral equivalents, the Townsend Quartzite and Lefroy Formation, in the western Officer Basin, the Kulail Sandstone (Close et al. 2005) and possibly the Dean Quartzite in the Bloods Range region of the Amadeus Basin, the Heavitree Quartzite and lateral equivalents in the Georgina and Amadeus Basin, and parts of the Callanna Group in the Adelaide Rift Complex could be either Tonian or Cryogenian, depending on whether there is a stratigraphic break between them and the overlying Bitter Springs Formation and its lateral equivalents and whether they pre-date or were intruded by the Gairdner Dyke Swarm. The Alinya Formation in Giles 1 (eastern Officer Basin), which seems to have an anomalous acritarch assemblage (Zang 1995; see below), and the Lamil and Throssell Range Groups along the northeastern Pilbara margins could be either younger than (Bagas et al. 2002) or older than the Cryogenian Tarcunyah Group. In Tasmania, the Rocky Cape Group (c. 1000–750 Ma) is probably at least in part of Tonian age (Black et al. 2004). The mafic Gairdner Dyke Swarm, which intrudes the Musgrave Province, Gawler Craton and Stuart Shelf (Parker et al. 1987; Preiss 1987; Cowley & Flint 1993) and cross-cuts underlying Mesoproterozoic successions, and its northwesterly extension, the Amata Suite, probably relate to extension during the early tectonic evolution of the Adelaide Rift Complex and the Centralian Superbasin. It has not been determined whether the dykes underlie or intrude basal successions in these areas, but it is probable that they slightly post-date the onset of deposition. The dykes are probable feeders for mafic lavas of the poorly dated Wooltana Volcanics near the top of the Callanna Group and other volcanic rocks, such as the Beda Volcanics and those in the Bitter Springs Formation (Zhao et al. 1994); however, more precise dating of these formations is required before concluding they are all the same age. A Stuart Shelf dyke gave a SHRIMP U – Pb age of 827 + 6 Ma on baddeleyite, and the Little Broken Hill Gabbro in the Willyama Inlier gave an age of 827 + 9 Ma (Wingate et al. 1998). Sm –Nd isochrons of 867 + 47 and 802 + 35 Ma (Zhao & McCulloch 1993; Zhao et al. 1994) were obtained from a dyke intruding the Stuart Shelf (Preiss 2000). Another of the dykes, which cross-cuts Mesoproterozoic units in the Western Musgrave Complex, gave a zircon 207Pb/206Pb age of 824 + 4 Ma (Glikson et al. 1996). These ages constrain the maximum age of deposition of Australian Cryogenian successions to c. 830 Ma. Few Australian Cryogenian dates can be considered reliable (Preiss 2000; Fanning & Link 2008). A thermal-ionization mass spectrometry (TIMS) U –Pb age of 802 + 10 Ma was reported from the Rook Tuff, in the lower part of the Curdimurka Subgroup of the Callanna Group in the Adelaide Rift Complex (Fanning et al. 1986). Recent dating of a bedded tuff in the Fifteenmile Group of the Coal Creek Inlier, Ogilvie Mountains, NW Canada at c. 811.5 Ma (Macdonald et al. 2010) is consistent with the Rook Tuff age, and apparently lies at the same stratigraphic and chemostratigraphic level. Reid (2009) recently obtained an age of 795 + 5 Ma using laser ablation – inductively coupled plasma mass spectrometry

117

(LA –ICPMS) on a porphyry that intrudes the Skillogalee Dolomite and which is now thought to be ‘penecontemporaneous with Skillogalee-age deposition’ (Drexel 2009). This conflicts with an age on an underlying rhyolite from the Boucaut Volcanics at the base of the Burra Group cited as giving a U –Pb SHRIMP zircon age of 777 + 7 Ma by Preiss (2000); however, the Boucaut Volcanics date was recently considered unreliable by Fanning & Link (2008). A U –Pb zircon date on a thin volcaniclastic layer in the lower Wilyerpa Formation in the Adelaide Rift Complex (deposited during the waning phase of the Sturt glaciation; Preiss et al. 2011), and previously referred to as being within the Merinjina Tillite, gave an age of 659 + 6 Ma (Fanning & Link 2008), and indicates an age of c. 660 Ma for the Sturt glaciation. This is consistent with Re –Os ages from black shales that closely succeed the Sturt glacials and their equivalents: 657.2 + 5.4 Ma in the Aralka Formation in the Amadeus Basin, 643 + 2.4 Ma on the Tindelpina Shale Member in the basal Tapley Hill Formation in the Adelaide Rift Complex (Kendall et al. 2006) and 641 + 5 Ma on a black shale above the Julius River Member in Tasmania (Kendall et al. 2007). A detrital zircon U –Pb SHRIMP age of 725 + 11 Ma from the middle Kanpa Formation, .200 m below the Steptoe –Wahlgu Formation disconformity (Nelson 2002), indicates that the Western Australian hiatus equivalent to the Sturt glaciation is also younger than 725 Ma. There is no firm evidence in Australia for the Sturt glaciation extending back much beyond 660 Ma. This is in contrast with recent dating from NW Canada, where ages of 716.5 Ma were obtained for glacial units believed to be equivalents of the Sturt glaciation (Kendall et al. 2009; Macdonald et al. 2010). The age of the younger (Elatina) glaciation (Williams et al. 2008), and hence the base of the Ediacaran Period, remains uncertain. The Ediacaran boundary lies at the base of the Nuccaleena Formation, so it would be sensible for the underlying Elatina diamictite and lateral equivalents to be included with units of Cryogenian age, even if this part of the succession is younger than the previous upper chronometric boundary of 650 Ma (Gradstein et al. 2004). The Elatina Formation is currently correlated with the Ghaub Formation of Namibia, dated at 635.5 + 1.2 Ma (Hoffmann et al. 2004) and the Nantuo Tillite of China, dated by an ash bed within the cap dolostone above the tillite at 635.2 + 0.6 Ma (Condon et al. 2005) because of the presence of similar cap carbonates. Lithologically, a succession on King Island (the Cottons Breccia and overlying Cumberland Creek Dolostone) closely resembles the Elatina and Nuccaleena Formations, and correlation between the two successions appears robust (Calver & Walter 2000; Preiss 2000; Hoffman et al. 2009). The King Island succession is intruded by a 575 + 4 Ma subvolcanic sill, and overlain by mafic volcanics with a Sm –Nd isochron age of 579 + 16 Ma (Meffre et al. 2004), although the timing of the intrusion relative to deposition is unclear. In northwestern Tasmania, the Croles Hill Diamictite, upsection of the Sturt-equivalent Julius River Member, is locally underlain by rhyodacite with a SHRIMP U –Pb age of 582 + 4 Ma (Calver et al. 2004), indicating that this could be a Gaskiers-age glaciation, that is, c. 580 Ma. If the Cottons Breccia and Elatina Formation are correlatives of the Croles Hill Diamictite, then the base of the Ediacaran (as defined by the base of the Nuccaleena Formation) could be c. 580 Ma, rather than 635 Ma. Better age constraints are required on the

Fig. 8.5. (Continued) Quartzite; UGE, Umberatana Group (of the Adelaide Rift Complex) equivalents; WDM, Wanapi Dolomite Member; WM, Woolnough Member; WV, Wantapella Volcanics; YCE, Younghusband Conglomerate equivalent. Geochronology: (1) 827 + 6 Ma (Wingate et al. 1998) and 824 + 4 Ma (Glikson et al. 1996), Gairdner Dyke Swarm; (2) 802 + 10 Ma, Rook Tuff (Fanning et al. 1986); (3) 797 + 5 Ma, Skillogalee Dolomite (Drexel 2009); (4) 725 + 11 Ma detrital age, uppermost Kanpa Formation (Nelson 2002); (5) 659 + 6 Ma, Wilyerpa Formation (Sturt Tillite correlative) (Fanning & Link 2008); (6) 657.2 + 5.4 Ma, Aralka Formation (Kendall et al. 2006); (7) 647.2 + 10 Ma and 645.1 + 4.8 Ma, Tindelpina Shale Member (Kendall et al. 2006); (8) 640.7 + 4.7 Ma (Kendall et al. 2007), Black River Dolomite; (9) 657 + 17 Ma detrital age, Marino Arkose Member (Ireland et al. 1998); (10) 575 + 3 Ma (Calver et al. 2004), intrusive in Yarra Creek Shale; (11) 582 + 4 Ma, rhyodacite flow beneath Croles Hill Diamictite (Calver et al. 2004); (12) 556 + 24 Ma detrital age, Bonney Sandstone (Ireland et al. 1998).

118

K. GREY ET AL.

Elatina Formation and Nuccaleena Dolomite to confirm their correlation with the Chinese and Namibian glaciations and determine whether only the Croles Hill Diamictite is a Gaskiers glaciation equivalent, or whether other Australian Elatina glacial deposits are also of this age. Chronometric constraints on Ediacaran successions are poor and restricted to a Rb –Sr whole-rock isochron of 588 + 35 Ma on the postglacial Bunyeroo Formation, based on samples above and below the level of the Acraman impact ejecta layer in drill hole SCYW 1a (Hill et al. 2008) and a single detrital zircon grain, dated at 556 + 24 Ma, from the lower part of the Rawnsley Quartzite (Preiss 2000). The ejecta layer, a significant stratigraphic tie-line, has not been dated but was estimated to be c. 580 Ma (Williams & Gostin 2005) based on a c. 635 Ma age for the Elatina glaciation. The undated glaciogenic Egan Formation of the Kimberley area appears to correlate with carbonate units (the Julie Formation of the Amadeus Basin, the Elkera Formation of the Georgina Basin, the basal Bonney Sandstone of the Adelaide Rift Complex and the Wilari Dolomite Member, and the Tanana Formation of the Officer Basin) that lie well above the cap carbonate and just below the first record of bilaterians in other basins (Grey & Corkeron 1998).

Biostratigraphy Biostratigraphic correlation is increasingly significant for Australian Neoproterozoic correlation, and is based on chert microfossils, palynology and stromatolite biostratigraphy. Palynology provides good biostratigraphic control as well as indicating palaeoenvironment and thermal maturity. Microfossil assemblages consist predominantly of leiospheres, filaments and mat fragments, mostly of conservative, long-ranging species with simple morphologies, but containing some more complex and time-diagnostic species (Schopf 1968; Schopf & Barghoorn 1969; Fairchild 1976; Oehler 1976; Jackson & Muir 1981; Jackson & van de Graaff 1981; Zang & Walter 1992; Zang 1995; Grey & Cotter 1996; Cotter 1997, 1999; Grey & Stevens 1997; Grey 1999a, 2005a; Hill et al. 2000a, b). Data have been obtained from at least 1000 samples from about 30 drill holes across Australia, several consisting of continuous core (Figs 8.1 & 8.6). Stromatolite biostratigraphy is based on .200 samples collected from .80 localities, together with .70 horizons examined in drill core (Figs 8.2, 8.3 and 8.7, Table 8.1). Only localities with identified stromatolites are recorded here; there are numerous samples that have yet to be identified. Core data can be tied to field outcrops using both stromatolite and acritarch biostratigraphy, and taxonomic ranges are consistent across the basins and correspond to the correlations determined by independent methods such as lithostratigraphy, seismic and well-log data and chemostratigraphy. Biostratigraphic correlation is more difficult globally because so few successions have so far been adequately documented. Although there are probably at least two hundred assemblages recorded in the literature (Knoll 1996), ages are often poorly constrained, so it is difficult to know precisely when in the Neoproterozoic (or in some cases, in the Proterozoic) some of the successions occur. Several reviews have been published, most notably a compilation of c. 3500 occurrences of Proterozoic and Early Cambrian microfossils and microfossil-like objects, based on 470 stratigraphic units and 316 published papers by Mendelson & Schopf (1992). From the taxonomic point of view, the catalogue of acritarchs published by Fensome et al. (1990) is invaluable, but did not include microfossils other than those considered to be acritarchs. Many significant publications were overlooked in this catalogue and it is difficult to extract taxa that fall within the age constraints of interest. Given the numerous revisions to stratigraphy of recent years, updating the records is a major undertaking and beyond the

scope of this review. Moreover, very few of the records contain adequate range charts, so it is very difficult to determine the relative stratigraphic positions and ranges of each of the species.

Palynology In Australia, identical palynological successions are present in Empress 1A and Lancer 1 and are known from several other western Officer Basin drill holes, particularly LDDH 1, Kanpa 1A and Yowalga 3 (Grey & Cotter 1996; Cotter 1997, 1999; Grey & Stevens 1997; Grey 1999a, 2005a; Hill et al. 2000a) (Fig. 8.1). Elements of the various assemblages have been recorded in other drill holes, but distributions and ranges have not yet been thoroughly documented (Grey, unpublished data). The oldest described Cryogenian assemblage is probably that from the Alinya Formation in Giles 1, eastern Officer Basin (Zang 1995). This assemblage is puzzling because it contains Trachyhystrichosphaera, Cymatiosphaeroides kullingii and Valeria lophostriata (normally found in Tonian or Mesoproterozoic successions), vase-shaped microfossils, assigned to Melanocyrillium sp. (also usually found in early Cryogenian assemblages), Cryogenian species similar to those in the Bitter Springs and Browne Formations, and spiny forms resembling acanthomorphs recorded from the Ediacaran (although none of the latter seems identical to other Australian Ediacaran species). This strange combination of taxa may indicate reworking of older rocks and possibly the presence of two disconformable successions. This assemblage requires reassessment. Vase-shaped microfossils are locally abundant in chert in the Black River Dolomite below the diamictitic Julius River Member in Tasmania (Saito et al. 1988), and are consistent with an age pre-dating the Sturt glaciation (Porter & Knoll 2000). Elsewhere in Australia, acritarch and microfossil assemblages show consistent distribution patterns. In the western Officer Basin, they are present in the Browne Formation and its lateral equivalents (Jackson & van de Graaff 1981; Grey & Cotter 1996; Cotter 1997, 1999; Grey & Stevens 1997; Grey 1999a, 2005a). Well-preserved microfossils from chert in the Browne Formation at the Madley Diapirs (Jackson & van de Graaff 1981, p. 34; Cotter 1997) show low species diversity and mainly represent permineralized cyanobacterial mat. Palynomorphs from the Browne Formation in Browne 1 and 2, Kanpa 1A, Dragoon 1 and Yowalga 2 and 3 include Myxococcoides sp., Siphonophycus sp., Leiosphaeridia spp. and Synsphaeridium spp., many of which are long ranging and have been recorded from younger and older successions (Cotter 1997, 1999). Some species of Simia and Satka colonialica may have more restricted stratigraphic distributions. Similar taxa were subsequently identified in other Officer Basin drill holes, including LDDH 1, Empress 1A and Lancer 1 (Grey & Cotter 1996; Grey 1999a, 2005a). This lower Buldya Group (Browne Formation) assemblage has many species in common with assemblages from the Gillen Member of the Bitter Springs Formation (Zang & Walter 1992). Palynomorphs from the upper Buldya Group (Hussar, Kanpa and Steptoe Formations) in Hussar 1, Lungkarta 1, Yowalga 2 and 3, Empress 1A and Lancer 1 (Grey & Cotter 1996; Cotter 1999; Grey 1999a, 2005a; Hill et al. 2000a; Grey et al. 2005) include Cerebrosphaera buickii, C. ananguae, Chuaria sp. cf. circularis, Coneosphaera sp., Eoentophysalis croxfordii, Eomicrocystis sp. cf. elegans, Leiosphaeridia crassa, L. sp. cf. exsculpta, L. jacutica, L. minutissima, L. tenuissima, L. ternata, Myxococcoides cantabrigiensis, Ostiana microcystis, Pterospermopsimorpha insolita, Satka colonialica, Simia annulare, Stictosphaeridium sinapticuliferum, ?Symplassosphaeridium sp., Synsphaeridium spp., Tasmanites sp., as well as filamentous microfossils Calyptothrix sp. cf. alternata, Clavitrichoides rugosus, Oscillatoriopsis amadeus, Oscillatoriopsis sp., Siphonophycus kestron, S. robustum, S. septatum, S. solidum and S. typicum. Although taxa from the previous

AUSTRALIAN CRYOGENIAN CORRELATIONS

119

Fig. 8.6. (a– f) Cerebrosphaera buickii from Spitsbergen, Australia and the Grand Canyon. (a, b) Svanbergfjellet Formation, Spitsbergen: HUPC 62713B, P-2945-47M (S-33-4), holotype (a); HUPC 62763, 86-G-33-2S, SEM (b). Photos in a and b are courtesy of N. Butterfield (specimens described in Butterfield et al. 1994). (c) Hussar Formation, Officer Basin, Empress 1, 808.2 m/2, L-38-4. (d) Hussar Formation, Officer Basin, Lancer 1, 605.14 m/2, W-44-3, Officer Basin. (e) Pirilyungka Formation, Vines 1, Officer Basin, 1685.33 m/2, L-57-2. (f ) BLD 4, Skillogalee Dolomite, Stuart Shelf, BLD 4, 1008.2 m, E-56-0. (g) PP12, Anama Siltstone, Stuart Shelf, PP12, 156.9 m, N-58-2. (h) ‘Finke beds’, Amadeus Basin, 1502.0 m, Q-33-0. (i) Lower Chuar Group (Galeros Fm), Grand Canyon, USA, SP14-63-14 (Nagy 2008, MSc thesis, photo courtesy of S. Porter).

assemblage persist, the more elaborate forms tend to be fewer in number, so assemblages are dominated by leiospheres, Siphonophycus spp., Synsphaeridium spp., Chuaria sp. cf. circularis and Cerebrosphaera buickii. In this assemblage, the distinctive acritarch Cerebrosphaera buickii consistently first appears about the middle of the Hussar Formation in the western Officer Basin (Grey & Cotter 1996; Cotter 1999; Hill et al. 2000a). In the Adelaide Rift Complex, C. buickii is present in the Anama Siltstone Member of the

Rhynie Sandstone, near the base of the Burra Group in drill hole PP 12 and the Skillogalee Dolomite (in drill hole BLD 4) in the middle Burra Group (Hill et al. 2000a). The same species was recently identified in the ‘Finke beds’ between the Johnnys Creek Member of the Bitter Springs Formation and the Areyonga Formation in Wallara 1 in the Amadeus Basin (Grey, unpublished observations). C. buickii is a key species because it has been recognized at about the same stratigraphic level across Australia (Hill

120

K. GREY ET AL.

Cambrian c. 600

c. 635

c. 660 c. 695 c. 800

Approximate age (Ma)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Stratigraphic package

GM LCM2/BrF

Minjaria pontifera

LCM2/ERD

Linella avis

LCM2/BrF/?CD equiv.

Kulparia alicia

LCM2/BrF

Jurusania nisvensis

LCM2/BrF/WWF

Inzeria intia

CD LCM2/BrF LCM2/BrF/SHF LCM2/BrF/SHF/KF/WWF/Yb/CD

Gymnosolen f.indet. Boxonia pertaknurra Basisphaera irregularis Acaciella australica

UBG

Tungussia n. form

UBG

Stomatolite n. form (pseudocolumnar)

UBG/UTG

THF

Omachtenia n. form Acaciella augusta

BgL

Inzeria conjuncta

BgL

Inzeria multiplex

BgL

Boxonia melrosa

AnF

Katavia costata

AnF

Tungussia etina

AnF

Linella f. indet. Inzeria f. indet. Linella munyallina

EF

Jurusania burrensis

EF

Kulparia kulparensis

?AF

Clast supported stromatolite

PF

Divergent branching stromatolite

TF

unnamed stromatolite

PiS & OF/WF (in cap carbonate)

Kotuikania juvensis = Elleria minuta

upper carbonate BdF

Tesca stewartii

upper carbonate BwF

Stromatolite form 1 Walter et al.

upper carbonate BwF

Eleonora boondawarica

upper carbonate BwF

Acaciella savoryensis

Basal BnS

Linella f. indet.

JF/Basal BnS/WDM/EkF/EgF

Tungussia julia

EgF

Stromatolite form Grey & Corkeron

EkF

Georgina howchini

Linella munyallina assemblage

EF EF

Inzeria multiplex assemblage

BgL

Conophyton n. form Tungussia wilkatanna

Baicalia burra assemblage

Baicalia burra

Elatina glaciation

BuG/UBG/UTG/JCb/WHF/JRM (clasts)

UBG/BuG

Acaciella australica assemblage

Parmites f. indet. Tungussia erecta

Sturt glaciation

PQ/RBb

Fig. 8.7. Range chart of Neoproterozoic stromatolite distributions in Australia (see Table 8.1 for guide to abbreviations and additional stratigraphic information). Details of the stratigraphic units mentioned here can be found by searching the Geoscience Australia Stratigraphic Names database on http://dbforms.ga.gov.au/pls/www/ geodx.strat_units.int.

AUSTRALIAN CRYOGENIAN CORRELATIONS

121

Table 8.1. Key to abbreviations in stromatolite range chart (Fig. 8.7) Symbol

Package

Stratigraphic unit

AnF AF BdF BgL BnS BrF BuG BwF CD EF EgF EkF ERD GM JCb JF JRM KF LCM2 OF PF PiS PQ RBb SHF TF THF UBG UTG WDM WF WHF WWF Yb

8 6 1 9 1 13 12 1 13 7 1 1 13 14 12 1 12 13 13 2 5 2 15 15 13 4 10 12 12 1 2 12 13 13

Angepena Formation, Upalinna Subgroup, Umberatanna Group, Adelaide Rift Complex Aralka Formation (Mount Conner, Fenn Gap & Pioneer Ssst clast), Amadeus Basin Boord Formation, western Amadeus Basin Brighton Limestone, Upalinna Subgroup, Umberatanna Group, Adelaide Rift Complex Bonney Sandstone, Pound Subgroup, Wilpena Group, Adelaide Rift Complex Browne Formation, lower Buldya Group, western Officer Basin Burra Group (Skillogalee Dolomiteand other units), Adelaide Rift Complex Boondawari Formation, western Officer Basin Coominaree Dolomite, Callanna Group, Adelaide Rift Complex & eastern Officer Basin Etina Formation, Upalinna Subgroup, Umberatanna Group, Adelaide Rift Complex Egan Formation, Louisa Downs Group, central Kimberley region Elkera Formation, Mopunga Group, Georgina Basin Eliot Range Dolomite, Ruby Plains Group, eastern Kimberley area Gillen Member, Bitter Springs Formation, Amadeus Basin Johnnys Creek beds’, Amadeus Basin Julie Formation, Amadeus Basin Julius River Member, Black River Dolomite, Togari Group, Smithton Basin Karara Formation, Tarcunyah Group, NE Pilbara Loves Creek Member unit 2, Bitter Springs Formation, Amadeus Basin Olympic Formation, Amadeus Basin Pertatataka Formation, Amadeus Basin Pioneer Sandstone, Amadeus Basin Paralana Quartzite, Arkaroola Subgroup, Callanna Group, Adelaide Rift Complex River Broughton beds, Curdimurka Subgroup, Callanna Group, Adelaide Rift Complex Skates Hills Formation, Sunbeam Group, western Officer Basin Trezona Formationm, Upalinna Subgroup, Umberatanna Group, Adelaide Rift Complex Tapley Hill Formation, Nepouie Subgroup, Umberatanna Group, Adelaide Rift Complex Upper Buldya Group (upper Hussar, Kanpa & Steptoe Formations), western Officer Basin Upper Tarcunyah Group Undifferentiated, Officer Basin, south east Pilbara Wilari Dolomite Member, Tanana Formation, Ungoolya Group, eastern Officer Basin Wahlgu Formation, western Officer Basin Wright Hill Formation, eastern Officer Basin Waltha Woora Formation, Tarcunyah Group, NE Pilbara Yackah beds, Plenty Group, Georgina Basin

Details of the stratigraphic units mentioned here can be found by searching the Geoscience Australia Stratigraphic Names database: http://dbforms.ga.gov.au/pls/www/geodx.strat_units.int.

et al. 2000a) and could be an important index species for global correlation (Fig. 8.6). It was previously recorded from the Svanbergfjellet Formation and overlying Draken Formation, Akademikerbreen Group, Spitsbergen (Knoll et al. 1991; Butterfield et al. 1994) and the Tanner, Jupiter and Carbon Canyon Members of the Galeros Formation, lower and middle Chuar Group, Grand Canyon, USA (Nagy 2008). Considering that there is good geochronological constraint on the first appearance (younger than 802 + 10 Ma for the Rook Tuff and older than 795 + 5 Ma for the Skillogalee Dolomite), and the coincidence or near coincidence with three other key markers (d13C, 87Sr/86Sr and the first appearance of the stromatolite Baicalia burra), the first appearance of C. buickii is a strong candidate for a boundary that could be correlated globally (see section ‘Chemostratigraphy’ below). Sturt glacial and post-glacial assemblages are poorly known throughout Australia. In the Adelaide Rift Complex only a few simple, poorly preserved leiospheres have been recorded (Timofeev 1966; Grey, unpublished data). Elsewhere, successions appear barren. In Western Australia, a hiatus with a karstified surface characterizes the interval. In Vines 1, there may be a transition into the Pirrilyungka Formation, a newly defined diamictite unit that is probably a Sturt Tillite equivalent, and is directly overlain by the Wahlgu Formation, a diamictite correlated with the Elatina Formation (Haines et al. 2008; Hill et al. 2011). C. buickii is present, possibly in situ, in the turbiditic Unit 1 at the base of the Pirrilyungka Formation (Haines et al. 2008), but is only present as sparse, reworked fragments in the diamictite.

Sparse, reworked fragments are present in the Chambers Bluff Tillite in Nicholson 2, eastern Officer Basin (Eyles et al. 2007), another probable Sturt glacial equivalent. Elsewhere diamictite and interglacial successions appear barren. The overlying Ediacaran acritarch succession is very well constrained lithostratigraphically across Australia, and closely matches correlations based on chemostratigraphy, seismic interpretation and stromatolite biostratigraphy, even though absolute dating remains poor (Grey 2005b; Willman et al. 2006; Willman & Moczydłowska 2008). Acritarch distributions are consistent and well constrained from above the Elatina glaciation up to the change of facies associated with the appearance of the bilaterian Ediacara fauna. Although biostratigraphic distributions are consistent across Australia, they seem to be inconsistent with some of the distributions reported from NW Canada (Macdonald et al. 2010). There does not seem to be anything equivalent to the Wynniatt Group assemblage in Australia. Vase-shaped microfossils appear sparse in the Australian successions, and there are no reports of C. buickii at the appropriate stratigraphic levels in north America, with the exception of the Chuar Group.

Stromatolites Stromatolites have been recorded at various levels throughout the Neoproterozoic of Australia (Walter 1972; Preiss 1972, 1973a, b,

122

K. GREY ET AL.

Fig. 8.8. (a– h) Taxa found in the Acaciella australica Stromatolite Assemblage: from Jay Creek, Amadeus Basin, Northern Territory (b, f, g); from Lancer 1, Officer Basin, Western Australia (c, d, e). (a) Bioherms of A. australica in the Skates Hills Formation, Skates Hills, Officer Basin, Western Australia. (b) A. australica, polished slab showing branching columns. (c) Basisphaera irregularis, drill core split face, 349 m. (d) Boxonia pertaknurra, drill core surface, 332.5 m. (e) Inzeria intia, drill core split face 333 m. (f ) Kulparia alicia, GSWA 109264C, polished slab showing branching columns. (g) Linella avis, thin section, GSWA 109267. (h) Minjaria pontifera, cut core face, Wallara 1, 1979 m, Amadeus Basin, Northern Territory.

AUSTRALIAN CRYOGENIAN CORRELATIONS

123

Fig. 8.9. Taxa found in the Baicalia burra Stromatolite Assemblage. All specimens from upper Buldya Group, Officer Basin, Western Australia; all core samples from Empress 1A. (b– f, h, i) Drill core split faces. (a) B. burra, GSWA 84708, upper Buldya Group, near Lake Throssell. (b) B. burra, core face, c. 775 m, Kanpa Formation. (c) B. burra, small columns, 510.8 m, Steptoe Formation. (d, e) Tungussia wilkatanna, Steptoe Formation: (d) 656.9 m (e), 513.5 m. (f ) Tungussia new form, Empress 1A, 1077.4 m, Hussar Formation. (g, h) Conophyton new form, (g) GSWA 139573, upper Buldya Group, near Constance Headland; (h); 779.0 m, Kanpa Formation. (i) Unnamed pseudocolumnar stromatolite, 520.7 m, Steptoe Formation.

ADELAIDE RIFT COMPLEX Peake and Denison Ranges

13

6

Belair Subgroup

8

1400

o

BASIN

1300

ce Arunta Provin

W1 D A3 B

900

A

150

o

400

H

300

8

C.b. Assemblage

G

0

2

4

6

8 act

Steptoe Fm

600

700

Kanpa Fm

J I

-2

1260

800

F

1200

Composite

Areyonga Formation

G

‘Finke beds’

1000

900

Aeolian deposition

1400

1500

TD 1625

Basement

W

EASTERN OFFICER BASIN Manya-5

518

ite

Stromatolite

Mudstone–siltstone

Calcareous siltstone

Evaporite

Dolomitic siltstone

Dolomite

Missing section

Limestone

Basalt

Depth (m)

om

Cadlareena Volcanics

at

ol

Sandstone

St r

Key

-2

0

2

4

6

500

400 372

1139

1300

Break in section

B or D A or C

300

200

Dump outcrop

YCE

1250

Assemblage

CDE

1200

TD 1333

Fig. 8.10.

4

6

8

?E

100 0

ROCKY CAPE GROUP

ADELAIDE RIFT COMPLEX Peake and Denison Ranges

D C

468

1150

Unconformity

8

?F

700

0

600

B A

2

MTS

s

Basement

act

D C

C (% )

0

G

Ass.

A.a.

1400

Woolnough Member

6

E

1300

D C B A

A.a.

1300

4

200 13

800

200

E

2

400

-2

1049

100

Lancer Member

0

300

1200

C (% )

0

500

AMADEUS BASIN

400

E

13

Thickness -2 (m)

Member

100

500

Hussar Fm

G

300

700

1000

1100

Woolnough Member

8

600

F

1100

6

Bungaree Quartzite

800

G

900

1000

4

1000

800

Hussar Fm

2

1100

900

900

0

1200

H

NW TASMANIA Arthur River section

200

Bluebush outcrop

Kanp

700

-2

8

112

100

Heavitree Quartzite

0

B A

Cadlareena Volcanics

Depth (m)

50

weathered outcrop

-2

104

100

0

B or D

0

K. GREY ET AL.

600

I H

Depth (m)

300

6

s

6

C (% )

4

ite

4

0 Unit A (mostly quartz sandstone)

13

Assemblage

2

400 Pa ly St no ro lo m gy at ol ite s

C (% )

500 Keene Basalt

Pa ly St no ro lo m gy at ol ite s

13

0

100

Empress 1A

Assemblage

Pa ly St no ro lo m gy at ol ite s -2

2

578

500

Lancer 1 Depth (m)

Undalya Quartzite

200

WESTERN OFFICER BASIN

ol

140o

0

Undalya Quartzite

at

130o

I

100

500

TASMANIA 120o

200

600

40o

KING ISLAND

J

300

om

Drillhole Field section

o

400

I

800 700

40

500

St r

UBMF

1000

A.a. Assemb.

ADELAIDE RIFT COMPLEX

CD

SD

600

Pa l St yno ro lo m gy at ol ite s

STUART SHELF

8

YC

30o

Gawler Craton

6

700

1100

UBPD

4

900

Member

o

C

M5 Eastern Officer Basin

2

800

W

OFFICER BASIN

C (% )

0

1007

J

1200 PEAKE & DENISON RANGES

Thickness (m) -2

Member

L

E

30

4

1528

Kalachalp

Western Officer Basin

2

T

20

0

C.b.

AMADEUS BASIN

C (% )

-2

CALLANNA GROUP Arkaroola Subgroup

GEORGINA 20o

13

Thickness (m)

Calthorinna Tillite

KIMBERLEY REGION

ADELAIDE RIFT COMPLEX Mid-North (Southern Flinders Ranges)

Pa ly Assemblage Str nol om og at y ol ite s

150o

140o

B.burra

o

Auburn Dolomite Member

130

124

120o

AUSTRALIAN CRYOGENIAN CORRELATIONS

1974, 1976, 1985, 1987; Griffin & Preiss 1976; Walter et al. 1979, 1994; Grey 1995, 1999b, 2005a; Grey & Stevens 1997; Hill et al. 2000a; Grey et al. 2005, unpublished data; Williams et al. 2007; Planavsky & Grey 2008), but discussion here is restricted mainly to those of demonstrated stratigraphic significance (Figs 8.2, 8.3 and 8.7–8.9, Table 8.1). Throughout Australia, pre-Sturtglaciation Cryogenian successions are characterized by two widespread, apparently time-restricted stromatolite assemblages: the lower Acaciella australica Stromatolite Assemblage (Figs 8.2, 8.7 and 8.8) and the upper B. burra Stromatolite Assemblage (Stevens & Grey 1997; Hill et al. 2000a; Figs 8.3, 8.7 and 8.9). The value of these two assemblages for biostratigraphy is clearly established by the presence of both assemblages in the same relative stratigraphic relationships in three drill holes across two basins. Two other assemblages, the Inzeria multiplex Stromatolite Assemblage and the Linella munyallina Stromatolite Assemblage (Fig. 8.7), are present in the interglacial succession, but so far are known only from the Adelaide Rift Complex. Stromatolites are less common in the Ediacaran succession and seem to be of more limited biostratigraphic application than the Cryogenian ones (Fig. 8.7, Table 8.1). The Acaciella australica Stromatolite Assemblage (Figs 8.2, 8.7 and 8.8, Table 8.1), epitomized by Unit 2 of the Loves Creek Member, Bitter Springs Formation, Amadeus Basin, is characterized by Acaciella australica and 12 associated forms (Stevens & Grey 1997; Hill et al. 2000a). In Western Australia, this assemblage is known from the Browne Formation (including in continuous core in Empress 1A and Lancer 1), the Skates Hills Formation of the western Officer Basin (Hill et al. 2000a; Grey 1995, 2005a; Grey et al. 2005) and unassigned dolomite in the Tarcunyah Group (Grey, unpublished data). The Acaciella australica Stromatolite Assemblage is widespread in the Amadeus Basin, including the Wallara 1 drillcore (Walter 1972; Walter et al. 1979; Grey, unpublished data). It is also present in the Coominaree Dolomite of the Adelaide Rift Complex, equivalent successions in the eastern Officer Basin, the Yackah beds of the Georgina Basin, and in the Eliot Dolomite in the eastern Kimberley (Preiss 1973b, 1987; Grey 1995; Grey & Blake 1999; Grey et al. 2005; Planavsky & Grey 2008). In South Australia, the A. australica Stromatolite Assemblage in the Coominaree Dolomite is older than the 802 + 10 Ma Rook Tuff. Its stratigraphic position relative to other biostratigraphic markers, d13C and 87Sr/86Sr excursions and the Sturt glaciation is welldocumented in Australia (Hill & Walter 2000; Hill et al. 2000a, b; Hill 2005; unpublished data). One form, Linella avis, was first described from the Maly Karatau Ridge, Chichkan Formation, Malokaroy Group and the Talass Range in Kazakhstan (Krylov 1967, 1975; Sergeev 1989; Grey & Blake 1999; Raaben et al. 2001), which is possibly equivalent in age to the Bitter Springs Formation (Sergeev 1989; pers. comm. 1996). Poorly documented forms from the Mackenzie Mountains of Canada appear to belong to this assemblage and require further investigation (Grey, unpublished data). In Western Australia, the overlying B. burra Stromatolite Assemblage (Figs 8.3, 8.7 and 8.9, Table 8.1) ranges from about the middle Hussar Formation, through the Kanpa Formation and into the overlying Steptoe Formation. The assemblage is

125

dominated by B. burra, which is geographically widely distributed. In the Kanpa Formation, B. burra is accompanied by Conophyton new form, and in both the Officer Basin and Adelaide Rift Complex by T. wilkatanna. The assemblage has been recorded in Empress 1A, Lancer 1, Kanpa 1A and Wallara 1, as well as being known from many Officer Basin field localities (Grey et al. 2005). B. burra and Conophyton form indet. have also been collected from an unassigned unit near the top of the Tarcunyah Group near Constance Headland in the northwestern Officer Basin (Grey, unpublished data). A previously unknown form, tentatively identified as Tungussia? form indet., is present in both Empress 1A and Lancer 1. Both B. burra and T. wilkatanna are present in the Burra Group of the Adelaide Rift Complex (Preiss 1972, 1974, 1985, 1987), supporting correlation between the upper Buldya Group and at least part of the Burra Group. B. burra has also been recorded in Tasmania (Griffin & Preiss 1976), as erratics in Sturt glacial deposits (Julius River Member) of the Black River Dolomite (Calver 1998; Grey, unpublished data). Unpublished data suggest that this assemblage is present in western USA and Canada in the Mackenzie Mountains, but none of the stromatolites in this area or in other areas of NW Canada has been formally identified. Testing whether or not the apparently abundant stromatolites show similar distribution patterns to those of Australia could provide a significant step forward in global correlation.

Chemostratigraphy In the last decade, advances in Neoproterozoic chemostratigraphic (d13C and 87Sr/86Sr) correlations throughout Australia have provided ties between the Australian Cryogenian and successions in Canada, Svalbard, Namibia and Mongolia (Calver & Lindsay 1998; Calver 2000; Calver & Walter 2000; Hill & Walter 2000; Hill et al. 2000b; Walter et al. 2000; McKirdy et al. 2001; Hill 2005, unpublished data). Figures 8.10 and 8.11 show a compilation of pre-Sturt glaciation, and Sturt glaciation and post-Sturt glaciation d13C stratigraphy of Australia, respectively. There are few reliable Australian Cryogenian 87Sr/86Sr ratios, because dolomite predominates over limestone. Despite minimal utility for Australia-wide correlations, they are extremely important globally (see section ‘Discussion’). There is no official terminology in d13C or 87Sr/86Sr chemostratigraphy. In the Neoproterozoic most authors focus on correlating negative parts of d13C curves, because Neoproterozoic rocks predominantly have positive d13C values and any negative d13C values occur only for relatively brief periods of time. This has led to the use of ‘excursion’ and ‘anomaly’, often preceded by ‘negative’, the connotation being that these are short-lived events. Even when not preceded by ‘negative’ it could confuse the reader into that interpretation. There are significant ‘positive’ anomalies or excursions, particularly in Cryogenian rocks, that are important for global correlation. The term ‘stage’ has been used to describe the long interval of negative d13C values in the Loves Creek Member of the Bitter Springs Formation (Halverson et al. 2005), but could equally be applied to long intervals of

Fig. 8.10. (Continued) Pre-Sturt glaciation d13C stratigraphy of Australia. Location of drill hole and outcrop sections shown in inset (see Fig. 8.4 for explanation of abbreviations). Horizontal lines in palynology and stromatolites columns indicate first appearances: (A.a) Acaciella australica; (B) Baicalia burra Stromatolite Assemblage; (C.b) Cerebrosphaera buickii. See text for an explanation of d13C correlation intervals A–J. Western Officer Basin: lithostratigraphy (Stevens & Apak 1999; Haines et al. 2004), d13C (Hill & Walter 2000; Hill 2005) and biostratigraphy (Grey 1999a, b, 2005a); eastern Officer Basin: lithostratigraphy (Morton 1997), d13C (Hill & Walter 2000) and biostratigraphy (Preiss 1987; Grey 1995); Peake and Denison Ranges: lithostratigraphy (Fairchild 1976; Ambrose et al. 1981), d13C (Hill & Walter 2000) and biostratigraphy (Preiss 1987; Hill et al. 2000a); southern Flinders Ranges: lithostratigraphy (Preiss 1987), d13C (Hill & Walter 2000) and biostratigraphy (Hill et al. 2000a); Amadeus Basin: lithostratigraphy (Hill et al. 2000b, and references therein), d13C (Hill et al. 2000b) and biostratigraphy (unpublished data); Tasmania: lithostratigraphy and d13C (Calver 1998). Abbreviations: CD, Coominaree Dolomite; CDE, Coominaree Dolomite equivalent; MTS, molar tooth structure; TD, total depth of drill hole; WM, Woolnough Member; YC, Younghusband Conglomerate; YCE, Younghusband Conglomerate equivalent.

126

K. GREY ET AL.

positive values. The problem is that ‘stage’ has lithostratigraphic and biostratigraphic connotations. In this manuscript we prefer the use of ‘d13C interval’ or just ‘interval’ to describe those parts of d13C curves that are significant for Australian and global correlation and that may include positive or negative d13C excursions, or both, or parts of d13C curves that do not have anomalously low or high values. This terminology is preferable for two reasons. The first is that it does not exclude those parts of d13C curves that are not anomalous; a long interval of moderately positive d13C values can be equally important for correlation. Second, in some cases it minimizes focusing on every peak or trough in d13C curves that may be a result of local or regional factors, and instead concentrates on the broader-scale d13C trends or patterns. Ten distinct d13C intervals (A–J) have been identified below the base of the Sturt glaciation in Australia based on correlations between at least two separate drill hole or outcrop sections (see also Hill & Walter 2000; Walter et al. 2000) (Fig. 8.10). Beneath interval A there is only a d13C stratigraphy in the Amadeus Basin (Gillen Member, Bitter Springs Formation) so no Australia-wide correlations are possible; however, this interval is potentially important for global correlation. Intervals A –D are older than c. 800 Ma and coincide with the A. australica Stromatolite Assemblage described above. The palynological and microfossiliferous records at this stratigraphic level across Australia are also consistent. The predominant chemostratigraphic feature at this stratigraphic level is the presence of 13C-depleted d13C values ( –4 to 0‰) (intervals B and D; best represented by Units 1 & 2 of the Loves Creek Member, Bitter Springs Formation, Amadeus Basin), which are separated by a comparatively brief interval of positive d13C values (c. þ6‰) (interval C). There are no reliable 87Sr/86Sr ratios at this stratigraphic level, but there are in older sediments, in the Gillen Member of the Bitter Springs Formation. Intervals E –J are younger than c. 800 Ma and older than c. 660 Ma, are characterized by predominantly positive d13C values (mostly þ2 to þ6‰), and coincide with the first appearance of C. buickii and the B. burra Stromatolite Assemblage. First records of these taxa occur in interval E in the Amadeus Basin and interval F in the western Officer Basin (Grey, unpublished data), and in the lowermost Burra Group in the Adelaide Rift Complex (this chapter), which probably corresponds with interval E (however, further d13C stratigraphy needs to be completed at this level in the Adelaide Rift Complex in order to verify this). d13C values of between –3 and 0‰ in the middle Hussar Formation may correlate with the Black River Dolomite in Tasmania (interval F); however, this correlation is tentative because the Arthur River section is not as thick as in Lancer 1 and Empress 1A in the western Officer Basin. It is probable that this part of the succession (interval F) is missing from the central (Wallara 1 locality) and northeastern (AS 27 and AS 28 localities) Amadeus Basin. Interval F has also not yet been identified in the Adelaide Rift Complex; further work needs to be completed to see whether it does exist, because it could potentially be important in strengthening Australian correlations. Within the ranges of C. buickii and B. burra, d13C values peak at about þ8‰ (interval H), immediately preceding an interval of 13C-depleted d13C values (interval I). d13C values then return to between þ4 and þ6‰ (interval J) just prior to the Sturt glacial deposits (best represented in the Kalachalpa Formation of the Peake and Denison Ranges. The positive shift in d13C values from between – 4 and 0‰ (interval D) to between þ2 to þ6‰ (interval E) coincides with a rise in 87Sr/86Sr from 0.7057 to 0.7063, as recorded in the Bitter Springs Formation (Fanning 1986, reproduced in Hill et al. 2000b, unpublished data). The positive shifts in d13C and 87 Sr/86Sr are followed by the first records of C. buickii and the B. burra Stromatolite Assemblage (interval E). Therefore, there is an approximate coincidence between the first appearance of C. buickii and positive shifts in d13C and 87Sr/86Sr at c. 800 Ma, which is consistent with data from Svalbard and Canada (see

section ‘Discussion and Global Correlation’). The next youngest well-preserved 87Sr/86Sr ratios in Australia are from within the Sturt glacial deposits in northwestern Tasmania, where the lowest ratio is 0.7063 (Calver 1998). In Sturt glacial deposits and post-Sturt –pre-Elatina glaciation (interglacial) deposits there are lithological, geochemical (d13C intervals K – P) and geochronological similarities between the Amadeus Basin, Stuart Shelf, Adelaide Rift Complex and northwestern Tasmania. The remainder of the interglacial succession, excluding Elatina glaciation deposits, is only known in the Adelaide Rift Complex, so no interbasinal comparison is possible. However, the Adelaide Rift Complex d13C record compares closely with global records (see section ‘Discussion and Global Correlation’). A rise in 87Sr/86Sr ratios from 0.7063 within the Sturt glacial deposits (northwestern Tasmania; Calver 1998) to 0.7071 in the interglacial succession (Adelaide Rift Complex; McKirdy et al. 2001) also compares closely with global records. For Elatina glaciation units there are data from dolomite beds – originally primary precipitates – within the Olympic Formation (d13C, –1 to 3.5‰; Kennedy et al. 2001; Skotnicki et al. 2008) and Pioneer Sandstone (d13C, –2 to 2‰; Calver 1995) that define an upward increasing trend. These d13C data are comparable to those within the Ghaub Formation in Namibia (Kennedy et al. 2001) (see section ‘Discussion and Global Correlation’).

Discussion and global correlation From correlations shown in Figures 8.10 and 8.11, a composite Australian Cryogenian chemostratigraphy (d13C and 87Sr/86Sr) is compared with other global successions including the first records of C. buickii and the stromatolite B. burra (Fig. 8.12). In Spitsbergen and the USA, only Baicalia has been identified. To avoid ambiguous d13C correlations between global successions, the tie-lines in Figure 8.12 (dark green dashed lines) are chosen to coincide with the most extreme values, which avoids the more nuanced fluctuations that may be a result of more localized (regional) heterogeneity. In this way it is possible to show similarities in d13C and 87Sr/86Sr that extend globally. Three ‘broad’ intervals of the global d13C curves are possible, which are consistent with radiometric dates, 87Sr/86Sr and palaeontological data.

Australia – Canada– Svalbard: .800 Ma This correlation is based largely on an extended stratigraphic interval where d13C values fall between –4 and 0‰ (intervals B & D in Australia; Fig. 8.10). The c. 830–800 Ma age for this anomaly is based on radiometric dates from Australia (see section ‘Geochronology’) and is consistent with a c. 780 Ma age for the Little Dal Basalt at the top of the Mackenzie Mountains Supergroup in Canada (Jefferson & Parrish 1989; Harlan et al. 2003) and a c. 811.5 Ma age for a bedded tuff in the Fifteenmile Group of the Coal Creek Inlier, Ogilvie Mountains, northwestern Canada (Macdonald et al. 2010), which lies just beneath the extended interval of negative d13C values. It also corresponds with a rise in seawater 87Sr/86Sr from 0.7055–0.7057 to 0.7062– 0.7063. A third independent line of evidence for the uniqueness of this part of the Cryogenian record is the first record of C. buickii. In Spitsbergen, C. buickii first occurs in the upper parts of the Lower Dolomite Member of the Svanbergfjellet Formation (Butterfield et al. 1994), which is in the uppermost part of the extended interval of negative d13C values between c. 830 and 800 Ma (intervals B and D in Australia; Fig. 8.10). In Australia, the first record of C. buickii is in slightly younger rocks but there are no known suitable siltstone facies in Australia at this time within which C. buickii is typically found. In the Grand Canyon, USA, the first record of C. buickii is in the lowermost Chuar

120o

140o

130o

150

o

KIMBERLEY REGION GEORGINA

o

AMADEUS BASIN

20

ADELAIDE RIFT COMPLEX

o

BASIN e Arunta Provinc

Western Officer Basin

Composite field and drillhole sections

LS

W1

13

30o

PEAKE & DENISON RANGES

Eastern Officer Basin

(m)

SC Gawler Craton

STUART SHELF

M

B2 O EC DC

30

Thickness

o

El Y

ADELAIDE RIFT COMPLEX

1000 km

KING ISLAND

40

F

130o

140o

150

o

STUART SHELF SCYW 1a drillhole 13 C (‰) Depth

M

Whyalla Sandstone

Wallara-1 drillhole 13

L

Pioneer Sandstone

Areyonga Formation

Ar

Depth (m) -4 1281

1300

1400

-2

C (‰) 0

2

4

L

K

Appila Tapley Hill Formation Tillite TSM

CENTRAL AMADEUS BASIN 100

A

0

2

4

6

BL

200

0

(m) -2 1140

1200

M

Etina Formation

UMBERATANA GROUP

Ringwood Member

400

300

Unamed member

Aralka Formation

?

4

6

8

10

-8

-6

-4

-2

0

2

4

6

8

-8

-6

-4

-2

0

2

4

6

8

10

Key

10

500

N

Sandstone

Glacial diamictite

Mudstone–siltstone

Diamictite

Dolomite

Ooids

Limestone

0 Brighton Limest.

4

562

500

2

O

Thickness -10 130

-8

-6

-4

-2

0

2

4

6

8

Missing section

10

Unconformity

100

13

Cdol (‰ VPDB)

0 Depth

-10 -8

0

-6

-4

-2

0

2

4

6

8

10 13

Ccarb (‰ VPDB) estimated from +

WD

1450

Corg

M 1000

NW TASMANIA Forest-1 drillhole

1300

1400

13

500 MCGM

2

13

L K

TSM

L K 2031

Depth (m) -6 790

Kanunnah Subgroup

1500

Wilyerpa Formation

0

Tapley Hill Formation

C (‰)

-2

0

0

880

13

-4

390

Thickness -10

Limbla Syncline field section -6

-2

AUSTRALIAN CRYOGENIAN CORRELATIONS

NE AMADEUS BASIN Thickness (m)

-4

P Thickness -10

o

TASMANIA 120o

-6

500

0

Drillhole Field section

o

Enorama Shale

40

C (‰)

-10 -8

800

JRM

OFFICER BASIN

TOGARI GP BRD

Musgrave Province

Trezona Fm.

20

900

C (‰)

-4

-2

0

2

4

L K

975

T.D. 1075m

1425

Johnnys Creek beds

127

Fig. 8.11. Sturt glaciation and post-Sturt glaciation d13C stratigraphy of the Australian Cryogenian. Location of drill hole and outcrop sections shown in inset (see Fig. 8.4 for explanation of abbreviations). See text for an explanation of d13C correlation intervals K–P. Stratigraphy and d13C data from the Amadeus Basin and Stuart Shelf (Walter et al. 2000), the Adelaide Rift Complex (Walter et al. 2000; McKirdy et al. 2001), and northwestern Tasmania (Calver 1995, 1998). In the Adelaide Rift Complex the following sections were used: drill hole Blinman 2 for the Wilyerpa (Sturt Tillite correlative) and Tapley Hill Formations; Depot Flat field section for the Brighton Limestone; field section east of the Oraparinna Diapir for the Etina Formation; Mallee Water field section for the Enorama Shale; Doodney’s Well Hills field section for the Trezona Formation. Abbreviations: A, Areyonga Formation; Ar, Aralka Formation; BL, Brighton Limestone; BRD, Black River Dolomite; El, Elatina Formation; JRM, Julius River Member; MCGM, Mount Caernarvon Greywacke Member; TSM, Tindelpina Shale Member; WD, Wockerawirra Dolomite; Y, Yudnamutana Subgroup.

AUSTRALIA

-6

-4

-2

2

4

6

-8

8

0.7071

650

-4

-2

2

4

6

0.7063

-8

8

-6

Sr/86Sr (0.70..)

13

60 70 80

0

-4

-2

2

4

6

8

-8

-6

60 70 80

13

60 70 80 0

-4

-2

0

2

4

6

-8

8

-6

13

-4

-2

0

2

4

6

8

W 0.7071

TW

0.7071

0.7072

0.7071

0.7068

0.7068

Rapit

S ,

-6

KEELE

T

6 7 8 5

13

60 70 80 0

MONGOLIA Sr/86Sr (0.70..)

87

87

ELBO

-8

Sr/86Sr (0.70..)

128

13

60 70 80

NAMIBIA

87

Sr/86Sr (0.70..)

T

Sr/86Sr (0.70..)

AGE (Ma) 635

SVALBARD

CANADA 87

87

675

CLG

T 0.7064

DRAKEN BACKLUNDTOPPEN ELBO

, 700

Wilyerp 725 illite

4

S Kalachalp 750

Kanp

USA 13

SVANBERGFJ

((

Johnnys ,

3 2

*

-6

-4

-2

0

2

4

6

8

-8

-6

-4

-2

0

2

4

6

8

0.7069

*

775

800

-8

0.7062

-8

*

-6

-4

-2

0

2

4

6

K. GREY ET AL.

’

0.7067

8

60 70 80

0.7063 0.7062

GP

825

0.7057

850

-8 -8

-6

60 70 80

-4

-2

0

2

4

6

8

-6

-4

-2

0

2

4

6

13

Ccarb (‰)

13

Corg (‰) C. buickii 1st record

* B. burra 1st record

((

Siliciclastic Carbonate/siliciclastic Carbonate Glacial

1

8

60 70 80 0.7055 0.7055

-8

-6

-4

-2

0

2

4

6

8

60 70 80

Fig. 8.12. Global d13C, 87Sr/86Sr and biostratigraphic correlation of the Australian Cryogenian. The circled numbers on the time scale correspond to geochronological ages in Figure 8.5. Australia: 87Sr/86Sr ratios from Fanning (1986) and reproduced in Hill et al. (2000b) (Gillen Member, unpublished data) (Unit 3, Loves Creek Member), Calver (1998) (Julius River Member, Black River Dolomite), and McKirdy et al. (2001) (Brighton Limestone and Trezona Formation). Canada: d13C data from Halverson (2006) (Little Dal Group and Coates Lake Group), and Hoffman & Schrag (2002) and James et al. (2001) (Windemere Supergroup). 87Sr/86Sr ratios from Narbonne et al. (1994), Kaufman et al. (1997) and Halverson et al. (2007b). c. 780 Ma age for the Little Dal Basalt (Jefferson & Parrish 1989; Harlan et al. 2003). Abbreviations: CLG, Coates Lake Group; Tw, Twitya Formation; Keele, Keele Formation. Svalbard: d13C data from Halverson et al. (2005, 2007a). 87Sr/86Sr ratios from Halverson et al. (2005, 2007b). First record of Cerebrosphaera buickii, upper Lower Dolomite Member, Svanbergfjellet Formation (Butterfield et al. 1994). Abbreviations: Elbo, Elbobreen Formation. Namibia: d13C data from Halverson et al. (2005). 87Sr/86Sr ratios from Yoshioka et al. (2003) and Halverson et al. (2007b). 746 Ma age from Hoffman et al. (1996) and 760 Ma age from Halverson et al. (2005). Mongolia: d13C and 87Sr/86Sr data from Shields et al. (2002). USA: d13C data from Dehler et al. (2005). First record of Cerebrosphaera buickii, Tanner Member, Galeros Formation, Chuar Group (S. Porter and Nagy, unpublished data). 770 Ma age from (Williams et al. 2003) and 742 Ma age from Karlstrom et al. (2000).

AUSTRALIAN CRYOGENIAN CORRELATIONS

Group, which has a maximum age of c. 770 Ma (Dehler et al. 2005, and references therein) and is therefore consistent with the approximate ages of the first records in Australia and Spitsbergen. Only limited palynological studies have been carried out on the Mackenzie Mountains successions, and have been mainly on Ediacaran to Cambrian samples (Baudet et al. 1989), but the Wynniatt Formation of the Shaler Supergroup in the Minto Inlier has 13 acritarch species (Butterfield & Rainbird 1998). A synthesis of lithostratigraphic studies of arctic Canada reveals close similarities between the Mackenzie Mountains and Shaler Supergroups, and more specifically between the Rusty Shale and Wynniatt Formation (Rainbird et al. 1996). The absence of C. buickii in the Wynniatt Formation, whose probable correlative, the Rusty Shale, is just beneath the extended interval of negative d13C values between c. 830 and 800 Ma, is consistent with the stratigraphic level of C. buickii in Spitsbergen. Although further palynology in Canada above this stratigraphic level is necessary to determine exactly where C. buickii first appears, the first record of this acritarch is globally consistent. A fourth, and potentially more controversial line of evidence is the first record of the B. burra Assemblage, which across Australia consistently first appears at about the same stratigraphic level as C. buickii. The stromatolite Baicalia has been reported in Svalbard (Halverson et al. 2007b) and the USA (Dehler et al. 2005), just above the first appearance of C. buickii, but has not been identified to form level. Together, d13C, 87Sr/86Sr and the first records of C. buickii and B. burra have the potential to improve global correlation and possibly inform a chronostratigraphic definition of the Cryogenian Period, and that c. 800 Ma could be a potential lower age limit for the co-occurrence of significant geochemical and palaeontological anomalies. The authors recognize, however, that more research, in particular on the palaeontological record, needs to be completed to strengthen these arguments.

Australia – Svalbard – Namibia – USA: 800– 700 Ma d13C values in this period are mostly moderately 13C-enriched with several short-lived intervals to between –2 and 0‰. The distinguishing feature of the d13C record is the occurrence of values of about þ8 to þ10‰ in Australia, Svalbard, Namibia, and possibly the USA. (The Chuar Group record is problematic because it is predominantly based on d13Corg. None of the d13Ccarb values exceed þ4‰ and many of the ‘calculated’ (d13Ccarb ¼ 1TOC – d13Corg) d13Ccarb values exceed þ10‰, the maximum known values in carbonate successions from Svalbard and Namibia.) A radiometric age of 760 Ma in Namibia places a maximum age constraint on the d13C peak. In Svalbard and Namibia there are two stratigraphic levels of þ8 to þ10‰ values, which makes precise correlation with Australia more ambiguous than lower in the succession. In the Coates Lake Group of Canada there are extremely 13 C-depleted d13C values that may correlate with the lower Polarisbreen Group of Svalbard, but it is difficult to be certain because in the Coates Lake Group d13C values rapidly shift to þ8‰ and there is no record of this in Svalbard or elsewhere. An equivalent magnitude d13C shift occurs between the two large Cryogenian glaciations on all continents; however, 87Sr/86Sr values of 0.7064 in the Coates Lake Group strongly suggest it is a pre-Sturt glacial unit. The 87Sr/86Sr record in Svalbard suggests a rise in values from 0.7062 to 0.7069 concomitant with the rise in d13C values to their peak of þ8 to þ10‰. In Svalbard, the 87Sr/86Sr values then decrease to 0.7067. Ratios of 0.7064 in the Coates Lake Group and 0.7063 in the Sturt glaciation in Tasmania suggest a further decline in seawater 87Sr/86Sr. A similar rise then decline in seawater 87Sr/86Sr is known in Greenland (Fairchild et al. 2000) and, more recently, Scotland (Sawaki et al. 2010), so a stratigraphic horizon at c. 760 Ma could be a strong candidate for chronostratigraphic definition of the Cryogenian Period because of the coincidence of maximum pre-Sturt glaciation

129

peaks in both d13C and 87Sr/86Sr within the stratigraphic range of C. buickii and B. burra.

Australia – Canada – Svalbard – Namibia – Mongolia: 660– 635 Ma The ‘interglacial’ part of the Cryogenian succession has a remarkably consistent d13C and 87Sr/86Sr record. In immediately postSturt glaciation sediments, d13C values mostly fall within the range of –4 to –2‰, with the exceptions of Australia ( –6 to –4‰) and Svalbard ( –1 to 0‰). With the exception of Svalbard (d13C between þ6 and þ7), d13C values then rise rapidly in all countries to between þ8 and þ10‰, and even higher in Mongolia. 87 Sr/86Sr values rise concomitantly from 0.7068 to 0.7071. In Australia, Canada and Namibia, d13C values then rapidly decline to between –6 and –10‰ (the Trezona ‘anomaly’; Halverson et al. 2005). These interglacial peaks and troughs are short-lived compared to the Ediacaran Shuram anomaly, and are more likely to represent a global ocean signal (Bristow & Kennedy 2008, and references therein). Unless evidence is presented to the contrary, the stratigraphic level at which the maximum interglacial peaks in d13C and 87Sr/86Sr occur is a candidate for further subdivision of the Cryogenian Period, between c. 650 and 635 Ma. A stratigraphic level within or close to the lithological boundaries of the Sturt glaciation is not appropriate because of the apparent range in radiometric ages for the Sturt glaciation globally (Fanning & Link 2004, 2008).

Conclusions The thick and well-preserved Cryogenian successions in several Australian basins can be correlated using a variety of means, and these techniques are most effectively used in combination. However, poor preservation or lack of appropriate facies often limits the applicability of biostratigraphy and chemostratigraphy, and the latter is often ambiguous owing to the non-uniqueness of patterns of variation. Chronometric constraints are particularly important, especially in testing global correlations, but such constraints are regrettably few. Recent dating shows that the Sturt glaciation ended later (c. 660 Ma: Kendall et al. 2007; Fanning & Link 2008) than presumed correlatives on other continents. That the same may be true of the Elatina glaciation is suggested (but not proved) by dating from Tasmania (Calver et al. 2004). More dates are sorely needed, including a focus on the age of the Elatina glacial, the close of which marks the end of the Cryogenian Period. As was the case with the base of the Ediacaran, it is probably not advisable to place the lower Cryogenian boundary at a horizon associated with glaciation (in this case, the Sturt glaciation), which is marked by major disconformities and unconformities. Additionally, although the Sturt glaciation appears to be consistent in age across Australia, it may not be synchronous with glacial units elsewhere that are normally correlated with it. The A. australica and B. burra Stromatolite Assemblages are Australia-wide markers, and unpublished data suggest they may be present on other continents. However, the global distributions of these readily correlatable Australian taxa have yet to be demonstrated. The first appearance of the acritarch C. buickii, coupled with the first appearance of the stromatolite B. burra, appears to be a significant marker, possibly globally, and should be considered for determining the positioning of a boundary. When global correlations are attempted with successions in Canada, Svalbard, Namibia, the USA and Mongolia there are consistencies between d13C and 87Sr/86Sr excursions beneath the Sturt glaciation (c. 660 Ma in Australia) and between the Sturt and Elatina glaciations, and the approximate ages at which these anomalies occur in other global successions are consistent with

130

K. GREY ET AL.

the Australian time scale. It thus appears that Cryogenian d13C and 87 Sr/86Sr excursions are of global extent and probably synchronous and, together with the traditional tool of period definition, first and last appearances of fossils, provides a way forward for further subdivision of the Neoproterozoic, at the base of the Cryogenian Period. There are two stratigraphic levels where it may be possible to make a chronostratigraphic definition of the base of the Cryogenian Period. The first is in rocks of c. 800 Ma (based on radiometric ages in Australia and Canada), where there is an interval of negative d13C values ( –4 to 0‰) that coincides with a rise in 87 Sr/86Sr ratios from 0.7055 to 0.7062 and the first appearance of the distinctive acritarch C. buickii. Three independent lines of evidence make this stratigraphic level a strong candidate for basal definition of the Cryogenian Period. The second is at c. 760 Ma (a radiometric age from Namibia), where the pre-Sturt glaciation d13C record peaks at about þ8‰ in Australia, Svalbard, Namibia and the USA. 87Sr/86Sr ratios from Svalbard show a concomitant increase from 0.7062 to 0.7069, and then a later decrease to 0.7064 (Coates Lake Group, Canada). There is no equivalent record of the 87Sr/86Sr rise from 0.7062 to 0.7069 in other countries due to incompatible lithologies, which weakens the case for boundary placement at this stratigraphic level. These age rocks coincide with the stratigraphic ranges of C. buickii and B. burra. The consistency of the data in numerous sections in Australia, and limited reports of their occurrence elsewhere, indicate that these two levels have potential for global boundary placement. The synchroneity of glaciations in Australia and elsewhere in the world remains unresolved, with new dates from NW Canada conflicting with the limited data from Australia (Kendall et al. 2009). Biostratigraphic data through this interval is also sparse, but between the Sturt and Elatina glaciations (c. 660–635 Ma) there is the possibility for further subdivision of the Cryogenian Period at a stratigraphic level that coincides with a peak in d13C of between þ8 and þ10‰ and 87Sr/86Sr of between 0.7071 and 0.7072.

Appendix 1 Taxonomic citations The names of authors of scientific names have been omitted in the text and text figures and instead are presented here. The names are the names of authors of taxa, not references, and therefore they are not necessarily cited in the reference list.

Spheroidal microfossils and acritarchs Cerebrosphaera ananguae Cotter 1999 Cerebrosphaera buckii Butterfield 1994 in Butterfield et al. 1994 Chuaria sp. cf. Ch. circularis Walcott 1899; emend. Vidal and Ford 1985 Coneosphaera sp. Luo 1991 Cymatiosphaeroides kullingii Knoll 1984, emend. Knoll et al. 1991 Eoentophysalis croxfordii (Muir 1976) Butterfield 1994 Eomicrocystis sp. cf. E. elegans Golovenok and Belova 1984 Leiosphaeridia crassa (Naumova 1949) Jankauskas 1989 Leiosphaeridia jacutica (Timofeev 1966) emend. Mikhailova and Jankauskas in Jankauskas et al. 1989 Leiosphaeridia minutissima (Naumova 1949) emend. Jankauskas in Jankauskas et al. 1989 Leiosphaeridia sp. cf. L. exsculpta (Timofeev 1969) emend. Mikhailova in Jankauskas et al. 1989 Leiosphaeridia tenuissima Eisenack 1958 Leiosphaeridia ternata (Timofeev 1966) emend. Mikhailova in Jankauskas et al. (1989) Melanocyrillium Bloeser 1979 ex Bloeser 1985 Myxococcoides cantabrigiensis Knoll 1982

Ostiana microcystis Hermann 1976 Pterospermopsimorpha insolita Timofeev 1969 Satka colonialica Jankauskas 1979 Simia annulare Timofeev 1969 emend. Mikhailova in Jankauskas et al. (1989) Stictosphaeridium sinapticuliferum Symplassosphaeridium Timofeev 1959 ex Timofeev 1969 Synsphaeridium Eisenack 1965 Tasmanites Newton 1875 Trachyhystrichosphaera Timofeev & German 1976 emend. Jankauskas et al. 1989 Valeria lophostriata (Jankauskas), Jankauskas 1982

Filamentous microfossils Calyptothrix sp. cf. Ca. alternata Jankauskas 1980 Clavitrichoides rugosus Mikhailova in Jankauskas et al. 1989 Oscillatoriopsis amadeus (Schopf & Blacic 1971) Butterfield 1994 Siphonophycus kestron Schopf 1968 Siphonophycus robustum (Schopf 1968) Knoll et al. 1991 Siphonophycus septatum (Schopf 1968) Knoll et al. 1991 Siphonophycus solidum (Golub 1979) emend. Butterfield 1994 Siphonophycus typicum (Hermann 1974) Butterfield et al. 1994

Stromatolites Acaciella australica (Howchin 1914) Walter 1972 Baicalia burra Preiss 1972 Basisphaera irregularis Walter 1972 Boxonia pertaknurra Walter 1972 Conophyton Maslov 1937 Elleria minuta Walter and Krylov in Walter et al. 1979 Inzeria intia Walter 1992 Inzeria multiplex Preiss 1973 Jurusania nisvensis Raaben 1964 Kulparia alicia (Cloud and Semikhatov 1969) Walter 1972 Linella avis Krylov 1967 Linella munyallina Preiss 1974 Tungussia Semikhatov 1962 Tungussia wilkatanna Preiss 1974 The ideas and figures presented in this manuscript have benefited from discussions with numerous researchers, but in particular J. Gehling, P. Haines, R. Hocking, W. Preiss and G. Shields. Field samples and data have been collected by many geologists for more than 100 years, especially M. Walter, W. Preiss, I. Williams, N. Planavsky and M. Stevens. K.G. publishes with the permission of the Executive Director of the Geological Survey of Western Australia and is an Honorary Associate at Monash University and the University of Western Australia. This paper is a contribution to International Geological Correlation Program 512: Neoproterozoic Ice Ages.

References Ambrose, G. J., Flint, R. B. & Webb, A. W. 1981. Precambrian and Paleozoic geology of the Peake and Denison Ranges. Geological Survey of South Australia Bulletin 50. Bagas, L., Grey, K. & Williams, I. R. 1995. Reappraisal of the Paterson Orogen and Savory Basin. Western Australia Geological Survey Annual Review 1994–95, 55 – 63. Bagas, L., Grey, K., Hocking, R. M. & Williams, I. R. 1999. Neoproterozoic successions of the northwestern Officer Basin: a reappraisal. Western Australia Geological Survey Annual Review 1998– 99, 39 – 44. Bagas, L., Camacho, A. & Nelson, D. R. 2002. Are the Neoproterozoic Lamil and Throssell Groups of the Paterson Orogen allochthonous? Geological Survey of Western Australia Annual Review, 2000/01, 45 – 52. Baudet, D., Aitken, J. D. & Vanguestaine, M. 1989. Palynology of uppermost Proterozoic and lowermost Cambrian Formations,

AUSTRALIAN CRYOGENIAN CORRELATIONS

central Mackenzie Mountains, northwestern Canada. Canadian Journal of Earth Sciences, 26, 129–148. Black, L. P., Calver, C. R., Seymour, D. B. & Reed, A. 2004. SHRIMP U –Pb detrital zircon ages from Proterozoic and Early Palaeozoic sandstones and their bearing on the early geological evolution of Tasmania. Australian Journal of Earth Sciences, 51, 885–900. Blake, D. H., Tyler, I. M. & Sheppard, S. 1998. Geology of the Ruby Plains 1:100 000 sheet area (4460), Western Australia. Australian Geological Survey Organisation, Canberra. Bristow, T. F. & Kennedy, M. J. 2008. Carbon isotope excursions and the oxidant budget of the Ediacaran atmosphere and ocean. Geology, 36, 863–866. Butterfield, N. J. & Rainbird, R. H. 1998. Diverse organic-walled fossils, including ‘possible dinoflagellates’, from the early Neoproterozoic of arctic Canada. Geology, 26, 963– 966. Butterfield, N. J., Knoll, A. H. & Swett, K. 1994. Paleobiology of the Neoproterozoic Svanbergfjellet Formation, Spitsbergen. Fossils and Strata, 34, 84. Calver, C. R. 1995. Ediacarian Isotope Stratigraphy of Australia. PhD thesis, Macquarie University, Sydney, Australia. Calver, C. R. 1998. Isotope stratigraphy of the Neoproterozoic Togari Group, Tasmania. Australian Journal of Earth Sciences, 45, 865– 874. Calver, C. R. 2000. Isotope stratigraphy of the Ediacarian (Neoproterozoic III) of the Adelaide Rift Complex, South Australia, and the overprint of water column stratification. Precambrian Research, 100, 121– 150. Calver, C. R. & Lindsay, J. F. 1998. Ediacarian sequence and isotope stratigraphy of the Officer Basin, South Australia. Australian Journal of Earth Sciences, 45, 513–532. Calver, C. R. & Walter, M. R. 2000. The late Neoproterozoic Grassy Group of King Island, Tasmania: correlation and palaeogeographic significance. Precambrian Research, 100, 299– 312. Calver, C. R., Black, L. P., Everard, J. L. & Seymour, D. B. 2004. U –Pb zircon age constraints on late Neoproterozoic glaciation in Tasmania. Geology, 32, 893– 896. Close, D. F., Edgoose, C. J. & Scrimgeour, I. R. 2005. The Tjauwata Group: a late Mesoproterozoic rift succession underlying the southwestern Amadeus Basin. Central Australian Basins Symposium: Petroleum and Minerals Potential, 16 –18 August 2005, Northern Territory Geological Survey, 39 – 40. Coats, R. P. & Preiss, W. V. 1980. Stratigraphic and geochronological reinterpretation of Late Proterozoic glaciogenic sequences in the Kimberley Region, Western Australia. Precambrian Research, 13, 181– 208. Condon, D., Zhu, M., Bowring, S., Wang, W., Yang, A. & Jin, Y. 2005. U –Pb ages from the Doushantuo Formation, China. Science, 308, 95 – 98. Corkeron, M. 2007. ‘Cap carbonates’ and Neoproterozoic glacigenic successions from the Kimberley region, north-west Australia. Sedimentology, 54, 871– 903. Corkeron, M. L. & George, A. D. 2001. Glacial incursion on a Neoproterozoic carbonate platform in the Kimberley. Geological Society of America Bulletin, 113, 1121– 1132. Cotter, K. L. 1997. Neoproterozoic microfossils from the Officer Basin, Western Australia. Alcheringa, 21, 247–270. Cotter, K. L. 1999. Microfossils from Neoproterozoic Supersequence 1 of the Officer Basin, Western Australia. Alcheringa, 23, 63 – 86. Cowley, W. M. & Flint, R. B. 1993. Epicratonic igneous rocks and sediments. In: Drexel, J. F., Preiss, W. V. & Parker, A. J. (eds) The Geology of South Australia, Volume 1, The Precambrian. South Australian Geological Survey Bulletin, 54, 142– 147. Dehler, C. M., Elrick, M., Bloch, J. D., Crossey, L. J., Karlstrom, K. E. & Des Marais, D. J. 2005. High-resolution d13C stratigraphy of the Chuar Group (ca. 770– 742 Ma), Grand Canyon: Implications for mid-Neoproterozoic climate change. Geological Society of America Bulletin, 117, 32 –45. Dow, D. B. & Gemuts, I. 1969. Geology of the Kimberley Region, Western Australia: The East Kimberley. Geological Survey of Western Australia Bulletin, 120, 135. Drexel, J. F. 2009. Review of the Burra Mine Project, 1980– 2008 – a progress report. South Australia. Department of Primary Industries and Resources Report Book, 2008/16, 76.

131

Dunster, J. N., Kruse, P. D., Duffett, M. L. & Ambrose, G. J. 2007. Geology and resource potential of the southern Georgina Basin. Northern Territory Geological Survey, Digital Information Package, DIP007. Eyles, C. H., Eyles, N. & Grey, K. 2007. Palaeoclimate implications from deep drilling of Neoproterozoic strata in the Officer Basin and Adelaide Rift Complex of Australia: a marine record of wet-based glaciers. Palaeogeography, Palaeoclimatology, Palaeoecology, 248, 291–312. Fairchild, I. J., Spiro, B., Herrington, P. M. & Song, T. 2000. Controls on Sr and C isotope compositions of Neoproterozoic Sr-rich limestones of East Greenland and North China. In: Grotzinger, J. P. & James, N. P. (eds) Carbonate Sedimentation and Diagenesis in the Evolving Precambrian World. SEPM Special Publication, 67, 297– 313. Fairchild, T. R. 1976. The geological setting and palaeobiology of a late Precambrian stromatolitic microflora from South Australia. PhD thesis, University of California, USA. Fanning, C. M. 1986. 87Sr/86Sr of gypsum/anhydrite and carbonate samples. Australian Mineral Development Laboratories Report G 6696/86 (unpublished). Fanning, C. M. & Link, P. K. 2004. U– Pb SHRIMP ages of the Neoproterozoic (Sturtian) glaciogenic Pocatello Formation, southeastern Idaho. Geology, 32, 881–884. Fanning, C. M. & Link, P. K. 2008. Age constraints for the Sturtian Glaciation; data from the Adelaide Geosyncline, South Australia and Pocatello Formation, Idaho, USA. In: Gallagher, S. J. & Wallace, M. W. (eds) Neoproterozoic Extreme Climates and the Origin of Early Metazoan Life. Selwyn Symposium of the GSA Victoria Division, September 2008, Geological Society of Australia Extended Abstracts, 91, 57– 62. Fanning, C. M., Ludwig, K. R., Forbes, B. G. & Preiss, W. V. 1986. Single and multiple grain U– Pb zircon analyses for the early Adelaidean Rook Tuff, Willouran Ranges, South Australia. Geological Society of Australia Abstracts, 15, 255– 304. Fensome, R. A., Williams, G. L., Barss, M. S., Freeman, J. M. & Hill, J. M. 1990. Acritarchs and fossil prasinophytes: an index to genera, species and infraspecific taxa. American Association of Stratigraphic Palynologists Foundation, Contributions Series 25, 771. GEOSCIENCE AUSTRALIA, Stratigraphic Names database. World Wide Web address: http://dbforms.ga.gov.au/pls/www/geodx. strat_units.int Glikson, A. Y., Stewart, A. J. et al. 1996. Geology of the western Musgrave Block, central Australia, with particular reference to the mafic– ultramafic Giles Complex. Australian Geological Survey Organisation Bulletin, 239, 206. Gorter, J., Grey, K. & Hocking, R. 2007. The petroleum exploration potential of the Australian Infracambrian (Ediacaran) of the Amadeus and Officer basins. Australian Petroleum Production and Exploration Association Journal, 47, 391–392 (extended abstract). Gradstein, F. M., Ogg, J. G., Smith, A. G., Bleeker, W. & Lourens, A. J. 2004. A new geologic time scale, with special reference to Precambrian and Neogene. Episodes, 27, 83 – 100. Grey, K. 1995. Neoproterozoic stromatolites from the Skates Hills Formation, Savory Basin, Western Australia, and a review of the distribution of Acaciella australica. Australian Journal of Earth Sciences, 42, 123– 132. Grey, K. 1999a. Appendix 7: Proterozoic palynology of samples from Empress 1A. In: Stevens, M. K. & Apak, S. N. GSWA Empress 1 and 1A well completion report, Yowalga Sub-basin, Officer Basin, Western Australia. Western Australia Geological Survey Record, 1999/4, 68 –69. Grey, K. 1999b. Appendix 8: Proterozoic stromatolite biostratigraphy of Empress 1A. In: Stevens, M. K. & Apak, S. N. GSWA Empress 1 and 1A Well Completion Report, Yowalga Sub-basin, Officer Basin, Western Australia. Western Australia Geological Survey Record, 1999/4, 70 –72. Grey, K. 2005a. Preliminary report on the Proterozoic biostratigraphy, Lancer 1, Officer Basin, Western Australia. In: Mory, A. J. & Haines, P. W. (eds) Lancer 1 Well Completion Report (interpretive papers), Officer and Gunbarrel Basins, Western Australia. Western Australia Geological Survey Record, 2005/4, 81.

132

K. GREY ET AL.

Grey, K. 2005b. Ediacaran palynology of Australia. Memoir of the Association of Australasian Palaeontologists, 31, 439. Grey, K. & Blake, D. 1999. Neoproterozoic (Cryogenian) stromatolites from the Wolfe Basin, east Kimberley, Western Australia: correlation with the Centralian Superbasin. Australian Journal of Earth Sciences, 46, 329– 341. Grey, K. & Corkeron, M. 1998. Late Neoproterozoic stromatolites in glacigenic successions of the Kimberley region, Western Australia: evidence for a younger Marinoan glaciation. Precambrian Research, 92, 65 –87. Grey, K. & Cotter, K. L. 1996. Palynology in the search for Proterozoic hydrocarbons. Western Australia Geological Survey Annual Review, 1995– 96, 70– 80. Grey, K. & Stevens, M. K. 1997. Neoproterozoic palynomorphs of the Savory Sub-basin, Western Australia, and their relevance to petroleum exploration. Western Australia Geological Survey Annual Review, 1996– 97, 49– 54. Grey, K., Hocking, R. M. et al. 2005. Lithostratigraphic nomenclature of the Officer Basin and correlative parts of the Paterson Orogen, Western Australia. Western Australia Geological Survey Report, 93, 89. Griffin, B. J. & Preiss, W. V. 1976. The significance and provenance of stromatolitic clasts in a probable Late Precambrian diamictite in northwestern Tasmania. Proceedings of the Royal Society of Tasmania Papers, 110, 111–127. Haines, P. W. 1988. Storm-dominated mixed carbonate/siliciclastic shelf sequence displaying cycles of hummocky cross-stratification, Late Proterozoic Wonoka Formation, South Australia. Sedimentary Geology, 58, 237–254. Haines, P. W. 1990. A late Proterozoic storm-dominated carbonate shelf sequence: the Wonoka Formation in the central and southern Flinders Ranges, South Australia. In: Jago, J. B. & Moore, P. S. (eds) The Evolution of a Late Precambrian –Early Palaeozoic Rift Complex: Adelaide Geosyncline. Geological Society of Australia Special Publication, 16, 177– 198. Haines, P. W., Mory, A. J., Stevens, M. K. & Ghori, K. A. R. 2004. GSWA Lancer 1 well completion report (basic data), Officer and Gunbarrel Basins, Western Australia. Geological Survey of Western Australia Record, 2004/10, 39. Haines, P. W., Hocking, R. M., Grey, K. & Stevens, M. K. 2008. Vines 1 revisited: are older Neoproterozoic glacial deposits preserved in Western Australia? Australian Journal of Earth Sciences, 55, 397– 406. Halverson, G. 2006. A Neoproterozoic chronology. In: Xiao, S. & Kaufman, A. (eds) Neoproterozoic Geobiology and Paleobiology. Topics in Geobiology 27. Springer, New York, 231–271. Halverson, G. P., Hoffman, P., Schrag, D. P., Maloof, A. C. & Rice, A. 2005. Towards a Neoproterozoic composite carbon isotope record. Geological Society of America Bulletin, 117, 1181– 1207. ¨ ., Maloof, A. C. & Bowring, S. A. Halverson, G. P., Duda´s, F. O 2007a. Evolution of the 87Sr/86Sr composition of Neoproterozoic seawater. Palaeogeography, Palaeoclimatology, Palaeoecology, 256, 103– 129. ¨. & Halverson, G. P., Maloof, A. C., Schrag, D. P., Duda´s, F. O Hurtgen, M. 2007b. Stratigraphy and geochemistry of a ca. 800 Ma negative carbon isotope interval in northeastern Svalbard. Chemical Geology, 237, 23 –45. Harlan, S., Heaman, L., LeCheminant, A. & Premo, W. 2003. Gunbarrel mafic magmatic event: a key 780 Ma time marker for Rodinia plate reconstructions. Geology, 31, 1053–1056. Hill, A. C. 2005. Stable isotope stratigraphy, GSWA Lancer 1, Officer Basin, Western Australia. In: Mory, A. J. & Haines, P. W. (eds) GSWA Lancer 1 Well Completion Report (Interpretive Papers), Officer and Gunbarrel Basins, Western Australia. Geological Survey of Western Australia Record, 2005/4, 1– 11. Hill, A. C. & Walter, M. R. 2000. Mid-Neoproterozoic (830–750 Ma) isotope stratigraphy of Australia and global correlation. Precambrian Research, 100, 181– 211. Hill, A. C., Cotter, K. L. & Grey, K. 2000a. Mid-Neoproterozoic biostratigraphy and isotope stratigraphy in Australia. Precambrian Research, 100, 281– 298.

Hill, A. C., Arouri, K., Gorjan, P. & Walter, M. R. 2000b. Geochemistry of marine and non-marine environments of a Neoproterozoic cratonic carbonate/evaporite: the Bitter Springs Formation, Central Australia. In: Grotzinger, J. P. & James, N. P. (eds) Carbonate Sedimentation and Diagenesis in the Evolving Precambrian. SEPM Special Publication, 67, 327–344. Hill, A. C., Haines, P. W., Grey, K. & Willman, S. 2008. New records of Ediacaran Acraman ejecta in drill holes from the Stuart Shelf and Officer Basin, South Australia. Meteoritics and Planetary Science, 42, 1883– 1891. Hill, A. C., Haines, P. W. & Grey, K. 2011. Neoproterozoic glacial deposits of central Australia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 677– 691. Hoffmann, K.-H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. U– Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817– 820. Hoffman, P. & Schrag, D. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129–155. Hoffman, P., Hawkins, D., Isachsen, C. & Bowring, S. 1996. Precise U–Pb zircon ages for early Damaran magmatism in the Summas Mountains and Welwitschia Inlier, northern Damara belt, Namibia. Communications of the Geological Survey of Namibia, 11, 47 – 52. Hoffman, P. F., Calver, C. R. & Halverson, G. P. 2009. Cottons Breccia of King Island, Tasmania: glacial or non-glacial, Cryogenian or Ediacaran? Precambrian Research, 172, 311– 322. Ireland, T. R., Flo¨ttmann, T., Fanning, C. M., Gibson, G. M. & Preiss, W. V. 1998. Development of the Early Paleozoic Pacific Margin of Gondwana from detrital zircon ages across the Delamerian Orogen. Geology, 26, 243– 246. Jackson, M. J. & Muir, M. D. 1981. The Babbagoola Beds, Officer Basin, Western Australia: correlations, micropalaeontology, and implications for petroleum prospectivity. BMR Journal of Australian Geology and Geophysics, 6, 81 – 93. Jackson, M. J. & van de Graaff, W. J. E. 1981. Geology of the Officer Basin. Bureau of Mineral Resources Bulletin, 206, 1 – 102. James, N., Narbonne, G. & Kyser, T. 2001. Late Neoproterozoic cap carbonates: Mackenzie Mountains, northwestern Canada: precipitation and global glaciation. Canadian Journal of Earth Science, 38, 1229– 1262. Jefferson, C. & Parrish, R. 1989. Late Proterozoic stratigraphy, U/Pb zircon ages and rift tectonics, Mackenzie Mountains, northwestern Canada. Canadian Journal of Earth Sciences, 26, 1784– 1801. Karlstrom, K. E., Bowring, S. A. et al. 2000. Chuar Group of the Grand Canyon: Record of breakup of Rodinia, associated change in the global carbon cycle, and ecosystem expansion by 740 Ma. Geology, 28, 619–622. Kaufman, A., Knoll, A. & Narbonne, G. 1997. Isotopes, ice ages and terminal Proterozoic Earth history. Proceedings of the National Academy of Sciences, 95, 6600–6605. Kendall, B., Creaser, R. A. & Selby, D. 2006. Re– Os geochronology of postglacial black shales in Australia: consequences for timing of the Sturtian glaciation. Geology, 34, 729–732. Kendall, B., Creaser, R. A., Calver, C. R., Raub, T. D. & Evans, D. A. D. 2007. Neoproterozoic paleogeography, Rodinia breakup, and Sturtian glaciation: constraints from Re– Os black shale ages from southern Australia and northwestern Tasmania. Geological Society of America Abstracts with Programs, 39, 335. Kendall, B. S., Creaser, R. A., Calver, C. R., Raub, T. D. & Evans, D. A. D. 2009. Correlation of Sturtian diamictite successions in southern Australia and northwestern Tasmania by Re– Os black shale geochronology and the ambiguity of ‘Sturtian’-type diamictitecap carbonate pairs as chronostratigraphic marker horizons. Precambrian Research, 172, 301–310. Kennard, J. M., Nicoll, R. S. & Owen, M. 1986. Late Proterozoic and Early Palaeozoic Depositional Facies of the Northern Amadeus Basin, Central Australia. 12th International Sedimentological Congress, Field Excursion 25B. Bureau of Mineral Resources, Canberra, 125.

AUSTRALIAN CRYOGENIAN CORRELATIONS

Kennedy, M. J., Christie-Blick, N. & Prave, A. R. 2001. Carbon isotopic composition of Neoproterozoic glacial carbonates as a test of paleoceanographic models for snowball Earth phenomena. Geology, 29, 1135– 1138. Knoll, A. H. 1996. Chapter 4: Archaean and Proterozoic palaeontology. In: Jansonius, J. & McGregor, D. C. (eds) Palynology: Principles and Applications. American Association of Stratigraphic Palynologists Foundation 1, Publishers Press, Salt Lake City, 51– 80. Knoll, A. H., Swett, K. & Mark, J. 1991. Paleobiology of a Neoproterozoic tidal flat/lagoonal complex: the Draken Conglomerate Formation, Spitsbergen. Journal of Paleontology, 65, 531–570. Knoll, A. H., Walter, M. R., Narbonne, G. M. & Christie-Blick, N. 2006. The Ediacaran Period: a new addition to the geologic time scale. Lethaia, 39, 13 – 30. Korsch, R. J. & Kennard, J. M. 1991. Geological and geophysical studies in the Amadeus Basin, central Australia. BMR Bulletin, 236, 511– 524. Krylov, I. N. 1967. Rifeyskie i nizhne-kembriyskie stromatolity Tyan’Shanya i Karatau [Riphean and Lower Cambrian stromatolites of Tien-Shan and Karatau]. Akademiya Nauk, SSSR, Trudy, Geologicheskii Institut Leningrad, 171, 76 (in Russian). Krylov, I. N. 1975. Stromatolity Rifeya i Fanerozoya SSSR. [Riphean and Phanerozoic stromatolites in the USSR]. Akademiya Nauk SSSR, Trudy, Geologicheskii Institut, 274, 243 (in Russian). Lindsay, J. F. 1993. Geological Atlas of the Amadeus Basin. Australian Geological Survey Organisation, 25 plates. Lindsay, J. F. 1995. Geological Atlas of the Officer Basin South Australia. Australian Geological Survey Organisation and Mines and Energy Department, South Australia, 30 plates. Lindsay, J. F. & Reine, K. 1995. Well-log data, Officer Basin, South Australia. Australian Geological Survey Organisation Canberra Record 1995/2, 13. Macdonald, F. A., Schmitz, M. D. et al. 2010. Calibrating the Cryogenian. Science, 327, 1241– 1243. McKirdy, D. M., Burgess, J. M. et al. 2001. A chemostratigraphic overview of the late Cryogenian interglacial sequence in the Adelaide Fold-Thrust Belt, South Australia. Precambrian Research, 106, 149– 186. Meffre, S., Direen, N. G., Crawford, A. J. & Kamenetsky, V. S. 2004. Mafic volcanic rocks on King Island, Tasmania: evidence for 579 Ma break-up in east Gondwana. Precambrian Research, 135, 177– 191. Mendelson, C. V. & Schopf, J. W. 1992. Proterozoic and selected Early Cambrian microfossils and microfossil-like objects. In: Schopf, J. W. & Klein, C. (eds) Evolution of the Proterozoic Biosphere – A Multidisciplinary Study. Cambridge University Press, New York, 865– 895. Morton, J. G. G. 1997. Lithostratigraphy and environments of deposition. In: Morton, J. G. G. & Drexel, J. F. (eds) Petroleum Geology of South Australia, Volume 3: Officer Basin. South Australian Department of Mines and Energy Resources Report Book, 97/19, 47– 86. Morton, J. G. G. & Drexel, J. F. (eds) 1997. Petroleum Geology of South Australia Volume 3: Officer Basin. South Australia, Department of Mines and Energy Resources Report Book, 97/19, 173. Mory, A. J. & Haines, P. W. 2005. GSWA Lancer 1 well completion report (interpretive papers), Officer and Gunbarrel Basins, Western Australia. Geological Survey of Western Australia Record, 2005/4, 90. Nagy, R. M. 2008. Microfossils from the Neoproterozoic Chuar Group, Grand Canyon, Arizona: taxonomy, paleoecological analysis and implications for life during the onset of Neoproterozoic glaciation. PhD thesis, University of California, Santa Barbara. Narbonne, G., Kaufman, A. & Knoll, A. 1994. Integrated chemostratigraphy and biostratigraphy of the Windermere Supergroup, northwest Canada: implications for Neoproterozoic correlations and the early evolution of animals. Geological Society of America Bulletin, 106, 1281–1292. Nelson, D. R. 2002. Compilation of geochronology data, 2001. Western Australia Geological Survey Record, 2002/2, 282. Oehler, D. Z. 1976. Transmission electron microscopy of organic microfossils from the Late Precambrian Bitter Springs Formation of

133

Australia: techniques and survey of preserved ultrastructure. Journal of Paleontology, 50, 90 –106. Parker, A. J., Rickwood, P. C. et al. 1987. Mafic dyke swarms of Australia. In: Halls, H. C. & Fahrig, W. F. (eds) Mafic Dyke Swarms. Geological Association of Canada, Special Papers, 34, 401– 417. Phillips, B. J., James, A. W. & Philip, G. M. 1985. The geology and hydrocarbon potential of the northwestern Officer Basin. APEA Journal, 25, 52– 61. Planavsky, N. & Grey, K. 2008. Stromatolite branching in the Neoproterozoic of the Centralian Superbasin, Australia: an example of shifting sedimentary and microbial control of stromatolite morphology. Geobiology, 6, 33– 45. Porter, S. M. & Knoll, A. H. 2000. Testate amoebae in the Neoproterozoic Era: evidence from vase-shaped microfossils in the Chuar Group, Grand Canyon. Paleobiology, 26, 360–385. Preiss, W. V. 1972. The systematics of South Australian Precambrian and Cambrian stromatolites, part I. Transactions of the Royal Society of South Australia, 96, 67– 100. Preiss, W. V. 1973a. The Systematics of South Australian Precambrian and Cambrian Stromatolites, Part II. Transactions of the Royal Society of South Australia, 97, 91 –125. Preiss, W. V. 1973b. Early Willouran stromatolites from the Peake and Denison Ranges and their stratigraphic significance. South Australian Department of Mines and Energy, Report Book, 73/208, 27. Preiss, W. V. 1974. The Systematics of South Australian Precambrian and Cambrian Stromatolites, Part III. Transactions of the Royal Society of South Australia, 98, 185–208. Preiss, W. V. 1976. Proterozoic stromatolites from the Nabberu and Officer Basins, Western Australia, and their stratigraphic significance. South Australia Geological Survey, Report of Investigations, 47, 51. Preiss, W. V. 1985. Stratigraphy and tectonics of the Worumba Anticline and associated intrusive breccias. South Australia Geological Survey Bulletin, 52, 85. Preiss, W. V. 1987. The Adelaide Geosyncline –late Proterozoic stratigraphy, sedimentation, paleontology and tectonics. South Australian Geological Survey Bulletin, 53, 438. Preiss, W. V. 1993. Neoproterozoic. In: Drexel, J. F., Preiss, W. V. & Parker, A. J. (eds) The Geology of South Australia, Volume 1, The Precambrian. South Australian Geological Survey Bulletin, 54, 171– 203. Preiss, W. V. 2000. The Adelaide geosyncline of South Australia and its significance in Neoproterozoic continental reconstruction. Precambrian Research, 100, 21 –63. Preiss, W. V., Walter, M. R., Coats, R. P. & Wells, A. T. 1978. Lithological correlations of Adelaidean glaciogenic rocks in parts of the Amadeus, Ngalia and Georgina Basins. Bureau of Mineral Resources Geology & Geophysics Australia Journal, 3, 43 –53. Preiss, W. V., Gostin, V. A., McKirdy, D. M., Ashley, P. M., Williams, G. E. & Schmidt, P. W. 2011. The glacial succession of Sturtian age in South Australia – the Yudnamutana Subgroup. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 701– 712. Raaben, M. E., Sinha, A. K. & Sharma, M. 2001. Precambrian Stromatolites of India and Russia: A Catalogue of Type-Form-Genera. Birbal Sahni Institute of Palaeobotany, Lucknow, 125. Rainbird, R. H., Jefferson, C. W. & Young, G. M. 1996. The early Neoproterozoic sedimentary Succession B of northwestern Laurentia: correlations and palaeogeographic significance. Geological Society of America Bulletin, 108, 454–470. Reid, A. 2009. Appendix G: U– Pb zircon dating of the porphyry. Burra Mine. In: Drexel, J. F. (ed.) Review of the Burra Mine Project, 1980– 2008 – A Progress Report. South Australia. Department of Primary Industries and Resources Report Book, 2008/16, 68– 75. Reid, P. W. & Preiss, W. V. 1999. Parachilna map sheet (second edition). South Australia Geological Survey, Geological Atlas 1:250 000 Series, Sheet SH54-13. Saito, Y., Tiba, T. & Matsubara, S. 1988. Precambrian and Cambrian cherts in northwestern Tasmania. Bulletin of the National Science

134

K. GREY ET AL.

Museum, Series C: Geology and Paleontology, National Science Museum, Tokyo, 14, 59– 70. Sawaki, Y., Kawai, T. et al. 2010. 87Sr/86Sr chemostratigraphy of Neoproterozoic Dalradian carbonates below the Port Askaig Glaciogenic Formation, Scotland. Precambrian Research, 179, 150– 164. Schopf, J. W. 1968. Microflora of the Bitter Springs Formation, Late Precambrian, Central Australia. Journal of Paleontology, 42, 651– 688. Schopf, J. W. & Barghoorn, E. S. 1969. Microorganisms from the Late Precambrian of South Australia. Journal of Paleontology, 43, 111– 118. Sergeev, V. N. 1989. Microfossils from transitional Precambrian – Phanerozoic strata of central Asia. Himalayan Geology, 13, 269– 278. Shields, G. A., Brasier, M. D., Stille, P. & Dorjnamjaa, D. 2002. Factors contributing to high d13C values in Cryogenian limestones of Western Mongolia. Earth & Planetary Science Letters, 196, 99 –111. Skotnicki, S. J., Hill, A. C., Walter, M. R. & Jenkins, R. 2008. Stratigraphic relationships of Cryogenian strata disconformably overlying the Bitter Springs Formation, northeastern Amadeus Basin, Central Australia. Precambrian Research, 165, 243–259. Smith, K. G. 1972. Stratigraphy of the Georgina Basin. Bureau of Mineral Resources Australia Bulletin, 111, 156. Stevens, M. K. & Apak, S. N. 1999. Empress 1 and 1A well completion report, Yowalga Sub-basin, Officer Basin, Western Australia. Western Australia Geological Survey, Record, 1999/4, 110. Stevens, M. K. & Grey, K. 1997. Skates Hills Formation and Tarcunyah Group, Officer Basin, carbonate cycles, stratigraphic position and hydrocarbon prospectivity. Western Australia Geological Survey Annual Review for 1996–97, 55 – 60. Timofeev, B. V. 1966. Mikropaleofitologicheskoe issledovanie drevnikh svit [Microphytological investigations of ancient formations]. Akademiya Nauk SSSR, Isdatelskvo Nauka [Science Publishing House], Moscow, 1 – 147. (Russian, published in English translation dated 1974 by British Library – Lending Division, Yorkshire, England, 214.) Townson, W. G. 1985. The subsurface geology of the western Officer Basin – results of Shell’s 1980– 1984 petroleum exploration campaign. APEA Journal, 25, 34 – 51. Walter, M. R. 1972. Stromatolites and the biostratigraphy of the Australian Precambrian and Cambrian. Palaeontological Association London, Special Papers in Palaeontology, 11, 190. Walter, M. R. 1980. Adelaidean and Early Cambrian stratigraphy of the southwestern Georgina Basin: correlation chart and explanatory notes. Bureau of Mineral Resources Australia Report, 214, 21. Walter, M. R. & Gorter, J. D. 1994. The Neoproterozoic Centralian Superbasin in Western Australia: the Savory and Officer Basins. In: Purcell, P. G. & Purcell, R. R. (eds) The Sedimentary Basins of Western Australia. Petroleum Exploration Society of Australia, West Australian Basins Symposium, Perth, WA, 1994, Proceedings, 851– 864. Walter, M. R., Krylov, I. N. & Preiss, W. V. 1979. Stromatolites from Adelaidean (Late Proterozoic) sequences in central and South Australia. Alcheringa, 3, 287– 305. Walter, M. R., Grey, K., Williams, I. R. & Calver, C. R. 1994. Stratigraphy of the Neoproterozoic to early Palaeozoic Savory Basin, Western Australia, and correlation with the Amadeus and Officer Basins. Australian Journal of Earth Sciences, 41, 533– 546. Walter, M. R., Veevers, J. J., Calver, C. R. & Grey, K. 1995. Neoproterozoic stratigraphy of the Centralian Superbasin, Australia. Precambrian Research, 73, 173–195.

Walter, M. R., Veevers, J. J., Calver, C. R., Gorjan, P. & Hill, A. C. 2000. Dating the 840– 544 Ma Neoproterozoic interval by isotopes of strontium, carbon and sulfur in seawater, and some interpretative models. Precambrian Research, 100, 371–433. Wells, A. T. & Moss, J. F. 1983. The Ngalia Basin, Northern Territory: Stratigraphy and structure. Australian Bureau of Mineral Resources Bulletin, 212, 88. Wells, A. T., Forman, D. J., Ranford, L. C. & Cook, P. J. 1970. Geology of the Amadeus Basin, central Australia. Bureau of Mineral Resources Australia Report, 100. Williams, G. E. & Gostin, V. A. 2005. Acraman– Bunyeroo impact event (Ediacaran), South Australia, and environmental consequences: twenty-five years on. Australian Journal of Earth Sciences, 52, 607– 620. Williams, G. E., Jenkins, R. J. F. & Walter, M. R. 2007. No heliotropism in Neoproterozoic columnar stromatolite growth, Amadeus Basin, central Australia: geophysical implications. Palaeogeography, Palaeoclimatology, Palaeoecology, 249, 80 – 89. Williams, G. E., Gostin, V. A., McKirdy, D. M. & Preiss, W. V. 2008. The Elatina glaciation, late Cryogenian (Marinoan Epoch), South Australia: sedimentary facies and palaeoenvironments. Precambrian Research, 163, 307– 331. Williams, I. R. 1992. Geology of the Savory Basin, Western Australia. Western Australia Geological Survey Bulletin, 141, 115. Williams, M., Crossey, L. J. et al. 2003. Dating sedimentary sequences: in situ U/Th-Pb microprobe dating of early diagenetic monazite and Ar – Ar dating of marcasite nodules; case study from Neoproterozoic black shales in the southwestern. U.S. Geological Society of America Abstracts with Programs, 35, 595. Willman, S. & Moczydłowska, M. 2008. Ediacaran acritarch biota from the Giles 1 drill hole, Officer Basin, Australia, and its potential for biostratigraphic correlation. Precambrian Research, 162, 498– 530. Willman, S., Moczydłowska, M. & Grey, K. 2006. Neoproterozoic (Ediacaran) diversification of acritarchs – a new record from the Murnaroo 1 drillcore, eastern Officer Basin, Australia. Review of Palaeobotany and Palynology, 139, 17 –39. Wingate, M. T. D., Campbell, I. H., Compston, W. & Gibson, G. M. 1998. Ion microprobe U–Pb ages for Neoproterozoic basaltic magmatism in south-central Australia and implications for the breakup of Rodinia. Precambrian Research, 87, 135–159. Yoshioka, H., Asahara, Y., Tojo, B. & Kawakami, S. 2003. Systematic variations in C, O and Sr isotopes and elemental concentrations in Neoproterozoic carbonates in Namibia: implications for glacial to interglacial transition. Precambrian Research, 124, 69 – 85. Zang, W. 1995. Early Neoproterozoic sequence stratigraphy and acritarch biostratigraphy, eastern Officer Basin, South Australia. Precambrian Research, 74, 119– 175. Zang, W. & Walter, M. R. 1992. Late Proterozoic and Early Cambrian microfossils and biostratigraphy, Amadeus Basin, central Australia. Memoirs of the Association of Australasian Palaeontologists 12, 1 –132. Zhao, J.-X. & McCulloch, M. T. 1993. Sm– Nd mineral isochron ages of Late Proterozoic dyke swarms in Australia: evidence for two distinctive events of mafic magmatism and crustal extension. Chemical Geology (Isotope Geoscience Section), 109, 341–354 Zhao, J.-X., McCulloch, M. T. & Korsch, R. J. 1994. Characterization of a plume-related 800 Ma magmatic event and its implications for basin formation in central-southern Australia. Earth & Planetary Science Letters, 121, 349– 367.

Chapter 9 A user’s guide to Neoproterozoic geochronology DANIEL J. CONDON1* & SAMUEL A. BOWRING2 1

NERC Isotope Geoscience Laboratories, British Geological Survey, Keyworth NG12 5GS, UK

2

Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, MA 02139, USA *Corresponding author (e-mail: [email protected]) Abstract: Geochronology is essential for understanding Neoproterozoic Earth history. Here we review the types of rocks and minerals that are used to date geologic events and the analytical protocols for the different radio-isotopic decay systems employed. We discuss the limitations and potential of these methodologies for dating Neoproterozoic stratigraphy, highlighting the major sources and magnitudes of uncertainties and the assumptions that often underpin them.

Crucial to understanding the nature and causes of glaciations, large perturbations of biogeochemical cycles and key evolutionary innovations that took place during the Neoproterozoic is our ability to precisely correlate and sequence disparate stratigraphic sections. Relative ages of events can be established within single sections or by regional correlation using litho-, chemo- and/or bio-stratigraphy. However, relative chronologies do not allow for testing of event synchroneity, robustness of correlations or calculation of rates and duration of change. At present, a major limitation to our understanding of Neoproterozoic history is the dearth of precise and accurate radio-isotopic dates. However, the dramatic increase in geochronological constraints over the past five years is cause for optimism. This chapter reviews the radio-isotopic techniques used for dating of Neoproterozoic stratigraphy. It is aimed at the many geologists, climate scientists, palaeobiologists and geophysicists who use Neoproterozoic geochronology, especially those less familiar with the process of obtaining a date from a rock or mineral. We provide an outline of the strengths, weaknesses and limitations of the different techniques, as well as critical evaluation of the assumptions that underpin the accuracy and precision of the calculated dates. We discuss specific examples where some of these assumptions are not valid, resulting in inaccurate dates assigned to certain stratigraphic levels. Overall, we emphasize the U –Pb (zircon) geochronology of volcanic ash beds. However, we also review Re – Os, Lu –Hf and Pb/Pb geochronology and their application to dating sedimentary rocks. Although the K – Ar and 40 Ar/39Ar methods are mainstays of Mesozoic and younger time scale calibration, there are few examples where they have been successfully applied to dating eruption ages of Neoproterozoic volcanic rocks. This is most likely due to the susceptibility of K-bearing minerals to alteration and open system behaviour related to metamorphism.

What can be dated? The absolute age of Neoproterozoic sedimentary successions can be determined via radio-isotopic dating of minerals separated from volcanic rocks. These minerals, such as zircon, are highly enriched in the parent isotope (U in the case of zircon, see below) and can be dated via radio-isotopic decay schemes (such as U –Pb, Re – Os and K –Ar/40Ar/39Ar techniques). As these minerals crystallize (or become ‘closed’ to the loss of daughter isotope) at the approximate time of the magmatic eruption, their age is assumed to approximate the depositional age of the volcanic rock. It is also possible in some cases to apply radio-isotopic dating

directly to sedimentary chemical precipitates and organic-rich components enriched in parent isotopes (U, Re, Lu) (see section ‘Whole-rock geochronometers (Re – Os, Lu –Hf, Pb/Pb)’).

Dating accessory minerals from volcanic rocks Zircon (ZrSiO4) is a common accessory mineral in silicic volcanic rocks ranging from lavas to air-fall tuffs to volcaniclastic sedimentary rocks and is a nearly ubiquitous component of most clastic sedimentary rocks. The refractory and durable nature of zircon over a wide range of geological conditions means that it is likely to retain its primary crystallization age even through subsequent metamorphic events. Silicic air-fall tuffs are the most common volcanic rocks in fossil-bearing sequences and are found in layers that range in thickness from a millimetres to many metres. They are commonly preserved in marine settings. In most of these rocks the primary volcanic ash has been altered, probably soon after deposition, to clay minerals in a process that does not affect zircon. What makes zircon ideal for U –Pb dating is the fact that, because U has a similar charge and ionic radius to Zr, it substitutes readily into the zircon crystal structure (in modest amounts, typically in the tens to hundreds of parts per million (ppm) range), whereas Pb has a different charge and larger ionic radius, leading to its effective exclusion from the crystal lattice. Therefore, at the time of crystallization (t0), there is effectively no Pb present in a crystal (although mineral and fluid inclusions may contain Pb) and the present-day Pb is the direct product of in situ U decay since t0 (see section ‘Uranium – lead’ for further details). An additional factor that makes zircon a robust chronometer is its high closure temperature (.900 8C) for Pb diffusion (Cherniak & Watson 2003), or the temperature below which U and Pb do not undergo significant thermally activated volume diffusion. This means that zircons tend to preserve their primary ages, even in volcanic rocks metamorphosed to amphibolites-facies conditions. The refractory and durable nature of zircon means that it is often recycled through crustal processes of erosion, metamorphism and magmatism, which, combined with its high closure temperature, means that it is possible to ‘inherit’ older zircon in newly formed igneous rocks. Silicic volcanic rocks are often the product of melting or assimilation of older zircon-bearing rocks and often contain inherited zircon. This phenomenon is most commonly observed as older cores surrounded by younger, magmatic zircon overgrowths. Such grains pose an analytical challenge, as the different domains should be analysed separately. This is best achieved using either microbeam techniques (see section ‘U –Pb microbeam techniques’), which employ the high spatial resolution

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 135– 149. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.9

136

D. J. CONDON & S. A. BOWRING

of focused ion beams or lasers to analyse different domains within a single crystal, or mechanical micro-sampling of single grains followed by conventional analysis. In addition, silicic volcanic rocks often contain inherited whole grains that range in age from only slightly older than the eruption to tens to hundreds of millions of years older. These zircons are the refractory residue from assimilation of older rocks in the magma or are incorporated during an explosive eruption. A further complication arises when one considers that the crystallization history of a magma is not necessarily an instantaneous event and, in some circumstances (such as largevolume silicic eruptions), zircons (and other minerals) may crystallize tens to hundreds of thousands of years prior to an eruption and/or over a protracted interval. Thus the zircons separated from an air-fall tuff may record a range of dates from the eruption age to millions of years prior. In such situations, high-precision single-grain analyses are required to isolate different populations and assign an eruption age to the sample. It is also possible to date other U-bearing accessory phases from volcanic ash beds such as monazite and titanite; however, these are much less common in their occurrence than zircon. Monazite [(LREE)PO4] is typical of peraluminous magmas and metamorphic rocks, but it is also sometimes found in volcanic ash beds (e.g. Roden et al. 1990). Monazite has a similarly high closure temperature (.900 8C) to zircon and can incorporate ThO2 at the percent level in addition to U into its crystal lattice, making it ideally suited to U – Th –Pb dating. Titanite (CaTiSiO5) is a common accessory phase in plutonic and metamorphic rocks, but somewhat less common in volcanic rocks. A potential advantage of titanite is that it has a lower closure temperature than zircon and monazite (c. 650 8C) and thus does not retain Pb until it cools below that temperature. Thus, in volcanic rocks, pre-eruptive residence is not a limiting factor for its use in dating volcanic eruptions. However, titanite also incorporates some initial Pb into its crystal structure that limits the precision of titanite U –Pb dates.

Dating chemical precipitates and organic residues Chemically precipitated rocks and organic residues from sedimentary rocks can be exploited as chronometers using the isochron approach, and calculated dates are interpreted as the time of precipitation and/or early diagenesis. The most commonly dated Neoproterozoic rocks using this approach include carbonates, phosphates and organic-rich shales, as they contain high concentrations of the different parent nuclides (e.g. Re, U and Lu). U– Pb (Pb/Pb) dating of carbonates. At the time of formation, car-

bonates incorporate U into their crystal lattice, typically with a concentration of several ppm (although in some cases this can be many tens of ppm), as well as (initial) Pb with concentrations typically in the parts per billion (ppb) range, and therefore have the potential for U –Pb dating. Carbonates form in a variety of terrestrial and marine environments, but most Neoproterozoic successions targeted for U –Pb dating are marine. Fluid-mediated recrystallization of carbonates is a common process during burial and diagenesis, and very few (if any) Neoproterozoic carbonates have escaped this process. The variable mobility of Pb and U means that it is common for the two to become decoupled, especially during recent exposure and weathering, in which case the U – Pb systematics are often unreliable. If the decoupling of the U and Pb occurred recently, then the Pb/Pb systematics are often relied upon for age information. Several studies have generated Pb/Pb isochrons from Neoproterozoic successions with variable success (Babinski et al. 1999, 2007; Folling et al. 2000; Barfod et al. 2002; Chen et al. 2004), and most demonstrate evidence for disturbance of the Pb/Pb systematics during burial/ metamorphic events (Babinski et al. 1999; Folling et al. 2000). A major issue related to Pb/Pb isochron dates is how to assess

closed-system behaviour, and it is common to use either a statistical measure of coherence (see section ‘Sources and types of uncertainty’) or pre-existing age constraints to aid in the interpretation. However, although closed-system behaviour results in statistical coherence, the inverse is not always true (see section ‘Uncertainties as a result of geologic complexity’). Recent studies have employed combined textural and Sr isotope analyses of the carbonates as an independent proxy for disturbance during fluid flow events (Babinski et al. 2007). For a recent review of U –Pb dating of carbonates see Rasbury & Cole (2009). Re– Os dating of organic-rich sediments. Both Re and Os are redox-

sensitive metals with multiple oxidation states. Re decays to Os with a half-life of c. 41 Ga (Table 9.1). Both metals can become concentrated within anoxic sediments via redox reactions near the sediment –water interface and also have been shown to be preferentially incorporated into the organic matter of shales (Creaser et al. 2002). Following deposition, both metals are often relatively immobile and can behave as a closed system through burial, diagenesis and low-grade metamorphism, allowing exploitation as a chronometer. In the past five years, the Re –Os geochronometer has been applied to several organic-rich Neoproterozoic stratigraphic intervals (Schaefer & Burgess 2003; Kendall et al. 2004, 2006), as well as several Phanerozoic intervals (Cohen et al. 1999; Creaser et al. 2002; Selby & Creaser 2005), and has given new hope for dating thick sedimentary sequences devoid of volcanic rocks. For a recent review of Re – Os dating of organic-rich sediments, see Kendall et al. (2009). U–Pb (Pb/Pb) and Lu–Hf dating of phosphates. Phosphate minerals are enriched (relative to the fluid from which they precipitate) in rare earth elements (REE), Th and U, and therefore have potential for geochronology using the Lu – Hf, Th –Pb and U –Pb decay schemes. Few studies have successfully exploited the Lu –Hf and U – Pb system for the precise dating of sedimentary phosphates (Barfod et al. 2002). In addition to dating ‘bulk’ phosphates, both monazite and xenotime (YPO4) are known to form during early diagenesis (Evans et al. 2002; Rasmussen 2005). Both minerals have highly favourable U(Th)/Pb concentrations, but their occurrence as very small grains or overgrowths makes them difficult to isolate and analyse. Xenotime occurs as syntaxial overgrowths on zircons, but due to their small size (a few tens of micrometres) and textural complexity, an in situ isotopic technique with a spatial resolution of ,10 mm is required to successfully date xenotime; this has so far only been achieved by ion microprobe. An alternative approach is the Th – U –total Pb electron microprobe technique (Williams et al. 2007). This technique has the disadvantage that it requires high concentrations of Th, U and Pb for detection, limiting it to phases such as monazite; however, it does have high spatial resolution (beam width down to ,5 mm) and therefore the ability to analyse very small grains in their petrographic context, but with uncertainties of several percent.

Table 9.1. Radiometric decay systems used in geochronology (see text for further details) Radioactive parent nuclide 238

U U 187 Re 176 Lu 87 Rb 40 K 235

Radiogenic daughter nuclide 206

Pb Pb 187 Os 176 Hf 86 Sr 40 Ar, 40Ca 207

Half-life (years)

4.468  109 7.038  108 4.16  1010 3.71  1010 4.944  1010 1.25  109

In some cases the ‘whole-rock geochronometers’ may be useful for obtaining temporal constraints on sedimentary successions devoid of volcanic material. In general, best-case uncertainties associated with calculated isochron dates are in the 0.5 –1% level, comparable to those of U –Pb (zircon) microbeam dates. Despite the relatively large uncertainties intrinsic to the dating of these types of rocks, the Pb/Pb isochron method has been applied and interpreted as an estimate of depositional ages (see above). Problems associated with isochron geochronometers are centred around the lack of an independent check on open-system behaviour.

Maximum and minimum age constraints Not all sedimentary successions are amenable to direct dating via radio-isotopic methods. In the absence of zircon-bearing volcanic rocks, or chemical sediments for isochron dating, another approach is to obtain maximum age constraints by dating detrital zircons. Detrital zircons in sedimentary rocks can range from being considerably older than the estimated depositional age they are contained within (often many hundreds of millions of years older) to close to the age of deposition, and in this latter case can provide useful constraints (Bingen et al. 2005). The youngest age determined in a detrital population gives a maximum depositional age for the unit. Minimum age constraints can be provided by overlying strata, which may contain age-diagnostic fossils, and/or cross-cutting igneous intrusions.

Radio-isotopic geochronometers A relatively small number of radioactive decay schemes are suitable for dating Neoproterozoic rocks and are listed in Table 9.1. Each of these is based upon the radioactive decay of a parent nuclide to a stable daughter nuclide. Obtaining an accurate date using these decay systems requires that (i) the decay constant of the parent nuclide is accurately and precisely determined; (ii) closed-system behaviour, which can be simply stated to mean that the parent/daughter ratio has only changed by radioactive decay; and (iii) the initial daughter nuclide, if present, can be precisely and accurately accounted for. In this section, we outline the basic principles of the radio-isotopic geochronometers in Table 9.1, separating the U –Pb system applied to U-bearing accessory minerals from those that use an isochron approach (Re – Os, Lu – Hf, Pb/Pb, etc.) applied to chemical precipitates and organic residues.

Uranium – lead U –Pb geochronology is often regarded as the gold standard of geochronology because, unlike all other chronometers, it exploits two independent decay schemes, 235U to 207Pb and 238U to 206Pb, and both the 238U and 235U decay constants are relatively precise and accurate (Jaffey et al. 1971). Two separate dates for a

206Pb/238U date =

ln (1+ 206Pb*/238U ) λ238

Pb*/

238

137

950 Ma

850 Ma

0.14

750 Ma

λ235

ln (1+ 207Pb*/235U )

206

Precipitation of xenotime also occurs during fluid and thermal events, which introduces the potential for complexity through multiple episodes of growth (Rasmussen 2005). In contrast to xenotime, diagenetic monazite tends to occur as impure nodules (up to 2 mm in diameter) in shales. Studies of Palaeozoic diagenetic monazite nodules have demonstrated the utility of the Pb/Pb and Th –Pb systems, although the U –Pb systematics have been perturbed, indicating recent U remobilization (Evans et al. 2002). Neither (diagenetic) monazite or xenotime U – Pb geochronology have been successfully applied to the dating of a Neoproterozoic sedimentary succession, although recent Th –U –total Pb investigation of monazite from the Adelaide Rift complex (Mahan et al. 2010) does show potential.

U

USER’S GUIDE TO NEOPROTEROZOIC GEOCHRONOLOGY

0.12

650 Ma 0.10 207Pb/206Pb date =

e

λ235t

–1

235U

λ238t

–1

238U

e

today

207

0.9

1.1

1.3

1.5

Pb*/

235

U

Fig. 9.1. U–Pb (Wetherill) concordia diagram for the age range 542– 1000 Ma. The grey band represents the concordia curve plotted to reflect the uncertainties in the 235U and 238U decay constants.

zircon based on each individual decay scheme may be calculated and plotted on a concordia diagram (Fig. 9.1). On a conventional (Wetherill) concordia diagram the X and Y axes are the ratio of radiogenic 207Pb to 235U and radiogenic 206Pb to 238U ratios, respectively, and the concordia curve represents the simultaneous solution of the decay equations for a given age. A third date, the 207 Pb/206Pb date, can be determined from only Pb isotopic measurements by knowing both the 235U and 238U decay constants and the present-day 235U/238U ratio, which is assumed to be 137.88 (Steiger & Jager 1977). Calculation of the U –Pb dates requires determination of the Pb* –U ratio (Pb* denotes radiogenic Pb). See Figure 9.1 for age equations. The different half-lives of 238U and 235U (c. 4.5 and c. 0.7 Ga, respectively) mean that by the Neoproterozoic time, much smaller amounts of 235U (relative to 238U) remained due to the higher decay rate, and smaller amounts 207Pb were produced per increment of time relative to 206Pb. The advantage of two independent chronometers in the same mineral is that it is possible to detect small amounts of open-system behaviour such as Pb loss or the inheritance of older Pb. This is a major factor in our ability to make reliable, high-precision age determinations, as we can evaluate whether a number of analyses represents a single time of mineral growth. It has been known for several decades that zircons often show evidence of post-crystallization Pb loss. This has the effect of lowering the U –Pb ratios and the derived dates. The analyses plot below the concordia curve and are termed normally ‘discordant’, with 207Pb/206Pb dates . 207 Pb– 235U dates . 206Pb – 238U dates (Fig. 9.2). Silver & Duetsch (1963) demonstrated that Pb loss was correlated with radiation damage. Since that time it has become widely appreciated that Pb is not lost by thermally activated volume diffusion but rather by fast-pathway diffusion from damaged parts of the crystal lattice. In order to minimize/eliminate the effects of postcrystallisation Pb loss, it is possible to pre-treat zircons to preferentially remove domains that have lost Pb, leaving zircon that has remained a closed system. One major approach involves physically abrading away the exterior portions of the zircons (Krogh 1982a), based on the observation that the outer portions were richest in U and thus susceptible to radiation damage and Pb loss. Krogh (Krogh 1982b) also demonstrated that the careful selection of the least magnetic zircons (diamagnetic) often corresponded to those with lowest U contents, radiation damage, and least affected by Pb loss. These approaches were widely applied

D. J. CONDON & S. A. BOWRING

720

Upper intercept or 207Pb/206Pb date: 713.65 ±0.70 (±4.5) Ma

slope = e λt -1

206

Pb/

238

U

138

206Pb/238U

date: 711.65 ±0.58 (±1.4) Ma

D*/Dref

0.1165

705

0.1155

t2 t1

207Pb/235U

date: 712.35 ±0.83 (±1.9) Ma

700 0.1145

to P/Dref

-lo

ss

Y-axis intercept = D0/Dref

Pb

695 0.1135

690 0.975

207

0.99

1.00

1.01

1.02

Pb/

235

U

Fig. 9.2. Schematic U–Pb concordia diagram illustrating the 206Pb– 238U, 207 Pb– 235U and 207Pb/206Pb dates that can be calculated. The data presented are a subset of analyses from sample WM54 (Bowring et al. 2007). Uncertainties are the 2s internal uncertainties, whereas those in parentheses are the 2s internal plus the systematic decay constant uncertainties: 0.11% for 238U and 0.14% for 235U (Jaffey et al. 1971). The grey band is the concordia line plotted to reflect the uncertainties in the U decay constants. Error ellipses (white) are plotted with 2s internal uncertainties, and black error bars represent 2s internal plus systematic decay constant uncertainties.

until the development of a new technique described as ‘chemical abrasion’ (CA-TIMS) (Mattinson 2005). This technique involves annealing zircon grains at 800–900 8C followed by partial dissolution in hydrofluoric acid. This method effectively ‘mines out’ or preferentially dissolves the higher U parts of the zircon that have been damaged by radiation and are thus susceptible to fast-pathway diffusion of Pb from the zircon crystal. In many cases these domains are irregular in shape and occur in the grain interiors. This method seems to offer the promise of effective elimination of open-system behaviour in most zircon. Microbeam techniques (see section ‘U– Pb microbeam techniques’) have not typically used pre-treatment techniques, as they assume that Pb loss is restricted to the exterior portions of grains, which they attempt to avoid during the in situ analyses. In addition, grains treated using this method are often extremely fragile and not amenable to mounting and polishing.

Fig. 9.3. Schematic isochron diagram illustrating (i) the situation at t0 where different samples from the same stratigraphic interval record a spread in parent(P)/daughter(Dref) ratios but a constant initial daughter isotopic composition. D0 is the initial amount of daughter isotope and D* is the radiogenic daughter isotope (from decay and initial) such that the daughter produced solely from decay (since t0) ¼ D* –D0.

with a slope equal to e lt –1 (Fig. 9.3). A typical isochron is plotted P/Dref on the X axis and D*/Dref on the Y axis, where P is the number of parent atoms in the sample, Dref is the number of atoms of a stable reference isotope of the daughter element and D* the total number of radiogenic daughter atoms plus initial atoms (D0) of the same isotope (amount of daughter atoms due to decay ¼ D* –D0) (Fig. 9.3). At the time of sample formation (t0), all samples would define a horizontal line. However, as the parent decays over time (t1, t2, etc.), each sample will evolve along a slope of –1, and samples with higher initial parent/daughter ratio (P/Dref) will be displaced most (Fig. 9.3), such that the isochron rotates to a positive slope. Assuming closedsystem behaviour since the formation of the sample, a linear regression through the points allows calculation of the slope and Y-intercept from which the age and the initial isotopic composition of the daughter can be determined (Fig. 9.3).

Analytical methodologies The majority of age constraints for Neoproterozoic strata have been obtained by U –Pb zircon dating, so most of this section is concerned with U –Pb analytical methods. The analytical methods used for other chronometers have some similarities with the ID-TIMS U –Pb method. The differences are discussed below.

Whole-rock geochronometers (Re – Os, Lu – Hf, Pb/Pb)

U– Pb methodologies

When rocks or minerals do not have extreme enrichment in the parent/daughter ratio but instead have a range, the isochron approach is often useful. The isochron method involves analyses of multiple cogenetic samples (minerals or subsamples of a rock from an identical stratigraphic level) and is used for systems where initial daughter atoms are present and there is the possibility of a range in the parent/daughter ratio. The Re –Os, Lu –Hf and U –Pb chronometers can be used to date carbonates, phosphates and organic-rich shales that incorporate both parent and daughter isotopes at the time of their formation. Ideally, the isochron approach allows determination of both an age and an initial isotopic composition of the daughter element. For multiple cogenetic samples to preserve the time of system closure, samples must begin with (i) a homogeneous initial daughter isotopic composition and (ii) a spread in the parent/daughter ratio such that, over time, different samples with different parent/daughter ratios will evolve and will define a straight line

There are two main approaches to U – Pb zircon geochronology: in situ ‘microbeam’ techniques and isotope dilution thermal ionization mass spectrometry (ID-TIMS). The major difference is that in ID-TIMS geochronology, zircon is dissolved and the U and Pb separated from the other elements before analysis, but in microbeam techniques the zircon is analysed using either a laser that ablates the sample prior to injection into a plasma or a focused ion beam that sputters the sample and generates secondary ions from a mineral mounted in epoxy and polished or a thin section. U –Pb ID-TIMS. ID-TIMS analyses of zircon (either as multi-grain fractions, single grains or grain fragments) involve dissolution of the zircon in the presence of tracer isotopes and are called isotope dilution. Because the tracer is added before dissolution, the tracer/sample isotope ratios stay constant, despite incomplete recovery during chemistry and low ionization efficiency during

USER’S GUIDE TO NEOPROTEROZOIC GEOCHRONOLOGY

analysis. For U –Pb ID-TIMS analyses, the most common tracers are 205Pb and 235U. 205Pb is an artificial isotope that does not occur in nature, whereas 235U occurs naturally, although an enriched form is used as a tracer. Natural U in the zircon is assumed to have 238U/235U ¼ 137.88 (Steiger & Jager 1977), so the tracer can be used to determine the number of moles of 238 U and 235U in a sample. Following dissolution, the sample undergoes chemical purification using anion exchange chemistry that allows separation of Zr and REEs from Pb and U, and Pb and U from one another. Following purification, Pb and U are analysed separately by thermal ionization mass spectrometry, where the ratios of sample isotopes (204Pb, 206Pb, 238U, etc.) to the tracer isotopes (205Pb, 235U) can be measured. As the amount of tracer isotope added to the sample is known, the number of atoms of each naturally occurring isotope in the sample can be determined based upon the measured ratio of sample isotope/tracer isotope ratio. After corrections for mass-dependent isotope fractionation, and the minor contribution of common Pb and U from the reagents, the tracer and labware, the sample 206Pb/207Pb, 206 Pb/238U and 207Pb/235U ratios can be determined and 206 Pb/207Pb, 206Pb/238U and 207Pb/235U dates calculated. See section ‘Calibrating tracers for isotope-dilution’ for further discussion of the isotope dilution technique. Optimization of this technique means that it is now possible to date zircons with ,10 pg radiogenic Pb with a precision of ,0.1% on the U –Pb ratio for single-grain analyses. However, it is a very labour-intensive technique, as each single U –Pb analysis takes several hours of mass spectrometry and chemical purification in an ultraclean laboratory environment, making it time-consuming to develop high-n datasets.

U– Pb microbeam techniques U –Pb geochronology by microbeam techniques has revolutionized geochronology over the past two decades. The two major techniques are secondary ion mass spectrometry (SIMS), typified by the SHRIMP (sensitive high resolution ion microprobe) and laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS). Both techniques (collectively termed ‘microbeam’ techniques) offer high-spatial-resolution analyses using either a focused ion beam to sputter a volume of zircon (SIMS) or a laser that is used to vaporize a volume of zircon (LA-ICP-MS). Microbeam techniques allow in situ analysis of very small volumes and thus high spatial resolution. A typical volume of zircon analysed by an ion probe is cylindrical, 20 –30 mm in diameter and several micrometres deep, with somewhat larger volumes for LA-ICP-MS (Kosler & Sylvester 2003). In addition, the analyses can be done relatively rapidly (many tens of analyses per day for LA-ICP-MS and tens for SIMS) and are amenable to automation. Furthermore, these techniques also allow analysis of other isotopes/elements of interest (Hf, O, REEs) and can be made on the same zircon grains in close proximity to the volume analysed for geochronology. Fundamental to the microbeam U – Pb zircon methods is use of a primary standard against which the U – Pb ratio of the unknown zircon is calibrated. For SIMS techniques this calibration involves analyses of standard zircon to develop a calibration curve for a known U –Pb ratio (which is determined via ID-TIMS analyses) and against which analyses of unknown zircons can be compared. This is achieved through analytical sessions where a standard zircon is repeatedly analysed interspersed with analyses of unknown zircons (this is termed sample-standard bracketing). During the course of a single analytical session, the measured ratio/date of a standard can drift by several percent. For LAICP-MS the approach is somewhat similar in that sample-standard bracketing is applied to determine the inter-elemental fractionation, which is then applied to the unknown zircons. In both

139

SIMS and LA-ICP-MS techniques, the 207Pb/206Pb ratio is a direct measurement; for SIMS mass-dependent fractionation appears to be minimal and the measured ratio is commonly used, whereas in LA-ICP-MS analyses mass-dependent fractionation is quantifiable and is corrected for, either using sample-standard bracketing and/or using a solution with known 205Tl/203Tl ratio to correct for mass bias on Pb isotopic ratios. For further details of microbeam techniques see Ireland & Williams (2003) for a review of SIMS U – Pb geochronology and Kosler & Sylvester (2003) for a review of LA-ICP-MS geochronology. There is a tradeoff between the benefit of the high spatial resolution provided by microbeam techniques and the precision of individual spot analyses. LA-ICP-MS and SIMS analyses are approximately an order of magnitude less precise than ID-TIMS (Ireland & Williams 2003; Kosler & Sylvester 2003). In situ techniques are without question essential tools for characterizing complex (zoned) zircons from volcanic and metamorphic rocks and for characterizing detrital populations, which in some cases can provide robust estimates of the maximum age of a sequence.

Isochron techniques Isochron techniques involve analysis of multiple samples assumed to be the same age, which have a spread in parent/daughter ratio and have remained closed systems. To maximize the probability that the samples are the same age and have the same initial isotopic composition it is preferable to design an appropriate sampling strategy. For example, Re – Os geochronology is often attempted on laminated organic-rich shales, so it is essential to have multiple samples from the same restricted stratigraphic interval (e.g. Kendall et al. 2004) to avoid samples that span an amount of time much larger than uncertainties. In some cases, such as working with core samples, this is not always possible (Schaefer & Burgess 2003; Kendall et al. 2006). Samples that integrate a significant amount of time could introduce scatter if there is temporal variation in the initial isotopic composition of the daughter element, and/or the sample represents an amount of time that far exceeds measurement uncertainty, especially in condensed sections. For U – Pb in carbonates, different cement domains may have very different U –Pb ratios, but is difficult to know a priori whether the cements are all exactly the same age and have the same initial ratios. In both cases, deviations from the assumption of the same age and initial ratio lead to increased scatter and uncertainties in calculated dates. Sample dissolution and purification techniques are similar to the procedures for U –Pb ID-TIMS. Before isotope ratio mass spectrometry, samples undergo dissolution and chemical purification. For multi-element systems (such as Re –Os, U –Pb and Lu –Hf ), isotopic tracers are added before dissolution for the isotope dilution (see above), whereas for systems where only daughter isotopes are measured (i.e. Pb/Pb), direct measurements of the isotope ratios are made. The isotopic composition is determined via thermal ionization mass spectrometry, although it is also possible to use solution-mode ICPMS for most elements. The accuracy and precision of isochron techniques is largely controlled by having a sufficient spread in initial parent/daughter ratio, closed-system behaviour and an identical initial isotopic composition for all samples. For precipitates such as carbonates and phosphates there is often no significant detrital input; however, this is not the case for organic-rich shales targeted for Re –Os where there is potential for significant concentrations of initial Os from multiple sources. This has been demonstrated in several studies (Creaser et al. 2002; Kendall et al. 2004); however, it is possible to limit the detrital Os contribution by selective dissolution of the organic component using a CrO3 – H2SO4 dissolution approach. Kendall et al. (2004) compared two dissolution methods (aqua regia v. CrO3 –H2SO4 dissolution) on greenschist facies organic-rich shale from the Old Fort Point

140

D. J. CONDON & S. A. BOWRING

Formation in Western Canada. Both dissolution techniques were used on the same powders. However, the aqua regia method yielded scattered data and a resulting ‘isochron’ regression with a mean square weighted deviation (MSWD) of 65 and a large age uncertainty (9%) in comparison to the CrO3 –H2SO4 dissolution method, which yielded an isochron with much less scatter (MSWD ¼ 1.2) and a relatively low uncertainty (0.8% 2s) (Kendall et al. 2004).

Sources and types of uncertainty Without an accurate estimation of total uncertainty, the radioisotopic age of a given rock or mineral is of limited value. For example, suppose a date of 618 Ma is reported for a detrital zircon from a unit that underlies a Marinoan-type glacial deposit. If the date is relatively precise (and accurate), +2 Ma, then it could be inferred that the onset of glacial sediment accumulation post-dates deposition of the detrital component at 618 + 2 Ma. On the other hand, if the 618 Ma date has an uncertainty of 100 Ma then the detrital zircon could be as old as 718 Ma or young as 518 Ma, making it of limited use. Consider the case where the 618 Ma is the 206Pb – 238U date with an uncertainty of 10 Ma, but the 207Pb/206Pb date is 650 + 400 Ma. Although the two dates overlap (and are therefore technically concordant), the lack of precision on the 207Pb/206Pb date renders it impossible to assess open-system behaviour (such as Pb loss or inheritance) within the limits imposed by the precision of the 207Pb/206Pb date, so the accuracy of the 206Pb– 238U date is not known; if the zircon has lost Pb, the 206Pb– 238U date is a minimum date. This latter example may seem extreme, but data of this type have been published and are often uncritically cited. For example, Ireland et al. (1998) published an extensive dataset of SHRIMP U –Pb zircon dates on detrital zircons from the Kanmantoo Group in Australia (it should be noted that it was not the intention of this study to constrain the timing of sediment accumulation). The units sampled included the Marino Arkose, where 50 detrital zircons were analysed, the majority of which were .1 Ga (n ¼ 48). Two grains yielded 206Pb– 238U dates of 649 + 17 and 655 + 17 Ma, which many researchers use to indicate that the Marino Arkose is c. 650 Ma (or younger) (Zhou et al. 2004; Halverson et al. 2005; Peterson et al. 2005). The 207Pb/206Pb dates associated with these two analyses are 470 + 440 Ma and 666 + 307 Ma, respectively. Although it is possible that the two zircons are indeed c. 650 Ma, it is not legitimate to assume the 206Pb – 238U date is an accurate estimate of the age of the zircons without considering both the 207Pb– 235U and 206Pb/207Pb uncertainties. In practical terms these two detrital zircon dates provide no significant constraint on the timing of sedimentation. It is advisable for consumers of geochronological data to understand the various sources of error and when one must consider the total uncertainty of a given date as opposed to its constituent parts. Although the uncertainty of each date contains an internal/random component in the total uncertainty, there are also components that are systematic, such as errors from the decay constants. When comparing ages determined by the same isotopic system these can be ignored, offering a potential increase in resolving power. In this section we review the different sources of uncertainties and the assumptions that underlie the often quoted (or not) errors. For more detailed treatment of uncertainties in geochronology the following articles are recommended: Ireland & Williams (2003), Stern & Amelin (2003), Schmitz & Schoene (2007), and various papers by Ludwig (1980, 1991, 1998, 2003).

Random/internal uncertainties Random/internal uncertainties are those relating to the measurement of isotopic ratios of the sample, standards and blanks, and are used in the derivation of errors of the radiogenic ratios.

These uncertainties are intrinsic to each analysis and represent the minimum uncertainty that must be considered. Most of these sources of random uncertainty relate to the mass spectrometry measurements and our ability to measure and reproduce isotopic ratios, such as corrections made for laboratory blank. Factors such as the electronic noise of detectors place a theoretical limit on the precision that can be achieved by detecting a certain number of ions over a finite period of time (counting statistics). However, for almost all geochronological applications, other factors such as correction for mass-dependent fractionation that occurs during sample ionization, and correction for laboratory blank and/or initial parent and daughter nuclide, dominate the analytical uncertainty budget, especially for small samples. The analytical uncertainties associated with U – Pb ID-TIMS dates have decreased substantially over the past decade. This is due in large part to a reduction in the common Pb levels introduced in the laboratory (laboratory blank) and correction for mass fractionation though use of a ‘double-spike’, where two tracer isotopes of the same element (202Pb/205Pb or 233U/235U, for example) are used for real-time mass fractionation correction, helping reduce the uncertainty in the mass-dependent fractionation. See Schmitz & Schoene (2007) and Bowring et al. (2006) for a more complete discussion of sources of uncertainty in U – Pb ID-TIMS analyses. In SIMS and LA-ICP-MS (in ‘dry’ mode) analyses, common Pb levels are intrinsically low and the uncertainty related to any correction for common Pb is usually insignificant compared to the uncertainty in the U –Pb normalization (see below). In ‘wet’ mode, as is typical for LA-ICP-MS U –Pb analyses, the correction for Pb blank and isobaric interferences on the Pb and U peaks can be a significant source of uncertainty, requiring a correction to be made (Horstwood et al. 2003). Microbeam U –Pb standardization. Microbeam U –Pb data are

acquired in analytical sessions where the analysis of unknown zircons are alternated with standards. The raw measured U –Pb ratio of the standard varies or ‘drifts’ during an analytical session due to slight changes in instrument parameters. Therefore there is an uncertainty associated with the Pb–U standardization that must be considered. This uncertainty is on the order of 1% (Stern & Amelin 2003), although the attributed magnitude is dependent on the frequency with which the standard is analysed. There are differences of opinion on how the uncertainty related to the U –Pb standardization is factored into the total uncertainty of a date. Some groups consider that the ‘standardization’ value is constant for a given session, so the uncertainty is systematic and need only be considered when comparing data collected in different analytical sessions. In this approach the session Pb– U uncertainty can be simply added to the weighted mean uncertainty in a manner analogous to the ID-TIMS tracer calibration uncertainty. Conversely, other groups consider that the reproducibility of the standard is a reflection of the true external reproducibility of all analyses and that it should be incorporated into individual analyses of unknown zircon (Ireland & Williams 2003; Stern & Amelin 2003). Typically, larger uncertainties for each analysis result in a reduction of the MSWD for weighted mean dates from multiple analyses, but has obvious implications for the identification of outliers (see discussion in Ireland & Williams 2003). Information regarding the approach taken to the standardization uncertainty is often recorded in the footnotes to the data table or in the data repository, but can be crucial when trying to precisely sequence rocks or calculate durations of events. Typical internal (2s) uncertainties for ID-TIMS 207Pb/206Pb and 206Pb – 238U zircon dates are c. 0.5 –0.2% and c. 0.1– 0.05%, respectively, and for microbeam techniques internal (2s) uncertainties on 207Pb/206Pb and 206Pb– 238U dates are c. 3 –5% and c. 1 –2%, respectively. The difference is about an order of magnitude, and for a 206Pb– 238U date of 600 Ma amounts to the difference in uncertainty on a single date between 0.3 and 0.6 Ma for ID-TIMS and 6 and 12 Ma for microbeam U –Pb dates.

USER’S GUIDE TO NEOPROTEROZOIC GEOCHRONOLOGY

Systematic/external uncertainties Systematic uncertainties are those related to a parameter that has a fixed value and is effectively constant for a suite of determinations, and is usually determined by experiments. An example of such a parameter used in U –Pb geochronology with systematic errors would include the 238U decay constant (Jaffey et al. 1971), which will affect all 238U – 206Pb dates and thus can be ignored when comparing any 238U derived dates. The level at which these systematic errors have to be considered depends upon their nature; if they are specific to a single laboratory (i.e. tracer calibration) then the systematic uncertainty has to be considered when comparing dates determined using a different tracer. Similarly, decay constant uncertainties affect all dates derived from that decay constant equally, so they only need to be considered when comparing dates derived from different decay constants (i.e. 238U – 206Pb and Re –Os dates). They can otherwise be ignored when calculating differences between dates based upon the same decay scheme. This distinction between random and systematic (or internal and external) types of uncertainty is important, as in some circumstances where the same parameter (decay constant, lab tracer) is used for calculating a suite of dates the effective uncertainty can be reduced. This results in increased temporal resolution at which synchronicity, rates and durations of proxy record evolution can be assessed. Decay constants. One source of systematic uncertainty that affects all radio-isotopic dates is that related to uncertainty in the decay constants (Table 9.1). Three approaches have been taken to determine the decay constants (the probability that a given atom will decay per unit of time) of the long-lived radionuclide: (i) direct counting; (ii) in-growth; and (iii) geological comparison. Direct counting involves the detection of alpha, beta or gamma activity relative to the total number of radioactive atoms. In-growth relies upon the quantification of a decay product that is accumulated from a quantity of high-purity parent nuclide over a welldefined period of time. Geological comparison involves analyses of cogenetic rocks or minerals with multiple chronometers, assuming that each chronometer should yield the same date. This approach has the potential for relative intercalibration of the decay constants, but accurate intercalibration requires that at least one decay constant is accurate and known with some precision (Mattinson 2000; Schoene et al. 2006). This is usually assumed to be the 238U and 235U due to the precision with which the decay constants have been determined (Jaffey et al. 1971) and the internal check provided by closed-system zircon analyses (Mattinson 2000; Schoene et al. 2006). The counting experiments of Jaffey et al. (1971) determined the 238 U and 235U decay constants with uncertainties of 0.11% and 0.14%, respectively. These values have been adopted for use in geochronology (Steiger & Jager 1977). The 187Re and 176Lu decay constants have been determined by both direct counting experiment and through geological comparison with the U –Pb system, and uncertainties are estimated at c. 0.4–0.5% (Scherer et al. 2001; Selby et al. 2007). The incorporation of decay constant uncertainties is becoming increasingly important as the internal precision of dates is reduced and multiple geochronometers are being used to investigate the same time intervals. The decay constant uncertainties for isochron dates are typically ,20% of the total uncertainty budget; in contrast, the uncertainties in the U decay constants are often .50% of the total uncertainty budget of U –Pb ID-TIMS dates (Fig. 9.2). The situation for the ID-TIMS U –Pb community is that 206Pb– 238U and 207Pb/206Pb dates often do not overlap within analytical precision, and the U decay constant uncertainties must be considered (Ludwig 2000; Begemann et al. 2001; Schoene et al. 2006). As the ‘consumer’ often uses these dates interchangeably, we are now seeing 206Pb– 238U and 207 Pb/206Pb age uncertainties presented as +X/Y/Z and +X/Z

141

respectively, where X is the analytical/internal uncertainty, Y is the analytical uncertainty plus the systematic tracer calibration uncertainty, and Z is the total uncertainty including X, Y and the decay constant uncertainties. This permits use of the data with the level of uncertainty that is appropriate to the problem being addressed. Age of primary standards for microbeam U –Pb dating. As discussed

above, U –Pb microbeam techniques rely upon measurement of the U – Pb ratio relative to standard minerals of known age. The U –Pb dates of these minerals are determined via ID-TIMS analyses with typical total uncertainties of 0.1 –0.3% that should be propagated into the total uncertainty of the final U/Pb microbeam date. Because the systematic uncertainty related to the age of the primary standard is about an order of magnitude less than the random errors related to the U – Pb determination of the unknown mineral, it is often not considered significant. Intra- and intercrystal homogeneity is a fundamental requirement of a zircon standard for microbeam U – Pb geochronology as the U – Pb ratio of the standard is considered invariant. Isotopic homogeneity is assessed by multiple ID-TIMS analyses on single crystals and/or crystal fragments (Black et al. 2003; Schmitz et al. 2003) with variability assessed on a microgram scale. In contrast, microbeam techniques require standards that are homogeneous on the submicrometre scale. At present, zircon standards are either chips of megacrysts (e.g. SL13 and 91500) or multicrystal mineral separates from plutonic rocks (e.g. Temora, R33). In general the zircon standards are relatively homogeneous at the level that can be detected by either microgram ID-TIMS analyses or nanogram SIMS analyses. However there have been issues with at least one of the megacryst standards (SL13), which is heterogeneous at the micrometre-scale (Ireland & Williams 2003). The fact that all zircon standards are natural means they are not perfect, as they are likely to be affected by zonation and/or Pb loss and/or other (matrix-related) differences. For most wellcharacterized zircon standards this variation occurs well below the level of quantification/detection and is not significant. Calibrating tracers for isotope-dilution. For isotopic analyses that use the addition of isotopic tracers (isotope dilution), the accuracy of the tracer calibration (isotopic composition and concentrations) has major control over the accuracy of the derived dates for a mineral or rock. Calibration of tracers is performed through admixing the tracer with another solution of known isotopic composition and, importantly, known purity. High-purity metals or salts (see Selby et al. 2007 for details of a Re –Os tracer, Wasserburg et al. 1981 for Sm– Nd, Schoene et al. 2006 for details of a U –Pb tracer calibration) are used as the basis of the gravimetric reference solutions against which the concentration of the tracer isotope can be determined. Therefore, the purity of the metal or salt, and the accuracy of the weighing prior to dissolution, controls the precision and accuracy of the calibration. This total uncertainty is typically estimated at c. 0.1%. For multi-element tracers, the elemental (i.e. U –Pb) ratio is fixed, so the uncertainty in the tracer calibration is systematic and can be ignored for the practical purposes of age determinations generated using the same tracer. This is particularly useful when attempting to determine the relative time difference between samples such as determining sediment accumulation rates (Bowring et al. 2007), or assessing synchroneity of events (Condon et al. 2005). At present, it is typical that each isotope laboratory has their own tracer. This means that the tracer uncertainty has to be considered when comparing dates with other laboratories and other techniques. Recently, the U –Pb ID-TIMS community has made an effort to eliminate this inter-laboratory uncertainty through the development and calibration of a large amount of 205Pb– 233U – 235U tracer for community use under the auspices of the EARTHTIME Initiative (Condon et al. 2007). As analytical uncertainties are reduced it is becoming apparent that inter-laboratory (or inter-session) variations are

142

D. J. CONDON & S. A. BOWRING

comparable to, or greater than, internal precision. Although much of this can be eliminated by a common tracer solution for U –Pb ID-TIMS, as discussed above, it is also desirable that along with analyses of unknown zircons, secondary standards (minerals or synthetic solutions) are analysed in order to provide an accurate assessment of long-term reproducibility and inter-laboratory agreement.

Calculating an age from multiple dates A significant proportion of temporal constraints for Neoproterozoic strata are derived from U –Pb dates on zircons from volcanic rocks. The final reported date and associated uncertainty are often weighted mean dates derived from a number (n) of individual dates on different zircons (or zircon sub-domains). This is the case for data acquired using both ID-TIMS and microbeam techniques. The weighted mean weights each individual analysis (such as a single SIMS spot or single-grain ID-TIMS analyses) according to its precision, so analyses with a low uncertainty contribute more to the weighted mean than those with high uncertainty. Importantly, the use of a weighted mean algorithm (or other averaging) is underpinned by the expectation of a single population with normally distributed errors (i.e. there is no correlation between precision of analyses and age). If the errors on the individual analyses are approximately equal (as is typical for microbeam U –Pb pffiffiffi data), then the weighted mean uncertainty is proportional to 1/ n. In this case, high-n datasets can be used to reduce the overall age uncertainty for data collected on a single population with normally distributed errors. If the uncertainties on the individual analyses are variable then the weighted mean uncertainty is controlled by the most precise analyses making the weighted mean date less proportional to n. High-n datasets are critical for assessing analytical v. geological scatter; however, the real limit on the precision is the analytical uncertainty of single (spot or grain) analyses, as this controls our ability to resolve real variation within a series of analyses due to ‘open-system’ behaviour. A common measure of the ‘coherence’ of a dataset is a statistical parameter called the MSWD (mean square of the weighted deviates; York 1966, 1967). A value of c. 1 indicates that the scatter in the data can be explained by analytical uncertainties alone. Values much less than 1 indicate that analytical uncertainties have been overestimated. Values greater than 1 can indicate either that the uncertainties have been underestimated or that another source of scatter, often called ‘geological’ scatter is present. Furthermore, the actual MSWD value for which the scatter of the data can be considered due to analytical factors alone is not restricted to a value of 1, but in fact varies according to the number of data points in the calculation (Wendt & Carl 1991). So, to be 95% confident that the scatter of the data is due to the analysis when n ¼ 5, an acceptable MSWD range would be 0.2 –2.2, but for n ¼ 25 this would be 0.6– 1.5 (Wendt & Carl 1991). Although not often explicitly stated, an MSWD of 1 does not necessarily mean there is a single (age) population. Rather, it indicates that if real (age) variation is present, it cannot be resolved within the precision of the individual analyses. It is not uncommon to see both isochron and weighted mean dates reported with MSWDs much greater than 1. These dates likely have little geological significance and should not be used.

Uncertainties as a result of geologic complexity Uncertainty as a result of geological complexity is the most difficult to quantify. The most common cause of excess scatter is opensystem behaviour resulting from either inheritance of older zircon and/or Pb loss. For U – Pb zircon analyses, reduced errors on single analyses often exposes fine-scale variability that may reflect protracted or punctuated crystallization of zircon crystals

in a magma chamber or the effects of very subtle open-system behaviour. Thus, high-precision analyses do not always transform into reduced uncertainties in calculated weighted mean dates. Complex U –Pb zircon systematic. In the past decade, errors associ-

ated with ID-TIMS analyses have been reduced by almost an order of magnitude. These reduced errors offer unprecedented precision, but also expose geological complexity at the ,0.1% level, sometimes resulting in scatter that exceeds analytical uncertainties. It is now common for a geochronologist to be faced with a population of zircon analyses that do not form a coherent cluster, and the crucial question is how to interpret the data to arrive at an eruption and/or depositional age. The advent of CA-TIMS pre-treatment for the elimination of Pb loss has been extremely important, as it gives one confidence that in many cases Pb loss need not be considered as a cause of excess scatter. Furthermore, for Neoproterozoic rocks, the concordia curve has a shallow enough slope, and the 207 Pb – 235U dates measured precisely enough to be able to evaluate discordance at the per mil level. However, this is not the case for microbeam U – Pb dates. As outlined above, microbeam U – Pb dates on volcanic rocks rely upon the averaging of a relatively high-n dataset (10 –20) of relatively imprecise (c. 2– 4%) U –Pb determinations to get a weighted mean date with a precision of c. 1% or less. Underpinning these lower uncertainties is the assumption of a single population with normally distributed errors. However, it is the low precision of each analysis combined with variability of the standard analyses that bracket unknowns that often precludes the detection of subtle amounts of Pb loss or inheritance. Stated another way, if the amount of Pb loss or inheritance is less than the precision of a single spot analysis, then it cannot be detected via normal statistical proxies (such as the MSWD), so the assumption of a normal distribution may be invalid (Fig. 9.4). If Pb loss is the main source of open-system behaviour, this will have the effect of lowering the 206Pb – 238U date on some analyses, as well as the weighted mean 206Pb– 238U date. If unrecognized, the younger dates could be misconstrued as the true crystallization age. To highlight the problem of assigning dates to samples from multiple U –Pb analyses, we compare two published studies on zircons from the same ash bed from the lower Doushantuo Formation of China. Zhang et al. (2005) obtained a weighted-mean 206 Pb – 238U age of 621 + 7 Ma (n ¼ 13, MSWD ¼ 1.13, does not include the 0.48% error in U – Pb calibration) from concordant U –Pb zircon analyses on their sample 04SC20A, which occurs c. 2 m above the cap carbonate in the Yangtze Gorge region. Condon et al. (2005) sampled the same ash bed (their sample YG04-2) and analysed the zircons by the ID-TIMS method. The ID-TIMS dataset showed a spread in U –Pb dates that formed a linear array below Concordia, indicating zero-age Pb loss. Three of the ten zircons analysed were ‘concordant’ and yield a weighted mean 206Pb– 238U date of 632.3 + 0.5 Ma (MSWD ¼ 1.15, does not include the 0.1% error in U –Pb tracer calibration). The two dates are statistically different and do not overlap, even when estimates of random and systematic error are considered; however, when compared at the precision of a single analysis, all the SHRIMP U –Pb dates overlap with the ID-TIMS dates (Fig. 9.4). A likely interpretation of this dataset is that a number of SIMS analyses were located on domains that had lost a small amount of Pb, resulting in a lower mean 206Pb– 238U date when all the analyses are included. The higher precision of the ID-TIMS analyses allows small amounts of Pb loss to be recognized and those data points not included in calculating the final 206Pb– 238U date. A third U –Pb date has been published for this same ash bed: Yin et al. (2005) obtained a weighted mean 206Pb – 238U age of 628.3 + 5.8 Ma (MSWD ¼ 0.86). This publication does not include any details of the U –Pb data such as error in U –Pb normalization, but it is presumed this weighted mean uncertainty does not include any systematic sources of uncertainty. This date

USER’S GUIDE TO NEOPROTEROZOIC GEOCHRONOLOGY

(a)

206

Pb/

238

U

680

640

0.104

620 0.100

600 0.096

0.092

560 0.5

0.6

0.8

0.9

207

235

207

235

Pb/

U

Pb/

238

U

(b)

206

635

0.1030

0.1025

625

0.1010 0.848

0.852

0.856

0.860

0.864

0.868

Pb/

U

(c) 660

640

620

600

580

SHRIMP analyses (Zhang et al. 2005)

ID-TIMS analyses (Condon et al. 2005)

Fig. 9.4. U –Pb zircon data on samples from the lower Doushantuo Formation (2 m above the cap carbonate), Yangtze Gorge region, China. (a) U– Pb concordia plot of SHRIMP U –Pb (zircon) dataset obtained on zircon from (Zhang et al. 2005). (b) U– Pb concordia plot of ID-TIMS U–Pb (zircon) dataset for the same ash bed (Condon et al. 2005). (c) Plot of single analysis 206 Pb – 238U dates of both the SHRIMP and ID-TIMS datasets. The grey bar reflects the weighted mean SHRIMP U–Pb date (all points included) and ID-TIMS date (three concordant analyses included). The error bar incorporates error in U–Pb calibration (SHRIMP and TIMS).

143

is indistinguishable from either the other SHRIMP date (Zhang et al. 2005) or the ID-TIMS date (Condon et al. 2005). As a second example of problems related to determining a sample age from a population of U –Pb (zircon) dates we present new ID-TIMS data on zircon from another sample that was previously analysed by SHRIMP to derive a U – Pb date (Fig. 9.5). Fanning and Link (Fanning & Link 2004) dated samples from the Neoproterozoic Pocatello Formation, Idaho, using the SHRIMP U –Pb method on zircon from a series of volcanic layers and clasts within diamictite. We obtained new ID-TIMS U – Pb zircon data on one of these samples, 06PL00, that had yielded a weighted mean SHRIMP 206Pb/238U date of 708 + 8 Ma (n ¼ 18, MSWD ¼ 1.7, includes additional 0.35% error in U/Pb normalization). For a population of this size an MSWD of 1.7 is at the upper limit of acceptability (Wendt & Carl 1991), possibly indicating scatter due to non-analytical causes, that is real age variation in the population. Six single zircon crystals have been analysed by ID-TIMS (Table 9.2 & Fig. 9.5) at the Massachusetts Institute of Technology following the analytical procedure outlined in Bowring et al. (2007). Three zircons are from the same mineral separate used for the SHRIMP study (06PL00) and three are from an attempted recollection (36PL05) at the same stratigraphic level as sample 06PL00 in the Scout Mountain Member of the Pocatello Formation, with the authors of the original study. One zircon (00PL06 z2) is concordant at c. 705 Ma, whereas the remaining five zircons form a population at c. 690 Ma. Within this population, four zircons give equivalent 206Pb– 238U dates with a weighted mean of 687.4 + 0.6/1.3 Ma, and the fifth zircon (34PL05 z4) gives a similar 207Pb– 206Pb date but a slightly younger 206Pb– 208U date, which we interpret to reflect Pb loss. This new ID-TIMS U – Pb date of 687.4 + 1.3 Ma (includes 1‰ tracer calibration error) is distinctly younger than the weighted mean SHRIMP U –Pb date of 705 + 8 Ma (includes 0.35% U –Pb calibration error) for the same sample and stratigraphic level (Fanning & Link 2004). However, when we look at the data in terms of single analysis data points, we can see that there are some similarities between the two datasets (Fig. 9.5): they both contain zircon 206Pb– 238U dates of c. 705 Ma and c. 690 Ma. The proportions of these different age populations is different, with the ID-TIMS dataset being dominated by c. 690 Ma grains, whereas this age population is subordinate in the SHRIMP dataset, most likely reflecting the low-n nature of the ID-TIMS dataset and perhaps differences in pre-selection of zircons prior to analysis based upon morphology and so on. Importantly, the relatively low precision of the SHRIMP U – Pb dates means that it is not possible to discern different populations at the ,3% level. The assumption of a single (age) population and resultant weighted mean calculation results in a U –Pb date that is c. 20 Ma older than the concordant ID-TIMS population, which is considered the most robust estimate of the age of the rock. These two case studies from the lower Doushantuo Formation, China, and the Scout Mountain Member, Idaho, serve to illustrate the problems related to the recognition of real scatter, be it due to Pb loss and/or a mixed age population, at a resolution that is less than the precision of single analysis. In such cases, exploiting high-n datasets to derive weighted mean dates with lower uncertainties can result in inaccurate age assignments for a given sample. This section is not meant to suggest all SHRIMP U –Pb data are inaccurate but to demonstrate that beyond the level of a single analysis uncertainty (i.e. in the calculation of weighted mean dates) assumptions are being made about the nature of the population. If these assumptions are valid then the weighted mean date and associated error will be accurate; however, if these assumptions are not valid (i.e. it is not a single age population with normally distributed errors), then the calculated weighted mean date and associated error may be an inaccurate reflection of the sample age. In situ techniques are without question essential tools for characterizing complex zircons from volcanic and

144

D. J. CONDON & S. A. BOWRING

metamorphic rocks and for rapid analysis of detrital populations. Ideally, microbeam techniques can be used to rapidly characterize a population of zircons by analysing a small volume of many zircons, which could then be followed by conventional highprecision geochronology of selected grains. Linear arrays v. isochrons. As outlined above, a limitation of the isochron approach is the lack of an independent check for opensystem behaviour, unlike the dual decay scheme of the U –Pb system. Most studies use a combination of bracketing age constraints and/or a statistical measure of coherence (MSWD or uncertainty) to assess whether the system has been perturbed; however, precision and amount of scatter cannot be used as a proxy for closed-system behaviour. For example, organic-rich sediments from the Aralka Formation, Australia, have been analysed for Re –Os geochronology using both the aqua regia and CrO3 – H2SO4 dissolution methods. Schaffer & Burgess (2003) used the aqua regia dissolution method and obtained a three-point isochron (samples integrated over 1.6 m stratigraphic thickness) of 592 + 14 Ma (MSWD  1). An expanded dataset collected over 10 m stratigraphic thickness yielded a nine-point regression and an age of 623 + 18 Ma (MSWD ¼ 5.2). Subsequent Re –Os analyses (on samples from a 2 m interval within the 10 m interval sampled by Schaefer & Burgess (2003) using the CrO3 – H2SO4 dissolution method yielded a 10-point isochron with an age of 657.2 + 5.4 Ma (2s internal uncertainties, MSWD ¼ 1.2) (Kendall et al. 2006). This difference is attributed to either a sampling and/or analytical artefact(s) related to sample digestion and/or sample-spike equilibration (Kendall et al. 2006, 2009). It is clear from the Aralka case study that one cannot use coherence of a dataset as a means to assess accuracy, especially with isochrons based upon low-n datasets (or sub-sets). The bottom line is that a suite of samples with the same initial ratio and a range of parent/daughter ratios that evolve in a closed system can yield an isochron; however, a suite of samples that define a linear array alone cannot be inferred to represent closedsystem behaviour, as simple mixing of two reservoirs can yield linear arrays.

Dating the Neoproterozoic Geochronology plays an important role in the independent integration of geographically disparate stratigraphic sections and therefore plays a key role in addressing questions pertaining to the correlation of glaciogenic intervals and their synthesis in terms of number of glacial events, their duration and geographic extent. Several recent studies have assessed the published age constraints for Neoproterozoic glaciations in order to address such questions, and we briefly summarize the current state of play in light of what we consider the key datasets that pertain to the tempo of glaciation, the development of early animals and other interesting events.

How many glacial intervals?

Fig. 9.5. U– Pb zircon data on samples from the Scout Mountain Member, Pocatello Formation, southern Idaho. (a) U– Pb concordia plot of SHRIMP U–Pb (zircon) dataset obtained on sample 00PL00 (Fanning & Link 2004). (b) New U –Pb concordia plot of ID-TIMS U–Pb (zircon) dataset on zircons from the 00PL00 mineral separate and a recollection from the same stratigraphic level (see Table 9.2). (c) Plot of single analysis 206Pb– 238U dates of both the SHRIMP and ID-TIMS datasets. The grey bar reflects the weighted mean SHRIMP U–Pb date (all points included) and ID-TIMS date (00PL06 z2a and 34PL05 z4 excluded) and error incorporates error in U–Pb calibration.

In the past ten years, the number and quality of geochronological data have increased considerably and (at least) three distinct glacial ‘intervals’ have been geochronologically delineated. These temporally distinct glacial intervals are also broadly distinct in terms of the physical characteristics of associated strata (Kennedy et al. 1998; Hoffman & Schrag 2002; Halverson et al. 2005). I. Ediacaran glaciations. Ediacaran-age (c. 635 to 542 Ma) glacial

deposits are preserved on at least three palaeocontinents. The Gaskiers Formation on the eastern Avalonian Peninsula has been directly dated by U –Pb (zircon) ID-TIMS dates on ash beds that

USER’S GUIDE TO NEOPROTEROZOIC GEOCHRONOLOGY

145

Table 9.2. ID-TIMS U-Pb data on zircons from 06PL00 and 34PL05 Concentrations

Age (Ma) 206

Fractions1

U (ppm)2

Pb (ppm)2

Th U3

Pb*/Pbc4

34PL05 z19 34PL05 z29 34PL05 z49 06PL00 z1 06PL00 z2 06PL00 z3

55 33 67 13 42 52

7 4 9 2 5 7

0.69 0.50 1.07 0.49 0.78 0.78

33.5 35.8 74.3 10.9 22.5 18.4

208

207

204

206

206

1935.4 2174.6 3915.9 672.6 1278.0 1052.3

0.2160 0.1559 0.3331 0.1541 0.2433 0.2426

0.0625 0.0626 0.0625 0.0625 0.0630 0.0625

Pb* Pb5

Pb Pb6

Pb Pb6

207

% err7 0.13 0.24 0.09 0.35 0.13 0.15

Pb U6

235

0.96940 0.97056 0.96463 0.96970 1.00320 0.96939

206

% err7 0.18 0.31 0.12 0.48 0.21 0.25

Pb U6

238

0.11254 0.11253 0.11203 0.11252 0.11551 0.11248

% err7

corr. coef.

0.12 0.19 0.08 0.30 0.16 0.19

0.696 0.655 0.690 0.676 0.779 0.795

206

Pb U8

238

687.49 687.44 684.51 687.37 704.68 687.13

207

Pb U8

235

688.15 688.75 685.69 688.31 705.43 688.15

207

Pb Pb8

206

690.35 693.04 689.56 691.35 707.82 691.53

1

z1, z2, etc. are fractions composed of single grains. Grains were annealed and leached prior to dissolution. See Bowring et al. (2007) for analytical procedures. Volumes estimated from images, absolute uncertainty is c. 50%. 3 Model Th–U ratio calculated from radiogenic 208Pb/206Pb ratio and 207Pb/206Pb age. 4 Ratio of radiogenic Pb (including 208Pb) to common Pb. 5 Measured ratio corrected for spike and fractionation only. Mass fractionation corrections were based on analysis of NBS-981 and NBS-983. Corrections of 0.25 + 0.04%/ amu (atomic mass unit) and 0.07 + 0.04%/amu were applied to single-collector Daly analyses and dynamic Faraday– Daly analyses, respectively. 6 Corrected for fractionation, spike and blank. All common Pb was assumed to be procedural blank. 7 Errors are 2s, propagated using the algorithms of Ludwig (1980). 8 Calculations are based on the decay constants of Jaffey et al. (1971). 9 34PL05 was collected within a few hundred metres of the original 06PL00 sample locality and at the same stratigraphic level within the Scout Mountain Member. See Bowring et al. (2007) for details of the analytical methods. 2

occur below, within and above the glacial Gaskiers Formation, constraining its age to c. 584 to 582 Ma with a maximum duration of 2.6 Ma (Bowring et al. 2003). Similar-age (,600–542 Ma) glacial successions occur elsewhere on Avalonia (Thompson & Bowring 2000), the NE margin of Laurentia (Condon & Prave 2000), Baltica (Bingen et al. 2005) and the Tarim Block, NW China (Xu et al. 2009); however, the age of these units is less well constrained, making it impossible to assess the synchronicity of Ediacaran glaciation(s). II. ‘Marinoan’ glaciations. Marinoan-type glacial deposits are con-

sidered those associated with cap carbonates with distinctive lithological and isotopic similarities (Kennedy et al. 1998; Hoffman et al. 2007) and they occur on nearly every palaeocontinent. A maximum age constraint for the Marinoan glaciation comes from China and Australia. In China, the top of pre-glacial Datangpo Formation is constrained by a U – Pb (zircon) SHRIMP date of 654.5 + 3.8 Ma (Shihong et al. 2008). In Australia, a Re –Os isochron age of 643 + 2.4 Ma (Kendall et al. 2006) was obtained from black shale of the Tindelpina Member at the base of the pre-Marinoan Tapley Hill Formation. Synglacial age constraints come from 635.5 + 1.2 Ma U– Pb (zircon) ID-TIMS date from an ash bed within the basinal facies of the Ghaub Formation (Hoffmann et al. 2004) and a similar 636.3 + 4.9 Ma U –Pb (zircon) SHRIMP date from the base of the Nantuo Formation (Shihong et al. 2008). Termination of this glacial episode is constrained by an ash bed within the cap carbonate of the Nantuo Tillite (Yangtze Platform), which has been dated at 635.2 + 0.6 Ma (Condon et al. 2005). Combined, these data indicate a duration of ,10 Ma and globally synchronous termination of the Marinoan glaciation at c. 635 Ma. It should be noted that there still remains a lack of consensus over the age of the actual Marinoan glacial in its type section and inferentially the base of the Ediacaran Period. III. Sturtian glaciation(s). Sturtian-type glacial deposits are typically defined as being stratigraphically lower than Marinaon-type deposits and are commonly associated with dark, organic-rich cap carbonates and sedimentary ironstones. On the Congo craton, the Chuos Formation overlies the Naauwpoort volcanics dated at 746 + 2 Ma (Hoffman et al. 1996), providing a robust maximum age constraint. In Oman, the glaciogenic Ghubrah

Formation is dated at c. 712 Ma (Bowring et al. 2007), although this glacial deposit, where dated, is not associated with a cap carbonate. U –Pb (zircon) SHRIMP dates from the Pocatello Formation, Idaho (Fanning & Link 2004), are commonly cited as age constraints for Sturtian glaciation. However, as we have outlined above, there are accuracy issues with some of these dates and, perhaps more importantly, their geological context has been questioned (Fanning & Link 2008). Similarly, commonly cited age constraints from the Edwardsburg Formation preserved in roof pendants of the Idaho Batholith (Lund et al. 2003) lack robust sedimentological and stratigraphic constraint in order to assess their relevance for the Sturtian glacial episode. Relevant age constraints from post-Sturtian and pre-Marinoan glacial deposits come from both Australia and China. In Australia, a Re –Os date of 657.2 + 5.4 Ma (Kendall et al. 2006) for the basal Aralka Formation provides a minimum constraint on the age of the underlying Areyonga glacial deposits in the Amadeus Basin, which is correlated with the type-Sturtian glacial deposits (Walter et al. 2000). In China, a U –Pb (zircon) ID-TIMS date of 663 + 4 Ma from the Datangpo Formation (Zhou et al. 2004) directly overlies the glacial Tiesiao Formation and its associated Mn-rich cap carbonate, again providing a minimum age constraint on a Sturtian-type glacial deposit. Thus, Sturtian-type glacial deposits are constrained to the interval c. 746– 663 Ma. However, the nature of the current dataset, with issues remaining about the glacial affinity/stratigraphic context of certain successions as well as those relating to the radio-isotopic dates themselves, means that it is not possible to infer anything about duration and/or synchronicity of the glacial deposits and, inferentially, the nature of the glacial episode(s) they reflect. IV. Pre-Sturtian? glaciations. Evidence for a pre-Sturtian glaciation (c. 740 –760 Ma) comes from a small number of sections in Namibia (Frimmel et al. 1996), Zambia (Key et al. 2001) and NW China (Xu et al. 2009); however, in each case, either the exact relationship of the dates to the glacial unit is inferred and a complicated local stratigraphy lessens the confidence that the dated unit directly constrains a glacial deposit (Frimmel et al. 1996; Key et al. 2001) and/or the glacial nature of the dated unit remains equivocal (Frimmel et al. 1996; Xu et al. 2009). As such, the community should remain circumspect about an older Cryogenian glaciation until more equivocal evidence is presented.

146

D. J. CONDON & S. A. BOWRING

Dating the earliest fossils

Other Neoproterozoic events

There is growing consensus on the development of Ediacaran-age organisms based upon the sequencing of key assemblages via high-precision U –Pb dating. We briefly review the nature of, and age constraints for, key assemblages/Fauna.

In addition to the climatic and biological events/developments that took place during the Neoproterozoic, there were also a series of other events that took place, including impact events and major isotope excursions. Age constraints for most of these events are based upon correlation-based inference and as such we do not discuss them here. One event, the Shuram/Wonoka carbon isotope excursion, is directly constrained by radio-isotopic dating and is discussed further below.

I. Avalon assemblage. The Avalon assemblage is preserved in

relatively deep-water marine settings within the Avalon zone of Newfoundland and England, and is characterized by rangemorphs, frond- and bush-shaped colonies, which lacked mobility. The earliest Ediacaran (macroscopic) fossil (Charnia, C. masoni, C. wardi), fronds some 2 m in length, occur in the Drook Formation (Narbonne & Gehling 2003) and are dated at c. 577 Ma (Bowring et al. 2003), some 5 Ma after the end of the Gaskiers glaciation. This assemblage extends in range to c. 560 Ma in Newfoundland and England; however, this upper limit requires more robust constraint. II. White Sea assemblage. The White Sea assemblage is preserved in

a shallow marine setting in the Vendian sections of the White Sea and the Ediacaran Member of Australia (Narbonne 2005). This assemblage includes fronds similar to those of the Avalon assemblage, but also includes segmented and discoid fossils (such as Kimberella) as well as horizontal traces indicating mobile bilaterians were present, marking the advent of muscularity recorded in the fossil record. In the White Sea sections, this assemblage is constrained to c. 555 Ma (Martin et al. 2000), probably extending to close to the base of the Cambrian. III. Nama assemblage. Nama assemblage fossils, typified by those of the Nama Group, Namibia, include rangemorphs (mainly multifrondate fronds) and horizontal traces similar to the White Sea assemblage, with an age c. 550 Ma to the base of the Cambrian. An important addition to the Nama assemblage are the weakly calcified metazoans, Cloudina and Namacalathus, representing a major evolutionary development (calcification) and providing a record of predation (Hua et al. 2003). These weakly calcified metazoans occur in many terminal Ediacaran successions (Oman, Namibia, China and Canada), and their age range is constrained from c. 550 Ma to the base of the Cambrian (Grotzinger et al. 1995; Amthor et al. 2003; Condon et al. 2005). IV. Doushantuo fauna. The Doushantuo fauna comprise phospha-

tized eggs and embryos (Xiao et al. 1998, 2000), as well as microscopic sponges (Li et al. 1998), and thus provides critical constraint upon the earliest animals. The age of the Doushantuo Formation is constrained to c. 635 –550 Ma (Condon et al. 2005). These fossils occur in the middle of the formation and are therefore likely c. 590 to 570 Ma, although this remains to be constrained by direct dating. The Miaohe biota of the upper Doushantuo Formation contains diverse macroalgal assemblages (Xiao et al. 2002). Although not part of the White Sea assemblage, the Miaohe biota is similar in age, being slightly older than the 550.1 + 0.6 Ma ash from the top of the Doushantuo Formation (Condon et al. 2005), perhaps indicating a temporal coincidence between the development of macroalgae and bilaterians (White Sea assemblage, see above). Recently Love et al. (2009) documented an abundance of a Demospongiae-specific biomarker from Cryogenian (c. 710– 635 Ma) sedimentary rocks in Oman. These data indicate a pre-Ediacaran age for the advent of multicellularity. Finally, the disappearance of these Ediacaran fossil assemblages (Rangemorphs, weakly calcified metazoans) is coincident with the base of the Cambrian and a negative d13C isotope excursion (Grotzinger et al. 1995; Amthor et al. 2003). The base of the Cambrian is dated at c. 541 Ma (Amthor et al. 2003; Bowring et al. 2007). The Early Cambrian fossil record documents a period of enhanced biomineralization and increased muscular development.

I. Shuram/Wonoka carbon isotope excursion. The Shuram/Wonoka carbon isotope excursion (CIE) is a pronounced large-magnitude (c. 15‰) d13C excursion that is recorded in Oman, Australia, China, Siberia, Namibia and several other locations (Burns & Matter 1993; Calver 2000; Condon et al. 2005; Melezhik et al. 2005, 2008). Termination of this event has been dated at c. 550 Ma in the Doushantuo Formation, where d13C cross from negative to positive (Condon et al. 2005). The onset (and inferentially the duration) has yet to be satisfactorily constrained, with detrital zircons from the base of the CIE providing maximum age constraints dated at c. 620 Ma (Bowring et al. 2007, 2009). The duration of this unique CIE is a matter of some debate, with estimates varying from c. 50 Ma (Le Guerroue et al. 2006) to c. 5– 11 Ma (Bowring et al. 2007). Constraining the timing of onset of the Shuram/Wonoka CIE is critically important to assessing hypotheses that implicate this event in oxygenation of oceans and early animal evolution (Condon et al. 2005; Fike et al. 2006).

Future directions Clearly, research on the Neoproterozoic Earth System would greatly benefit from more and, perhaps more importantly, better dates. The number of possible dates is limited by the identification of suitable sections/materials for analyses: ash bed bearing successions for U– Pb (zircon) dating, organic-rich sediments for Re – Os analyses, and so on. There is undoubtedly a wealth of material to be analysed from intervals of interest. Timely identification of targets among the background of similar-looking material, which will not yield meaningful age constraints, requires a change in approach, with the development of better screening protocols, in order to maximize efficiency From the U –Pb zircon perspective we propose there is great utility in breaking down the microbeam v. ID-TIMS barrier. We have attempted to highlight in this chapter that these two analytical methods have complementary attributes: the microbeam U –Pb techniques are capable of high sample throughput, a factor of c. 10 –20 greater than U –Pb ID-TIMS, whereas the ID-TIMS U –Pb analyses are c. 10– 20 times more precise. Sample throughput for ID-TIMS analyses is an issue, and what may not be appreciated by the non-practitioner is that for every dated sample that makes it into the literature, there will have been many more that have been analysed and yielded no meaningful chronological information, most likely due to a predominance of inherited/xenocrystic material. An optimized approach for an enhanced U –Pb (zircon) based chronology for the Neoproterozoic would involve a first-phase microbeam U – Pb dating, where samples are effectively ‘screened’, prior to a second phase involving ID-TIMS U –Pb dating of samples that yield microbeam U –Pb dates of interest. Where samples contain mixed-age zircon populations it would be possible to either remove zircons from a resin mount of ID-TIMS analyses or, preferably, to simply mount the zircons on double-sided sticky tape, making removal for ID-TIMS analysis straightforward. This approach is not suitable for SIMS U – Pb analyses, but is easily adaptable for LA-ICP-MS analyses. Combining the efficiency of the microbeam techniques with the precision of ID-TIMS would be the most effective plan of attack for improving

USER’S GUIDE TO NEOPROTEROZOIC GEOCHRONOLOGY

the chronology of Neoproterozoic strata, and this would require a collaborative approach between the various groups involved.

Conclusions There are a number of radio-isotopic dating techniques that can be used to constrain the age of Neoproterozoic sedimentary successions These include U –Pb dating of minerals (predominantly zircon) from interlayered volcanic rocks, and Re – Os or Pb/Pb isochron dating of organic-rich sediments or carbonates. Most of the Neoproterozoic time scale is calibrated by U –Pb zircon dates from air-fall tuffs, but this dataset is augmented by a growing number of Re –Os and Pb/Pb isochron dates. Each of the radio-isotopic systems has different strengths and weaknesses. U –Pb (zircon) dating is considered the premier geochronometer; however, it is of limited use in successions devoid of zircon-bearing volcanics. Realizing the goal of a robust and highly resolved temporal framework for the Neoproterozoic will require exploitation of all these different methodologies and refined searches for volcanic air fall deposits.

All dates involve some subjective interpretation Dates interpreted as depositional ages are usually based on multiple analyses from the same stratigraphic horizon. In the case of U –Pb (zircon) dates from volcanic rocks, the reported age is usually based on a weighted mean 206Pb – 238U date in which a variable number of single 206Pb– 238U dates (either single-grain or single-spot analyses) are weighted (based upon their associated uncertainty) and a mean calculated. Underpinning this mean date, and its lower uncertainty, is the assumption of a single population with normally distributed errors, an assumption that is not always valid. Isochron dates are similarly based upon the linear regression of a number of data points that are assumed to be cogenetic, have a common initial daughter isotopic composition, and the samples analysed have acted as a ‘closed system’ since their formation. Although most attempt to produce age constraints with the lowest possible uncertainty, the best measure of high-precision age constraints is the precision of a single analysis. It is this precision that controls the evaluation of ‘open-system’ behaviour and thus the accuracy of the final date. Although the coherence of a dataset is often used as a proxy for closed-system behaviour, the coherence is limited by the precision of the individual analyses.

The uncertainty of a date is as important as the date itself (Ludwig 2003) The total uncertainty of a radio-isotopic date comprises random (or internal) and systematic (or external) components. The random/ internal uncertainties are related to the measurement of the isotopic ratios and the corrections applied. Systematic uncertainties are those related to the uncertainty in absolute value of various constant parameters used in the calculation of either an isotopic ratio or in the calculation of the date itself. Analytical uncertainties should reflect the ability to reproduce a given isotopic ratio and represent the minimum uncertainty that may be considered. For microbeam U –Pb dates, the standard calibration is best considered as a non-systematic uncertainty and should be incorporated into each individual spot U –Pb date uncertainty (Ireland & Williams 2003). If comparing dates generated using different techniques or using different calibration materials (such as mineral standards for microbeam dates or isotopic tracer for ID-TIMS), then the systematic uncertainties related to these calibrations must be

147

considered. An additional systematic uncertainty arises from decay constant uncertainties. These uncertainties have been determined experimentally or assessed via geological comparison with another decay scheme. There is considerable variation in the published literature regarding the treatment of the constituent parts of the total uncertainty. In some cases, such as when using ID-TIMS 206Pb– 238U dates generated using a single isotopic tracer solution, certain components can be ignored (in this case tracer calibration uncertainty and 238U decay constant uncertainty). This can be useful when attempting to determine sediment accumulation rates (Bowring et al. 2007) or assess the synchronicity of events (Condon et al. 2005). The constituent of the total uncertainty that is most difficult to quantify is that related to the geology. This can be considered at the scale of relating a given mineral to a stratigraphic level (i.e. do the magmatic zircons date the exact age of a given air-fall tuff?). However, a first-order uncertainty is often the relationship of a dated unit to the section of interest. Regardless of how robust the radio-isotopic data are, if the geological context of a dated sample is poorly constrained, the date is of little use.

A robust, highly resolved chronology for the Neoproterozoic is yet to be fully realized Although there has been a marked increase in the number of radioisotopic age constraints for Neoproterozoic sedimentary successions in the past decade, there is still much we do not know, limiting our ability to test correlations and rate dependent hypotheses. This is in part due to the lack of suitable material to date in key sections, although it is also due to the reproducibility of some of the radio-isotopic techniques and systems employed. Accelerated progress could be made through increased collaboration within the community, so the most appropriate and efficient methodologies are applied to the intervals of interest. We thank E. Arnaud, M. Babinbski, J. Etienne, M. Horstwood, S. Kamo, B. Kendall, A. Maloof and T. Prave for constructive reviews. This chapter is an outgrowth of a continuing research programme supported by the NSF, the NASA Astrobiology Program and NERC. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Amthor, J. E., Grotzinger, J. P., Schroder, S., Bowring, S. A., Ramezani, J., Martin, M. W. & Matter, A. 2003. Extinction of Cloudina and Namacalathus at the Precambrian – Cambrian boundary in Oman. Geology, 31, 431– 434. Babinski, M., Van Schmus, W. R. & Chemale, F. 1999. Pb/Pb dating and Pb isotope geochemistry of Neoproterozoic carbonate rocks from the Sao Francisco basin, Brazil: implications for the mobility of Pb isotopes during tectonism and metamorphism. Chemical Geology, 160, 175– 199. Babinski, M., Vieira, L. C. & Trindade, R. I. F. 2007. Direct dating of the Sete Lagoas cap carbonate (Bambui Group, Brazil) and implications for the Neoproterozoic glacial events. Terra Nova, 19, 401– 406. Barfod, G. H., Albarede, F., Knoll, A. H., Xiao, S. H., Telouk, P., Frei, R. & Baker, J. 2002. New Lu– Hf and Pb/Pb age constraints on the earliest animal fossils. Earth and Planetary Science Letters. 201, 203–212. Begemann, F., Ludwig, K. R. et al. 2001, Call for an improved set of decay constants for geochronological use. Geochimica et Cosmochimica Acta, 65, 111– 121. Bingen, B., Griffin, W. L., Torsvik, T. H. & Saeed, A. 2005. Timing of Late Neoproterozoic glaciation on Baltica constrained by detrital zircon geochronology in the Hedmark Group, south-east Norway. Terra Nova, 17, 250–258.

148

D. J. CONDON & S. A. BOWRING

Black, L. P., Kamo, S. L., Allen, C. M., Aleinikoff, J. N., Davis, D. W., Korsch, R. J. & Foudoulis, C. 2003. TEMORA 1: a new zircon standard for Phanerozoic U– Pb geochronology. Chemical Geology, 200, 155–170. Bowring, S., Myrow, P., Landing, E. & Ramezani, J. 2003. Geochronological constraints on terminal Neoproterozoic events and the rise of metazoans. Geophysical Research Abstracts, 5, 219. Bowring, S. A., Schoene, B., Crowley, J. L., Ramezani, J. & Condon, D. J. 2006. High-precision U –Pb zircon geochronology and the stratigraphic record: progress and promise. The Paleontological Society Papers, 12, 25– 46. Bowring, S. A., Grotzinger, J. P., Condon, D. J., Ramezani, J., Newall, M. & Allen, P. A. 2007. Geochronologic constraints of the chronostratigraphic framework of the Neoproterozoic Huqf Supergroup, Sultanate of Oman. American Journal of Science, 307, 1097– 1145. Bowring, S. A., Grotzinger, J. P., Condon, D. J., Ramezani, J. & Newall, M. J. 2009. Reply to Comment: Oman Chronostratigraphy. American Journal of Science, 309, 91– 96. Burns, S. J. & Matter, A. 1993. Carbon isotopic record of the latest Proterozoic from Oman. Eclogae Geologicae Helvetiae, 86, 595–607. Calver, C. R. 2000. Isotope stratigraphy of the Ediacaran (Neoproterozoic III) of the Adelaide Rift Complex, Australia, and the overprint of water column stratification. Precambrian Research, 100, 121–150. Chen, D. F., Dong, W. Q., Zhu, B. Q. & Chen, X. P. 2004. Pb/Pb ages of Neoproterozoic Doushantuo phosphorites in South China: constraints on early metazoan evolution and glaciation events. Precambrian Research, 132, 123– 132. Cherniak, D. J. & Watson, E. B. 2003. Diffusion in Zircon. Reviews in Mineralogy and Geochemistry, 53, 113– 143. Cohen, A. S., Coe, A. L., Bartlett, J. M. & Hawkesworth, C. J. 1999. Precise Re–Os ages of organic-rich mudrocks and the Os isotope composition of Jurassic seawater. Earth and Planetary Science Letters, 167, 159–173. Condon, D. J. & Prave, A. R. 2000. Two from Donegal: Neoproterozoic glacial episodes on the northeast margin of Laurentia. Geology, 28, 951– 954. Condon, D., Zhu, M. Y., Bowring, S., Wang, W., Yang, A. H. & Jin, Y. G. 2005. U– Pb ages from the Neoproterozoic Doushantuo Formation, China. Science, 308, 95– 98. Condon, D., Schoene, B., Bowring, S., Parrish, R., McLean, N., Noble, S. & Crowley, Q. 2007. EARTHTIME; isotopic tracers and optimized solutions for high-precision U–Pb ID-TIMS geochronology. Eos, Transactions, American Geophysical Union, 88, Fall Meeting Supplement, Abstract V41E-06. Creaser, R. A., Sannigrahi, P., Chacko, T. & Selby, D. 2002. Further evaluation of the Re– Os geochronometer in organic-rich sedimentary rocks: a test of hydrocarbon maturation effects in the Exshaw Formation, Western Canada Sedimentary Basin. Geochimica et Cosmochimica Acta, 66, 3441– 3452. Evans, J. A., Zalasiewicz, J. A., Fletcher, I., Rasmussen, B. & Pearce, N. J. G. 2002. Dating diagenetic monazite in mudrocks: constraining the oil window? Journal of the Geological Society, 159, 619– 622. Fanning, C. M. & Link, P. K. 2004. U –Pb SHRIMP ages of Neoproterozoic (Sturtian) glaciogenic Pocatello Formation, southeastern Idaho. Geology, 32, 881–884. Fanning, C. M. & Link, P. K. 2008. Age constraints for the Sturtian glaciation: data from the Adelaide Geosyncline, South Australia and Pocatello Formation, Idaho, USA, Selwyn Symposium Vol. 91. Geological Society of Australia Abstracts, Melbourne, Geological Society of Australia, 57 – 62. Fike, D. A., Grotzinger, J. P., Pratt, L. M. & Summons, R. E. 2006. Oxidation of the Ediacaran Ocean. Nature, 444, 744–747. Folling, P. G., Zartman, R. E. & Frimmel, H. E. 2000. A novel approach to double-spike Pb/Pb dating of carbonate rocks: examples from Neoproterozoic sequences in southern Africa. Chemical Geology, 171, 97 –122. Frimmel, H. E., Klotzli, U. S. & Siegfried, P. R. 1996. New Pb/Pb single zircon age constraints on the timing of Neoproterozoic glaciation and continental break-up in Namibia. Journal of Geology, 104, 459– 469.

Grotzinger, J. P., Bowring, S. A., Saylor, B. Z. & Kaufman, A. J. 1995. Biostratigraphic and geochronological constraints on early animal evolution. Science, 270, 598–604. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Toward a Neoproterozoic composite carbonisotope record. Geological Society of America Bulletin, 117, 1181– 1207. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129–155. Hoffman, P. F., Hawkins, D. P., Isachsen, C. E. & Bowring, S. A. 1996. Precise U–Pb zircon ages for early Damaran magmatism in the Summas Mountains and Welwitschia Inlier, northern Damara Belt. Communications of the Geological Survey of Namibia, 11, 47– 53. Hoffman, P. F., Halverson, G. P., Domack, E. W., Husson, J. M., Higgins, J. A. & Schrag, D. P. 2007. Are basal Ediacaran (635 Ma) post-glacial ‘cap dolostones’ diachronous? Earth and Planetary Science Letters, 258, 114– 131. Hoffmann, K. H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. U– Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817– 820. Horstwood, M. S. A., Foster, G. L., Parrish, R. R., Noble, S. R. & Nowell, G. M. 2003. Common-Pb corrected in situ U– Pb accessory mineral geochronology by LA-MC-ICP-MS. Journal of Analytical Atomic Spectrometry, 18, 837– 846. Hua, H., Pratt, B. R. & Zhang, L.-Y. 2003. Borings in Cloudina shells: complex predator– prey dynamics in the terminal Neoproterozoic. Palaios, 18, 454– 459. Ireland, T. R. & Williams, I. S. 2003. Considerations in zircon geochronology by SIMS. Reviews in Mineralogy and Geochemistry, 53, 215– 241. Ireland, T. R., Flottmann, T., Fanning, C. M., Gibson, G. M. & Preiss, W. V. 1998. Development of the early Paleozoic Pacific margin of Gondwana from detrital-zircon ages across the Delamerian orogen. Geology, 26, 243–246. Jaffey, A. H., Flynn, K. F., Glendenin, L. E., Bentley, W. C. & Essling, A. M. 1971. Precision measurement of half-lives and specific activities of 235U and 238U. Physics Reviews, C4, 1889– 1906. Kendall, B. S., Creaser, R. A., Ross, G. M. & Selby, D. 2004. Constraints on the timing of Marinoan ‘Snowball Earth’ glaciation by 187 Re/187Os dating of a Neoproterozoic, post-glacial black shale in Western Canada. Earth and Planetary Science Letters, 222, 729– 740. Kendall, B., Creaser, R. A. & Selby, D. 2006. Re– Os geochronology of postglacial black shales in Australia: constraints on the timing of ‘Sturtian’ glaciation. Geology, 34, 729–732. Kendall, B., Creaser, R. A. & Selby, D. 2009. 187Re– 187Os Geochronology of Precambrian Organic-Rich Sedimentary Rocks. Geological Society, London, Special Publications, 326, 85 –107. Kennedy, M. J., Runnegar, B., Prave, A. R., Hoffmann, K. H. & Arthur, M. A. 1998. Two or four Neoproterozoic glaciations? Geology, 26, 1059–1063. Key, R. M., Liyungu, A. K., Njamu, F. M., Somwe, V., Banda, J., Mosley, P. N. & Armstrong, R. A. 2001. The western arm of the Lufilian Arc in NW Zambia and its potential for copper mineralization. Journal of African Earth Sciences, 33, 503. Kosler, J. & Sylvester, P. J. 2003. Present trends and the future of zircon in geochronology: laser ablation ICPMS. Zircon, 53, 243– 275. Krogh, T. E. 1982a. Improved accuracy of U– Pb zircon ages by the creation of more concordant zircon systems using an air abrasion technique. Geochimica et Cosmochimica Acta, 46, 637– 649. Krogh, T. E. 1982b. Improved accuracy of U –Pb zircon dating by selection of more concordant fractions using a high-gradient magnetic separation technique. Geochimica et Cosmochimica Acta, 46, 631– 635. Le Guerroue, E., Allen, P. A., Cozzi, A., Etienne, J. L. & Fanning, M. 2006. 50 Myr recovery from the largest negative d13C excursion in the Ediacaran ocean. Terra Nova, 18, 147– 153. Li, C. W., Chen, J. Y. & Hua, T. E. 1998. Precambrian sponges with cellular structures. Science, 279, 879– 882. Love, G. D., Grosjean, E. et al. 2009. Fossil steroids record the appearance of Demospongiae during the Cryogenian period. Nature, 457, 718– 721.

USER’S GUIDE TO NEOPROTEROZOIC GEOCHRONOLOGY

Ludwig, K. R. 1980. Calculation of uncertainties of U– Pb isotope data. Earth and Planetary Science Letters, 46, 212–220. Ludwig, K. R. 1991. Isoplot – a plotting and regression program for radiogenic isotope data. USGS Open File Report, 91– 445. Ludwig, K. R. 1998. On the treatment of concordant uranium –lead ages. Geochimica et Cosmochimica Acta, 62, 665– 676. Ludwig, K. R. 2000. Decay constant errors in U– Pb concordia-intercept ages. Chemical Geology, 166, 315–318. Ludwig, K. R. 2003. Mathematical-statistical treatment of data and errors for 230Th/U geochronology. Reviews in Mineralogy and Geochemistry, 52, 631– 656. Lund, K., Aleinikoff, J. N., Evans, K. V. & Fanning, C. M. 2003. SHRIMP U–Pb geochronology of Neoproterozoic Windermere Supergroup, central Idaho: implications for rifting of western Laurentia and synchroneity of Sturtian glacial deposits. Geological Society of America Bulletin, 115, 349–372. Mahan, K. H., Wernicke, B. P. & Jercinovic, M. J. 2010. Th–U –total Pb geochronology of authigenic monazite in the Adelaide rift complex, South Australia, and implications for the age of the type Sturtian and Marinoan glacial deposits. Earth and Planetary Science Letters, 289, 76– 86. Martin, M. W., Grazhdankin, D. V., Bowring, S. A., Evans, D. A. D., Fedonkin, M. A. & Kirschvink, J. L. 2000. Age of Neoproterozoic bilatarian body and trace fossils, White Sea, Russia: implications for metazoan evolution. Science, 288, 841– 845. Mattinson, J. M. 2000. Revising the ‘gold standard’ – the uranium decay constants of Jaffey et al. 1971. EOS, AGU Fall Meeting Supplement Abstact V61A-02. Mattinson, J. M. 2005. Zircon U –Pb chemical abrasion (‘CA-TIMS’) method: combined annealing and multi-step partial dissolution analysis for improved precision and accuracy of zircon ages. Chemical Geology, 220, 47 –66. Melezhik, V. A., Fallick, A. E. & Pokrovsky, B. G. 2005. Enigmatic nature of thick sedimentary carbonates depleted in 13C beyond the canonical mantle value: the challenges to our understanding of the terrestrial carbon cycle. Precambrian Research, 137, 131–165. Melezhik, V. A., Roberts, D., Fallick, A. E. & Gorokhov, I. M. 2008. The Shuram– Wonoka event recorded in a high-grade metamorphic terrane: insight from the Scandinavian Caledonides. Geological Magazine, 145, 161– 172. Narbonne, G. M. 2005. The Ediacara biota: Neoproterozoic origin of animals and their ecosystems. Annual Review of Earth and Planetary Sciences, 33, 421– 442. Narbonne, G. M. & Gehling, J. G. 2003. Life after snowball: the oldest complex Ediacaran fossils. Geology, 31, 27– 30. Peterson, K. J., McPeek, M. A. & Evans, D. A. D. 2005. Tempo and mode of early animal evolution: inferences from rocks, Hox, and molecular clocks. Paleobiology, 31, 36– 55. Rasbury, E. T. & Cole, J. 2009. Directly dating geologic events: U–Pb dating of carbonates. Reviews of Geophysics, 47, RG3001, doi: 10.1029/2007RG000246. Rasmussen, B. 2005. Radiometric dating of sedimentary rocks: the application of diagenetic xenotime geochronology. Earth-Science Reviews, 68, 197. Roden, M. K., Parrish, R. R. & Miller, D. S. 1990. The absolute age of the Eifelian Tioga Ash Bed, Pennsylvania. Journal of Geology, 98, 282– 285. Schaefer, B. F. & Burgess, J. M. 2003. Re– Os isotopic age constraints on deposition in the Neoproterozoic Amadeus Basin: implications for the ‘Snowball Earth’. Journal of the Geological Society, 160, 825– 828. Scherer, E., Munker, C. & Mezger, K. 2001. Calibration of the lutetium – hafnium clock. Science, 293, 683– 687. Schmitz, M. D. & Schoene, B. 2007. Derivation of isotope ratios, errors, and error correlations for U–Pb geochronology using 205 Pb-235U-(233U)-spiked isotope dilution thermal ionization mass spectrometric data. Geochemistry Geophysics Geosystems, 8, Q08006, doi: 10.1029/2006GC001492. Schmitz, M. D., Bowring, S. A. & Ireland, T. R. 2003. Evaluation of Duluth Complex anorthositic series (AS3) zircon as a U– Pb geochronological standard: new high-precision isotope dilution thermal ionization mass spectrometry results. Geochimica et Cosmochimica Acta, 67, 3665–3672.

149

Schoene, B., Crowley, J. L., Condon, D. J., Schmitz, M. D. & Bowring, S. A. 2006. Reassessing the uranium decay constants for geochronology using ID-TIMS U– Pb data. Geochimica et Cosmochimica Acta, 70, 426– 445. Selby, D. & Creaser, R. A. 2005. Direct radiometric dating of the Devonian-Mississippian time-scale boundary using the Re –Os black shale geochronometer. Geology, 33, 545– 548. Selby, D., Creaser, R. A., Stein, H. J., Markey, R. J. & Hannah, J. L. 2007. Assessment of the Re-187 decay constant by cross calibration of Re –Os molybdenite and U– Pb zircon chronometers in magmatic ore systems. Geochimica et Cosmochimica Acta, 71, 1999–2013. Shihong, Z., Ganqing, J. & Yigui, H. 2008. The age of the Nantuo Formation and Nantuo glaciation in South China. Terra Nova, 20, 289– 294. Silver, L. T. & Duetsch, S. 1963. Uranium – lead isotopic variations in zircon: a case study. Journal of Geology, 71, 721– 758. Steiger, R. H. & Jager, E. 1977. Subcommission on geochronology – convention on use of decay constants in geochronology and cosmochronology. Earth and Planetary Science Letters, 36, 359– 362. Stern, R. A. & Amelin, Y. 2003. Assessment of errors in SIMS zircon U–Pb geochronology using a natural zircon standard and NIST SRM 610 glass. Chemical Geology, 197, 111–142. Thompson, M. D. & Bowring, S. A. 2000. Age of the Squantum ‘tillite’, Boston Basin, Massachusetts: U –Pb zircon constraints on terminal Neoproterozoic glaciation. American Journal of Science, 300, 630– 655. Walter, M. R., Veevers, J. J., Calver, C. R., Gorjan, P. & Hill, A. C. 2000. Dating the 840– 544 Ma Neoproterozoic interval by isotopes of strontium, carbon, and sulfur in seawater, and some interpretative models. Precambrian Research, 100, 371–433. Wasserburg, G. J., Jacobsen, S. B., Depaolo, D. J., Mcculloch, M. T. & Wen, T. 1981. Precise determination of Sm/Nd ratios, Sm and Nd isotopic abundances in standard solutions. Geochimica et Cosmochimica Acta, 45, 2311– 2323. Wendt, I. & Carl, C. 1991. The statistical distribution of the mean squared weighted deviation. Chemical Geology: Isotope Geoscience, 86, 275– 285. Williams, M. L., Jercinovic, M. J. & Hetherington, C. J. 2007. Microprobe monazite geochronology: understanding geologic processes by integrating composition and chronology. Annual Review of Earth and Planetary Sciences, 35, 137–175. Xiao, S. H., Zhang, Y. & Knoll, A. H. 1998. Three-dimensional preservation of algae and animal embryos in a Neoproterozoic phosphorite. Nature, 391, 553– 558. Xiao, S. H., Yuan, X. L. & Knoll, A. H. 2000. Eumetazoan fossils in terminal Proterozoic phosphorites? Proceedings of the National Academy of Sciences of the United States of America, 97, 13 684– 13 689. Xiao, S. H., Yuan, X. L., Steiner, M. & Knoll, A. H. 2002. Macroscopic carbonaceous compressions in a terminal Proterozoic shale: a systematic reassessment of the Miaohe biota, south China. Journal of Paleontology, 76, 347– 376. Xu, B., Xiao, S. et al. 2009. SHRIMP zircon U– Pb age constraints on Neoproterozoic Quruqtagh diamictites in NW China. Precambrian Research, 168, 247–258. Yin, C. Y., Tang, F. et al. 2005. U–Pb zircon age from the base of the Ediacaran Doushantuo Formation in the Yangtze Gorges, South China: constraint on the age of Marinoan glaciation. Episodes, 28, 48– 49. York, D. 1966. Least squares fitting of a straight line. Canadian Journal of Physics, 44, 1079– 1086. York, D. 1967. The best isochron. Earth and Planetary Science Letters, 2, 479– 482. Zhang, S. H., Jiang, G. Q., Zhang, J. M., Song, B., Kennedy, M. J. & Christie-Blick, N. 2005. U –Pb sensitive high-resolution ion microprobe ages from the Doushantuo Formation in south China: constraints on late Neoproterozoic glaciations. Geology, 33, 473– 476. Zhou, C., Tucker, R., Xiao, S., Peng, Z., Yuan, X. & Chen, Z. 2004. New constraints on the ages of Neoproterozoic glaciations in south China. Geology, 32, 437– 440.

Chapter 10 Modelling the Snowball Earth YVES GODDE´RIS1*, GUILLAUME LE HIR2 & YANNICK DONNADIEU3 1

LMTG, CNRS-Observatoire Midi-Pyre´ne´es, 14 avenue Edouard Belin, 31400 Toulouse, France 2

IPGP, 4 place Jussieu, 75252 Paris, France

3

LSCE, CEA-CNRS, Orme des Merisiers, 91191Gif-sur-Yvette, France *Corresponding author (e-mail: [email protected])

Abstract: We review most of the modelling studies performed to date to understand the initiation and melting of a Snowball Earth, as well as to describe the glacial environment during the glaciation itself. All the described scenarios explaining the onset of glaciation rely on a sufficient decrease in the concentrations of atmospheric greenhouse gases (GHGs), typically resulting from the equatorial palaeogeography of the late Proterozoic. It is still heavily debated whether or not the oceanic ice cover was thick during the glaciation itself. However, a consensus has arisen that the most climatically stable scenarios imply the existence of a globally frozen ocean, with a thick ice cover caused by the flowing of high-latitude sea-ice glaciers towards the equator. Depending on the characteristics of the ice, a thin ice layer may have persisted along the equator, but this numerical solution is rather fragile. During the snowball event itself, model results suggest the existence of wet-based continental glaciers. Some parts of the continents may have remained ice-free. From the modelling perspective, the most significant problem in the snowball hypothesis, particularly in its ‘hard snowball’ version (the most stable numerically), is the melting phase. With improved modelling, the CO2 threshold required to melt the snowball is much higher than initially thought, significantly above 0.29 bar. Indeed, because of the very cold conditions prevailing at the surface of the Earth during the glacial event, the atmosphere becomes vertically isothermal, strongly limiting the efficiency of the greenhouse effect. This melting problem is further highlighted by geochemical modelling studies that show that weathering of the oceanic crust might be an active sink of CO2 during the glacial event, limiting the rise in atmospheric CO2. The solution might be found by considering the input of dark dust from catastrophic volcanic eruptions that would efficiently decrease the albedo of the ice. Finally, modelling studies also explore the aftermath of the glaciation. The world might have been drier than initially anticipated, resulting in the persistence of the supergreenhouse effect for at least one million years after the melting phase.

The Snowball Earth hypothesis is an exciting research field for climate modellers. First, it requires the use of diverse models to explore the salient aspects of the theory, from global geochemical models to ice sheet and climate models (Table 10.1). Second, coupling of these models is often required, which results in the construction of new numerical tools that can be applied to other palaeoclimate problems. Finally, the modelling of such an extreme environment leads scientists to push their numerical models to their physical limits, which is always instructive in terms of the numerical and physical behaviour of these models. The first modelling studies on the ice-albedo instability that underpins the Snowball Earth hypothesis were performed independently by William Sellers and Mikhail Budyko (1969), with a global Energy Balance Model (EBM, or 0D model). With their respective models, they demonstrated that reducing the current solar constant by even a few percent drives a rapid transition of the Earth climate into an ice-covered state, characterized by a high albedo (0.85) and global mean temperature of –100 8C (Sellers 1969). Modelling of the nonlinear response of the ice line as a function of the incoming absorbed energy demonstrates the existence of three stables states of the Earth climate system: (1) ice-free (e.g. the warm Cretaceous), (2) partially ice-free (e.g. the Phanerozoic glaciations and the present-day state of the Earth system) and (3) globally ice-covered (e.g. the Neoproterozoic Snowball Earth). Historically, the first two of these states have been the focus of most climatic modelling. However, since the publication of Hoffman et al. (1998), significant efforts have been made to understand the third stable state with respect to the Neoproterozoic glaciations. Below we present a summary of the results and the main modelling efforts carried out to date to understand the causes of Neoproterozoic Snowball glaciations, the environment during glaciation, and the mechanisms able to trigger deglaciation. We will focus on hypotheses that have been quantified using complex

numerical models – that is, models not restricted to first-order mass balance calculations, but rather accounting for the physics of a snowball event. For this reason, we do not attempt to discuss all the hypotheses proposed to explain multiple equatorial glaciations.

How to initiate a snowball glaciation? Thus far, three scenarios have been proposed and numerically tested to explain the onset of a snowball glaciation. They all rely on large drops in the partial pressure of greenhouse gases (GHGs). Scenario 1: It has been suggested that the partial pressure of atmos-

pheric methane might have been quite high during the Proterozoic (Pavlov et al. 2003), reaching up to 100– 300 ppmv because of intense activity of methanogenic bacteria in an anoxic ocean. The progressive oxidation of the atmosphere might have led to the collapse of the atmospheric methane stock in the Neoproterozoic, leaving the atmosphere with a low CO2 content, thus triggering global glaciations before pCO2 could self-adjust via the silicate-weathering feedback (Walker et al. 1981). This mechanism has not been fully tested to date and requires specific but unvalidated assumptions: the rise in O2 and subsequent collapse in CH4 must have occurred during the Neoproterozoic, and the global glaciation must have followed this event within no more than a few million years, based on the inferred residence time of carbon in the ocean – atmosphere system (Franc¸ois & Godde´ris 1998). Should this time span be longer, the imbalance between solid Earth degassing and continental silicate weathering would force CO2 to rise and compensate for the methane collapse. The two following scenarios rely on the palaeogeographic setting of the Neoproterozoic. This is still a matter of intense debate, but there is now a general consensus about the dominantly

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 151– 161. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.10

Table 10.1. A summary of all modelling studies of the Neoproterozoic ice ages (with type of model employed and reference) Tools

Authors

Note

Snowball Earth inception EBM EMIC GEO AGCM EMIC AGCM-GEO AGCM-GEO AGCM AGCM EMIC EMIC-GEO EMIC OAGCM AGCM-EMIC EMIC-GEO AGCM-ISM OAGCM AGCM EMIC EBM-GEO EMIC OAGCM AGCM-ISM EMIC EMIC AGCM OAGCM GEO OAGCM AGCM EBM-ISM/AGCM AGCM AGCM AGCM AGCM EBM EBM-GEO EBM EBM

Rose & Marshall (2009) Micheels & Montenari (2008) Godde´ris & Donnadieu (2008) Ishiwatari et al. (2007) Lewis et al. (2007) Godde´ris et al. (2007) Peltier et al. (2007) Romanova et al. (2006) Pavlov et al. (2005) Donnadieu et al. (2004b) Donnadieu et al. (2004a) Lewis et al. (2004) Poulsen & Jacob (2004) Ramstein et al. (2004) Donnadieu et al. (2004c) Pollard & Kasting (2004) Peltier et al. (2004) Jenkins (2004) Stone & Yao (2004) Godde´ris et al. (2003) Lewis et al. (2003) Poulsen (2003) Baum & Crowley (2003) Bendtsen (2002) Bendtsen & Bjerrum (2002) Donnadieu et al. (2002) Poulsen et al. (2002) Schrag et al. (2002) Poulsen et al. (2001) Baum & Crowley (2001) Hyde et al. (2000) Chandler & Sohl (2000) Jenkins (2000) Jenkins & Smith (1999) Jenkins & Frakes (1998) Crowley & Baum (1993) Marshall et al. (1988) Budyko (1969) Sellers (1969)

Ice-albedo instability Slushball conditions Comment on Peltier (2007) Ice-albedo instability Sea-ice dynamics Continental drift and CO2 decrease CO2 scenario for a slushball Test of the climatic factors Impact of interstellar dust Sea-ice dynamics Continental drift and CO2 decrease Test of the climatic factors Sea-ice dynamics Review Continental drift and CO2 decrease Glacial deposits Ice-albedo instability Ice-albedo instability Ice-albedo instability Igneous provinces and CO2 decrease Ocean dynamics Climatic feedbacks Climatic feedbacks Climatic feedbacks Climatic feedbacks Testing the high obliquity Climatic feedbacks Methane and CO2 decrease Climatic feedbacks Climatic feedbacks Slushball solution Climatic feedbacks Climatic feedbacks Climatic feedbacks Climatic feedbacks Climatic feedbacks Ice-albedo instability Ice-albedo instability Ice-albedo instability

During the Snowball Earth GEO GEO EBM þ ice-shelf flow EBM þ ice-shelf flow EBM þ ice-shelf flow EBM þ ice-shelf flow GCM-ISM Complex 1D ice model Spectral model Sea ice model

Le Hir et al. (2008b) Le Hir et al. (2008a) Warren & Brandt (2006) Pollard & Kasting (2006) Goodman (2006) Pollard & Kasting (2005) Donnadieu et al. (2003) Goodman & Pierrehumbert (2003) Warren et al. (2002) McKay (2000)

CO2 evolution, carbon cycling Seawater composition Sea-ice thickness Sea-ice thickness Comment Sea-ice thickness Continental ice dynamics Sea-ice thickness Sea-ice thickness Sea-ice thickness

Melting the Snowball Earth AGCM-dust model AGCM AGCM EMIC AGCM AGCM EBM-ISM EBM with CO2 ice clouds

Le Hir et al. (2010) Abbot & Pierrehumbert (2010) Le Hir et al. (2007) Lewis et al. (2006) Pierrehumbert (2005) Pierrehumbert (2004) Crowley et al. (2001) Caldeira & Kasting (1992)

Explicit effect of volcanic ash on snow/ice albedo Explicit effect of volcanic ash on snow/ice albedo Inversion of the atmospheric vertical thermal gradient Sensitivity to snow and ice albedo values Inversion of the atmospheric vertical thermal gradient Inversion of the atmospheric vertical thermal gradient Near-Snowball Earth CO2 ice cloud effect

Tools Aftermath of the Snowball Earth GCM-GEO/weathering model GEO

Year/ Authors / Journal

Le Hir et al. (2009) Higgins & Schrag (2003)

Note

Weathering rates, climate, CO2 restoring Carbon isotopic signal

Energy balance models (EBMs) only compute the radiative energy budget of the atmosphere. The transport is implicitly included as well as the water cycle. Radiative convective models (RCMs) include a description of the radiative energy budget of the atmosphere, and of the upward transport of energy by convection. General circulation models (GCMs) resolve explicitly the transport of energy and matter within the atmosphere (AGCM) and the ocean (OGCM) on a three-dimensional grid and physically compute the water cycle. Earth system Models of Intermediate Complexity (EMIC) bridge the gap between the simple RCM, the EBM and the highly complex GCMs. They couple simplified atmospheric and oceanic models and are generally designed for long-term simulations (104 to 105 years). GEO stands for global biogeochemical cycle models and ISM for ice-sheet model.

MODELLING SNOWBALL EARTH

153

low-latitude location of most of the continental blocks from 800 Ma to 600 Ma (Evans 2000; Meert & Powell 2001; Torsvik et al. 2001; Trindade et al. 2003; Macouin et al. 2004; Meert & Torsvik 2004; Hoffman & Li 2009). Scenario 2: This scenario (Schrag et al. 2002) was developed to

explain the unusual relationship between the onset of glaciation and a large negative d13C (carbonate) anomaly that preceded it (Halverson et al. 2002). High pre-glacial d13C values (.þ5‰) were assumed to result from a combination of high phosphorous flux to the oceans and efficient burial of organic carbon in the deltas of large equatorial watersheds from continents clustered in the tropics. Analogous present-day deltas include those of the Amazon and southeastern Asian rivers such as the Irrawady, Ganges and Salween. However, it is worth noting that the organic matter being buried in the Amazon and Bengal fans is today partly of terrestrial origin, which would not have been the case in the Neoproterozoic. Furthermore, the high sedimentation rates and resulting efficient burial of organic matter in southeastern Asia (up to 100% of the amount of organic carbon reaching the sediment) is very high, not because of the low-latitude location of these catchments, but because of their source in the Himalayas (Galy et al. 2007). Thus, it is not clear that Neoproterozoic palaeogeography alone could account for the unusually efficient organic carbon burial. The subsequent decrease in d13C prior to the onset of snowball glaciation is interpreted to be the result of a protracted release of CH4 derived from methanogenesis within the previously buried organic carbon pool. In the Schrag et al. (2002) model, after available reservoirs for methane clathrate on continental margins were filled, CH4 began to leak to the anoxic water column and then to the atmosphere, where it drove an increase in CH4 partial pressure. The gradual substitution of CO2 by CH4 then generated a severe vulnerability in global climate because CH4 has much shorter residence times than CO2 and is not similarly regulated through silicate weathering. Thus, if this situation persisted for several hundreds of thousands of years, then analogous to the previous scenario, a sharp decrease in the CH4 flux would precipitate a sudden cooling event. This model for triggering a Snowball Earth has the advantage of being the only one that can account for the major negative d13C anomalies preceding Cryogenian glaciations (Halverson et al. 2005; Prave et al. 2009). However, this scenario has not been thoroughly modelled, and some aspects remain unresolved, such as what triggered the collapse in the CH4 flux and whether O2 partial pressure was sufficiently low to allow the accumulation of high concentrations of atmospheric methane. Scenario 3: The prominent tectonic event of the Neoproterozoic was the break-up of the Rodinia supercontinent. The breakup started c. 800 Ma ago, when continental masses were distributed in the tropics (Li et al. 2003). Donnadieu et al. (2004a) calculated the climatic and geochemical impact of this break-up, through the development of a numerical model coupling a 2.5D climate model (Petoukhov et al. 2000) and a global biogeochemical cycle model (Godde´ris & Joachimski 2004). The use of an explicit climate model allows the subsequent increase in rainfall and runoff above the continents to be captured as the breakup proceeds. As a result, CO2 consumption through continental weathering is stimulated, driving atmospheric CO2 levels towards the threshold required to initiate a snowball glaciation (Fig. 10.1). This hypothesis therefore allows atmospheric CO2 to decrease to the threshold level without invoking a change in CO2 degassing. Donnadieu et al. (2004a) computed a decrease in atmospheric CO2 from 1830 ppmv down to 510 ppmv during the breakup of Rodinia and a concomitant drop in global mean annual temperature from 10.8 8C to 2 8C, assuming a solar constant reduced by 6% relative to the present. Importantly, the drastic cooling produced in this model is the result of dominantly longitudinal break-up in the

Fig. 10.1. CO2 consumption by continental weathering rate for a supercontinental (Rodinia) and a dispersed configuration in 104 mol of Ca2þ or Mg2þ/km2/a, for a fixed 6.5 PAL of CO2. Source: Donnadieu et al. (2004a).

low latitudes, in contrast to the latitudinal break-up of Pangaea, which did not trigger widespread glaciation. However, Rodinia break-up alone is not sufficient to initiate a snowball. Indeed atmospheric CO2 stabilizes above the required threshold (a value dependent on the model used, see next section). However, the possible solution lies in the geodynamics of supercontinental break-up, which is heralded by the onset of large magmatic provinces as a result of the accumulation of heat below large continental assemblages (Courtillot et al. 1999) and subsequent release through mafic eruptions. Specifically, a mantle superplume, starting at 830 Ma and lasting some 85 million years, is associated with the early stages of Rodinia break up (Li et al. 1999; Li et al. 2003). Tholeiitic magmatism has been identified between 825 and 755 Ma, especially in Australia (Wingate et al. 1998; Wingate & Giddings 2000) and northwestern Laurentia (Park et al. 1995), but also in south China (Li et al. 1999) and on the Congo craton (Key et al. 2001). Collectively, these magmatic events appear to define a plume time-cluster based on the database of Ernst and Buchan (2002). The most widespread flood basalt event probably occurred on Laurentia, with a minimum size of 3  106 km2 and possibly equivalent in area to the original Siberian Traps. The onset of continental flood basalts results first in a massive but short-lived release of CO2 into the atmosphere, leading to an enhanced greenhouse effect. However, a few million years after the end of the eruptive phase, this CO2 excess is rapidly consumed by intensified silicate weathering. In about 4 million years, the CO2 level stabilizes at a value lower than the pre-eruption level because highly weatherable basalt on the surface consumes CO2 (Dessert et al. 2001). The efficiency of this long-term cooling effect depends on the surface area of the large igneous province as well as on the precipitation and temperature where the traps occur (Godde´ris et al. 2003). Donnadieu et al. (2004a) have demonstrated that the weathering of freshly erupted basaltic surface as a disaggregating Rodinia drifts towards the humid equatorial regions is sufficient to trigger a snowball glaciation. Strictly

154

Y. GODDE´RIS ET AL.

speaking, this scenario can only apply to the first snowball glaciation (i.e. Sturtian episode) and does not explicitly account for the negative d13C anomaly that precedes this glaciation (Prave et al. 2009). Nevertheless, this study indicates that a globally cool climate is a predictable consequence of the low-latitude break-up of Rodinia.

The ice and climate of the Snowball Earth The onset of the snowball: ice-albedo instability Decrease in atmospheric GHGs will always cool global climate. However, even with very low GHG levels, is the Snowball Earth a plausible climatic end-member? A first set of numerical studies focused on the possibility of sea ice reaching the equator. Simple energy balance models all display a climatic instability once the ice line reaches about 308 latitude (Budyko 1969). This runaway ice-albedo feedback occurs once the surface radiation entering the open ocean regions cannot balance the heat loss by reflection of sunlight from low-latitude ice-covered regions, resulting in rapid ice growth and global cooling. However, initial coupled ocean – atmosphere general circulation models did not follow this behaviour prescribed by EBMs. Using the coupled Fast Ocean Atmosphere Model (FOAM), which accounts for ocean dynamics, Poulsen (2003) performed a simulation with a unique atmospheric pCO2 fixed at 140 ppm, a solar constant reduced by 7%, an idealized equatorial supercontinent and a sea-ice thickness of 1000 m. He forced the ice line to be located at 108 latitude. Once the prescribed condition for the location of the ice line was released, it did not continue to the equator but rather retreated to higher latitudes, .408N and S. This result was used to argue that a full snowball climate is climatically untenable (Poulsen et al. 2001; Poulsen 2003). However, Lewis et al. (2004) demonstrated that this instantaneous sea-ice retreat was caused by the absence of sea-ice dynamics and was an artefact of the way the energy required to melt the sea ice was treated. Using a similar model, Lewis et al. (2007) then obtained the surprising result that the ocean completely froze over when pCO2 fell below 3000 ppmv. The potential of the onset of a full snowball glaciation in a coupled dynamic model has been confirmed by a study performed with an Earth Model of Intermediate Complexity (EMIC) that took into account sea-ice thermodynamics and ocean dynamics, where two realistic continental configurations were tested (supercontinent Rodinia and a disaggregated Rodinia along the equator; Donnadieu et al. 2004b). Probably the most important conclusion arising from this study was that climatic instability exists for both continental configurations, meaning that ocean dynamics cannot prevent the onset of a Snowball Earth. However, they also found that the CO2 threshold required to trigger a snowball is (i) highly dependent on the continental configuration, being higher for the dispersed continental configuration, and (ii) below the threshold values calculated with an atmospheric model, confirming the important role of the ocean dynamics on the onset of a Snowball Earth, but also emphasizing the role of atmospheric meridional heat transport. Indeed, the Hadley cells strengthen as the ice moves equatorward in response to decreasing atmospheric CO2 pressure, bringing more and more heat from the equator to the ice front, thus slowing its advance. But this strong negative feedback collapses once the ice line reaches 308 latitude. Indeed, as the ice line is now located on the descending branch of the Hadley cell, the cold air in contact with the ice is now efficiently transported back towards the equator along the lower meridional branch of the cell (Bendsten 2002). Air temperatures at the equator then rapidly decrease to zero, shutting down the climatic negative feedback, allowing the onset of snowball glaciation. Heat transport in the ocean is more complex and largely depends on the continental configuration. Nevertheless, Donnadieu et al.’s (2004b) key finding is that heat transport during the advance of

the ice and the final collapse into a snowball state is mainly driven by the atmosphere, and the ocean plays only a minor role. The snowball CO2 thresholds were computed to be 90 ppmv for a supercontinental configuration and 150 ppmv for the dispersed configuration along the equator. These levels are much lower than initially expected from atmospheric model simulations (Jenkins & Smith 1999). Similar conclusions regarding the critical role played by the Hadley transport once the ice line reaches 308 latitude were obtained recently with a general circulation model (GCM) coupled to a thermodynamic sea-ice model, assuming an idealized supercontinent located at the equator (Poulsen & Jacob 2004).

During the Snowball Earth: the slushball theory as a solution to explain the survival of photosynthetic life? How can climate models that predict a totally ice-capped ocean be reconciled with palaeontological evidence that establishes the persistence of photosynthetic activity throughout the snowball event? Hyde et al. (2000) proposed a scenario where the equatorial ocean remains ice-free during snowball events. This ‘tropical oasis’ solution is colloquially referred to as the ‘slushball’ hypothesis and has gained traction amongst geologists who see evidence for open ocean conditions during the presumed snowball glaciations (e.g. Leather et al. 2002; Allen & Etienne 2008). This result was obtained by coupling a simple energy balance model and an icesheet model under a solar constant reduced by 6% and assuming a very low CO2 level of 140 ppmv. Under such conditions, massive icecaps develop on continents, as well as thick ice sheets on the global ocean. Then, the calculated icecaps were specified as boundary conditions of a GCM, in which the CO2 partial pressure was prescribed at 700 ppmv. The GCM output showed that the ocean remains free of ice below 258 latitude. However, this climatic solution is not self-consistent: if the glaciers were allowed to respond to the 700 ppm pCO2, they would retreat. And if CO2 was reduced to 140 ppmv, the equatorial ocean would freeze. This slushball simulation is therefore not robust, because the threshold for the onset of a Snowball Earth simulated with a GCM is generally far below 700 ppmv (Poulsen 2003; Donnadieu et al. 2004a) and the stability range of the slushball solution. An additional argument against the slushball model comes from sea-ice dynamics. As demonstrated by Lewis et al. (2007), a sea-ice margin reaching low latitudes without driving a total collapse of the ocean is only obtained in simulations omitting the sea-ice dynamic process. In oceanic models including sea-ice dynamics, the ice-albedo instability feedback does not allow a stable slushball solution.

During the Snowball Earth: a clear equatorial thin-ice as a solution for life? Based on observations of lakes in the Antarctic Dry Valley, McKay (2000) formulated an alternative scenario to explain the survival of photosynthesis. Using a simple method to reconstruct the energy balance of the ice cover, he showed that if the mean annual temperature at the equator was around –30 8C, then the equilibrium thickness of the bare sea ice in this area might have been as little as a few metres, assuming an ice albedo of ,0.75 (McKay 2000). This result was obtained assuming low absorption efficiency of the ice within the visible part of the solar spectrum, so that heat can be transferred to the water below the ice, thus limiting the growth of the ice sheet through seawater freezing. Such a low absorption coefficient for the ice requires a slow rate of freezing, which prevents the incorporation of air bubbles and keeps the ice relatively clear. However, such clear ice is inconsistent with high albedo, which requires a high bubble density. The calculated

MODELLING SNOWBALL EARTH

thickness is critically dependent on the ice albedo, exceeding several tens of metres if the ice albedo is .0.8. The thin-ice model has been challenged by Warren et al. (2002), who described the absorption of solar energy as a function of wavelength and depth, accounting for the scattering of radiation by air bubbles inside the ice. They found that the absorption of incident energy within the first centimetres of the ice sheet was much higher than previously expected (McKay 2000), particularly in the near-infrared wavelength band (0.7 –3 mm). Furthermore, the visible part of the spectrum is largely scattered by air bubbles, brine inclusions and cracks. Heat is thus not efficiently transferred to the water below the ice. As a consequence, their results suggest that the ocean was everywhere capped by at least several hundred metres of ice. In Warren et al.’s (2002) model, very low albedos (,0.4) of the ice would be required to allow a thin-ice solution, implying a very low air bubble density. Such albedo values are not realistic under the tropical sun, because it would warm the ice above the melting point. Another argument supports the thick-ice hypothesis. The latitudinal air-temperature gradient in a snowball environment would produce thicker sea ice at higher latitudes than around the equator. This is not dynamically stable, so ice would flow from high to low latitudes as sea glaciers reducing the thickness gradient, but without forming a uniform thickness everywhere due to slow flow rates (Goodman & Pierrehumbert 2003). The question of thick v. thin ice remains open. A recent modelling study coupling an EBM with an ice-shelf model demonstrates the possibility of the existence of areas of thin ice (,3 m thick) below a latitude of 208 (Pollard & Kasting 2005) (Fig. 10.2). Climate simulations of the Snowball Earth conditions predict net positive surface accumulation (precipitation –evaporation) of a few mm/a in mid to high latitudes. Snow thus accumulates on the ice, strongly increasing the albedo and allowing the growth of thick sea ice. But in low latitudes, the net surface accumulation is negative, the sea ice remains free of snow, and the thin-ice solution is valid. Pollard & Kasting (2005) calculated that the snow limit is located around a latitude of 128. Below this latitude, the sea ice is a marine ice accumulated through basal freezing, much less bubbly and clearer than the snow-derived glacial ice. The highlatitude thick ice (150 m thick) flows from high latitude at a rate of 800 m/a and eventually crosses the 128 latitude limit. Because the precipitation –evaporation (P– E) budget becomes negative, the snow accumulated on this ice starts to sublimate or melt. As a result, up to 5 W m22 is absorbed by the ice, and internal melting occurs between latitudes of 128 and 108. Thin ice may thus be maintained at low latitude, where the albedo can be as low as 0.45, despite the flowing of high-latitude sea-ice glaciers.

Fig. 10.2. Sea-ice thickness during a snowball in the Pollard & Kasting (2005) model, as a function of latitude. The dashed line indicates bubbly ice, and the solid line clear ice. Flowing of sea glaciers is accounted for.

155

Furthermore, Pollard & Kasting (2005) argue that, in any case, low-latitude large lakes and confined seas like the modern Mediterranean, protected from planetary-scale sea-glacier flow by surrounding continents, would maintain areas with thin ice cover. This model has been challenged by Goodman (2006) and Warren & Brandt (2006). Goodman (2006) noted that the atmospheric and oceanic water cycle must be internally balanced, because the glaciation lasts for millions of years, to avoid any concentration of the water mass into the atmosphere or the ocean. Because Pollard & Kasting (2005) calculated that the ice sheet was formed by freezing of seawater and that net sublimation transfers water from the ice sheet to the atmosphere in the tropics, either some meteoric ice would have to return water to the ocean (which did not happen in the model) or else marine ice could not reach the surface, and all of the ice-sheet surface would consist of meteoric ice. However, meteoric ice is bubbly and opaque, with a high albedo, thus counteracting the internal melting process proposed by Pollard & Kasting (2005). In conclusion, whether the thin-ice solution can be maintained or not is still a matter of debate. We can only conclude that very clear thin ice is a fragile solution, and can potentially be formed only where meteoric ice is absent. To date, model studies have reconstructed the Snowball Earth as a hostile environment for photosynthetic marine algae and associated food chains, with no ice-free equatorial ocean, nor tropical thin-ice. However, the first description of the snowball theory suggested the existence of cracks in the sea ice, whether it was thin or not (Hoffman et al. 1998). In a cold environment, the equatorial sea ice would form cracks if its speed exceeds 50 m/a. Newly formed sea ice in cracks will contain brine channels, which today host a variety of organisms including phototrophs. Moreover, if the ice was thin, winds and tidal forces were probably strong enough to generate leads (Pollard & Kasting 2005). Also, active hydrothermal areas in shallow water should maintain open waters in their direct vicinity. Hence the existence of several types of refugia required to explain the persistence of photosynthesis cannot be ruled out from numerical simulations alone. Indeed, large populations of microscopic organisms can populate a small volume in which light remains available. They can preserve a relative diversity even if the remaining oasis is small (Knoll 2003).

During the Snowball Earth: glacial deposits and continental ice behaviour The behaviour of the terrestrial glacial regime during the Neoproterozoic glaciations is still a matter of debate. Some Snowball Earth detractors claim that the glacial sequences cannot be explained with the Snowball Earth scenario. Indeed, the near shutdown of the hydrological cycle simulated by climatic models, once the Earth is entirely glaciated, stands in contrast to the requirements for active, wet-based continental ice sheets to produce the observed thick glacial deposits. The extent and dynamics of the continental ice has been estimated through the forcing of an icesheet model with the output of a GCM in snowball conditions (Donnadieu et al. 2003). The GCM (LMDz) was run assuming 1 PAL (preindustrial atmospheric level, or 280 ppmv) of CO2, a solar constant reduced by 6%, and a realistic palaeogeographic configuration including an estimate of the continental relief (Greenvillian-aged mountains, with a prescribed elevation of c. 2000 m, relics of the last major orogenic cycle). With those forcings, the ice line rapidly reaches 308 latitude, reaching the instability threshold and resulting in the onset of a snowball state. The ice-sheet model (Ritz et al. 2001) is then run using the climate fields generated by the GCM. The three main conclusions of this study were as follows. First, thick continental ice-sheets build up within a few hundred thousand years, with a mean of 2500 m and a maximum thickness of 5000 m above Laurentia and Antarctica. The ice sheets nucleate on high relief and in less

156

Y. GODDE´RIS ET AL.

than 50 ka spread across the continental interior. Second, the model predicts a final relative fall in sea level of 200 m. However, the drop in sea level is highly dependent on several competing phenomena. Glacioeustasy is the dominant control; for example, with an average thickness of ice on all continents of 2000 m, the fall will be c. 500 m, assuming an emerged continental surface of one-quarter the area of the oceans. However, glacio-isostatic rebound, hydro-isostatic rise of the seafloor (due to a reduction in water load) and the tectonic subsidence balance the initial glacio-eustatic fall, reducing the relatively low sealevel fall despite a very significant land ice volume. Performing similar GCM – ice sheet model simulations, but without prescribing any relief on the continents, Pollard & Kasting (2004) also show that thick ice sheets build up on continents in the tropics, but the location of the foci of the ice sheets is determined by the dynamically controlled regions of positive P– E. Third a large area of the calculated continental ice sheet is wetbased, so basal sliding is expected to occur. It is particularly active in the equatorial area where the calculated sliding speed can reach several tens of m/a. If the sea ice in the equatorial area was thin (see above discussion), such sliding of the ice sheets may explain the main features of glacial deposits accumulated during the Snowball Earth event (Donnadieu et al. 2003). Some parts of the continents remain ice-free, even after the ice sheet reaches a steady-state volume after 400 ka in the ice-sheet model. Such ice-free areas may act as dust sources, darkening the snow over large surfaces through aeolian deposition, helping to maintain areas of thin ice on the ocean (Pollard & Kasting 2005).

Melting the Snowball Earth The commonly accepted scenario for the melting of the Snowball Earth is from Kirschvink (1992) and Hoffman et al. (1998). However, the solution for escape from a snowball was first proposed by Walker et al. (1981) in their seminal paper on silicate weathering and the long-term stabilization of global temperatures. Before the publication of the Snowball Earth hypothesis, they had already proposed a solution to the melting problem: ‘If global glaciations were to occur, the rate of silicate weathering should fall very nearly to zero, and carbon dioxide should accumulate in the atmosphere at whatever rate it is released from volcanoes. Even the present rate of release would yield 1 bar of carbon dioxide in only 20 million years. The resultant large greenhouse effect should melt the ice cover in a geological short period of time.’

This general prescription for the recovery from a snowball remains essentially valid. Indeed, the snowball hypothesis fundamentally relies on the strong reduction of the carbon sink during the snowball. Organic carbon burial was certainly maintained at a very low level during the event, because most of the ocean was covered by thick ice. Also, silicate weathering was strongly inhibited by a drastic decrease in the continental water cycle and the low temperatures. As a result, in the absence of rains to scrub the atmosphere of CO2, the carbon originating from the degassing of the solid Earth should accumulate in the atmosphere to very high levels, until a threshold for deglaciation is reached. Caldeira & Kasting (1992) calculated the oft-cited figure of 450 times the present-day (i.e. 1992) CO2 level (or 0.126 bar) as the threshold where CO2 levels counteract the strong ice-albedo feedback and initiate melting. At the present-day degassing rate of 6.8  1012 mol/a (a value chosen to balance the present-day continental silicate weathering; Gaillardet et al. 1999), it would take only 3.5 million years to reach this threshold. However, as in all facets of modelling of the Snowball Earth, the situation is more complicated.

The melting problem: the climatic modelling argument The threshold value of 0.126 bar CO2 for initiating snowball melting was calculated with an EBM assuming the present-day

insolation. This threshold rises to 0.29 bar with the same model if a realistic decrease of 6% in the solar constant is used. It would then take 8 million years to melt the hard snowball, still assuming the present-day degassing rate, without accounting for possible dissolution of carbon into the ocean (Higgins & Schrag 2003), which would increase the time required to accumulate threshold CO2 levels. Furthermore, an energy balance model is probably not the best tool with which to explore high CO2 climate, because too many parameterizations used in EBMs (such as cloud forcing) are calibrated on the present climate. Working with an atmospheric GCM (the FOAM GCM) and increasing CO2 partial pressure in a fully glaciated world, Pierrehumbert (2004) showed that 0.29 bar is strictly a minimum value for the threshold, and most likely it is much higher. Through an extrapolation of the CO2 radiative forcing function, he showed that 2 bar of CO2 might have been required to melt the snowball, corresponding to an unrealistic glacial duration of .50 million years. These simulations were carried out assuming an idealized rectangular supercontinent and present-day orbital parameters. The greenhouse efficiency G can be defined as the difference between the amount of energy emitted by the Earth surface and the amount of energy lost as long-wave radiation at the top of the atmosphere, the outgoing long-wave radiation (OLR): G ¼ sTS4  OLR

(1)

where Ts is the average surface temperature. OLR is a function of the effective planetary temperature at the top of the atmosphere, Te, and is equal to sTe4 (for the modern Earth, Te ¼ 255 K). Today, the atmosphere is not isothermal, and the effective temperature is lower than the temperature at the surface. Consequently G is positive. During the snowball event, in the winter hemisphere, the atmosphere is isothermal with height, because of inefficient solar heating of the surface (high albedo) and convection due to absence of a warm ocean. The air-temperature profile as a function of height even displays an inversion at ground level, due to the extremely cold conditions at the Earth surface. As a result, Te is almost equal to Ts, and the greenhouse efficiency approaches 0. Adding CO2 to the system will thus not efficiently warm the Earth, as long as the temperature profile remains isothermal. The outgoing long-wave radiation might even exceed the black-body radiation at the Earth surface, and the net greenhouse effect is slightly negative for the winter hemisphere. In the summer hemisphere, the low tropopause limits the vertical temperature gradient, reducing the efficiency of the greenhouse effect, which is compounded by the virtual lack of water vapour feedback at such cold temperatures. Furthermore, the cloud greenhouse effect is reduced, because the high-altitude clouds (contributing the most to the greenhouse effect) are thin or absent because of the very low water vapour concentration in the cold upper troposphere in the summer hemisphere, and at any level in the winter hemisphere (Pierrehumbert 2004). A solution to this problem may be found in reduction of the albedo. Decreasing the albedo would increase the amount of solar flux absorbed at the surface and promote the onset of a normal vertical temperature profile of the atmosphere in the winter hemisphere and increase the efficiency of the greenhouse effect. Causes of the reduction in albedo are not yet defined, but the thin-ice solution or the presence of dirty snow generated by volcanic eruptions or windblown dust from deserts are reasonable mechanisms (Pierrehumbert 2004). Expanding on the work of Pierrehumbert (2004, 2005), Le Hir et al. (2010) explored the impact of a large volcanic eruption in the final stage of the snowball glacial event when atmospheric CO2 had already reach 0.2 bar. Huge eruptions, similar to the Toba eruption at 74 ka, occur once or twice per million years and release up to 2  1016 g of dust into the atmosphere. Given the expected duration of the glacial event (close to 10 million years), it seems plausible that a major volcanic eruption occurred

MODELLING SNOWBALL EARTH

157

during the final glacial stage. A GCM (LMD) was used that included an improved calculation of the albedo of the snow, including the effect of aging of the snow and dust input. Le Hir et al. (2010) showed that the sudden release of 2  1016 g of dust into the atmosphere decreases the albedo of the surface along the equator by about 0.2 units, but actually increases the albedo by about 0.1 units between 208 and 408 latitude. The decrease in albedo along the equator is due to dust accumulation in an area where net ablation of the snow cover is predicted. Temperature rises along the equator, activating Hadley cell circulation and leading to an increase in fresh snow precipitation above 208 latitude. Dust is buried under the fresh snow and albedo increases. Overall, the mean annual temperature between latitudes of –10 and þ108 rises by about 6 8C, reaching values as high as –2 8C. Although a volcanic eruption does not seem to be sufficient to generate the dramatic melting of the Snowball Earth, dust input does seem to be a plausible mechanism to reduce the required CO2 threshold.

The melting problem: the geochemical modelling argument Problems also arise from geochemical modelling. Using a model describing the global carbon and alkalinity cycles during the Snowball Earth event, Le Hir et al. (2008b) calculated that an open water area of 3000 km2 during glaciation is large enough to allow efficient diffusion of atmospheric carbon into the ocean such that the atmosphere and the ocean are in equilibrium with respect to carbon at the million-year timescale. Present-day emerged geothermal environments occupy a surface with an area of c. 1.5  106 km2 (Dessert et al. 2003). Hence, during global glaciations, some small open-water areas, easily in excess of 3000 km2, may be maintained by heat production and lava flows. For open water to be maintained, two conditions must be fulfilled. First, the hydrothermal zone must not be too deep, otherwise their heat would be diffused over an area so large that it cannot melt the overlying sea ice. Second, the hydrothermal zone must not be too shallow on the continental shelf where the heat flux would only melt the sea ice in the immediate vicinity, resulting in a lake above the thick sea ice which may be disconnected from the global ocean under the thick sea ice (Goodman 2006). If only 2% of geothermal environments had fulfilled these conditions, then massive dissolution of CO2 should have occurred, acidifying the ocean. Even assuming the existence of a buffering mechanism through seafloor carbonate dissolution, the pH of the water drops to about 6 within 3 to 4 million years after the onset of the Snowball glaciation, with atmospheric CO2 levels reaching more than 0.1 bar (Le Hir et al. 2008b). Such acidification of the waters enhances the dissolution rate of the oceanic basaltic crust by a factor of more than 4, even assuming a glacial thermohaline circulation fixed at 1% of its present-day value. Seafloor weathering consumes exospheric carbon, trapping it as carbonates in the veins of the oceanic crust (Alt & Teagle 1999). This means that all carbon sinks are not drastically reduced during the glacial event, thus partly counteracting the rise in atmospheric CO2 thought to be required to melt the snowball. Le Hir et al. (2008b) calculated that within 30 million years (thought to be the maximum duration of a snowball event), atmospheric CO2 cannot rise above 0.24 bar, a level below the predicted melting threshold determined by Pierrehumbert (2004). Thus, these results further exacerbate the problem of whether a snowball could actually be melted (Fig. 10.3).

Aftermath of the Snowball Earth Few studies have dealt with the global environment and climate during the melting of a snowball. The extreme greenhouse atmosphere expected to be typical of the melting phase is not well

Fig. 10.3. Snowball duration as a function of the CO2 threshold needed to initiate melting. The stars correspond to estimated levels with various models and boundary conditions (stars), assuming accumulation of CO2 in the atmosphere, and no diffusion in the ocean. S is the solar constant normalized to the present-day value. The triangle indicates a simulation where atmospheric CO2 is allowed to dissolve into the ocean through cracks, assuming the 0.29 bar threshold (EBM simulation with a solar constant reduced by 6%). The square further assumes consumption of CO2 by seafloor weathering, buffering the rise in CO2 and hence delaying the melting.

understood, and no investigations have been performed with complex numerical models to date. Three studies explore the geochemical global cycles in the aftermath of the snowball. Two of them are extremely simple (box models) (Higgins & Schrag 2003; Le Hir et al. 2008a). In both cases, the principle difficulty in simulating the post-glacial environment arises from the quantification of the weathering rates of the continental surfaces under a very warm climate. Weathering during the melting phase is crucial for at least two reasons (Higgins & Schrag 2003): (i) cap-carbonate accumulations are thought to be the consequence of the transfer of alkalinity to the ocean from the extremely fast dissolution of continental carbonates exposed to acid rain under several hundred times the present-day CO2 atmospheric pressure, and (ii) the rate at which continental silicate rocks dissolve controls the time span needed to restore normal climatic conditions, because silicate weathering consumes atmospheric CO2 and stores it as carbonate sediments on the seafloor. However, weathering is a function of air temperature and continental runoff. It can be safely assumed that temperatures are high under 0.3 bar of atmospheric CO2 once the melting is completed, but global precipitation patterns in such a climate are not easy to predict. Lee Kump (personal communication, cited by Higgins & Schrag 2003) suggested a maximum sevenfold increase in continental runoff compared to the present-day global value. However, a recent extensive numerical study in which a GCM was coupled to a transport-reactive model simulating weathering rates demonstrated that such an increase is not realistic (Le Hir et al. 2009). Indeed, the total amount of energy that can be used to evaporate waters from the ocean and to produce continental rainfall in a supergreenhouse environment (post-snowball) is limited by the total incoming solar energy. Le Hir et al. (2009) showed that above 0.11 bar of CO2, all incoming solar radiation is used as latent heat. Accounting for a realistic geographical distribution of the continents and an explicit modelling of the water cycle, the model predicts that the continental runoff cannot exceed 1.2 times the present-day runoff under more than 0.11 bar of CO2. Nevertheless, assuming a sevenfold increase in the continental runoff, Higgins & Schrag (2003) suggested that continental weathering (carbonate þ silicate rocks) may have increased by a factor of 50 directly after melting, relative to present-day values. As a

Y. GODDE´RIS ET AL.

158

result, normal values of atmospheric CO2 (around several hundred ppmv) would be restored in 200 ka, starting from 0.126 bar. In their study, Le Hir et al. (2009) found that CO2 consumption through continental weathering cannot exceed 10 times the present-day value, even accounting for the high surface area of rock flour covering the continents, because weathering rates are inhibited by the moderate increase in precipitation. Consequently, it takes at least one million years for atmospheric CO2 to return to pre-glacial values in response to silicate weathering in the post-snowball environment. One of the most striking features of the post-snowball environment is the accumulation of up 18.5 m (on average) of cap dolostones on continental shelves, characterized by very low d13C (c. –4‰, Pee Dee Belemnite Standard (PDB)). Heretofore, modelling this peculiar phase of the snowball sequence has proven challenging, because modelling the melting of huge icecaps under very high CO2 levels is not straightforward. For that reason, no explicit modelling has been performed and significant simplification is required. Regarding the d13C signature of cap carbonates, Higgins & Schrag (2003) calculated that the nadir of the negative d13C excursion is observed 10 ka after the resumption of continental weathering. This decrease in d13C directly after melting initiation is linked to the decrease in the carbon isotopic fractionation between CO2 and carbonates, itself forced by a prescribed rise in sea surface temperature of 30 8C. In their model, with a very high prescribed weathering flux, cap-dolostone accumulation occurs in a few ka. In a similar study, coupling a radiative –convective climate model to a global biogeochemical cycle model, Le Hir et al. (2008a) found that the most negative d13C value was attained 100 ka after the resumption of continental weathering, suggesting a longer time span for cap-dolostone accumulation. This longer time span was also an outcome of the more extensive modelling in Le Hir et al. (2009). Accounting for the total amount of water released by the melting of the huge icecap (200  103 km3; Donnadieu et al. 2003), and estimating the total flux of Mg2þ and Ca2þ released by weathering during the transgressive phase with a reactive-transport model, Le Hir et al. (2009) calculated that only 50 cm of pure dolostone would accumulate on the shelves in the first 10 ka, far below the estimated worldwide average thickness of 18.5 m (Hoffman et al. 2007). This slow rate of accumulation suggests that cap-dolostone deposition may have spanned longer than 10 ka, collaborating arguments for a longer timescale for cap-dolostone deposition based on the existence of multiple magnetic reversal events measured in various cap dolostones (Trindade et al. 2003; Raub & Evans 2006). However, a corollary of this conclusion is that, because the cap dolostone was strictly deposited during the post-glacial melting phase (Hoffman et al. 2007), it requires that deglaciation was also protracted, and it is not clear if this is physically feasible. Therefore, the timescale of melting and cap-dolostone precipitation remains an unresolved question, awaiting careful modelling of the melting phase.

Discussion and conclusions Do the climate models teach us something about the causes and consequences of the Neoproterozoic climatic events? The answer is surely yes if the models have a physical basis. They allow us to explore the space of parameters of the snowball glacial events, the variability of these parameters being fixed by field studies or by the known physical behaviour of the climate. The answer is no if they are mere mathematical manipulations performed without heed to realistic physical parameters and geological constraints. The most important results of the modelling studies can be summarized as follows: (1)

Onset of snowball glaciations. At least two Cryogenian glaciations were global in extent (Macdonald et al. 2010) and

(2)

(3)

(4)

preceded by negative d13C anomalies (Prave et al. 2009), supporting the idea that snowball events are based on major perturbations to the carbon cycle, most likely very low pCO2. The general tectonic context of the Neoproterozoic is probably the main precondition for the onset of global glaciation, facilitating atmospheric CO2 pumping by continental weathering. The break-up of the Rodinia supercontinent triggered an increase in continental weathering through enhanced humidity above continental surfaces that led to a generally cool climate during the Neoproterozoic. As long as the continents were dispersed along the equator, climate remained cool. Within this cool context, additional triggers, such as the weathering of magmatic provinces erupted during the initial phase of the break-up, or atmospheric methane collapse, were required to initiate snowball glaciations. During the snowball. First, sea ice is allowed to reach the equator once the CO2 threshold for initiating a runaway ice-albedo feedback is reached. Whether this sea ice was thick or thin is still debatable, but the thin-ice solution cannot be excluded, thus potentially allowing a vast refuge for primary producers. Like modern ice sheets, snowball ice sheets on the continents grew to several kilometres in thickness and were wet-based, consistent with the thick accumulation of glacial deposits during the glaciations. Finally, massive dissolution of atmospheric CO2 into the ocean occurred if even a small surface area of the ocean (3000 km2) remained open. In this scenario the ocean may have been largely acidified during the snowball events, with a pH as low 6, promoting the increase in the consumption of carbon by the oceanic crust by a factor of 4. Hence, with such an efficient carbon sink, the CO2 accumulation would not be linear, but asymptotic, approaching around 0.25 bar of atmospheric CO2. Melting the snowball. Many unresolved questions remain concerning snowball modelling, but the most important of them concerns the initiation of melting. A significant conclusion of the modelling studies is that a very high CO2 level is required to melt the snowball, probably .0.29 bar. This level might never be reached, because carbon is consumed by the enhanced dissolution of the basaltic oceanic crust. Efforts should focus on this problem, exploring processes that might reduce the albedo of the Earth and enable melting at lower CO2 levels. In the aftermath of the snowball glaciations. The alkalinity required to accumulate the cap dolostone results from the weathering of continental carbonate during the supergreenhouse following the snowball event. This post-snowball environment needs to be explored in more detail. It seems that the most prominent and rapid environmental perturbations occurred during the onset and subsequent relaxation of the supergreenhouse in the direct aftermath of the snowball. Modelling studies contradict earlier assertions of a highly vigorous hydrological cycle during the supergreenhouse event, with runoff only as much as 20% higher compared to its present-day value. Consequently, weathering rates were also lower, and rather than a few hundred thousand years (Hoffman et al. 1998; Higgins & Schrag 2003), it most likely took .1  106 years for atmospheric CO2 to return to pre-glacial values. These results also suggest that capdolostone accumulation endured on the order of 100 ka.

If the conditions for initiation of the Snowball Earth seem to be reasonably well understood (Table 10.1), much work remains to resolve the melting and climate throughout the deglaciation. Correctly simulating the duration of deglaciation with fully coupled climate and ice-sheet models could provide important constraints on global cap-dolostone precipitation. However, nature is more complex than models, and it is unrealistic to expect that models

MODELLING SNOWBALL EARTH

will ever capture the full complexity of the Neoproterozoic ice ages. Nevertheless, models provide important clues and some quantitative constraints on the controversial question of whether these catastrophic events really did occur. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Abbott, D. S. & Pierrehumbert, R. T. 2010. Mudball: surface dust and snowball Earth deglaciation. Journal of Geophysical Research, 115, doi: 10.1029/2009JD012007. Allen, P. A. & Etienne, J. L. 2008. Sedimentary challenge to Snowball Earth. Nature Geoscience, 1, 817– 825. Alt, J. C. & Teagle, D. A. H. 1999. The uptake of carbon during the alteration of ocean crust. Geochimica Cosmochimica Acta, 63, 1527– 1535. Baum, S. K. & Crowley, T. J. 2001. GCM response to late Precambrian (590 Ma) ice-covered continents. Geophysical Research Letters, 28, doi: 10.1029/2000GL011557. Baum, S. K. & Crowley, T. J. 2003. The snow/ice instability as a mechanism for rapid climate change: a Neoproterozoic snowball Earth model example. Geophysical Research Letters, 30, doi: 10.1029/ 2003GL017333. Bendtsen, J. 2002. Climate sensitivity to changes in the solar insolation in a simple coupled climate model. Climate Dynamics, 18, 595– 609. Bendtsen, J. & Bjerrum, C. J. 2002. Vulnerability of climate on Earth to sudden changes in insolation. Geophysical Research Letters, 29, doi: 10.1029/2002GL014829. Budyko, M. I. 1969. The effect of solar radiation variations on the climate of the Earth. Tellus, 21, 611–619. Caldeira, K. & Kasting, J. F. 1992. Susceptibility of the early Earth to irreversible glaciation caused by carbon dioxide clouds. Nature, 359, 226– 228. Chandler, M. A. & Sohl, L. E. 2000. Climate forcings and the initiation of low-latitude ice sheets during the Neoproterozoic Varanger glacial interval. Journal of Geophysical Research – Atmospheres, 105, 20 737– 20 756. Condon, D., Zhu, M., Bowring, S., Wang, W., Yang, A. & Jin, Y. 2005. U –Pb ages from the Neoproterozoic Doushantuo Formation, China. Science, 308, 95 – 98. Courtillot, V., Jaupart, C., Manighetti, I., Tapponnier, P. & Besse, J. 1999. On causal links between flood basalts and continental breakup. Earth and Planetary Science Letters, 166, 177–195. Crowley, T. J. & Baum, S. K. 1993. Effect of decreased solar luminosity on Late Precambrian ice extent. Journal of Geophysical Research, 98, 16 723– 16 732. Crowley, T. J., Hyde, W. T. & Peltier, W. R. 2001. CO2 levels required for deglaciation of a ‘Near-Snowball’ Earth. Geophysical Research Letters, 28, 283– 286. Dessert, C., Dupre´, B., Franc¸ois, L. M., Schott, J., Gaillardet, J., Chakrapani, G. J. & Bajpai, S. 2001. Erosion of Deccan Traps determined by river geochemistry: impact on the global climate and the 87Sr/86Sr ratio of seawater. Earth and Planetary Science Letters, 188, 459– 474. Dessert, C., Dupre´, B., Gaillardet, J., Franc¸ois, L. M. & Alle`gre, C. J. 2003. Basalt weathering laws and the impact of basalt weathering on the global carbon cycle. Chemical Geology, 202, 257– 273. Donnadieu, Y., Ramstein, G., Fluteau, F., Besse, J. & Meert, J. 2002. Is high obliquity a plausible cause for Neoproterozoic glaciations? Geophysical Research Letters, 29, doi: 10.1029/2002GL015902. Donnadieu, Y., Fluteau, F., Ramstein, G., Ritz, C. & Besse, J. 2003. Is there a conflict between the Neoproterozoic glacial deposits and the snowball Earth interpretation: an improved understanding with numerical modeling. Earth and Planetary Science Letters, 208, 101– 112. Donnadieu, Y., Godde´ris, Y., Ramstein, G., Ne´delec, A. & Meert, J. G. 2004a. Snowball Earth triggered by continental break-up through changes in runoff. Nature, 428, 303– 306.

159

Donnadieu, Y., Ramstein, G., Fluteau, F., Roche, D. & Ganopolski, A. 2004b. The impact of atmospheric and oceanic heat transport on the sea ice-albedo instability during the Neoproterozoic. Climate Dynamics, 22, 293– 306. Donnadieu, Y., Ramstein, G., Godde´ris, Y. & Fluteau, F. 2004c. Global tectonic setting and climate of the Late Neoproterozoic: a climate–geochemical coupled study. In: Jenkins, G. S., McMenamin, M. A. S., McKay, C. P. & Sohl, L. (eds) The Extreme Proterozoic: Geology, Geochemistry, and Climate. Geophysical Monograph 146. American Geophysical Union, Washington, DC, 79 – 89. Ernst, R. E. & Buchan, K. L. 2002. Maximum size and distribution in time and space of mantle plumes: evidence from large igneous provinces. Journal of Geodynamics, 34, 309–342. Evans, D. A. D. 2000. Stratigraphic, geochronological, and paleomagnetic constraints upon the Neoproterozoic climatic paradox. American Journal of Science, 300, 347–433. Franc¸ois, L. M. & Godde´ris, Y. 1998. Isotopic constraints on the Cenozoic evolution of the carbon cycle. Chemical Geology, 145, 177–212. Gaillardet, J., Dupre´, B., Louvat, P. & Alle`gre, C. J. 1999. Global silicate weathering and CO2 consumption rates deduced from the chemistry of the large rivers. Chemical Geology, 159, 3– 30. Galy, V., France-Lanord, C., Beyssac, O., Faure, P., Kudrass, H. & Palhol, F. 2007. Efficient organic carbon burial in the Bengal fan sustained by the Himalayan erosional system. Nature, 450, 407– 410. Godde´ris, Y. & Joachimski, M. M. 2004. Global change in the late Devonian: modelling the Frasnian–Famennian short-term carbon isotope isotope excursions. Palaeogeography, Palaeoclimatology, Palaeoecology, 202, 309– 329. Godde´ris, Y. & Donnadieu, Y. 2008. Carbon cycling and snowball Earth. Nature, 456, doi: 10.1038/nature07653. Godde´ris, Y., Donnadieu, Y. et al. 2003. The Sturtian glaciation: fire and ice. Earth and Planetary Science Letters, 211, 1 –12. Godde´ris, Y., Donnadieu, Y. et al. 2007. Coupled modeling of global carbon cycle and climate in the Neoproterozoic: links between Rodinia breakup and major glaciation. Comptes Rendus Geoscience, 339, 212–222. Goodman, J. C. 2006. Through thick and thin: marine and meteoric ice in a ‘Snowball Earth’ climate. Geophysical Research Letters, 33, doi: 10.1029/2006GL026840. Goodman, J. C. & Pierrehumbert, R. T. 2003. Glacial flow of floating marine ice in ‘Snowball Earth’. Journal of Geophysical Research, 108, doi: 10.1029/2002JC001471. Halverson, G. P., Hoffman, P. F., Schrag, D. P. & Kaufman, A. J. 2002. A major perturbation of the carbon cycle before the Ghaub glaciation (Neoproterozoic) in Namibia: prelude to snowball Earth? Geochemistry, Geophysics, Geosystems, 3, doi: 10.1029/ 2001GC000244. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Towards a Neoproterozoic composite carbon isotope record. Geological Society of America Bulletin, 117, 1181–1207. Higgins, J. A. & Schrag, D. P. 2003. Aftermath of a snowball Earth. Geochemistry Geophysics Geosystems, 4, doi: 10.1029/2002GC000403. Hoffman, P. F. & Li, Z.-X. 2009. A palaeogeographic context for Neoproterozoic glaciation. Palaeogeography, Palaeoclimatology, Palaeoecology, 277, 158– 172. Hoffman, P. F., Kaufman, A. J., Halverson, G. P. & Schrag, D. P. 1998. A Neoproterozoic Snowball Earth. Science, 281, 1342– 1346. Hoffman, P. F., Halverson, G. P., Domack, E. W., Husson, J. M., Higgins, J. A. & Schrag, D. P. 2007. Are basal Ediacaran (635 Ma) post-glacial ‘cap dolostones’ diachronous? Earth and Planetary Science Letters, 258, 114–131. Hyde, W. T., Crowley, T. J., Baum, S. K. & Peltier, W. R. 2000. Neoproterozoic ‘snowball Earth’ simulations with a coupled climate/ice sheet model. Nature, 405, 425– 429. Ishiwatari, M., Nakajima, K., Takehiro, S. & Hayashi, Y.-Y. 2007. Dependence of climate states of gray solar atmosphere on solar constant: from the runaway greenhouse to the snowball states. Journal of Geophysical Research, 112, doi: 1029/2006JD007368. Jenkins, G. S. 2000. Global climate model high-obliquity solutions to the ancient climate puzzles of the Faint-Young Sun Paradox and

160

Y. GODDE´RIS ET AL.

low-altitude Proterozoic Glaciation. Journal of Geophysical Research, 105, 7357– 7370. Jenkins, G. S. 2004. High obliquity as an alternative hypothesis to early and late Proterozoic extreme climate conditions. In: Jenkins, G. S., McMenamin, M. A. S., McKay, C. P. & Sohl, L. (eds) The Extreme Proterozoic: Geology, Geochemistry, and Climate. Geophysical Monograph 146. American Geophysical Union, Washington, DC, 73– 78. Jenkins, G. S. & Frakes, L. A. 1998. GCM sensitivity test using increased rotation rate, reduced solar forcing and orography to examine low latitude glaciation in the Neoproterozoic. Geophysical Research Letters, 25, 3525–3528. Jenkins, G. S. & Smith, S. R. 1999. GCM simulations of snowball Earth conditions during the Late Proterozoic. Geophysical Research Letters, 26, 2263–2266. Key, R. M., Liyungu, A. K., Njamu, F. M., Somwe, V., Banda, J., Mosley, P. N. & Armstrong, R. A. 2001. The western arm of the Lufilian Arc in NW Zambia and its potential for copper mineralization. Journal of African Earth Science, 33, 503–528. Kirschvink, J. L. 1992. Late Proterozoic low-latitude global glaciation: the snowball earth. In: Schopf, J. W. & Klein, C. (eds) The Proterozoic Biosphere. Cambridge University Press, Cambridge, 51 –52. Knoll, A. H. 2003. Life on a Young Planet. Princeton University Press, New Jersey. Leather, J., Allen, P. A., Brasier, M. D. & Cozzi, A. 2002. Neoproterozoic snowball Earth under scrutiny: evidence from the Fiq glaciation of Oman. Geology, 30, 891–894. Le Hir, G., Ramstein, G., Donnadieu, Y. & Pierrehumbert, R. T. 2007. Investigating plausible mechanisms to trigger a deglaciation from a hard snowball Earth. Comptes Rendus Geoscience, 339, 274– 287. Le Hir, G., Godde´ris, Y., Donnadieu, Y. & Ramstein, G. 2008a. A geochemical modelling study of the evolution of the chemical composition of seawater linked to a snowball glaciation. Biogeosciences, 5, 253–267. Le Hir, G., Godde´ris, Y., Ramstein, G. & Donnadieu, Y. 2008b. A scenario for the evolution of the atmospheric pCO2 during a snowball Earth. Geology, 36, 47 –50. Le Hir, G., Donnadieu, Y. et al. 2009. The snowball Earth aftermath: exploring the limits of continental weathering processes. Earth and Planetary Science Letters, 277, 453–463. Le Hir, G., Donnadieu, Y., Krinner, G. & Ramstein, G. 2010. Toward the snowball Earth deglaciation. Climate Dynamics, 35, 285–297. Lewis, J. P., Weaver, A. J., Johnston, S. T. & Eby, M. 2003. Neoproterozoic ‘snowball Earth’: dynamic sea ice over a quiescent ocean. Paleoceanography, 18, doi: 10.1029/2003PA000926. Lewis, J. P., Eby, M., Weaver, A. J., Johnston, S. T. & Jacob, R. L. 2004. Global glaciation in the Neoproterozoic: reconciling previous modelling results. Geophysical Research Letters, 31, doi: 10.1029/ 2004GL019725. Lewis, J. P., Weaver, A. J. & Eby, M. 2006. Deglaciating the snowball Earth: sensitivity to surface albedo. Geophysical Research Letters, 33, doi: 10.1029/2006GL027774. Lewis, J. P., Weaver, A. J. & Eby, M. 2007. Snowball versus slushball Earth: dynamic versus nondynamic sea ice? Journal of Geophysical Research, 112, doi: 10.1029/2006JC004037. Li, Z. X., Li, X. H., Kinny, P. D. & Wang, J. 1999. The break up of Rodinia: did it start with a mantle plume beneath South China? Earth and Planetary Science Letters, 173, 171–181. Li, Z. X., Li, X. H., Kinny, P. D., Wang, J., Zhang, S. & Zhou, H. 2003. Geochronology of Neoproterozoic syn-rift magmatism in the Yangtze Craton, South China and correlations with other continents: evidence for a mantle superplume that broke up Rodinia. Precambrian Research, 122, 85 – 109. Macdonald, F. A., Schmitz, M. D. et al. 2010. Calibrating the Cryogenian. Science, 327, 1241–1243. Macouin, M., Besse, J., Ader, M., Gilder, S., Yang, Z., Sun, Z. & Agrinier, P. 2004. Combined paleomagnetic and isotopic data from the Doushantuo carbonates, South China: implications for the ‘snowball Earth’ hypothesis. Earth and Planetary Science Letters, 224, 387– 398.

Marshall, H. G., Walker, J. C. G. & Kuhn, W. R. 1988. Long-term climate change and the geochemical cycle of carbon. Journal of Geophysical Research, 93, 791– 801. McKay, C. P. 2000. Thickness of tropical ice and photosynthesis on a snowball Earth. Geophysical Research Letters, 27, 2153–2156. Meert, J. G. & Powell, C. M. 2001. Assembly and break-up of Rodinia: introduction to the special volume. Precambrian Research, 110, 1– 8. Meert, J. G. & Torsvik, T. H. 2004. Paleomagnetic constraints on Neoproterozoic ‘Snowball Earth’ continental reconstructions. In: Jenkins, G. S., McMenamin, M. A. S., McKay, C. P. & Sohl, L. (eds) The Extreme Proterozoic: Geology, Geochemistry, and Climate. Geophysical Monograph 146. American Geophysical Union, Washington, DC, 5 –11. Micheels, A. & Montenari, M. 2008. A snowball Earth versus a slushball Earth : results from Neoproterozoic climate modeling sensitivity experiments. Geosphere, 4, 401–410. Park, J. K., Buchan, K. L. & Harlan, S. S. 1995. A proposed giant dyke swarm fragmented by the separation of Laurentia and Australia based on paleomagnetism of ca.780 Ma mafic intrusions in western North America. Earth and Planetary Science Letters, 132, 129– 139. Pavlov, A. A., Hurtgen, M. T., Kasting, J. F. & Arthur, M. A. 2003. Methane-rich Proterozoic atmosphere. Geology, 31, 87 –90. Pavlov, A. A., Toon, O. B., Pavlov, A. K., Bally, J. & Pollard, D. 2005. Passing through a giant molecular cloud: ‘snowball’ glaciations produced by interstellar dust. Geophysical Research Letters, 32, doi: 10.1029/2004GL021890. Peltier, W. R., Tarasov, L., Vettoretti, G. & Solheim, L. P. 2004. Climate dynamics in deep time: modeling the ‘snowball bifurcation’ and assessing the plausibility of its occurrence. In: Jenkins, G. S., McMenamin, M. A. S., McKay, C. P. & Sohl, L. (eds) The Extreme Proterozoic: Geology, Geochemistry, and Climate. Geophysical Monograph 146. American Geophysical Union, Washington, DC, 107–124. Peltier, W. R., Liu, Y. & Crowley, J. W. 2007. Snowball Earth prevention by dissolved organic carbon remineralization. Nature, 450, 813– 818. Petoukhov, V., Ganopolski, A., Brovkin, V., Claussen, M., Eliseev, A., Kubatzki, C. & Rahmstorf, S. 2000. CLIMBER-2: a climate system model of intermediate complexity. Part I: model description and performance for present climate. Climate Dynamics, 16, 1 – 17. Pierrehumbert, R. T. 2004. High levels of atmospheric carbon dioxide necessary for the termination of global glaciation. Nature, 429, 646– 649. Pierrehumbert, R. T. 2005. Climate dynamics of a hard Snowball Earth. Journal of Geophysical Research, 110, doi: 10.1029/2004JD005162. Pollard, D. & Kasting, J. F. 2004. Climate-ice sheet simulations of Neoproterozoic glaciation before and after the collapse to snowball Earth. In: Jenkins, G. S., McMenamin, M. A. S., McKay, C. P. & Sohl, L. (eds) The Extreme Proterozoic: Geology, Geochemistry, and Climate. Geophysical Monograph 146. American Geophysical Union, Washington, DC, 91 –105. Pollard, D. & Kasting, J. F. 2005. Snowball Earth: a thin-ice solution with flowing sea glaciers. Journal of Geophysical Research, 111, doi: 10.1029/2004JC002525. Pollard, D. & Kasting, J. F. 2006. Reply to comment by Stephen G. Warren and Richard E. Brandt on ‘Snowball Earth: A thin-ice solution with flowing sea glaciers’. Journal of Geophysical Research 111, C09017, doi: 10.1029/2006JC003488. Poulsen, C. J. 2003. Absence of a runaway ice-albedo feedback in the Neoproterozoic. Geology, 31, 115– 118. Poulsen, C. J. & Jacob, R. L. 2004. Factors that inhibit snowball Earth simulation. Paleoceanography, 19, doi: 10.1029/2004PA001056. Poulsen, C. J., Pierrehumbert, R. T. & Jacob, R. L. 2001. Impact of ocean dynamics on the simulation of the Neoproterozoic ‘snowball Earth’. Geophysical Research Letters, 28, 1575–1578. Poulsen, C. J., Jabob, R. L., Pierrehumbert, R. T. & Huynh, T. T. 2002. Testing paleogeographic controls on a Neoproterozoic snowball Earth. Geophysical Research Letters, 29, doi: 10.1029/ 2001GL014352.

MODELLING SNOWBALL EARTH

Prave, A. R., Fallick, A. E., Thomas, C. W. & Graham, C. M. 2009. A composite C-isotope profile for the Neoproterozoic Dalradian Supergroup of Scotland and Ireland. Journal of the Geological Society, London, 166, 1 –13. Ramstein, G., Donnadieu, Y. & Godde´ris, Y. 2004. Proterozoic glaciations. Comptes Rendus Geoscience, 336, 639–646. Raub, T. D. & Evans, D. A. D. 2006. Magnetic reversals in basal Ediacaran cap carbonates: a critical review. Eos, Transactions American Geophysical Union, 87, abstract GP41-02. Ritz, C., Rommelaere, V. & Dumas, C. 2001. Modeling the evolution of Antarctic ice sheet over the last 420 000 years: implications for altitude changes in the Vostok region. Journal of Geophysical Research, 106, 31 943– 31 964. Romanova, V., Lohmann, G. & Grosfeld, K. 2006. Effect of land albedo, CO2, orography, and oceanic heat transport on extreme climates. Climate of the Past, 2, 31 –42. Rose, B. E. J. & Marshall, J. 2009. Ocean heat transport, sea ice, and multiple climate states: insights from energy balance models. Journal of Atmospheric Sciences, 66, 2828– 2843. Schrag, D. P., Berner, R. A., Hoffman, P. F. & Halverson, G. P. 2002. On the initiation of a Snowball Earth. Geochemistry, Geophysics, Geosystems, 3, doi: 10.1029/2001GC000219. Sellers, W. D. 1969. A global climatic model based on the energy balance of the Earthatmosphere system. Journal of Applied Meteorology, 8, 392– 400. Stone, P. H. & Yao, M. S. 2004. The ice-covered Earth instability in a model of intermediate complexity. Climate Dynamics, 22, 815– 822.

161

Torsvik, T. H., Carter, L. M., Ashwal, L. D., Bhushan, S. K., Pandit, M. K. & Jamtveit, B. 2001. Rodinia refined or obscured: paleomagnetism of the Malani igneous suite (NW India). Precambrian Research, 108, 319–333. Trindade, R. I. F., Font, E., D’Agrella-Filho, M. S., Nogueira, A. C. R. & Riccomini, C. 2003. Low-latitude and multiple geomagnetic reversals in the Neoproterozoic Puga cap carbonate, Amazon craton. Terra Nova, 15, 441–446. Walker, J. C. G., Hays, P. B. & Kasting, J. F. 1981. A negative feedback mechanism for the long-term stabilization of Earth’s surface temperature. Journal of Geophysical Research, 86, 9776–9782. Warren, S. G. & Brandt, R. E. 2006. Comment on ‘Snowball Earth: a thin-ice solution with flowing sea glaciers’ by D. Pollard and J.F. Kasting. Journal of Geophysical Research, 111, C09016, doi: 10.1029/2005JC003411. Warren, S. G., Brandt, R. E., Grenfell, T. C. & McKay, C. P. 2002. Snowball Earth: ice thickness on the tropical ocean. Journal of Geophysical Research, 107, doi: 10.1029/2001JC001123. Wingate, M. T. D., Campbell, I. H., Compston, W. & Gibson, G. G. 1998. Ion microprobe U– Pb ages for Neoproterozoic basaltic magmatism in south-central Australia and implications for the breakup of Rodinia. Precambrian Research, 87, 135–159. Wingate, M. T. D. & Giddings, J. W. 2000. Age and paleomagnetism of the Mundine Well dyke swarm, Western Australia: implications for an Australia– Laurentia connection at 755 Ma. Precambrian Research, 100, 335–357.

Chapter 11 The record of Neoproterozoic glaciation in the Taoude´ni Basin, NW Africa G. A. SHIELDS-ZHOU1*, MAX DEYNOUX2 & LAWRENCE OCH1 1

Department of Earth Sciences, University College London, Gower Street, London, WC1E 6BT, UK 2

Laure´lie, 12270 Bor et Bar, France

*Corresponding author (e-mail: [email protected]) Abstract: The Taoude´ni Basin covers over 1 000 000 km2 of the West African Craton, bounded by Pan-African orogenic belts. Four supergroups separated by craton-scale unconformities are recognized, with Neoproterozoic glaciogenic deposits occurring at the base of Supergroup 2. The Jbe´liat Group occurs along a continuous, 1300-km-long, narrow belt from the Adrar region of Mauritania to the eastern limit of the Hank in Algeria and comprises thin glacial drift capped widely by periglacial polygonal structures, with more complex glacial sequences preserved in palaeo-depressions. A thicker, variously marine and continental glaciogenic succession can be found in southern parts, while fully marine, glacially influenced successions are only known from the extreme SW of the basin. The ‘triad’ sequence of diamictites overlain by barite-bearing ‘cap’ dolostones and then by green shales and/or bedded cherts (silexites) is ubiquitous and has long been used to correlate the Supergroup 1/2 boundary across the basin and into the surrounding orogenic belts. The bedded cherts commonly show a volcanic influence and are cemented by early marine calcite at their base at Adrar, Mauritania. Although fossil-based age constraints are scarce and ambiguous, regional tectonic events indicate that ‘triad’ deposition occurred between the Bassaride (665– 655 Ma) and Dahomeyide (610–580 Ma) orogens. Recent U –Pb zircon studies of ignimbrite tuffs provide a minimum age for the glaciation of c. 600 Ma. Correlation of supergroup 2 glacial deposits with the c. 635 Ma end-Cryogenian (‘Marinoan’) glaciation is likely and is supported by limited carbon and strontium isotope data. Barite is commonly found within the cap carbonate and may relate to methane seepage and/or unusual oceanographic conditions after deglaciation. Several studies have attributed sequence complexity within the post-glacial succession to isostatic reequilibration. The Taoude´ni Basin represents a rare Neoproterozoic example of terrestrial tillites and associated periglacial facies.

On the West African Craton, the Neoproterozoic and Palaeozoic sedimentary cover can be subdivided into four supergroups (or megasequences; e.g. Deynoux et al. 2006) bounded by cratonscale unconformities (Trompette 1973). The lithostratigraphic ‘triad’ association (diamictite-cap dolostone-bedded chert), which forms the base of Supergroup 2, has long been used as a marker horizon (Zimmermann 1960; Leprun & Trompette 1969). The first detailed studies of the glacial sedimentology of this region were conducted during the 1970s, and what follows relies heavily on over three decades of study by French researchers who recorded the extraordinary preservation of glacial and periglacial features across vast areas of arid North Africa (e.g. Deynoux 1982; Deynoux et al. 2006). This chapter mainly focuses on some better known sections of the Taoude´ni Basin that have recently been the subject of isotopic studies (Alvaro et al. 2007; Shields et al. 2007a, b). Owing to the large size of the Taoude´ni Basin, these descriptions are necessarily oversimplified and the reader is referred to the published literature for more detailed information on these and other (e.g. Bassaride, Rokelide, AntiAtlas, Hoggar-Iforas and Dahomeyide belts, and the Tindouf Basin) Neoproterozoic glaciogenic successions of the West African craton (Deynoux 1980, 1985; Deynoux & Trompette, 1981; Proust & Deynoux 1994; Deynoux et al. 2006). Glaciogenic deposits of the neighbouring Volta Basin (Fig. 11.1) are presumed to be correlative with those of the Taoude´ni Basin Supergroup 2 and will also be considered here.

Structural framework and basin setting The West African Craton has been a tectonic entity since at least 1600 Ma and is rimmed by mobile belts that became active during late Proterozoic and Palaeozoic times and are related to the Transaharan suture zone (Fig. 11.1). On the craton, the sedimentary cover is preserved in widespread, variably connected basins containing upper Proterozoic and Palaeozoic deposits of which the largest is the Taoude´ni Basin, forming a thin (3 km on

average), continuous tabular blanket over distances of 1000– 1500 km (Fig. 11.1). The Taoude´ni Basin was possibly a less rigid part of the platform, and is rimmed by the Mauritanide and Hoggar-Iforas fold belts. No substantial sediment thickening occurs in the central part of the basin, but extra subsidence has caused some thickening in peripheral troughs and sub-basins, such as the Gourma Basin to the east and the Madina-Kouta and Bove´ basins to the west, which were involved in the Pan-African and/or Hercynian fold belts. With the exception of the area fringing the peripheral fold belts, the Taoude´ni Basin has been affected by regional-scale (epeirogenic) tilting, which has generated low angular unconformities, and the reactivation of basement faults frequently intruded by diabase sills and dykes. The sedimentary deposits are generally devoid of tectonism, while metamorphism is limited to the thermal effects of diabase intrusions in the proximal host material that took place when the Atlantic Ocean began opening during the Jurassic Period. In this regard, some sedimentary rocks may have experienced locally high-grade, hydrothermal alteration. In spite of their age, the sediments have undergone only modest burial diagenesis because the sedimentary pile is relatively thin and the platform has remained above sea level since the end of the Carboniferous Period. The Volta Basin thickens eastward, where it becomes gently folded before disappearing beneath the Dahomeyide thrust belt. According to sparse contextual information (Deynoux et al. 2006), the Volta Basin can be considered either to be a single foreland basin or as part of the Pan-African passive margin overlain by a foreland basin made up of the Tamale Supergroup, lateral equivalents of which can be found in intermontane, molasse-filled grabens of the Dahomeyide belt (Affaton et al. 1991).

Stratigraphy On the West-African platform, the Neoproterozoic and Palaeozoic sedimentary cover can be subdivided into four supergroups

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 163– 171. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.11

164

G. A. SHIELDS-ZHOU ET AL.

Fig. 11.1. Simplified geological map of the Taoude´ni Basin and adjacent areas in NW Africa (Source: Deynoux et al. 2006).

(megasequences) separated by craton-scale unconformities (Trompette 1973). Supergroup 1 rests with major unconformity upon metamorphic and granitic basement across the Taoude´ni Basin and contains sedimentary strata of Mesoproterozoic age (Clauer et al. 1982; Rooney et al. 2010). Supergroup 2 begins with glacial deposits resting with an erosional and slightly angular unconformity upon Supergroup 1 or directly upon the basement. The glacial deposits are capped by a thin, but continuous unit of calcareous dolomite, which contains barite in places. Bedded cherts and green shales with a thin but continuous limestone bed at their base generally overlie this cap-dolostone unit and form the uppermost part of the so-called ‘triad’ of diamictitecap dolostone-bedded chert. Supergroup 3 consists of Late Ordovician glacial deposits and post-glacial Silurian shales, whereas Supergroup 4 corresponds to unconformably overlying Devonian and Carboniferous strata. In the Adrar region of Mauritania, glacial deposits of Supergroup 2 are unconformably underlain by early Neoproterozoic

strata of the Char, Atar and Assabet el Hassiane Groups (Fig. 11.2). The Atar Group and overlying siliciclastic rocks of the Assabel-el-Hassiane Group in Mauritania experienced minor tilting during Pan-African events before being peneplained by subsequent uplift and erosion to form a largely flat, sub-glacial substrate (Trompette 1973; Deynoux, 1980). Glacial deposits and their ‘cap dolostone’ form the Jbe´liat Group, which outcrops nearly continuously for 1300 km along a narrow belt from the Adrar region to the eastern limit of the Hank in Algeria (Fig. 1). The Adrar cap carbonate package consists of one or two dolostone units, with an intervening siliciclastic package of up to 40 metres thickness, and a laterally extensive, thin limestone bed that disconformably overlies the uppermost dolostone forming the base of the bedded cherts and shales of the Te´niagouri Group. In the Walidiala Valley, which straddles the Guinea-Senegal border at the extreme southwestern margins of the Taoude´ni Basin, Supergroups 1 and 2 overlie basement rocks of the Kenieba Inlier and are intruded and protected from erosion by

NEOPROTEROZOIC GLACIATION IN THE TAOUDE´NI BASIN

165

Fig. 11.2. Stratigraphy of the northern part of the Taoude´ni Basin in the Adrar area and the Hank-Fersiga area (after Deynoux et al. 2006). Ages are from Rooney et al. 2010 and Lahonde`re et al. 2005.

plateau-forming diabase sills of Jurassic age. Supergroup 1 consists of intertidal to supratidal, coarse-grained, red wackestones and intervening siltstone forming the Se´gou Group. The lower part of Supergroup 2 is represented by the Mali Group, which rests on the Mesoproterozoic Se´gou Group with deep erosional unconformity. The Mali Group (Fig. 11.3) consists of siliciclastic, mostly siltstone units and comprises in its basal part the regionally correlative triad of glaciogenic strata, dolostone and bedded chert. The Mali Group has been subdivided into two formations (Culver & Hunt 1991): the Hassanah Diallo Formation, 50 –120 m thick,

which represents the glacially related deposits, and the Nandoumari Formation, up to 130 m thick, which comprises at its base quartz arenites overlain by dolostone, siltstone and bedded chert (Shields et al. 2007a). The southern region of the Taoude´ni Basin stands out due to the presence below the triad and associated craton-wide unconformity of a 400–500-m-thick complex succession showing glacial influence (Fig. 11.4). Both triad and underlying glacially influenced strata were initially assigned to the ‘Bakoye Group’ (Simon 1979); however, the marked erosional and angular unconformity below

Fig. 11.3. Schematic geological map and stratigraphic column of the Mali Group of the southwestern Taoude´ni Basin in the Walidiala Valley (modified after Shields et al. 2007a).

166

G. A. SHIELDS-ZHOU ET AL.

Fig. 11.4. Stratigraphy and main depositional environments of the glacially related deposits in the southwestern part of the Taoude´ni Basin in Mali (Tambaoura plateaux) (modified from Deynoux et al. 2006).

the diamictites of the triad suggests that the underlying succession records a distinct history. Deynoux et al. (1989, 2006) subsequently subdivided the Bakoye Group into a Koniakari Group, corresponding to the diamictites and post-glacial carbonate unit of the triad, and a Wassangara Group, corresponding to the underlying, at least partly glacially influenced succession. The Wassangara Group forms most of the Tambaoura and Mandingue Plateaux in western Mali and part of the Afolle´ Massif in southern Mauritania (Fig. 11.1). It rests with an erosional and locally angular unconformity upon the Souroukoto Group. The Volta Basin (Fig. 11.5) succession rests with major unconformity on the Eburnean basement of the Leo Shield. The lower (Bombouaka) Supergroup is up to 1000 m thick, comprising largely siliciclastic rocks interspersed with thin calcareous beds in its middle part, and is in general reminiscent of the lower part of Supergroup 1 in the southern part of the Taoude´ni Basin. The middle (Pendjari or Oti) Supergroup is much thicker (2500 – 4000 m) and begins with glaciogenic deposits at its base, resting unconformably on various parts of the Bombouaka Supergroup or directly on basement. The triad of diamictite, dolostone and chert passed upwards into green turbiditic siltstones and shales, which contain phosphorite deposits. The upper (Tamale) Supergroup (500 m thick) is mainly exposed in Ghana and lies unconformably on the Pendjari Supergroup. It comprises terrestrial coarse-grained siliciclastic deposits with intercalated fine-grained and calcareous horizons.

Glaciogenic deposits and associated strata Palaeogeographic reconstructions of the triad-associated glaciation of NW Africa imply an ice sheet centred towards the north

of the Reguibat Shield (Fig. 11.1) with inferred glacial movement southward on the platform and laterally towards oceanic troughs or basins located at the present position of the Pan-African belts. Accordingly, continental glaciation is recorded in the northern part of the Taoude´ni Basin by a thin, irregular (0 –50 m thick) veneer of terrestrial tillites with subordinate proglacial outwash deposits preserved in limited shallow depressions. Towards the south of the platform, the glacial drift thickens (150 –200 m), showing marine influence in small intracratonic basins in western Mali (Kayes area) and becoming wholly marine at the margins of oceanic troughs bordering the Pan-African Bassaride Belt in eastern Senegal and Guinea. Along the northern margin of the Taoude´ni Basin, glacial drift comprises just a few metres of terrestrial tillite, with polygenic and often striated pebbles to boulders overlying striated pavements. Locally, a complex facies association is preserved in smooth large-scale palaeo-depressions such as in the Jbe´liat area in Adrar, Mauritania (Fig. 11.1). In these palaeo-depressions, the glacial deposits thicken up to 50 m and comprise distinct glacial sequences suggesting at least three glacial advances and retreats, with terrestrial tillites, outwash sandstones and various glacial features such as striated clasts and pavements, roches moutonne´es, lacustrine varves with dropstones, and remarkable polygonal structures and sand wedges related to permafrost just below the cap dolostone (Deynoux & Trompette 1976; Deynoux 1980, 1982, 1985). These deposits represent probably one of the only wellpreserved, purely continental records of a Pre-Pleistocene ice sheet (Eyles & Januszczak 2004). Although true tillites are rarer in the southern part of the Taoude´ni Basin, diamictites of the Koniakari Group and associated sandstones and shales form an up to 200-m-thick marine to continental glacial succession well exposed in the Kayes area of Mali (Rossi et al. 1984; Deynoux et al. 1991). In areas of only marine sedimentation like the Walidiala Valley of Senegal, glacial influence can only be recognized due to the occurrence of dropstones between debris flow deposits of the Pelel Member and in shales of the overlying Diagoma Member (Culver & Hunt, 1991; Shields et al. 2007a). The post-glacial cap-carbonate succession in Adrar has been described recently by Shields et al. (2007b) and Alvaro et al. (2007), and only a brief summary is given here. The generally buffcoloured cap dolostone is either well-bedded or brecciated with thin and discontinuous intervening stringers of black chert. Bedding surfaces are commonly erosional at all scales, that is, scoured, rather than gradational or crinkled as with microbialites, and show crossbedding only rarely. When passing to the brecciated facies, beds become laterally and vertically folded and disrupted. Fissures are common and are generally filled with ferroan dolomite, black chert, white quartz or barite. A second generation of fissures is filled with very coarse ferroan calcite spar. Microscopically, the Jbe´liat dolostone comprises alternating grain sizes from dolomicrite to dolosparite with poorly developed or residual cement made up of a mixture of clay, cryptocrystalline silica, siderite and pyrite. The cap dolostone splits locally into two distinct stratigraphic horizons providing space for 40 (Adrar) to 80 m (Hank) of onlapping green to purple shales and siltstones including sandy dolostone intercalations and passing upward in the Hank region into sandy tidal deposits and eolian sandstone (Moussine-Pouchkine & Bertrand-Sarfati 1997). Along the Atar cliff, the upper dolostone horizon becomes progressively sandy until it forms a bank up to 7 m thick made up entirely of coarse-grained to granular, trough crossbedded, dolomitic sandstone with isolated lenses of finer dolostone. Barite is associated with the brecciated facies at Jbe´liat, Mauritania (mostly in the upper cap-dolostone unit) and at Pont de Kabate´ in Mali (Shields et al. 2007b), and appears as palisadic crystals filling cracks or cavities, as well as monocrystalline geodes and laminated domes up to 30 cm across. In the Jbe´liat area and part of the Atar cliff area, over c. 30– 40 km of the cap-dolostone exposure, a persistent

NEOPROTEROZOIC GLACIATION IN THE TAOUDE´NI BASIN

2°W 12°

0°

2°E

100 km

167

Panafrican internal units Suture zone Kara struct. Unit Atacora struct. Unit

Panafrican external units

Buem struct. Unit

10°

Tamale Megasequence A Flyschoid succession

KFZ

8°

B

Tillite and cap dolostone

Peniari or Oti Megasequence

Yemboure Gr. Fosse aux lions Gr.

Bambouaka Megasequence

Dapaong Gr. 6°

c Atlanti

A

Ocean

Eburnean basement Thrust

B

50 km

30– 50-cm-thick horizon of limestone is intercalated between the cap dolostone and bedded cherts/siltstones of the overlying Te´niagouri Group. It is characterized by its purple to grey-green colour and its brecciated texture defined by oriented, millimetreto centimetre-scale angular flakes of detrital, largely volcanogenic material and authigenic barite (Hoffman & Schrag 2002; Shields et al. 2007b), cemented by early marine calcite. At both Jbe´liat and the Atar Cliffs area, this limestone horizon may comprise reverse-graded, calcite nodules or spherulites, which exhibit cores of volcanogenic chert. The contact of the purple limestone unit with the underlying cap dolostone is generally sharp and marked by an erosional hardground surface; however, its contact with the overlying cherts is transitional. In the marine succession at Walidiala Valley, the Bowal Member dolostone similarly consists of centimetre-to-decimetre bedded, internally laminated dolostone and dolomitic siltstone beds, overlying similar siltstone containing abundant lonestones (dropstones). The dolostone unit shows widespread slumping associated with the arrival of debris comprising volcanogenic diamict that underwent pervasive dolomitization after redeposition. The bedded dolostone unit is characterized by various forms of brittle deformation: fractures, tepee-like buckling and discontinuous fitted brecciation with created space being filled by early chert and/or by dolospar. In undeformed beds, this finegrained dolostone exhibits millimetre-scale layering defined by mechanical laminations with no textural evidence for any microbial influence. Microscopically, the Bowal Member dolostone comprises equigranular grains of dolomite (generally .90%) with accessory quartz and biotite grains of similar size.

Fig. 11.5. Simplified geological map and a synthetic cross-sections of the Volta Basin and Dahomeyide belt (modified after Deynoux et al. 2006).

Individual beds have sharp or erosional bases and fine upwards from a surface of detrital dolomite/quartz lag of c. 100 mm size to dolomicrite of ,10 mm in diameter (Shields et al. 2007a). The Volta Basin tillite fills an erosional and slightly angular unconformity of glacial origin with striated glacial pavements along NW –SE trending palaeo-valleys. Its modest thickness (c. 2 m) and geological context suggests that this deposit represents a ‘true’ terrestrial tillite, which is overlain by a ,6 m cap dolostone (Porter et al. 2004). At Bwipe (Ghana), the cap dolostone unit is thicker, comprising up to 12 m of finely laminated, ‘microbial’ dolostone (Nedelec et al. 2007).

Boundary relations with overlying and underlying non-glacial units Tilting during Pan-African orogenic events and largely pre-glacial peneplanation caused glacial and glacially influenced units of Supergroup 2 to be deposited unconformably upon marine sedimentary rocks of Supergroup 1 throughout the Taoude´ni Basin, and apparently across NW Africa, including in the Volta Basin. In the Adrar region of Mauritania, the top of the glacial drift is marked by highly indurated, black, coarse-grained sandstone, which forms patterned ground with polygonal structures and sand wedges typical of permafrost (Deynoux 1982). Laminated dolostone drapes the immediately post-glacial topography or rests directly upon the underlying and commonly striated periglacial substrate in both Taoude´ni and Volta basins.

168

G. A. SHIELDS-ZHOU ET AL.

In the Walidiala Valley section in Senegal, glaciomarine influence wanes up section as dropstones become both smaller and rarer in mudstones of the Diagoma Member, the top 10 m of which seem to be devoid of any dropstones. In places, channels cut into these mudstones and are filled with coarse-grained, crossbedded sandstone and gravels of the Tanague´ Member. A 2– 7 m thick, silty cap dolostone, the Bowal Member, caps the glaciogenic sequence and rests erosionally on the underlying siltstone due to slumping or possibly conformably on coarse-grained siliciclastics (Shields et al. 2007a).

Chemostratigraphy Only limited isotope work has been carried out in Neoproterozoic strata of the Taoude´ni and Volta basins. Earlier studies reported Sr- and C-isotope data from the type section of the unconformably underlying Atar Group, in Mauritania (Veizer et al. 1983; Fairchild et al. 1990), but only recently have studies focused on post-glacial strata across the basin (Alvaro et al. 2007; Shields et al. 2007a, b). Despite pervasive recrystallization and multiple phases of cementation, which makes primary carbon-isotope trends (e.g. the Jbe´liat section, Mauritania, and the Goumare´ section, Mali) hard to recognize, published results reveal a consistent picture of a post-glacial cap dolostone unit marked by low d13CPDB (averaging – 4‰) trending up-section to more negative values; this is most clearly seen in the Walidiala Valley sections of Senegal and the Koniakari section, Mali. Lowermost d13C values (– 6‰) are reported from a limestone unit at the base of the Te´niagouri Group, which overlies postglacial cap-dolostone strata in the Atar Cliffs area of Mauritania (Alvaro et al. 2007). Shields et al. (2007b) trace a trend to more positive values through this limestone unit with d13C as high as þ3.7‰ in places. High-resolution isotopic data from four sections of the Volta Basin (Burkina Faso: Arli, Kodjari, Koundjouari; and Ghana: Bwipe) show a similarly decreasing upward trend (Porter et al. 2004; Nedelec et al. 2007). Positive correlation between d13C and d18O appears to be characteristic (Porter et al. 2004; Shields et al. 2007a). Barite samples from Mali and Mauritania have been analysed for Sr and S isotopes (Shields et al. 2007b). 87Sr/86Sr ratios are consistent between barite deposits (0.70773 –0.70814) with the results of leaching experiments indicating that Sr in the barite was derived from seawater with 87Sr/86Sr ¼ 0.7077 –8. In contrast, barite d34S compositions range widely, exhibiting values between þ20‰ and þ46‰. This indicates that barite sulphate derives from a marine source with d34S of about þ20‰, and that it has undergone considerable isotopic fractionation due to sulphate reduction (Shields et al. 2007b). Sr isotope compositions have also been reported for the cap dolostone at Bwipe (Volta Basin), where they range from 0.7061 to 0.7073 (Nedelec et al. 2007). Such low values are atypical for basal Ediacaran dolostones, which generally exhibit variably radiogenic ratios (e.g. Yoshioka et al. 2003) that are unlikely to be representative of ambient seawater.

Other characteristics Economic deposits associated with the glaciogenic strata are limited. Barite deposits are commonly associated with the cap dolostone and are of high purity but are too localized to be of economic interest. The Atar and Assabet el Hassiane groups and their equivalents below the glacial level in Mauritania and Mali (as well as the Silurian shales above the Late Ordovician glacial deposits) have attracted interest as petroleum source rocks. Despite a gas show in the 1970s, only four wells are known to have been drilled to date (Craig et al. 2009). Banded haematite cherts (BIFs) are found within shales of the Wassangara Group (Taoudeni Basin, Mali),

sandwiched between terrestrial strata showing fluvial, eolian and glacial influence (Deynoux et al. 1991). Simple discoidal impressions were reported from strata below glaciogenic deposits by Bertrand-Sarfati et al. (1995) in the Hank region of northern Mauritania; however, these are not necessarily of Ediacaran age, as simple circular impressions are also known from Cryogenian strata in Canada and older Proterozoic strata elsewhere (e.g. Cruse & Harris 1994). The discovery of Cambrian-type small shelly fossils, including the protogastropod Aldanella attleborensis in dolostone scree in Walidiala Valley, Senegal (Culver et al. 1988) caused some to infer a Cambrian age for the Taoude´ni Basin glaciation (Bertrand-Sarfati et al. 1995; Evans 2000). Although the block was believed to derive from the post-glacial Bowal Member cap dolostone, subsequent investigations have not been able to confirm the findings and recent publications tend to play down their significance (Porter et al. 2004; Deynoux et al. 2006; Alvaro et al. 2007; Shields et al. 2007a, b).

Palaeolatitude and palaeogeography There are no firm palaeomagnetic constraints on the position of the West African craton during the Neoproterozoic. Existing data are consistent with a position at mid to high latitudes; however, the primary nature of all published results can be questioned (Evans 2000).

Geochronological constraints Early geochronological studies of the 1970s and 1980s applied K –Ar and Rb –Sr isotope techniques to mudstone units on the West African craton and demonstrated a late Neoproterozoic age for the ‘Infracambrian’ glaciogenic units of Supergroup 2 (Clauer et al. 1982; Clauer & Deynoux 1987). Subsequent work has largely confirmed this viewpoint, although some confusion was injected by the discovery (Culver et al. 1988) of Cambriantype fossils in dolostone scree of the Walidiala Valley, Senegal (see discussion above). The most robust age constraints derive from U – Pb studies of zircons from bedded tuffs within the lower Te´niagouri Group, Mauritania (Lahonde`re et al. 2005). These new age constraints (609.7 + 5.5 Ma, U –Pb TIMS zircon and 606 + 6 Ma, U –Pb SHRIMP zircon) provide a minimum age for the glacial Jbe´liat Group and its cap dolostone of about 600 Ma, which is consistent with correlation with the comparably well-dated c. 635 Ma glaciogenic successions of South China (Condon et al. 2005) and Namibia (Hoffmann et al. 2004). Deynoux et al. (2006) discuss existing age constraints in more detail and conclude that triad deposition across the Taoude´ni Basin, and likely also in the Volta Basin, took place between the Bassaride orogen (Pan-African I, 665– 655 Ma) and the Dahomeyide orogen (610 – 580 Ma). Isotope studies of post-glacial dolostone units across the basin (Shields et al. 2007a, b; Alvaro et al. 2007) support this interpretation (see Discussion below), and fossils are scarce and do not provide additional age constraints. Although some glaciogenic successions of the Taoude´ni Basin may show multiple glacial advances and retreats, for example, in Mali (Deynoux et al. 1991), there is at present no firm evidence for any older, midCryogenian (‘Sturtian’) glaciation. According to Villeneuve (1988) and Villeneuve & Corne´e (1994), the glacially influenced Wassangara Group, which lies unconformably below the triad in the Kayes area of Mali (Proust & Deynoux 1994), also postdates the collision between the West African craton and a western Senegalese block at 660 Ma (Pan-African 1 orogeny). The above constraints permit regional correlation of Neoproterozoic to Cambrian strata of the West African craton and surrounding Pan-African belts (Deynoux et al. 2006, fig. 14).

NEOPROTEROZOIC GLACIATION IN THE TAOUDE´NI BASIN

Discussion The glacial formations on the rim of the northern Taoude´ni Basin consist of terrestrial tillites and proglacial outwash sandstones, which are capped in the Adrar region of northern Mauritania by a persistent horizon of large-scale polygonal structures and sand wedges. These polygonal structures are evidence that a prolonged periglacial period with temperatures low enough for the development of permafrost (Deynoux 1982) followed the retreat of glaciers in this part of the craton. Analogous polygonal structures and sand wedges can also be found 500 km farther south in the Afolle´ Massif, south of Kiffa, indicating that permafrost was not restricted to the Adrar region but extended over a large part of the Taoude´ni Basin. Post-glacial cap dolostones of the northern Taoude´ni Basin mark an abrupt change in the regional depositional environment from terrestrial to marine. The cap dolostones and overlying shales (with calcitic cherts at the base) of the Te´niagouri Group are considered to relate to deglacial to post-glacial eustatic transgression, respectively. However, the presence of intervening shales and sandstones, levelling palaeoreliefs (see above) as observed in the Adrar and Hank regions, highlights the complexity of this transgression. According to Aı¨t-Kaci Ahmed & Moussine-Pouchkine (1994), and BertrandSarfati et al. (1997), in the Hank Algerian border region (Guettatira-Grizim-Fersiga area), the cap carbonates and intervening shales, which contain intercalations of phosphatic grainstones, were deposited during the glacioeustatic transgression. Owing to the balance between glacioeustasy and isostatic rebound, the cap carbonates were locally subjected on topographic highs to subaerial exposure with the development of a complex phosphatic profile, including mini-stromatolites, while sedimentation continued in the depressions. Intervening green shales pass progressively upward into sandy tidal deposits and eolian sandstones in the Hank. These sandstones are overlain by a transgressive surface covered by a few metres of conglomeratic and phosphatic sandstones that preceded the large-scale development on the whole platform of shales equivalent to the Te´niagouri Group in Mauritania. Accordingly, the post-glacial cap-carbonate succession across the Taoude´ni Basin may represent a condensed deposit that relates to the onset of the glacioeustatic transgression, with the transgressive and highstand system tracts preserved only in palaeodepressions. As in the Hank, the cap dolostone in the Adrar reveals a complex, relative sea-level record, with the post-glacial ‘purple’ limestone representing a condensation horizon that marks the beginning of a second, marine transgression, apparently coinciding with regional volcanism. Mineral relationships confirm that regional volcanism occurred after palaeo-topographic highs had for a time become emergent, allowing karst dissolution to occur. Carbonate saturation levels were high enough to grow calcite crystals rapidly on and in caverns immediately beneath the seafloor before glassy debris could disintegrate. The transition from dolomite to calcite precipitation is similar to other such transitions around the world (Hoffman & Schrag 2002), and could be correlative based on the co-occurrence of barite and isotopic trends. At the base of the Neoproterozoic-age Mali Group of the southwestern Taoude´ni Basin, debris flows and turbidite-like, sandy units of the Pelel Member pass upward into siltstone and shale of the Diagoma Member. These two units represent the progressive evolution from some portion of a fan delta fed by a nearby ice shelf to a more distal environment disturbed only by the occasional fallout from passing icebergs. The appearance of coarse-grained, crossbedded sandstone beds and gravels of the overlying Tanague´ Member heralds a return to a shallower, fluvially influenced environment before abrupt transgression caps the glaciogenic succession. The succession in the Walidiala Valley could represent therefore the stratigraphic expression of a glacial retreat in a proximal glaciomarine environment affected by glacioeustasy and isostatic rebound (Shields et al. 2007a).

169

The transgressive unit consists of a regionally extensive, 2 – 7-m-thick, silty dolostone, the Bowal Member, which is isotopically and petrographically indistinguishable from c. 635 Ma capdolostone units elsewhere in NW Africa and worldwide (Shields et al. 2007a). A large volcaniclastic debris flow has caused slumping and soft-sediment deformation within the cap dolostone of the Bowal Member. The widespread association of pyroclastic deposits with cap dolostone in geographically distant successions of the Taoude´ni Basin and neighbouring Hoggar-Iforas Belt (Deynoux et al. 2006) implies that volcanism and deglaciation were roughly contemporaneous across a large area. Early cementation is demonstrated in this case by pervasive dolomitization of the debris flow matrix and supports the notion of unusually high carbonate saturation levels during the aftermath of latest Cryogenian glaciation (Hoffman & Schrag 2002). The origin of the barite mineralization in the post-glacial carbonate succession was specifically addressed by Shields et al. (2007b), who considered that barite formed during the second marine transgression but before deposition of the overlying Te´niagouri Group shales and bedded cherts. Isotopic evidence suggests that barite precipitated due to the mixing of sulphate-rich seawater with a more reducing Ba-rich fluid. The consistent association of barite occurrences throughout the basin with terrestrial periglacial deposits (Deynoux 1980, 1982) and the sporadic occurrence of the barite permit the suggestions that barite mineralization was controlled by local seepage of permafrost methane following deglaciation. Such a scenario implies a relatively short time interval between deglaciation and barite genesis, which is somewhat supported by the general agreement between barite Sr- and S-isotope values and those of other basal Ediacaran cap-carbonate successions (Shields et al. 2007b). The common association of capdolostone to limestone transitions with barite mineralization could, however, point to a global oceanographic origin for the barite, whereby deeper Ba-rich seawater beneath a persistent pycnocline mixed with sulphate-rich surface seawater after deglaciation. The negative d13C trend shown in Taoude´ni Basin post-glacial dolostones is matched by identical trends in correlative sections of the neighbouring Volta Basin (Porter et al. 2004; Nedelec et al. 2007) and worldwide (Kennedy 1996; Halverson et al. 2004), all from dolostone units immediately overlying glaciogenic strata interpreted to belong to the latest Cryogenian (commonly referred to as ‘Marinoan’) glaciation. The affiliation of the Taoude´ni Basin triad with other diamictite – cap dolostone associations from around the world is consistent with the abovementioned geochronological constraints and Sr- and S- isotope data, thus lending further support to the notion of a global deglaciation at c. 635 Ma. In this regard, the limestone unit reported from above post-glacial cap dolostones in Adrar, Mauritania (Alvaro et al. 2007; Shields et al. 2007b) may be equivalent to isotopically similar post-glacial limestones around the world (cf. Hoffman & Schrag 2002). The anomalously low 87Sr/86Sr ratios reported by Nedelec et al. (2007) remain inexplicable; they could relate to diagenetic isotopic exchange with juvenile volcanic minerals or to the fact that contemporaneous seawater 87Sr/86Sr was significantly lower than previously believed. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Affaton, P., Rahaman, M. A., Trompette, R. & Sougy, J. 1991. The Dahomeyide orogen: tectonothermal evolution and relationships with the Volta Basin. In: Dallmeyer, R. D. & Le´corche´, J. P. (eds) The West African Orogens and Circum-Atlantic Correlatives. Springer-Verlag, New York, 107–122. Aı¨t-Kaci Ahmed, A. & Moussine-Pouchkine, A. 1994. Les formation cambriennes de Fersiga (Sud-Ouest du Tanezrouft): nouvelle

170

G. A. SHIELDS-ZHOU ET AL.

interpre´tation de la se´dimentation glaciaire et post-glaciaire sur le Craton Ouest Africain. Bulletin du Service ge´ologique d’Alge´rie, 5, 3 –21. ´ lvaro, J. J., Macouin, M., Bauluz, B., Clausen, S. & Ader, M. A 2007. The Ediacaran sedimentary architecture and carbonate productivity in the Atar Cliffs, Adrar, Mauritania: palaeoenvironments, chemostratigraphy and diagenesis. Precambrian Research, 153, 236– 261. Bertrand-Sarfati, J., Moussine-Pouchkine, A. Q., Amard, B. & Aı¨t Kaci Ahmed, A. 1995. First Ediacaran fauna found in western Africa and evidence for an Early Cambrian glaciation. Geology, 23, 133– 136. Bertrand-Sarfati, J., Flicoteaux, R., Moussine-Pouchkine, A. & Aı¨t Kaci Ahmed, A. 1997. Lower Cambrian apatitic stromatolites and phospharenites related to the glacio-eustatic cratonic rebound (Sahara, Algeria). Journal of Sedimentary Research, 67, 957– 974. Clauer, N. & Deynoux, M. 1987. New information on the probable isotopic age of the Late Proterozoic glaciation in West Africa. Precambrian Research, 37, 89– 94. Clauer, N., Caby, R., Jeanette, D. & Trompette, R. 1982. Geochronology of sedimentary and metasedimentary Precambrian rocks of the West African Craton. Precambrian Research, 18, 53– 71. Condon, D., Zhu, M., Bowring, S., Wang, W., Yang, A. & Jin, Y. 2005. U –Pb ages from the Neoproterozoic Doushantuo Formation, China. Science, 308, 95 –98. Craig, J., Thurow, J., Thusu, B., Whitham, A. & Abuttaruma, Y. 2009. Global Neoproterozoic Petroleum Systems: The Emerging Potential in North Africa. Geological Society, London, Special Publications, 326. Cruse, T. & Harris, L. B. 1994. Ediacaran fossils from the Stirling Range Formation, Western Australia. Precambrian Research, 67, 1 –10. Culver, S. J. & Hunt, D. 1991. Lithostratigraphy of the Precambrian – Cambrian boundary sequence in the southwestern Taoudeni Basin, West Africa. Journal of African Earth Sciences, 13, 407–413. Culver, S. J., Pojeta, J. & Repetski, J. E. 1988. First record of Early Cambrian shelly microfossils from West Africa, Geology, 16, 695– 599. Deynoux, M. 1980. Les formations glaciaires du Pre´cambrien terminal et de la fin de l’Ordovicien en afrique de l’Ouest. Deux exemples de glaciation d’inlandsis sur une plate-forme stable. Travaux du Laboratoire des Sciences de la Terre St-Jerome, Marseille (B), 17, 554. Deynoux, M. 1982. Periglacial polygonal structures and sand wedges in the late Precambrian glacial formations of the Taoudeni Basin in Adrar of Mauritania (West Africa). Palaeogeography, Palaeoclimatology, Palaeoecology, 39, 55– 70. Deynoux, M. 1985. Terrestrial or waterlain glacial diamictites? Three case studies from the Late Precambrian and Late Ordovician glacial drifts in West Africa. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 97– 141. Deynoux, M. & Trompette, R. 1976. Late Precambrian mixtites: glacial and/or nonglacial? A discussion dealing especially with the mixtites of West Africa. American Journal of Science, 276, 1302– 1315. Deynoux, M. & Trompette, R. 1981. Late Precambrian tillite of the Taoudeni Basin, West Africa. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, New York, 123– 134. Deynoux, M., Marchand, J. & Proust, J. N. 1989. Notice explicative de la carte ge´ologique du Mali occidental az 1/200.000. Feuilles Kankossa, Kayes, Kossanto. Re´publique du Mali, Direction Nationale de la Ge´ologie et des Mines, Bamako. Klo¨ckner Industrie-Anlagen, Duisburg, 54– 81. Deynoux, M., Proust, J. N. & Simon, B. 1991. Late Proterozoic glacially controlled shelf sequences in western Mali (West Africa). Journal African Earth Sciences, 12, 181–198. Deynoux, M., Affaton, P., Trompette, R. & Villeneuve, M. 2006. Pan-African tectonic evolution and glacial events registered in Neoproterozoic to Cambrian cratonic and foreland basins of West Africa. Journal of African Earth Sciences, 46, 397– 426. Evans, D. A. D. 2000. Stratigraphic, geochronological, and paleomagnetic constraints upon the Neoproterozoic climatic paradox. American Journal of Science, 300, 347– 433.

Eyles, N. & Januszczak, N. 2004. ‘Zipper-rift’: a tectonic model for Neoproterozoic glaciations during the breakup of Rodinia after 750 Ma. Earth Science Reviews, 65, 1– 73. Fairchild, I. J., Marshall, J. D. & Bertrand-Sarfati, J. 1990. Stratigraphic shifts in carbon isotopes from Proterozoic stromatolitic carbonates (Mauritania): influences of primary mineralogy and diagenesis. American Journal of Science, 290A, 46– 79. Halverson, G. P., Maloof, A. C. & Hoffman, P. 2004. The Marinoan glaciation (Neoproterozoic) in northeast Svalbard. Basin Research, 16, 297– 324. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129–155. Hoffmann, K.-H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. U– Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817– 820, doi:10.1130/G20519.l. Kennedy, M. J. 1996. Stratigraphy, sedimentology, and isotopic geochemistry of Australian Neoproterozoic postglacial cap dolostones: deglaciation, d13C excursions and carbonate precipitation. Journal of Sedimentary Research, 66, 1050– 1064. Lahonde`re, D., Roger, J. et al. 2005. Notice explicative des cartes ge´ologiques a` 1/200,000 et 1/500,000 de l’extre`me sud de la Mauritanie. DMG, Ministe`re des mines et de l’industrie, Nouakchott, Rapport BRGM/RC-54273-FR, 610. Leprun, J.-C. & Trompette, R. 1969. Subdivision du Voltaı¨en du massif de Gobnangou (Re´publique de Haute-Volta) en deux se´ries discordantes se´pare´es par une tillite d’aˆge e´ocambrien probable. Comptes Rendus de l’Acade´mie des Sciences, Paris, 269, 2187– 2190. Moussine-Pouchkine, A. & Bertrand-Sarfati, J. 1997. Tectonosedimentary subdivisions in the Neoproterozoic to Early Cambrian cover of the Taoudenni Basin (Algeria, Mauritania, Mali). Journal African Earth Sciences, 24, 425– 443. Nedelec, A., Affaton, P., France-Lanord, C., Charriere, A. & Alvaro, J. 2007. Sedmentology and chemostratigraphy of the Bwipe Neoproterozoic cap dolostones (Ghana, Volta Basin): a record of microbial activity in a peritidal environment. Comptes Rendus Geoscience, 339, 223– 239. Porter, A. M., Knoll, A. H. & Affaton, P. 2004. Chemostratigraphy of Neoproterozoic cap carbonates from the Volta Basin, West Africa. Precambrian Research, 130, 99 – 112. Proust, J. N. & Deynoux, M. 1994. Marine to non-marine sequence architecture of an intracratonic glacially related basin. Late Proterozoic of the West African platform in western Mali. In: Deynoux, M., Miller, J. M. G., Domack, E. W., Eyles, N., Fairchild, I. J. & Young, G. M. (eds) Earth’s Glacial Record. Cambridge University Press, Cambridge, 121– 145. Rooney, A. D., Selby, D., Houzay, J.-P. & Renne, P. R. 2010. Re –Os geochronology of Mesoproterozoic sediments from the Taoudeni basin, Mauritania: implications for basin-wide correlations, supercontinent reconstruction and Re– Os systematics of organic-rich sediments. Earth and Planetary Science Letters, 289, 486– 496. Rossi, P., Deynoux, M. & Simon, B. 1984. Les formations glaciaires du Pre´cambrien terminal et leur contexte stratigraphique (formations pre´ et post-glaciaires et dole´rites permiennes(?) du massif du Kaarta (dans le bassin de Taoude´ni au Mali occidental (Afrique de l’Ouest). Sciences Ge´ologiques Bulletin, Strasbourg, 37, 91 – 106. Shields, G. A., Deynoux, M., Culver, S. J., Brasier, M. D., Affaton, P. & Vandamme, D. 2007a. Neoproterozoic glaciomarine and cap dolostone facies of the southwestern Taoude´ni Basin (Walidiala Valley, Senegal/Guinea, NW Africa). Comptes Rendus de l’Acade´mie des Sciences: Geosciences, 339, 186– 199. Shields, G. A., Deynoux, M., Strauss, H., Paquet, H. & Nahon, D. 2007b. Barite-bearing cap carbonates of the Taoude´ni Basin, northwest Africa: sedimentary and isotopic evidence for methane seepage from permafrost after a Neoproterozoic glaciation. Precambrian Research, 153, 209–235. Simon, B. 1979. Essai de synthe`se sur les formations se´dimentaires de la partie occidentale du Mali. Rapport Ine´dit Laboratoire Ge´ologie Dynamique Univ. Aix-Marseille III, 133.

NEOPROTEROZOIC GLACIATION IN THE TAOUDE´NI BASIN

Trompette, R. 1973. Le Pre´cambrien supe´rieur et le Pale´ozoique infe´rieur de l’Adrar de Mauritanie (bordure occidentale du bassin de Taoudeni, Afrique de l’Ouest). Un exemple de se´dimentation de craton, Etude stratigraphique et se´dimentologique. Travaux du Laboratoire des Sciences de la Terre St-Jerome, Marseille (B), 7. Veizer, J., Compston, W., Clauer, N. & Schidlowski, M. 1983. 87Sr/86Sr in Late Proterozoic carbonates: evidence for a ‘mantle’ event at 900 Ma ago. Geochimica et Cosmochimica Acta, 47, 295– 302. Villeneuve, M. 1988. Evolution compare´e du bassin de Taoude´ni et de la chaine des Mauritanides en Afrique de l’Ouest. Comptes Rendus Acade´mie des Sciences, Paris, 307, 663– 668.

171

Villeneuve, M. & Corne´e, J. J. 1994. Structure, evolution and palaeogeography of the West African craton and bordering belts during the Neoproterozoic. Precambrian Research, 69, 307– 326. Yoshioka, H., Asahara, Y., Tojo, B. & Kawakami, S. 2003. Systematic variations in C, O, and Sr isotopes and elemental concentrations in Neoproterozoic carbonates in Namibia: implications for glacial to interglacial transition. Precambrian Research, 124, 69– 85. Zimmermann, M. 1960. Nouvelle subdivision des se´ries ante´gothlandiennes de l’Afrique occidentale (Mauritanie, Soudan, Se´ne´gal), Rap. 21st International Geological Congress, Copenhagen, 8, 26– 36.

Chapter 12 Neoproterozoic glaciogenic diamictites of the Katanga Supergroup, Central Africa SHARAD MASTER1* & MAREK WENDORFF2 1

Economic Geology Research Institute, School of Geosciences, University of the Witwatersrand, P. Bag 3, WITS 2050, Johannesburg, South Africa 2

AGH University of Science and Technology, Faculty of Geology, Geophysics and Environmental Protection, al. A. Mickiewicza 30, 30-059 Krakow, Poland *Corresponding author (e-mail: [email protected]) Abstract: Glaciogenic sediments of the Katanga Supergroup are represented by two units. The syn-rift Grand Conglomerat Formation (,765 + 5 Ma to .735 + 5 Ma) occurs within the Nguba Group, and the Petit Conglomerat Formation defines the base of the Kundelungu Group deposited in the earliest foreland basin of the Lufilian orogenic belt located between the Congo and Kalahari cratons. Their glacial origin is inferred on the basis of the following features: the common and widespread occurrence of thick polymictic conglomerates and diamictites with faceted and striated clasts, massive structure, abundant poorly sorted fine-grained matrix, and the presence of planarlaminated shales (laminites) with dropstones. Glaciomarine facies associations prevail over most of the geographic extent of both units, but at the northern periphery of the depository, continental glacial facies are present. The glaciomarine units are succeeded by carbonates: the Kakontwe Limestone and ‘Calcaire Rose’ respectively. The clasts in the glaciogenic units are of extrabasinal and intrabasinal provenance. Lower boundaries, conformable in the basin centre, evolve to unconformities in the marginal areas to the N and S. The palaeomagnetic evidence suggests deposition in low latitudes.

The Neoproterozoic – Lower Palaeozoic Katanga Supergroup of Central Africa (Katanga Province of Democratic Republic of Congo, Zambia, and eastern Angola) is exposed in the Lufilian belt (part of the continental system of the Pan-African orogenic belts of Africa), and also forms a less deformed plateau molasse/foreland sequence over the Congo Craton (Fig. 12.1). Katangan sediments were initially deposited in intracratonic rifts related to early Neoproterozoic extension along the southern margin of the Congo Craton. Continental break-up accompanied by mafic volcanism led to the formation of major unconformities, and development of a passive margin (Nguba Group). Subsequent deformation in late Neoproterozoic to early Palaeozoic orogenic belts (Lufilian-Zambezi), and syntectonic sedimentation in associated foreland basins, was related to the collision of the Congo and Kalahari Cratons during the amalgamation of the Gondwana Supercontinent (Hanson 2003; Wendorff 2005a, b; Master et al. 2005). Regional mapping of Katanga (at a scale of 1:500 000) first commenced with the work of Franz Edward Studt, Jules Cornet and Henri Buttgenbach (Studt et al. 1908; Studt 1908, 1913). The first identification of glacial diamictites in Katanga was made by Stutzer (1911, 1913a) and by Grosse (1912), but both authors misidentified these Katangan diamictites as belonging to the late Carboniferous Dwyka glaciation. Robert (1912a, b) first identified glacial diamictites belonging to the Kundelungu System in the Katanga region on the basis of striated pebbles. Hennig (1915) regarded the glacial beds of Central Africa identified by Stutzer (1911, 1913a) and Grosse (1912) as being much older than the Dwyka. Different facies were identified in the Katangan diamictites by Delhaye (1920). As a consequence of more detailed regional mapping, at a scale of 1:200 000, by van Doorninck (1928) and by Maurice Robert from 1926 to 1931, in which diamicitites were used extensively as easily recognized regional marker horizons, it was finally established by Robert (1933), Gysin (1934) and Grosemans (1935) that there are two diamictites, the Grand Conglomerat and Petit Conglomerat, which were regarded as forming, respectively, the base of the Lower and Upper ‘Series’ (‘Serie Infe´rieur’ and ‘Serie Supe´rieur’), constituting the upper part of the Kundelungu ‘System’. Van Doorninck (1928) and Robert (1940a, b, 1947) argued for a glacial

origin of these diamictites based on a number of characteristics, such as the great lateral continuity and thickness of these beds, and the extrabasinal origin and faceted and striated nature of the clasts. The Grand Conglomerat Formation has also been designated as unit Ki1.1 (Franc¸ois 1973) and as the Mwale Formation, or Ng1.1 (Batumike et al. 2007), while the Petit Conglomerat has been referred to as the Ks1.1 (Franc¸ois 1973), and the Kyandamu Formation or Ku1.1 (Batumike et al. 2007). The two diamictites are developed over a very extensive area, covering at least 65 000 km2, occurring roughly in the region between latitudes 88 and 138S, and longitudes 248 and 298E (Figs 12.2 & 12.3).

Structural and stratigraphic framework The Katanga Supergroup was deposited on a basement comprising the Palaeoproterozoic Lufubu Metamorphic Complex and the Bangweulu Block (Rainaud et al. 2005a; De Waele et al. 2006), and the Mesoproterozoic Kibaran Belt (Kokonyangi et al. 2006). Traditionally, the Katanga Supergroup has been subdivided into the Roan, Lower and Upper Kundelungu Groups (Cailteux et al. 1994; Kampunzu & Cailteux 1999). According to the recently revised stratigraphy (Wendorff 2003, 2005b, c; Wendorff & Key 2009), the Katanga Supergroup is subdivided into the following five groups: Roan (basal group), Nguba, Kundelungu, Fungurume and Plateau (Fig. 12.2). Not all researchers in the region have accepted the revised stratigraphy (e.g. Cailteux et al. 2007; Batumike et al. 2006, 2007). The Roan and Nguba groups record two distinct rifting stages resulting from early Neoproterozoic extension within Rodinia. The Kundelungu, Fungurume and Plateau groups were deposited in the succeeding foreland basins related to collision of the Congo and Kalahari cratons during Gondwana assembly. The Grand Conglomerat Formation occurs within the Nguba Group and records syn-rift glaciation after initial basaltic volcanism. The succeeding Petit Conglomerat Formation formed after the first orogenic event to affect the southern part of the Lufilian belt in what is now Zambia.

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 173– 184. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.12

174

S. MASTER & M. WENDORFF

Fig. 12.1. Position of the Lufilian arc in the Pan-African orogenic belts system in central and southern Africa. Box outlines the area shown in Fig. 12.2.

The Roan Group (Fig. 12.4a), deposited in the first Katangan rift basin (,880 Ma; Armstrong et al. 2005), nonconformably overlies variably eroded pre-Katangan basement and forms a continuous transgressive succession from terrigenous clastic sediments

(siliciclastic unit, or the Mindola Subgroup) at the base, grading upwards into a mixed association of siliciclastic and carbonate strata, the Kitwe Subgroup, succeeded by a carbonate platform sequence, the Bancroft Subgroup, which prograded from the south (Binda 1994). Major uplift (765 Ma) in the southern part of the Roan rift basin (Zambia) terminated deposition of the Roan Group platform carbonates and led to the opening of the Nguba rift (Wendorff 2005b). Wendorff (2005b, c) has interpreted the breccias of the Mufulira Formation (lower Mwashya) as syn-rift olistostromes, derived from uplifted Roan strata as products of mass-wasting and deposited by sediment-gravity flows at the base of the Mwashya Subgroup. Northward expansion of the Nguba rift, beyond the northern margin of the Roan rift, resulted in progradation of the olistostromes and their nonconformable deposition upon pre-Katangan basement in what is now the fold-thrust belt region in the Democratic Republic of Congo (DRC). The succeeding Mwashya strata are composed of terrigenous siliciclastic rocks, silicified oolitic/pisolitic grainstones, algal dolomites and ironstones (middle Mwashya), overlain by shales with siltstones grading upwards to black shales deposited under increasingly anoxic conditions (upper Mwashya). Basaltic volcaniclastic rocks and lavas mostly form subordinate interbeds in the Mwashya Subgroup, but are locally significant. For example, a broad belt of basaltic lava can be traced in a NE direction from eastern Angola across NW Zambia and into southern DRC (Unrug 1987, 1988; Kampunzu et al. 1993, 2000; Tembo et al. 1999; Key et al. 2002). In the centre of the basin, the upper

Fig. 12.2. Regional geology of the Lufilian belt (modified from Porada 1989; Porada & Berhorst 2000; Wendorff 2003). Localities discussed in the text: L, Lwaio Mission; M, Mwinilunga; Ks, Kansanshi; MF, Musonda Falls; Ma, Mansa; M-K, Makonga-Kibambale. 1– 10, positions of stratigraphic logs shown in Fig. 12.5.

NEOPROTEROZOIC GLACIOGENIC DIAMICTITES OF THE KATANGA SUPERGROUP

175

Fig. 12.3. Distribution of the Grand Conglomerat and Petit Conglomerat diamictites in Katanga (Democratic Republic of Congo) and in adjacent areas of Zambia (modified after Cahen & Lepersonne 1979).

Fig. 12.4. (a) Simplified version of revised stratigraphy of the Katangan succession based upon syntectonic conglomerate complexes and unconformity-bounded megasequences (Wendorff 2005b; Wendorff & Key 2009). Ages after Key et al. (2002) and Master et al. (2005), and references therein. Wavy lines indicate major unconformities. (b) Previous/traditional subdivisions of the Katangan and lithostratigraphic correlation between the DRC and Zambia (simplified from Cailteux et al. 1994). Ore horizons: *Cu in Lower Roan in Zambia; **Cu– Co in the Mines Group in the DRC. Glaciogenic horizons: GC, Grand Conglomerat; PC, Petit Conglomerat. Note that the Mines Group in the DRC consists of allochthonous megablocks, which resulted from dismemberment of nappes/thrust sheets with the Roan Group strata and gravity-driven emplacement of the mineralized Roan megablocks into the foreland basin (the nappes originated in the south of the orogen (now Zambia) and were thrust towards the foreland region in the north (today’s DRC) during the Pan-African orogenesis). Therefore the Mines Group does not exist as an individual stratigraphic unit (Wendorff 2003, 2005b), and the megablocks are embedded in the Fungurume Group strata, as shown in (a).

176

S. MASTER & M. WENDORFF

boundary of the Mwashya Subgroup is transitional into the Grand Conglomerat (Binda & Van Eden 1972; Bodiselitsch et al. 2005) and is defined by the appearance of dropstones in massive or laminated black shales typical of the underlying upper division of Mwashya Subgroup. Mafic igneous rocks that occur within the Roan, Mwashya and Grand Conglomerat have been interpreted as indicators of rifting continuing throughout the deposition of the Roan and Nguba groups (Tembo et al. 1999; Kampunzu et al. 2000). The northward propagation of the rifting is attributed to the formation of the Kundelungu Aulacogen, which was filled with up to 7 km of Neoproterozoic sediments, including the two glacial diamictites, the Grand and Petit Conglomerates (Dumont & Hanon 1997; Master 2007). Sedimentary strata that overlie the Grand Conglomerat include the Kakontwe Limestone Formation. Haematitic and jaspilitic banded Fe-formations are known to occur both in the Mwashya Subgroup below the Grand Conglomerat and in Nguba Group strata above it (Robb et al. 2003, 2004). Inversion from an extensional to a compressional tectonic regime occurred in the south of the Lufilian belt after the deposition of the Grand Conglomerat but before the Petit Conglomerat was deposited (Wendorff 2005b). West of the Kafue Anticline in Zambia, the Petit Conglomerat rests unconformably on a folded succession of the Roan-Nguba groups (Wendorff & Key 2009). It defines the base of the Kundelungu Group, which is an infill of the first foreland basin (Wendorff 2005b), itself superposed on the former Kundelungu aulacogen (Master 2007), and is composed of marine sandy shales, shales and dolomites, with thick proximal conglomerate complexes occurring in the south of the Lufilian belt. The Petit Conglomerat Formation is overlain by a pink dolomite (the Calcaire Rose). The succeeding Fungurume Group fills the second foreland basin formed in the northern part of the fold-thrust region of the Lufilian belt in the DRC (Wendorff 2003, 2005b) and contains synorogenic conglomerates and megablocks derived from nappes composed of the older Katangan strata uplifted to the south and thrust northwards (compare Fig. 12.4a, b). The lower boundary is unconformable on folded Kundelungu strata, and the succession evolves from sedimentary olistostromes through transitional shallow marine and continental redbeds to shallow marine sequences of siliciclastic and carbonate strata. The degree of deformation gradually decreases between the northern marginal part of the external fold-thrust belt and the succeeding undeformed Plateau Group (also known as the Biano Group) further to the north. Continental arkoses and shales of the latter unit were deposited in the youngest foreland basin, extending to the north of the Fungurume Group foreland (Figs 12.2 & 12.4). 40Ar/39Ar dating of detrital muscovite from the Plateau Group shows that these sedimentary rocks were deposited after 573 + 5 Ma (Master et al. 2005). The deformation of the Katanga Supergroup to form the Lufilian arc occurred during the collision of the Kalahari and Congo cratons, in the time interval from c. 590 to c. 512 Ma, followed by a protracted period of post-orogenic uplift and cooling extending to c. 483 Ma (Porada & Berhorst 2000; John et al. 2003, 2004; Rainaud et al. 2005b). During the Lufilian Orogeny, up to 150 km of crustal shortening occurred (Porada & Berhorst 2000; Jackson et al. 2003). Thick-skinned thrusting resulted in the formation of basement-cored domes in the Domes Region (Daly et al. 1984) (Fig. 12.2), while thin-skinned thrusting resulted in the external fold-thrust belt (Jackson et al. 2004). Former Roan rift basins suffered tectonic inversion during basin deformation (Selley et al. 2005). In the folded Lufilian fold-thrust belt, the metamorphic overprint ranges from greenschist to amphibolite grade, with local development of talc-kyanite whiteschists (John et al. 2004), and the diamictites are deformed and cleaved, with recrystallization of matrix constituents to biotite-bearing schists (Gysin 1934; Key et al. 2002; Broughton et al. 2002; Master et al. 2005; Rainaud et al. 2005b). However, in the external parts of the

Lufilian arc, and in the northern parts of the Katangan basin, the metamorphic grade is very low, and sedimentary structures are very well preserved (Master et al. 2005).

Glaciogenic deposits and associated strata Grand Conglomerat Formation, Nguba Group The diamictite of the Grand Conglomerat has long been interpreted as glaciogenic (Cahen 1978). The evidence for a glacial origin rests on the common and widespread occurrence (over an area of around 65 000 km2) of thick (up to a maximum of 1200 m) polymictic deposits containing subrounded to subangular faceted clasts (ranging from gravel-sized fragments to boulders .1 m3 in volume) with striations, sometimes in multiple sets (in more than 20 localities); the generally massive, poorly sorted nature of the diamictite with a fine-grained matrix supporting the coarser clasts; and the presence of associated planar-laminated shales with dropstones (van Doorninck 1928; vanden Brande 1936; Cahen 1954, 1963, 1978; Franc¸ois 1973; Binda & van Eden 1972; Dumont & Cahen 1977; Wendorff & Key 2009). Striated clasts have been described by many authors and illustrated by Stutzer (1911), Studt (1913), Grosse (1918), Robert (1946) and Cahen (1963). The clasts in the Grand Conglomerat diamictite are overwhelmingly composed of quartzite and vein quartz, but also include a variety of other lithologies, such as granites, gneisses, mica-schists, porphyries and basic rocks (van Doorninck 1928; Robert 1940a; Master et al. 2005). Franc¸ois (1973), working in the western end of the Lufilian arc, recorded clasts in the Grand Conglomerat made up predominantly (.75%) of quartzites (fine-grained, coarse-grained, rarely micro-conglomeratic) derived from the Kibaran Belt. In the Makonga and Kibambale regions (Fig. 12.3) in the NW Katangan Basin, adjacent to the Kibaran Belt, Dumont & Cahen (1977) recorded clasts of quartzites, biotite granite, phyllite, graphic pegmatite, tourmalinite, quartz, calcic marbles, jasper, carnelian agate, black silicified pisolites, black chert, amygdaloidal spilitic lavas and dolerite. Clasts made of oolitic cherts were recorded by Cornet (1897), Studt (1913), Stutzer (1913b), Guillemain (1913), Grosse (1918) and Dumont & Cahen (1977), especially in the vicinity of Mwashya in the northern part of the Lufilian arc. In the central Lufilian arc, most clasts have rounded edges, but completely rounded clasts are absent (van Doorninck, 1928). In the Luapula beds, clasts in clast-supported cobble and boulder conglomerates are well rounded, but clasts in matrix-supported poorly sorted pebbly sandstones and conglomerates are subangular to subrounded (Abraham 1959). A study of clast shapes in the Grand Conglomerat diamictites from two localities in the Likasi district was made by Museu (1987), who found that the quartzite clasts (c. 70% of the population) show a very poor sorting, but the median clast size diminished from NE towards the SW. More than 80% of the quartzite clasts were rounded or subrounded, and a flattening index showed values typical for torrential environments; hence these diamictites were interpreted by Museu (1987) to have been deposited in fluviatile or glaciofluvial environments. Petrographic studies of the Grand Conglomerat were initiated by Beck and Wagner, as reported by Stutzer (1911). Grosse (1918) and van Doorninck (1928) described the tillite as having a matrix of extremely fine-grained, ash-grey, almost opaque clayey material, with small (up to 0.75 mm), splinters and more rounded fragments of predominantly quartz, with sporadic rare fine-grained fragments of granite, graphic granite, quartz porphyry, feldspars, muscovite, chlorite, quartzite, magnetite, hornblende and other minerals. The petrography of the Grand Conglomerat at Kipushi indicates that it contains numerous clasts derived from a mixed plutonic (granitic and amphibolitic) and metavolcanic (quartz porphyry schist) terrain, such as the Lufubu Metamorphic Complex (Master et al. 2005). Detailed

NEOPROTEROZOIC GLACIOGENIC DIAMICTITES OF THE KATANGA SUPERGROUP

modern petrographic studies of the Grand Conglomerat diamictites have found microfabrics and microstructures in these diamictites that are comparable with those found in modern glaciogenic tills (Delpomdor 2007; Delpomdor et al. 2008).

Facies in the Grand Conglomerat The Grand Conglomerat sediments represent a broad range of continental, shallow marine and deep marine facies controlled by the sedimentary environment, ongoing tectonic evolution of the depository, climatic variations and changes in ice dynamics.

177

Proximal facies of a .400-m-thick gravelly fan-delta succession at Mushishima, west of the Kafue Anticline in Zambia (Figs 12.5 & 12.6), contain two interlayers of the glaciogenic sediments (Wendorff & Key 2009). The fan-delta conglomerates are massive, matrix-supported and form thick amalgamated beds that locally contain dish structures testifying to rapid deposition and early post-depositional fluid escape. Clast-supported conglomerate layers are rare. Together with pebbles, cobbles and solitary boulders of quartzite, arkose and siltstone, there occur clasts of carbonate rocks derived from the Roan Group. A more distal facies association of a proglacial fan delta occurs at Kansanshi to the west (Fig. 12.2), where a c. 200-m-thick

Fig. 12.5. Lithostratigraphic sections discussed in this chapter (from Wendorff & Key 2009, and references therein). Sections 2– 5 are modified from Wendorff (2005b), section 6 from Andersen & Unrug (1984), section 7 after Key et al. (2002), sections 8 and 9 from Franc¸ois & Cailteux (1981), and sections 10 and 11 after Dumont & Cahen (1977). The thicknesses of lithostratigraphic units in section 7 are approximate due to variable dips and poor exposure.

178

S. MASTER & M. WENDORFF

Fig. 12.6. Schematic SW to NE cross-sections through the Katangan Basin during deposition of the Grand Conglomerat, showing the Nguba Rift superposed on the earlier Roan Rift (see Fig. 12.2 for locations). (a) Eastern region – Kafue Anticline and Bangweulu Block (Fig. 12.5), from Mushishima to Mansa-Lueba on the Luapula River. (b) Western region – Zambia and DRC (Fig. 12.5), from Lwaio-Mwinilunga to Kibambale-Makonga.

glaciogenic succession is composed of coarse glacial rain-out (dropstones up to 10 cm across) of Roan and Mwashya derivation and muddy-silty suspension deposits laid down simultaneously with intermittent sediment-gravity flows ranging from gravelly debris flows to sandy turbidites, and associated with subordinate traction currents (Wendorff & Key 2009). Bed thickness usually ranges from a few centimetres to a few tens of centimetres. There are also rare sequences of c. 1-m-thick matrix- to clasts-supported conglomerate beds grading upwards to intervals of pebbly mudstone over 1 m thick, which may represent very thick pebbly turbidites/megaturbidites (Postma et al. 1988). Reworking of bottom sediment by intermittent weak traction currents is expressed by solitary silty current ripplemarks embedded in massive or laminated mudstone beds. Dropstones may occur in any of these sedimentary facies; their origin as ice-rafted melt-out debris is especially obvious when they occur as outsized clasts loaded into very thin beds and finely laminated siltstones and mudstones, the clast diameter often exceeding the thickness of the encompassing strata. The textural and structural sedimentary features of this facies association suggest subaqueous deposition as a proglacial fan delta or apron (Eyles & Eyles 1992; Wendorff & Key 2009). The facies associations and regional facies gradients between Mushishima and Kansanshi suggest the presence of a prominently uplifted southern margin of the Nguba rift in what is now Zambia (Wendorff 2005b). This uplifted margin is thought to have been the source area supplying the basin adjacent to the north with large amounts of detritus produced by glacial erosion of the Roan and Mwashya lithologies (Wendorff & Key 2009). In the Chambishi Basin, in borehole MJZC/9, the 26-m-thick Grand Conglomerat is conformable upon black shales and

turbidites (109-m-thick) of the underlying Mwashya Subgroup. Diamictites of the Grand Conglomerat are interbedded with turbidites, and are interpreted to have formed by sediment-gravity flow processes in a glaciomarine basin (Master et al. 2005). In the Itawa area near Ndola, Zambian Copperbelt, there are great thickness variations of the glaciomarine facies of the Grand Conglomerat (Binda & Van Eden 1972). On an east –west section line the thickness varies from 6 m to a maximum of 67 m, and back to 21 m, over a distance of just 1.5 km. Isopach contours of the Grand Conglomerat, based on its thickness in the 14 boreholes given in fig. 1 of Binda & Van Eden (1972), indicate the existence of a .5-km-long, c. 1-km-wide, trough-like feature trending SSE –NNW, which may have been a submarine canyon. In borehole section IT 28 at Itawa (Fig. 12.5, Section 5) the Grand Conglomerat is c. 150 m thick and contains subangular to subrounded rain-out clasts varying from granules to boulders of siliciclastic and dolomitic sedimentary rocks derived from the Roan-Mwashya succession. Characteristic for this section is a rain-out diamictite composed of grey and black silty and sandy mudstone often modified by remobilization and redeposition within the basin as debris flows (Wendorff & Key 2009). Numerous slump-folds (Binda & Van Eden 1972) and slump-generated debris flows that occur in both dropstone-rich and dropstonedevoid intervals suggest deposition on a palaeoslope unstable because of rapid deposition or syndepositional tectonic movements of the basin floor, or both processes combined (Eyles & Januszczak 2004; Wendorff & Key 2009). The uppermost part of the Grand Conglomerat Fm. at Itawa is a C –U sequence from laminite to thin turbidites representing Ta,e and Tc,e sequences of Bouma intervals. It is interpreted as distal facies

NEOPROTEROZOIC GLACIOGENIC DIAMICTITES OF THE KATANGA SUPERGROUP

deposited in a non-ice contact region of the basin (Wendorff & Key 2009). A 26-m-thick massive diamictite represents the Grand Conglomerat Fm. in borehole Ks 17 at Mokambo, to the east of Itawa. At the NE margin of the depository, in the Luapula River valley (Bangweulu Block region; Fig. 12.5, Section 6), glaciofluvial sandstone and pebbly sandstone with large-scale cross-bedded sets, indicating palaeocurrent flow towards the SW and south, is interbedded within a massive diamictite. In the NW part of the external fold-thrust belt in Zambia between Mwinilunga and Lwaio Mission (Fig. 12.5, Section 7; Fig. 12.6), a sequence of marine conglomerates, laminites and dropstone-bearing siltstones and shales occurs. The clasts represent basement granite and gneiss and fragments derived from the Roan units (Wendorff & Key 2009). At Lufunfu, in the NW part of the Lufilian Belt in the DRC (Fig. 12.5, Section 8; Fig. 12.6), the glaciomarine massive diamictite with a carbonaceous, slightly calcareous mudstone matrix reaches a thickness of 950 m (Franc¸ois 1973). The thickness of the glaciogenic strata decreases southwards, as dominant clast size decreases from pebble to granule and a conglomerate and several sandstone interbeds wedge out. These facies trends indicate that the source of the glaciogenic material supplied to the northern part of the basin was located to the north of the Nguba rift. At Tombolo, some 100 km north of Lufunfu (Fig. 12.5, Section 9; Fig. 12.6), the diamictite reaches a maximum thickness of 1200 m. In the Makonga-Kibambale area located within the Plateau region, which was a graben/aulacogen basin during deposition of the Nguba Group (Unrug 1987; Wendorff 2005b; Master 2007; Fig. 12.3, Fig. 12.5 (Sections 10 and 11), Fig. 12.6), c. 150 km north of the Lufilian fold-thrust belt, the Grand Conglomerat is represented by a massive diamictite with interbeds of coarse-grained, cross-bedded, feldspathic wacke (Dumont & Cahen 1977). Faceted and striated pebbles and cobbles of igneous and metamorphic rocks are derived from the Kibaran basement to the north and clasts of sedimentary rocks originated locally by scouring of the underlying Mwashya Subgroup. The glacial strata are intercalated with and overlain by basaltic pillow lavas and volcanic breccias, and intruded by dolerite sills. The pillow lavas imply a generally subaqueous environment, but it is not certain whether the glacial deposition between Makonga and Kibambale occurred in a marine or continental setting (Dumont & Cahen 1977). In summary, the nature of the sedimentary facies and their trends suggest that the Grand Conglomerat glaciogenic sediments were derived from both the southern and northern margin of the basin. Stratigraphy in the southern region (at the southern margin of the Nguba rift) is characterized by interbeds of the glacial sediments within a coarse-clastic, high-energy fan-delta association and abundance of clasts derived from the Roan and Mwashya units (Fig. 12.6). The presence of the fan-delta facies both below and above the glaciogenic strata, and the deposits of a proglacial fan delta, imply a strong uplift of the southern shoulder of the Nguba rift. This is consistent with the existence of a possible submarine channel and considered here as the reason why the Grand Conglomerat in the southern region has characteristics of a deep marine deposit adjacent to a steep palaeoslope, for example, a considerable proportion of mass-flow phenomena, ranging from coarse clastic debris flows to megaturbidites to small-scale low-density turbidites, as well as abundant slump beds and slump-initiated debris flows composed entirely of fine-grained intrabasinal components. At the northern margin, on the other hand, the continental glaciogenic deposits grade southward towards the open marine basin into glaciomarine melt-out sediments (Fig. 12.6). These interfinger laterally with, and pass upwards into, proximal sandstone interbeds of marginal marine origin. The proximal, postglacial clastic (terrigenous) facies become progressively finer within the marine basin further to the south, and in the distal regions of

179

the basin, grade into carbonates of the Kakontwe Limestone Formation. This pattern suggests that the marginal marine strata originated in a low-topography graded shelf during the synand immediately post-glacial period. On the other hand, the maximum thicknesses of the Grand Conglomerat in the NW region of the Nguba rift at Tombolo suggest subsidence rates three times higher than in the south and several times higher than in the rifted margin adjacent to the north. This situation suggests extremely high sedimentation rates, which must have balanced the high subsidence to maintain a low-gradient palaeotopography of the shelf. Such high sedimentation rates may be explained by extremely effective supply of the clastic material during deglaciation, derived from the Kibaran Belt further to the NW, where considerable topographic relief remained (Cahen & Lepersonne 1979).

Petit Conglomerat Formation, Kundelungu Group The Petit Conglomerat Formation forms the base of the Kundelungu Group, and may overlie the Nguba Group with an erosional unconformity (Wendorff 2003). The Petit Conglomerat diamicitite is, like the Grand Conglomerat, thought to be of glacial origin, based on the abundant and widespread presence of faceted and striated clasts of both intrabasinal and extrabasinal origin (Vanden Brande 1936; Cahen 1978). It is overlain consistently by a carbonate unit, the ‘Calcaire Rose’ or ‘Dolomie Rose de Lusele’, which consists of a 5–10-m-thick finely and regularly bedded pink dolomite (Franc¸ois 1973; Dumont & Cahen 1977). The Petit Conglomerat Formation has also been called the Kayandamu Formation (Ku1.1; Batumike et al. 2007). Cahen (1978) distinguished two facies in the Petit Conglomerat: a southern diamictite facies with small (,2 cm) clasts, and a northern mixed diamictite and conglomerate facies, with large clasts (up to 1 m granite clasts described by Grosemans 1935) of varied compositions. The thickness of the Petit Conglomerat also varies from north to south: it is thickest in the north central part of the Katangan basin (up to 80 m in the Lukafu area; Vanden Brande 1936). In the northern part of the western Lufilian Arc, Franc¸ois (1973) records its thickness varying from 50 to 35 m, while in the southern part of the same area, its thickness varies from 40 to 35 m. In the Kipushi area further southeast, it is only 24 m (Master et al. 2005). The Petit Conglomerat formation thins towards the northwestern margin of the Katangan basin: it is 25 m thick in the Makonga area, adjacent to the Kibaran Belt. On the northeastern margin of the basin, where they onlap against the basement rocks of the Bangweulu Block, the equivalent rocks of the Petit Conglomerat Formation in the Luapula Beds are up to 100 m thick, but here they consist of a series of alternating conglomerates and pebbly sandstones (Abraham 1959). The conglomerates have a coarse, poorly sorted angular to subangular arkosic sandy matrix, with well-rounded cobbles and boulders (up to 40 cm in diameter) made of granites and rhyolite porphyries from the Bangweulu Block basement, as well as quartzites and shales from the Mporokoso Group. The pebbly sandstones are trough crossbedded, having a poorly sorted subangular arkosic matrix, with scattered angular to subangular pebbles, many of which exhibit a triangular form and faceted appearance, similar to glaciogenic clasts described by Von Engelin (1930). Because of their textural and compositional immaturity, and the evidence that the pebbles have undergone little transport, these beds were interpreted by Abraham (1959) as periglacial outwash gravels. Immature, feldspathic, pebbly sandstones, which are interbedded within the conglomerate complex are trough crossbedded with ripple marks and peculiar cracks with polygonal outlines in plan view, resemble patterned ground typical of periglacial regions (Washburn 1969). These were interpreted by Daily & Cooper (1976) to be a relict three-dimensional network of ice veins in a permafrost environment.

180

S. MASTER & M. WENDORFF

In the northern regions, clasts recorded in the Petit Conglomerat consist of quartz, granites, basic rocks, agates, amygdaloidal lavas, rhyolites, quartzites, siliceous oolites, sandstones and shales, including many faceted and striated clasts (Grosemans 1935; Vanden Brande 1936; Batumike et al. 2006). Many of the clasts originated from the adjacent Kibaran Belt to the NW and from the Kibambale volcanic complex to the north (Dumont & Cahen 1977). In the Kapulo area of NE Katanga, just north of Lake Mweru, a distinctive and heterogeneous suite of clast types (maximum diameter, 30 cm) includes quartzite, rhyolite, porphyries, alaskites and rare clasts of gneiss, mica schists, metaconglomerates and pisolitic black cherts (Andre´ 1976; Cahen 1978) derived from the adjacent Bangweulu Block to the east (Master et al. 2005). Recent petrographic and geochemical studies on the Petit Conglomerat and other sedimentary rocks of the Nguba Group (Batumike et al. 2006, 2007) support the north –south facies variations, and a derivation from the Kibaran Belt and Bangweulu Block to the NW and NE of the Katangan basin. Petrographic studies of the Petit Conglomerat from Kipushi Mine were made by Master et al. (2005). Here the Petit Conglomerat consists of a fine-grained biotitic siltstone with a few scattered clasts, averaging about 0.5 mm across, but ranging up to a maximum size of 5 mm. The clasts consist of quartz, carbonate, shale, chert and altered orthoclase. These clasts are mainly of intrabasinal derivation, with some contribution from basement granitoids (orthoclase). The presence in this rock of acritarchs (of planktonic origin), similar to acritarchs previously described from Kundelungu beds, indicates that the rock was deposited in glaciomarine environment (Master et al. 2005). Recent detailed petrographic studies of the Petit Conglomerat diamictites from the Kiaka anticline indicate that they show microfabrics and microstructures comparable to those found in modern glaciogenic tills (Delpomdor 2007; Delpomdor et al. 2008).

Boundary relations with overlying and underlying non-glacial units The lower contact between the Grand Conglomerat and the underlying strata of the Mwashya Subgroup varies from being completely conformable, for example, in the Itawa area (Binda & Van Eden 1972), in the central part of the Chambishi Basin (Bodiselitsch et al. 2005) and in the central Lufilian arc (van Doorninck 1928), to disconformable onto the lower or middle Mwashya Subgroup in the western part of the Chambishi Basin (Wendorff & Key 2009), in the northern Lufilian arc (Robert 1940a), and in the Kibambale area (Dumont & Cahen 1977). In the Luapula Beds, the Grand Conglomerat rests nonconformably on the crystalline granitic basement of the Bangweulu Block, and with an angular unconformity on metavolcanic rocks of the Luapula Porphyries and quartzites of the Mporokoso Group (Thieme 1970, 1971; Andersen & Unrug 1984). The Grand Conglomerat is overlain conformably by carbonate strata of the Kakontwe Limestone Formation or by correlative terrigenous ‘fines’ in the regions proximal to the south and north margins of the Nguba Rift (Fig. 12.5). The Petit Conglomerat Formation rests with a disconformity or low-angle unconformity on upper Nguba Group strata in the DRC (Robert 1940a) and with a pronounced angular unconformity on the strongly folded Nguba succession in Zambia (Wendorff & Key 2009). In the Luapula Beds, it rests with a disconformity or low-angle unconformity on the Nguba Group and also transgressively onlaps onto the crystalline granitic basement, which it overlies nonconformably. The Petit Conglomerat Formation is overlain conformably by carbonate of the Lusele Formation (Dolomie Rose), or by shales where the carbonates are not developed.

Chemostratigraphy Reconnaissance carbon and oxygen isotope studies of Katangan carbonate rocks have been made by Master & Verhagen (unpublished data, 1993), and by Bodiselitsch (2004). Carbon isotopic values (d13C) in the Roan Group recorded by Master & Verhagen (unpublished data, 1993) range between þ2.2 and þ4.2‰ Vienna Pee Dee Belemnite (VPDB), while the values in the Mwashya Subgroup show a negative trend with stratigraphic height, ranging from þ2.8 to – 1.2‰ VPDB. A sample of carbonate from the matrix of the Grand Conglomerat had a d13C value of –4.6‰ VPDB. In the Kakontwe carbonate overlying the Grand Conglomerat, the d13C values show a stratigraphic upward positive trend, increasing from þ2.7 to þ5.5, while in the upper Nguba Group, values are between –0.2 to þ1.8‰ VPDB (Master & Verhagen, unpublished data, 1993). These isotopic values correlate well with the secular trends for the composite Neoproterozoic d13C record both preceding and following the Sturtian glaciation (Halverson et al. 2005). The carbon and oxygen isotopic data of Bodiselitsch (2004) remain unpublished; however, Bodiselitsch et al. (2005) reported that d13C values in the Calcaire Rose and other carbonate rocks overlying the Petit Conglomerat at Kipushi (DRC) were consistently in the range of –2 to –4‰ VPDB. Bodiselitsch et al. (2005) obtained detailed major and trace element geochemical profiles from drill cores that intersect the Grand and Petit Conglomerat diamictites and overlying carbonate rocks. They found significant Ir anomalies at or near the base of the carbonate rocks immediately overlying both the Grand Conglomerat and Petit Conglomerat diamictites at Kipushi, and also above the Grand Conglomerat at Chambishi. They interpreted the Ir anomaly to be due to platinum group elements (PGE)-enriched meteoritic debris (mainly in the form of interplanetary dust particles, or IDPs), which had accumulated on ice caps on a frozen earth during prolonged Cryogenian glaciations (i.e. during the Snowball Earth model of Kirschvink 1992; Hoffman et al. 1998), and which were released into the sediment during melting of the ice. They used the magnitude of the sharp Ir anomaly above the Petit Conglomerat diamictite to estimate a minimum time of 12 Ma for the duration of the ice cap, assuming a meteoritic flux similar to that which prevailed in the later Phanerozoic.

Other characteristics Economic deposits The Katangan glaciogenic diamictites and associated over- and underlying strata contain several economic mineral deposits, some of which are currently being exploited. Within the Mwashya strata below the Grand Conglomerat diamictite, there are two horizons of haematite and jaspilite, some outcrops of which have been mined for iron (Jamotte 1947). A mafic pyroclastic horizon within the Mwashya Subgroup is the host rock for stratabound copper – cobalt mineralization at the Shituru Mine in Katanga (Lefebvre 1974; Cailteux et al. 2007). The Grand Conglomerat diamictite itself is mineralized with disseminated copper sulphides and oxides in a few places in DRC: at Tombolo, NNE of Kolwezi; at the Lonshi mine in the Pedicle of Katanga and at Kamoa, 20 km west of Kolwezi, where a giant high-grade stratiform orebody was discovered in 2009. The periglacial conglomerates associated with the Grand Conglomerat in the Luapula Beds of northern Zambia contain disseminated copper mineralization in several places (Thieme 1970). The Kakontwe carbonate, which overlies the Grand Conglomerat, is mineralized with epigenetic breccia-fill Zn –Pb– Cu –Ge –Ga mineralization at the Kipushi Mine, DRC, as well as in the much smaller Kengere and Lombe deposits, which were exploited in the past (see review in Batumike et al. 2007). There is also a stratiform Cu –Co deposit, Tantara, consisting primarily of secondary copper carbonates and silicates,

NEOPROTEROZOIC GLACIOGENIC DIAMICTITES OF THE KATANGA SUPERGROUP

which is hosted by this stratigraphic unit (Reintjens 1935). In South Katanga, near the Zambian border, there are important iron deposits hosted by the Kakontwe limestone. The giant Kansanshi Cu –Au deposit in northern Zambia is hosted by flat-lying carbonates and schists of the Kakontwe Formation (Broughton et al. 2002). Ba –Fe mineralization occurs in the Kapumba deposit NW of Likasi, DRC, where barite-specularite veins cut the Grand Conglomerat and overlying dolomites and shales of the Nguba Group (Intiomale & Mbuyi 1997); these appear similar to the barite veins cutting cap carbonates in Mauritania (Shields et al. 2007). The Lusele or Calcaire Rose carbonate above the Petit Conglomerat diamictite contains copper mineralization at the Sokoroshi I and II deposits (Reintjens 1935). The Lusele carbonate in the Luapula Beds contains Pb–Zn mineralization (Thieme 1970). The metallic mineralizations associated with the Neoproterozoic glacial intervals in the Katangan and elsewhere have been speculatively linked to global anoxia, and the generation of reduced metalliferous fluids from associated reduced strata (Robb et al. 2003, 2004).

Characteristics of carbonate rocks overlying the diamictites The Kakontwe dolostone overlying the Grand Conglomerat contains several very distinctive features. At its base, where it is locally referred to as the ‘serie tigre´’, it has a laminated appearance, resembling algal lamination, and at Kipushi it also shows roll-up structures identical to those which have been described from the base of the Rasthof carbonate overlying the Chuos diamictite in Namibia (Hoffman et al. 1998). These basal carbonates also commonly show soft-sediment deformation features, such as convolute bedding and isoclinal folds, which have been observed at Kipushi and at Musoshi (Master, unpublished data). These structures resemble similar features seen in cap carbonates overlying Neoproterozoic glacial diamictites in Brazil (Nogueira et al. 2003) and in Mauritania (Master 2004, unpublished data; Shields et al. 2007).

Palaeolatitude and palaeogeography Meert (2003) inferred that at c. 750 Ma, the present-day southern part of the Congo Craton was at palaeolatitudes of between 5 and 108S, based on a palaeopole for the Mbozi Complex in southern Tanzania, for which a SHRIMP U –Pb zircon age of 748 + 6 Ma has been obtained (Mbede et al. 2004). Wingate et al. (2010) obtained palaeomagnetic data on the 800-m-thick 765 + 5 Ma (SHRIMP U –Pb zircon age) Luakela volcanics of the Lwavu Formation, Mwashya Subgroup (Key et al. 2002), based on alternating field (AF) and thermal demagnetization of 65 samples from nine sites. Of three components found, the one component, carried by magnetite and regarded as primary, coincides with the Mbozi palaeopole obtained by Meert et al. (1995). Because the overlying Grand Conglomerat diamictite is bracketed in age between 765 + 5 and 735 + 5 Ma (see below), it was most likely deposited in near-equatorial low latitudes. There is abundant palaeogeographic evidence from around the Congo Craton, in the form of extensive stromatolitic carbonate platforms, that it was situated in tropical palaeolatitudes during much of the Neoproterozoic Era (Verbeek 1970; Cahen 1973b, 1982; Alvarez 1993; Evans 2000). Thus although there are no direct palaeomagnetic studies of the Petit Conglomerat, it was also likely deposited in tropical palaeolatitudes.

Geochronological constraints Key et al. (2002) obtained a SHRIMP U –Pb zircon crystallization age of 765 + 5 Ma for the Luakela lavas from the Lwavu

181

Formation, Mwashya Subgroup. This age provides a maximum age constraint on the overlying diamictites correlated with the Grand Conglomerat Formation. Key et al. (2002) also obtained a SHRIMP U –Pb zircon age of 735 + 5 Ma from a volcanic breccia within strongly deformed strata at Mwinilunga (M in Fig. 12.2, & Fig. 12.5, Section 7), interpreted as being part of the Nguba beds overlying the Grand Conglomerat. Although its stratigraphic position is not entirely certain, the dated volcanic breccia is spatially associated with carbonate rocks that are not found beneath the Grand Conglomerat in the undeformed succession immediately to the west (LM, Lwaio Mission in Fig. 12.2). These carbonate rocks are correlated with the Kakontwe Limestone Formation (Wendorff & Key 2009). It is therefore concluded that the Grand Conglomerat was most likely deposited prior to 735 + 5 Ma and definitely after the 765 + 5 Ma Mwashya volcanism. The youngest detrital zircons from the Grand Conglomerat diamictite at Kipushi Mine, DRC, yielded an imprecise age of 729 + 50 Ma (Master et al. 2005), consistent with the above age constraints. From the available radiometric data, the age of the Petit Conglomerat is not yet well constrained, and is only bracketed between 735 + 5 Ma, the age of volcanics in the West Lunga Formation in the Nguba Group (Key et al. 2002), and c. 620 Ma, the age of uraninites from veins in thrust zones that affect the Katangan stratigraphy to the top of the Kundelungu Group (Cahen 1973a). However, on the basis of regional correlations and their shared tectonic history as Neoproterozoic rift to passive margin sequences on the southern edge of the Congo craton, the Petit Conglomerat and the overlying Calcaire Rose may be equivalent to the Ghaub diamictite and Maieberg Formation in the Otavi Group of the Damara Orogen, Namibia, which are dated at 635.5 + 1.2 Ma (Hoffmann et al. 2004). Palynological work on the Nguba Group, including laminated dropstone-bearing shales of the Grand Conglomerat, was carried out by Binda (1972a, b, 1977). Although several microfossil forms were identified (including three new varieties of the genusform Fibularix and seven new types of sphaeromorph acritarch), these had little biostratigraphic usefulness, except in indicating a Neoproterozoic age. Vavrdova & Utting (1974) described actritarchs from the Luapula beds, which they ascribed to a Lower Palaeozoic age; however, their age attribution is also considered to be unreliable, and of little stratigraphic significance (Binda 1977).

Discussion No striated pavements have ever been observed beneath the two Katangan diamictites thought to be of glacial origin, mainly because the underlying rocks consisted largely of unconsolidated sediments. On the basis of clast composition, and detrital zircon ages, the glaciers appear to have fed into the Katangan Basin from the flanking basement regions to the NW (Kibaran Belt) and to the NE (Bangweulu Block) (Master et al. 2005). Palaeocurrent data from associated periglacial (glaciofluvial) sandstones in the Luapula Beds of north Zambia indicate palaeocurrents trending towards the SW. Distinctive feldspar-phyric porphyritic granitoid clasts within the Grand Conglomerat in Kipushi district are glacial indicators that can be traced to a source in the Luina Dome 150 km away, indicating a WSW pre-tectonic transport direction for these glacial erratics (Master et al. 2005). Possible submarine channels within the glaciomarine facies of the Grand Conglomerat in the Itawa area of Zambia trend NNW. The distribution and thickness variations of facies, and indications of source areas, show that transport of the glaciers was globally generally north to south, with many local fluctuations. The size distribution and nature of clasts in the Petit Conglomerat indicate a north –south transport direction, corresponding to the diminution in size and abundance of extrabasinal clasts (derived from the Kibaran Belt and the Bangweulu Block).

182

S. MASTER & M. WENDORFF

However, a facies gradient and clast composition in the syn-glacial fan-delta facies to the west of the Kafue Anticline in Zambia suggest clast derivation and palaeotransport direction towards the north, away form the uplifted southern shoulder of the Roan-Nguba rift basin (Wendorff & Key 2009).

Conclusions The Grand and Petit Conglomerat diamictites of the Katanga Supergroup in Central Africa, which were deposited in a variety of continental, shallow and deep marine facies, show abundant evidence of a glaciogenic origin. This evidence includes the occurrence of (i) poorly sorted subangular to subrounded clasts of intra- and extrabasinal origin, ranging in size from millimetric grains to boulders with a volume of several cubic metres, sitting in a fine-grained matrix; (ii) faceted and striated clasts from numerous localities; (iii) microfabrics and microstructures diagnostic of processes associated with modern tillites; and (iv) associated sedimentary rocks indicative of glacial or periglacial conditions, including laminated facies with dropstones, and sandstones with ‘patterned ground’ orthogonal crack networks. Their widespread occurrence (over an area of .65 000 km2) and great thickness (up to 1200 m) make the Katangan diamictites some of the most important Neoproterozoic glaciogenic units. Despite their relatively poor geochronological constraints, the Grand and Petit Conglomerat diamictites can be reasonably correlated, respectively, with the Chuos and Ghaub glaciogenic diamictites of the Damara orogenic belt in Namibia, especially considering similarities in terms of C-isotope chemostratigraphy, textures in overlying carbonate units, and their common tectonic positions on the southern rifted margin of the Congo Craton. The presence of the Katangan diamictites within a major metallogenic district, the Central African Copperbelt, has resulted in many drill cores being made available through the diamictites and their over- and underlying carbonate strata. These cores have facilitated the very detailed geochemical studies that resulted in the discovery of Ir anomalies in the carbonate units overlying the diamictites (Bodiselitsch et al. 2005), which has contributed significantly to invigorating the debate on the Snowball Earth model for Neoproterozoic glaciations. We thank P. Binda, R. Key, P. Hoffman and F. Delpomdor for discussions; J. Cailteux, F. Mbuyi and B. Ngoie for guidance in the field; and J. Batumike for reprints. We are indebted to R. Hanson, J. Arthurs and G. Halverson for their comprehensive reviews, which have helped considerably in improving the manuscript. We are grateful to G. Halverson for his invitation to produce this review, and for his patience and encouragement during the editorial process. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Abraham, D. 1959. The stratigraphical and structural relationship of the Kundelungu System, Plateau Series and basement rocks in the MidLuapula valley, Northern Rhodesia. D. Phil. thesis, University of Leeds. Alvarez, Ph. 1993. Un mode`le de lagune d’aˆge Prote´rozoı¨que supe´rieur: le Schisto-calcaire du Congo. Journal of African Earth Sciences, 17, 75 –87. Andersen, L. S. & Unrug, R. 1984. Geodynamic evolution of the Bangweulu Block, northern Zambia. Precambrian Research, 25, 187– 212. Andre´, L. 1976. Etude ae´rophotomorphologiques et pe´trographique du Katangien de la mosaı¨que controˆle´e de Kapulo au Shaba. Me´moire de licence, Univ. Libre de Bruxelles, Belgium. Armstrong, R. A., Master, S. & Robb, L. J. 2005. Geochronology of the Nchanga Granite, and constraints on the maximum age of the Katanga Supergroup, Zambian Copperbelt. Journal of African Earth Sciences, 42, 32 –40.

Batumike, M. J., Kampunzu, A. B. & Cailteux, J. L. H. 2006. Petrology and geochemistry of the Neoproterozoic Nguba and Kundelungu Groups, Katangan Supergroup, southeast Congo: implications for provenance, palaeoweathering and geotectonic setting. Journal of African Earth Sciences, 44, 97 – 115. Batumike, M. J., Cailteux, J. L. H. & Kampunzu, A. B. 2007. Lithostratigraphy, basin development, base metal deposits, and regional correlations of the Neoproterozoic Nguba and Kundelungu rock successions, central African Copperbelt. Gondwana Research, 11, 432– 447. Binda, P. L. 1972a. Preliminary observations on the palynology of the Precambrian Katanga Sequence. Geologie en Mijnbouw, 51, 315– 319. Binda, P. L. 1972b. Microfossils from the Lower Kundelungu (Late Precambrian) of Zambia. 24th International Geological Congress, Montreal, Section 1, Precambrian Geology, 179–186. Binda, P. L. 1977. Microfossils from the Lower Kundelungu (Late Precambrian) of Zambia. Precambrian Research, 4, 285–306. Binda, P. L. 1994. Stratigraphy of Zambian Copperbelt orebodies. Journal of African Earth Sciences, 19, 251– 264. Binda, P. L. & Van Eden, J. G. 1972. Sedimentological evidence for the origin of the Precambrian Great Conglomerate (Kundelungu Tillite), Zambia. Palaeogeography, Palaeoclimatology, Palaeoecology, 12, 151– 168. Bodiselitsch, B. 2004. Geochemical and stable isotope investigations on ‘Snowball Earth’ samples from the Lufilian tectonic arc, D. R. Congo and Zambia, and on the late Eocene sediment samples from Massignano, Italy. PhD thesis (unpublished), University of Vienna, Vienna, Austria. Bodiselitsch, B., Koeberl, C., Master, S. & Reimold, W. U. 2005. Estimating duration and intensity of Neoproterozoic snowball glaciations from Ir anomalies. Science, 308, 239–242. Broughton, D. W., Hitzman, M. W. & Stephens, A. J. 2002. Exploration history and geology of the Kansanshi Cu –(Au) deposit, Zambia. Society of Economic Geologists, Special Publication, 9, 141– 153. Cahen, L. 1954. Ge´ologie du Congo Belge. Vaillant-Carmanne, Lie`ge. Cahen, L. 1963. Glaciations anciennes et de´rive des continents. Annales de la Socie´te´ ge´ologique de Belgique, 86, B19 –B84. Cahen, L. 1973a. L’uraninite de 620 m.a. post-date tout le Katangien, mise au point. Muse´e Royale d’Afrique Centrale, Tervuren (Belgique), De´partement de Ge´ologie et Mine´ralogie, Rapport Annuel, 1972, 35– 38. Cahen, L. 1973b. Corre´lations du certains se´ries du Pre´cambrien supe´rieur du Zaı¨re a` la lumie`re de l’e´tude des stromatolithes et des donne´es ge´ochronologie radiome´trique. Muse´e Royale d’Afrique Centrale, Tervuren (Belgique), De´partement de Ge´ologie et Mine´ralogie, Rapport Annuel, 1972, 38 – 51. Cahen, L. 1978. Les mixtites ante´-cambriennes de l’est du Zaı¨re: mise au point interimaire. Muse´e Royale d’Afrique Centrale, Tervuren (Belgique), De´partement de Ge´ologie et Mine´ralogie, Rapport Annuel, 1977, 33 –64. Cahen, L. 1982. Geochronological correlation of the Late Precambrian sequences on and around the stable zones of Equatorial Africa. Precambrian Research, 18, 73 –86. Cahen, L. & Lepersonne, J. 1979. Upper Proterozoic diamicitites of Shaba (formerly Katanga) and neighbouring regions of Zambia. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 162–166. Cailteux, J., Binda, P. L. et al. 1994. LithostratI˙graphical correlation of the Neoproterozoic Roan Supergroup from Shaba (Zaire) and Zambia, in the central African copper-cobalt metallogenic province. Journal of African Earth Sciences, 19, 265– 278. Cailteux, J. L. H., Kampunzu, A. B. & Lerouge, C. 2007. The Neoproterozoic Mwashya-Kansuki sedimentary rock succession in the central African Copperbelt, its Cu –Co mineralisation, and regional correlations. Gondwana Research, 11, 414–431. Cornet, J. 1897. Observations sur le terrains ancien du Katanga faites au cours de l’expe´dition Bia-Francqui (1891– 1893). Annales de la Socie´te´ Ge´ologique de Belgique, 24, Me´moires, 1896–1897, 25 – 191.

NEOPROTEROZOIC GLACIOGENIC DIAMICTITES OF THE KATANGA SUPERGROUP

Daily, B. & Cooper, M. R. 1976. Clastic wedges and patterned ground in the Late Ordovician–Early Silurian tillites of South Africa. Sedimentology, 23, 271– 227. Daly, M. C., Chakraborty, S. K. et al. 1984. The Lufilian arc and Irumide belt of Zambia: results of a traverse across their intersection. Journal of African Earth Sciences, 4, 311– 318. Delhaye, F. 1920. Les variations de facies du conglome´rat infe´rieur du Syste`me du Kundelungu au Katanga. Annales de la Socie´te´ Ge´ologique de Belgique, Publications relatives au Congo Belge et aux re´gions voisins, 43, 1919–1920, 19– 27. Delpomdor, F. 2007. Etude des de´poˆts diamictitiques du Ne´oprote´rozoı¨que supe´rieure en Re´publique De´mocratique du Congo et au sud-ouest du Rwanda. Unpublished report, Muse´e Royal d’Afrique Centrale, Tervuren, Belgium. Delpomdor, F., Tack, L. & Pre´at, A. 2008. Microstructures in the Neoproterozoic tillites around the Congo River Basin (CRB), Democratic Republic of the Congo (DRC) – comparison with the Karoo tillites from the Dekese borehole in the CRB. Extended Abstract, 22nd Colloqium of African Geology, 2 –6 November 2009, Hammamet, Tunisia. De Waele, B., Lie´geois, J.-P., Nemchin, A. A. & Tembo, F. 2006. Isotopic and geochemical evidence of Proterozoic episodic crustal reworking within the Irumide belt of south-central Africa, the southern metacratonic boundary of an Archaean Bangweulu Craton. Precambrian Research, 148, 225–256. Dumont, P. & Cahen, L. 1977. Les complexes conglomeratiques de la bordure sud-orientale de la Chaine Kibarienne et leurs relations avec les couches Katangiennes de l’Arc Lufilienne, Rapport Annuelle 1977. Musee´ royal de l’Afrique centrale, De´partment de Ge´ologie et Mine´ralogie, Tervuren, Belgique, 111–135. Dumont, P. & Hanon, M. 1997. Le plateau des Kundelungu, pale´ograben ou aulacoge`ne. In: Charlet, J.-M. (ed.) Proceedings of the International Cornet Symposium ‘Strata-bound Copper Deposits and Associated Mineralizations’ (Mons, 5 – 9 September 1994). Acade´mie Royale des Sciences d’Outre-Mer, 51– 69. Evans, D. A. D. 2000. Stratigraphic, geochronological, and paleomagnetic constraints upon the Neoproterozoic climatic paradox. American Journal of Science, 300, 347–433. Eyles, N. & Eyles, C. H. 1992. Glacial depositional systems. In: Walker, R. G. & James, N. P. (eds) Facies Models: Response to Sea Level Change. Geological Association of Canada, St. John’s, 73 – 100. Eyles, N. & Januszczak, N. 2004. ‘Zipper-rift’: a tectonic model for Neoproterozoic glaciations during the breakup of Rodinia after 750 Ma. Earth-Science Reviews, 65, 1– 73. Franc¸ois, A. 1973. L’extremite´ Occidentale de l’Arc Cuprife`re Shabien, Etude Ge´ologique. Ge´camines, Lubumbashi. Franc¸ois, A. & Cailteux, J. 1981. La couverture katangienne entre les socle de Nzilo et de la Kabompo, Re´publique du Zaı¨re, re´gion de Kolwezi. Annales de la Muse´e royal d’Afrique centrale, Sciences ge´ologiques, 87, 50. Grosemans, P. 1935. Contribution a` l’etude du conglome´rat de base (petit conglome´rat) du Kundelungu supe´rieur. Annales de Services des Mines, Comite´ Spe´cial du Katanga, 5, 38 –57. Grosse, E. 1912. Dwykakonglomerat und Karroosystem im Katanga. Zeitschrift der Deutsche Geologischer Gesellschaft, Monatsberichte, 64, 320– 321. Grosse, E. 1918. Grundlinien der Geologie und Petrographie des o¨stlichen Katanga. Neues Jahrbuch fu¨r Mineralogie, Geologie und Pala¨ontologie Beilage, 42, 272–419. Guillemain, C. 1913. Zur Geologie von Katanga. Zeitschrift der Deutsche Geologischer Gesellschaft, Monatsberichte, 65, 304– 328. Gysin, M. 1934. Les tillites me´tamorphiques du Kundelungu de la Haute Lufira (Congo Belge). Compte Rendu se´ances, Socie´te´ de Physique et de l’Histoire naturelle, Gene`ve, 51, 218–221. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Toward a Neoproterozoic composite carbon-isotope record. Geological Society of America Bulletin, 117, 1181– 1207. Hanson, R. E. 2003. Proterozoic geochronology and tectonic evolution of southern Africa. In: Yoshida, M., Windley, B. F. & Dasgupta, S. (eds) Proterozoic East Gondwana: Supercontinent Assembly,

183

Breakup. Geological Society, London, Special Publications, 206, 427– 463. ¨ quatorial- und Su¨dHennig, E. 1915. Die Glazialerscheinerungen in A afrika. Geologische Rundschau, 6, 154–165. Hoffman, P. F., Kaufman, A. J., Halverson, G. P. & Schrag, D. P. 1998. A Neoproterozoic snowball earth. Science, 281, 1342– 1346. Hoffmann, K.-H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. U–Pb zircon dates for the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817– 820. Intiomale, M. M. & Mbuyi, K. 1997. Le gisement Ba –Fe de Kapumba (Shaba, Zaı¨re). Muse´e Royale d’Afrique Centrale, Tervuren (Belgique), De´partement de Ge´ologie et Mine´ralogie, Rapport Annuel 1995–1996, 183– 186. Jackson, M. P. A., Warin, O. N., Woad, G. M. & Hudec, M. R. 2003. Neoproterozoic allochthonous salt tectonics during the Lufilian orogeny in the Katangan Copperbelt, central Africa. Geological Society of America Bulletin, 115, 314– 330. Jamotte, A. 1947. Esquisse ge´ologique de la re´gion de Kasenga. Bulletin de l’Institut Royale Coloniale Belge, 18, 461– 476. John, T., Schenk, V., Haase, K., Scherer, E. & Tembo, F. 2003. Evidence for a Neoproterozoic ocean in south central Africa from MORB-type geochemical signatures and P –T estimates of Zambian eclogites. Geology, 31, 243– 246. John, T., Schenk, V., Mezger, K. & Tembo, F. 2004. Timing and PT evolution of whiteschist metamorphism in the Lufilian arc-Zambezi belt orogen (Zambia): implications for the assembly of Gondwana. Journal of Geology, 112, 71– 90. Kampunzu, A. B. & Cailteux, J. 1999. Tectonic evolution of the Lufilian Arc (Central Africa Copperbelt) during the Neoproterozoic Pan-African orogenesis. Gondwana Research, 2, 401– 421. Kampunzu, A. B., Kanika, M., Kapenda, D. & Tshimanga, K. 1993. Geochemistry and geotectonic setting of late Proterozoic Katangan basic rocks from Kibambale in Central Shaba (Zaire). Geologische Rundschau, 82, 619– 630. Kampunzu, A. B., Tembo, F., Matheis, G., Kapenda, D. & HuntsmanMapila, P. 2000. Geochemistry and tectonic setting of mafic igneous units in the Neoproterozoic Katangan basin, Central Africa: implications for Rodinia breakup. Gondwana Research, 3, 125– 153. Key, R. M., Liyungu, A. K., Njamu, F. M., Somwe, V., Banda, J., Mosley, P. M. & Armstrong, R. A. 2002. The western end of the Lufilian arc in NW Zambia and its potential for copper deposits. Journal of African Earth Sciences, 33, 503–528. Kirschvink, J. L. 1992. Late Proterozoic low-latitude glaciation: the snowball Earth. In: Schopf, J. W. & Klein, C. (eds) The Proterozoic Biosphere. Princeton University Press, New York, 51 – 52. Kokonyangi, J. W., Kampunzu, A. B., Armstrong, R., Yoshida, M., Okudaira, T., Arima, M. & Ngulube, D. A. 2006. The Mesoproterozoic Kibaride belt (Katanga, SE D.R. Congo). Journal of African Earth Sciences, 46, 1– 35. Lefebvre, J. J. 1974. Mineralisations cupro-cobaltiferes associe´es aux horizons pyroclastiques situe´s dans le faisceau supe´rieur de la Serie de Roan, a` Shituru, Shaba, Zaire. In: Bartholome´, P., de Magne´e, I, Evrard, P. & Moreau, J. (eds) Gisements stratiformes et provinces cuprife`res. Socie´te´ Ge´ologique de Belgique, Lie`ge, 103– 122. Master, S. 2007. Neoproterozoic evolution of the Kundelungu Plateau, Katanga Supergroup, Central Africa: from aulacogen to foreland basin. Extended Abstract, IGCP 485 and IGCP 497 Joint Conference, Chouaı¨b Doukkali University, El Jadida, Morocco, 28 November – 5 December 2007, 62 – 65. Master, S., Rainaud, C., Armstrong, R. A., Phillips, D. & Robb, L. J. 2005. Provenance ages of the Neoproterozoic Katanga Supergroup (Central African Copperbelt), with implications for basin evolution. Journal of African Earth Sciences, 42, 41 –60. Mbede, E. I., Kampunzu, A. B. & Armstrong, R. A. 2004. Neoproterozoic inheritance during Cainozoic rifting in the western and southwestern branches of the East African Rift System: evidence from carbonatite and alkaline intrusions. International Commission on Earth Sciences in Africa Conference: The East African Rift, Addis Ababa, Abstracts. Meert, J. 2003. A synopsis of events related to the assembly of eastern Gondwana. Tectonophysics, 362, 1– 40.

184

S. MASTER & M. WENDORFF

Meert, J. G., van der Voo, R. & Ayub, S. 1995. Paleomagnetic investigation of the Neoproterozoic Gagwe lavas and Mbozi complex, Tanzania and the assembly of Gondwana. Precambrian Research, 74, 225– 244. Museu, M. 1987. Conside´rations sur l’origine du Grand Conglome´rat de base du Kundelungu infe´rieur au Shaba (Re´publique du Zaı¨re). Muse´e royale de l’Afrique centrale, Tervuren (Belgique), De´partement de Ge´ologie et Mine´ralogie, Rapport Annuel 1985–1986, 165– 168. Nogueira, A. C. R., Riccomini, C., Sial, A. N., Moura, C. A. V. & Fairchild, T. R. 2003. Soft-sediment deformation at the base of the Neoproterozoic Puga cap carbonate (southwestern Amazon craton, Brazil): confirmation of rapid icehouse to greenhouse transition in snowball Earth. Geology, 31, 613– 616. Porada, H. 1989. Pan-African rifting and orogenesis in southern to equatorial Africa and eastern Brazil. Precambrian Research, 44, 103– 136. Porada, H. & Berhorst, V. 2000. Towards a new understanding of the Neoproterozoic – Early Palaeozoic Lufilian and northern Zambezi Belts in Zambia and the Democratic Republic of Congo. Journal of African Earth Sciences, 30, 727–771. Postma, G., Nemec, W. & Kleinspehn, K. L. 1988. Large floating clasts in turbidites: a mechanism for their emplacement. Sedimentary Geology, 58, 47– 61. Rainaud, C., Master, S., Armstrong, R. A. & Robb, L. J. 2005a. Geochronology and nature of the Palaeoproterozoic basement in the Central African Copperbelt, with regional implications. Journal of African Earth Sciences, 42, 1 –31. Rainaud, C., Master, S., Armstrong, R. A., Phillips, D. & Robb, L. J. 2005b. Monazite U– Pb dating and 40Ar/39Ar thermochronology of metamorphic events in the Central African Copperbelt during the Pan-African Lufilian orogeny. Journal of African Earth Sciences, 42, 183– 199. Reintjens, E. 1935. Les gisements cuprife`res du Katanga et de la Rhode´sie septentrionale. Comite´ Special du Katanga, Annales des Services des Mines, 6, 10 – 19. Robb, L. J., Master, S., Armstrong, R. A., Rainaud, C. & Greyling, L. 2003. Timing of Cu–Co and Pb –Zn mineralisation in the Central African Copperbelt: a link to Neoproterozoic glaciations? Transactions of the Institution of Mining and Metallurgy, Section B, Applied Earth Sciences, 112, B164 –B166. Robb, L., Master, S., Armstrong, R. A., Greyling, L. & Rainaud, C. 2004. Neoproterozoic glaciations and the link to Cu –Co and Pb –Zn mineralization in the Central African Copperbelt. Geoscience Africa 2004 Conference, Abstracts Volume 2, University of the Witwatersrand, Johannesburg, 12 –16 July 2004, 550– 551. Robert, M. 1912a. La stratigraphie du syste`me du Kundelungu au Katanga. Annales de la Socie´te´ Ge´ologique de Belgique, Publications relatives au Congo Belge et aux re´gions voisins, 39, 1911– 1912, fasc. I, 5 – 8. Robert, M. 1912b. Le syste`me du Kundelungu au Katanga. Annales de la Socie´te´ Ge´ologique de Belgique, Publications relatives au Congo Belge et aux re´gions voisins, 40, 1912–1913, 213– 275. Robert, M. 1933. Le syste`me du Kundelungu au Katanga. Bulletin de l’Institut royale colonial belge, 4, 436 –440. Robert, M. 1940a. La glaciation du Kundelungu au Katanga (Congo Belge). Report of the 17th International Geological Congress, Moscow, USSR, 1937, 6, 99 –113. Robert, M. 1940b. Contribution a` la ge´ologie du Katanga. Le Syste`me du Kundelungu et le Syste`me Schisto-dolomitique (1re partie). Me´moire de l’Institut royale colonial belge, Science Naturelle et Me´decine, Coll. in-4o, 6, 108. Robert, M. 1946. Le Congo Physique, troisie`me e´dition, revue et comple´te´e. Vaillant-Carmanne, Lie`ge. Robert, M. 1947. Les traces de glaciation et les pe´riodes climatiques glaciaires au Katanga et en Afrique australe. Bulletin de la Socie´te´ belge de Ge´ologie, Pale´ontologie et Hydrologie, 56, 62 – 76. Selley, D., Broughton, D. et al. 2005. A New Look at the Geology of the Zambian Copperbelt. Society of Economic Geologists, Inc., Tulsa, 100th Anniversary Volume, 965– 1000. Shields, G., Deynoux, M., Strauss, H., Paquet, H. & Nahon, D. 2007. Barite-bearing cap dolostones of the Taoude´ni Basin, northwest Africa: sedimentary and isotopic evidence for methane seepage after a Neoproterozoic glaciation. Precambrian Research, 153, 209–235.

Studt, F. E. 1908. Carte ge´ologique du Katanga (1/500.000e) et notice explicative. Annales du Muse´e de Congo belge, se´r. 2, 1, 5– 16. Studt, F. E. 1913. The geology of Katanga and Northern Rhodesia: an outline of the geology of South Central Africa. Transactions of the Geological Society of South Africa, 16, 44– 106. Studt, F. E., Cornet, J. & Buttgenbach, H. 1908. Carte ge´ologique du Katanga et notes descriptives. Annales Muse´e du Congo, Bruxelles, Ge´ologie, Ge´ophysique, Mine´ralogie & Pale´ontologie, se´rie 2, Katanga, 1, 94. ¨ ber Dwyka Konglomerat im Lande Katanga, BelStutzer, O. 1911. U gisch Kongo. Zeitschrift der Deutsche Geologischer Gesellschaft, Monatsberichte, 63, 626– 629. ¨ ber glaziale Konglomerate im Lande Katanga, BelStutzer, O. 1913a. U gisch Kongo. Zeitschrift der Deutsche Geologischer Gesellschaft, Monatsberichte, 65, 114– 117. ¨ ber den geologischen Aufbau der su¨do¨stlichen Stutzer, O. 1913b. U Katanga. Jahresbericht der Freiberger Geologischer Gesellschaft, VI, 41 – 47. Tembo, F., Kampunzu, A. B. & Porada, H. 1999. Tholeiitic magmatism associated with continental rifting in the Lufilian Fold Belt of Zambia. Journal of African Earth Sciences, 28, 403– 425. Thieme, J. G. 1970. The geology of the Mansa area: Explanation of Degree Sheet 1128, parts of NW Quarter and NE Quarter. Report of the Geological Survey of Zambia, Lusaka, 26, 37. Thieme, J. G. 1971. The geology of the Musonda Falls area: Explanation of Degree Sheet 1028, SE Quarter. Report of the Geological Survey of Zambia, Lusaka, 32, 25. Unrug, R. 1987. Geodynamic evolution of the Lufilian arc and the Kundelungu aulacogen: Angola, Zambia and Zaire. In: Matheis, G. & Schandelmeier, H. (eds) Current Research in African Earth Sciences; 14th Colloquium of African Geology, Extended Abstracts. Balkema, Rotterdam, 117–120. Unrug, R. 1988. Mineralization controls and source of metals in the Lufilian fold belt, Shaba (Zaire), Zambia and Angola. Economic Geology, 83, 1247– 1258. Vanden Brande, P. 1936. Etudes ge´ologiques dans le feuille Lukafu. Comite´ Special du Katanga, Annales des Services des Mines, 6, 51–69. van Doorninck, N. H. 1928. De Lufilische Plooing in den Boven Katanga (Belgischen Congo). G. Naeff, ’s-Gravenhage. Vavrdova, M. & Utting, J. 1974. Lower Paleozoic microfossils from the Luapula beds of the Mansa area. Records of the Geological Survey of Zambia, 12, 81– 89. Verbeek, T. 1970. Ge´ologie et lithologie du Lindien (Pre´cambrien supe´rieur du Nord de la Re´publique De´mocratique du Congo). Annales de la Muse´e Royale d’Afrique Central, se´ries In-88, Sciences ge´ologiques, 66, 311. Von Engelin, O. D. 1930. Type form of facetted and striated glacial pebbles. American Journal of Science, 19, 9– 16. Washburn, A. L. 1969. Patterned ground in the Mesters Vig district, northeast Greenland. Biuletyn Peryglacjalny, Ło´dz´, 18, 259–330. Wendorff, M. 2003. Stratigraphy of the Fungurume Group– evolving foreland basin succession in the Lufilian fold-thrust belt, Neoproterozoic –Lower Palaeozoic, Democratic Republic of Congo. South African Journal of Geology, 106, 17 –34. Wendorff, M. 2005a. Coarse clastic markers of Rodinia breakup and Gondwana assembly in the Lufilian belt, Pan-African orogen of Central Africa. In: Wingate, M. T. D. & Pisarevsky, S. A. (eds) Supercontinents, Earth Evolution Symposium. Geological Society of Australia Abstracts, Perth, 101. Wendorff, M. 2005b. Evolution of Neoproterozoic – Lower Palaeozoic Lufilian arc, Central Africa: A new model based on syntectonic conglomerates. Journal of the Geological Society, London, 162, 5– 8. Wendorff, M. 2005c. Sedimentary genesis and lithostratigraphy of Neoproterozoic megabreccia from Mufulira, Copperbelt of Zambia. Journal of African Earth Sciences, 42, 61– 81. Wendorff, M. & Key, R. M. 2009. The relevance of the sedimentary history of the Grand Conglomerat Formation (Central Africa) to the interpretation of the climate during a major Cryogenian glacial event. Precambrian Research, 172, 127–142. Wingate, M. T., Pisarevsky, S. A. & De Waele, B. 2010. Paleomagnetism of the 765 Ma Luakela volcanics in Northwest Zambia and implications for Neoproterozoic positions of the Congo Craton. American Journal of Science, 310, 1333–1344.

Chapter 13 Neoproterozoic sequences of the West Congo and Lindi/Ubangi Supergroups in the Congo Craton, Central Africa JENNY TAIT1*, FRANCK DELPOMDOR2, ALAIN PRE´AT2, LUC TACK3, GIJS STRAATHOF1 & VALENTIN KANDA NKULA4 1

School of Geosciences, University of Edinburgh, Edinburgh EH9 3JW, UK

2

Department of Earth Sciences and Environmental Sciences, University of Brussels, 1050 Brussels, Belgium 3

Department of Geology and Mineralogy, Royal Museum for Central Africa (RMCA), Tervuren, Belgium

4

De´partement des Sciences de la Terre, Faculte´ des Sciences, Universite´ de Kinshasa (UNIKIN), Democratic Republic of the Congo *Corresponding author (e-mail: [email protected]) Abstract: The focus of this chapter is the West Congo Supergroup in the West Congo Belt (WCB), which extends along the western margin of the Congo Craton from Gabon in the north to northern Angola in the south, and the Lindi/Ubangi Supergroup of the Lindian and Fouroumbala – Bakouma Basins exposed on the northern margin of the craton. In both regions, up to two distinct diamictite horizons have been recognized, the younger of which is often associated with carbonate rocks. Geochronological constraints are generally rather poor, many of the deposits lack modern sedimentological analysis, and the glacial versus non-glacial genesis of the diamictites is a matter of debate in the literature. However, recent studies suggest a periglacial influence of diamictite deposition, particularly for the sequences in the WCB. The stratigraphy of the various basins is described, available geochemical and geochronological information collated, and recent work regarding the periglacial nature of the diamictites discussed. Finally, an updated chronostratigraphic correlation between the basins is presented. However, much more work is required, particularly in the Neoproterozoic basins on the northern margin of the Congo Craton, and more accurate geochronological constraints are required before the Neoproterozoic palaeogeography and depositional environments of the western and northern Congo Craton can be fully understood.

The Congo Craton (Fig. 13.1), which is defined here as the central African landmass that amalgamated at the time of Gondwana assembly (c. 550 Ma; De Waele et al. 2008), is encircled by Neoproterozoic sedimentary basins (Fig. 13.1, see also Hoffman 2011; Master & Wendorff 2011). Delhaye & Sluys (1923– 1924) first identified Precambrian diamictites of assumed glacial origin in the Democratic Republic of Congo (DRC), which Lepersonne (1951) subsequently interpreted in the West Congo Belt (WCB) as representing two distinct diamictite horizons termed the Lower and Upper Diamictite Formations. The stratigraphy of the WCB was first described in detail in Gabon by Hudeley (1966), in the Popular Republic of Congo by Dadet (1969), in the DRC by Cahen (1978) and in Angola by Stanton et al. (1963), and all available data were synthesized by Trompette (1994). The glacial origin of the late Neoproterozoic WCB diamictites was the centre of much debate in the early 1970s (see Kro¨ner & Carreira 1973; Schermerhorn 1974; Cahen & Lepersonne 1981). More recently, these deposits have been the subject of facies and microstructural analysis (Delpomdor 2007a, b, c; Delpomdor et al. 2008) and a periglacial influence suggested. Geochronological and palaeomagnetic data for the West Congo Supergroup and Lindi/Ubangi Supergroup are scant, and ages and palaeolatitudinal constraints are largely derived indirectly by correlation (e.g. Evans 2000). Recent chemostratigraphic studies on the Haut-Shiloango and Schisto-Calcaire Subgroups by Frimmel et al. (2006), Poidevin (2007) and Pre´at et al. (2011) support general lithological correlation between the Upper Diamictite and Lower Diamictite Formations with the glaciogenic Chuos and Ghaub Formations in the Kaoko-Damara belt to the south, and the Petit and Grand Conglome´rat Formations in the ZambeziLufilian belt to the SE.

Structural framework and basin setting The Congo Craton comprises several Archaean nuclei (Fig. 13.1), which welded together during the Eburnean orogeny c. 2.1– 1.8 Ga

(Pinna et al. 1996; De Waele et al. 2006, 2008; Noce et al. 2007; Delor et al. 2008), and subsequently remained stable and a coherent block throughout Late Palaeoproterozoic and Mesoproterozoic times (Tack et al. 2006, 2008, 2009). The Early Neoproterozoic was marked by rifting along the western margin of the Congo Craton related to the break-up of Rodinia and opening of the Adamastor Ocean (Tack et al. 2001), followed by the passive margin-type sedimentation of the West Congolian Group. The western margin of the Congo Craton collided with the active Sa˜o Francisco margin, thus forming the Arac¸uai – West Congo Orogen (AWCO), including the Brasiliano Arac¸uai belt now preserved adjacent to the Sa˜o Francisco Craton in Brazil (Pedrosa-Soares et al. 2008), and the WCB in central Africa. Late Neoproterozoic deformation in the WCB involved thrusting of Palaeoproterozoic basement rocks (the c. 2.1 Ga Kimezian Supergroup) onto the Neoproterozoic sequences of the West Congo Supergroup, while to the east the external foreland basin sequences of the West Congo Supergroup lie unconformably on the Archaean cratonic basement. Opening of the Atlantic in the Cretaceous split the Arac¸uai – WCB into two parts, the Brazilian side of which inherited two-thirds of the AWCO, including all Neoproterozoic ophiolitic slivers, the entire magmatic arc, the suture zone and syn- (c. 585–560 Ma) to post- (c. 530–490 Ma) collisional magmatism (Pedrosa-Soares et al. 2008). The northern margin of the Congo Craton is marked by the Late Neoproterozoic east – west trending Sergipano-Central African Belt, which stretches from northern Brazil, through southern Cameroon and the Central African Republic (CAR). This south-verging belt marks the boundary between the Central African Mobile Zone to the north, and the northern margin of the Congo and Sa˜o-Francisco cratons. This northwesternmost part of the Congo Craton is represented by the Archaean Ntem complex, which is overlain by Proterozoic metasedimentary and volcanic rocks (Lower Dja Series), and has been overthrust from the north by the Pan African Oubanguide Nappe of the Central African Belt. Nappe emplacement in southern Cameroon is constrained by 620– 610 Ma ages obtained from

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 185– 194. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.13

186

J. TAIT ET AL.

20°E

lt

30°E

CUB Ubangui

ian Be

Legend Phanerozoic to Recent cover Neoproterozoic

Kibaran LT Ub en Belt dia nB el t l e Bangweulu t erb p p L o Block C M Katanga elt B de i Lufilian m Iru h e r n elt Belt ut e B o d S mi rI u S.M. Zambezi Belt

Us a

Kasai Block

Angola Block

20°S

Palaeoproterozoic belts Palaeoproterozoic with possible Archaean basement (unexposed) Archaean cratons

Zimbabwe block

Damara & Kaoko Belts

8°E

10°E

12°E

14°E

LIBREVILLE

The West Congo Supergroup is exposed along the western margin of the Congo Craton in the Bas-Congo Basin of the DRC, and the Nyanga Basin of SW Gabon (Fig. 13.2). Up to two Neoproterozoic diamictite units, interpreted as being glaciogenic in origin, have been identified. The stratigraphy of these two sub-basins of the WCB are described below.

10°E 12°E 14°E 16°E 18°E 20°E 2°N

GABON

CONGO

2°S 4°S

2°S

ATLANTIC OCEAN

SW Gabon Basin N YA N BA GA SI -NI N AR

DRC

6°S

ATLANTIC OCEAN

8°S

ANGOLA I Bas-Congo Basin

4°S

500

Fig. 13.1. Geological sketch map of Sub-Saharan Africa (modified after De Waele et al. 2008) and showing the Congo-Uganda Block (CUB) and the Neoproterozoic sedimentary basins in the NE. LA, Lake Albert; LV, Lake Victoria; CKB, Choma Kalomo Belt; MB, Magondi Belt.

Stratigraphy of the West Congolian Group

0°

N

km

0

Limpopo

plutonic rocks (Toteu et al. 2006), while a minimum age of 628 + 12 Ma has been obtained from syn-collisional granites in the Sergipano Belt of Brazil (Oliveira et al. 2006; Bueno et al. 2009). Pin & Poidevin (1987) suggest that the Central African Belt represents a deeply eroded orogenic belt, whereas Abdelsalam et al. (2002) interpret it as the result of continent – continent collision between the northern Congo Craton and the Saharan Metacraton.

0°

Lurio Foreland

MB

CKB

Mesoproterozoic belts

N

lt

10°S

LV

Tanzania craton

zis

Araçuai Belt

Ru

Bambui

MbujiMayi

West West Congo Congo Belt Kimezian Belt

NE Kibaran Belt

Itombwe

CRB

Sao Francisco craton

LA

Ruw. Belt

Lindi

Gabon craton

0°S

40°E

Fouroumbala-Bakouma Basin

Mozambique Belt (East African Orogen)

an Be

lt

l Afric

Be

Serg

ntra ipe-Ce

ran

10°E

ga

0°E

KINSHASA

LEGEND POST KAROO KAROO SUPERGROUP INKISI GROUP WEST CONGOLIAN GROUP MAYUMBIAN GROUP ZADINIAN GROUP KIMEZIAN ARCHAEAN NOQUI-TYPE GRANITES

6°S 100

1

200

300 km

8°S

LUANDA

Fig. 13.2. Geological sketch map of the West Congo Belt (after Frimmel et al. 2006).

THE WEST CONGO AND LINDI/UBANGI SUPERGROUPS

The Bas-Congo Basin, DRC The Neoproterozoic West Congo Supergroup of the DRC is divided into the Zadinian, Mayumbian and West Congolian groups. The most complete Neoproterozoic sequence of the West Congolian

West-Congolian Group SW Gabon

West-Congolian Group Bas-Congo, DRC

187

Group, with two diamictite units, is exposed in the Bas-Congo Basin (Fig. 13.3). Overlying the c. 1000–910 Ma volcanoclastic Zadinian and Mayumbian Groups (Tack et al. 2001; Fig. 13.2), the West Congolian Group (originally described in detail by Cahen 1978) corresponds to passive margin siliciclastic and

Lindi Supergroup Fouroumbala-Bakouma CAR

Lindi/Ubangi Supergroup Lindi, DRC

† 566 Ma

MPIOKA SUBGROUP

SCHISTOGRESEUX GROUP

Dialanga Formation

BANALIA GROUP

Bili Formation

SC4 SC3 SC2 SC1

BAKOUMA SERIES

C5

SCHISTOCALCAIRE GROUP

†† 575 Ma

NIARI TILLITE BOUENZA FORMATION

C4

x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

Bakouma Formation Alolo Formation

BONDO TILLITE FORMATION

SCHISTOCALCAIRE SUBGROUP

ARUWIMI GROUP

Mbiana Formation

Galamboge Formation

C3

KEMBENAKANDO SERIES

C2

Kole Formation Mamungi Formation

LOKOMA GROUP

C1 Bobwamboli Formation

UPPER DIAMICTITE FORMATION

(?) AKWOKWO TILLITE

Asoso Formation

†† 645 Ma (Sh8)

†† 730-750 Ma

BOUGBOULOU SERIES

x

x x

x x

x

x

x

x x

x x

x

x x

x x

x x

x

x x

x

x x

x

x

x x

x x

x

x x

x x

x

x x

x x

x

x x

ITURI GROUP

Penge Formation

x

x x

x x

x x

x

x

LOWER TILLITE FORMATION

† 650 Ma (Sh6)

HAUT SHILOANGO SUBGROUP

x

Lenda Formation

x x

LEGEND DOLOMITE v v v v v

LOWER DIAMICTITE FORMATION

LIMESTONE/ ARGILLACEOUS LIMESTONE v

DIAMICTITE/ BASALT-DOLERITE SILL SHALE

† 923 ± 43 Ma

SANDSTONE-QUARTZITE/ CONGLOMERATE SANSIKWA SUBGROUP

x x x x x x x x x x x x

PALAEOPROTEROZOIC/ARCHAEAN OOLITE, CYANOBACTERIAL MATS, STROMATOLITE

v v v

††† 910 Ma

MAYUMBIAN GROUP

PANAFRICAN TECTONIC UNCONFORMITY † Frimmel et al. (2006) †† Poidevin (2007) ††† Tack et al. (2001)

Fig. 13.3. Stratigraphic logs of Neoproterozoic sequences from the western and northern margins of the Congo Craton. Modified from Delpomdor et al. (2011).

188

J. TAIT ET AL.

carbonate platform deposits and is particularly well developed in the foreland part of the WCB where it is only gently folded and unmetamorphosed. The sedimentary rocks have been subdivided into four subgroups: the Sansikwa, Haut-Shiloango, SchistoCalcaire and the Mpioka Subgroups as described below. The uppermost Inkisi ‘Subgroup’ is no longer considered to be part of the Neoproterozoic sequences (Frimmel et al. 2006 and references therein) and considered post-Pan African in age (see discussion further below). The Sansikwa Subgroup forms the base of the West Congolian Group and comprises siliciclastic rocks representing a continental rift depositional environment. The Lower Diamictite Formation marks the top of the Sansikwa Subgroup, and contains up to 400 m of diamictite with interbedded sands and shales (Cahen & Lepersonne 1981). Tholeiitic basalts, sometimes including pillows and hyaloclastic breccia, are interlayered in the diamictite (De Paepe et al. 1975; Kampunzu et al. 1991) with accompanying feeder dykes and sills intruded in the underlying Sansikwa Subgroup. Their emplacement age has yet to be confirmed, as the 40Ar/39Ar resetting age of 566 + 42 Ma is thought to represent regional greenschist facies metamorphism during Pan African orogeny and amalgamation of western Gondwana (Frimmel et al. 2006). The Sansikwa Subgroup is overlain by 700–800-m-thick, predominantly siliciclastic rocks of the Haut-Shiloango Subgroup. Conglomerates, quartzarenites, argillites and shales make up the lower two-thirds of the Haut-Shiloango Subgroup. The Sekelolo Limestone forms the upper part and consists of dominantly subtidal, medium- to fine-bedded limestones with a minor clastic component. The overlying Upper Diamictite Formation, which is recognized throughout the length of the WCB, from southern Gabon (the Niari Formation) through to northern Angola, is up to 200 m thick in the Bas-Congo Basin (Cahen & Lepersonne 1981). It is overlain by a sequence of up to 12-m-thick, finely laminated pink dolomitic carbonates, forming the lowermost part of the Schisto-Calcaire Subgroup, which share many lithological characteristics with basal Ediacarian carbonates elsewhere. The remainder of the Schisto-Calcaire Subgroup consists of a range of dominantly shallower-water carbonate shelf facies, including oolites, diverse stromatolites, evaporites and cherts (Alvarez 1995). The Schisto-Calcaire Subgroup shows a transition from east to west between dolomitic intertidal facies with stromatolitic biostromes to supratidal facies with evaporitic needles, ooids and cross-bedding (Delpomdor 2007c; Pre´at et al. 2010, 2011). These two facies are separated by a stromatolitic reef barrier (Trompette 1994). The Schisto-Calcaire Subgroup is overlain by the Mpioka Subgroup, c. 1000-m-thick siliciclastic succession with conglomerates, quartzoarenites and argillites, interpreted as being a late-orogenic molasse deposit. The Mpioka Subgroup is recognized throughout the WCB, and the uppermost subgroup has been locally affected by Pan African Late Neoproterozoic/Early Cambrian deformation. In this regard, it is analogous to the Mulden Group along the southwestern margin of the Congo Craton.

The SW Gabon Basin (or the Nyanga-Niari Basin) The West Congo Supergroup of Gabon is exposed in the Nyanga synclinorium (Ge´rard 1958; Dadet 1969), and is subdivided into several informal units. The oldest sedimentary rocks comprise the Bouenza Formation (Fig. 13.3), a fluvial sequence up to 100 m thick (Prian 2008), which records erosion of the Palaeoproterozoic and Mesoarchaean basement. The sandstones are massive with clear cross-bedding and ripple cross-lamination, suggesting palaeocurrents from SW to NE. These are overlain by the fluvio-glacial deposits of the Niari Group (or Niari Tillite, Fig. 13.3). The Bouenza Formation and the Niari Group record

lacustrine-fluvio-glacial infilling of extensional zones in the Central African basement (Alvarez 1995). On the east flank of the basin, the Niari Group rests on the Bouenza Formation with a non-angular unconformity. The Niari diamictites are overlain by the Schisto-Calcaire Group (or Carbonate Subgroup; Prian 2008) which is correlated with the Schisto-Calcaire Subgroup of the Bas-Congo Basin (DRC) and thought to be Late Neoproterozoic in age. Again, precise ages from the Gabon sequences are lacking, and the most reliable ages and correlations are summarized in Figure 13.3. The SchistoCalcaire Group of the SW Gabon Basin is predominantly a carbonate sequence composed of four formations consisting of calcareous to dolomitic shales with an uppermost sandy shale-siltstone unit with interbedded limestones. Carbonate deposition of the upper part was under similar sedimentological environments to those of the Bas-Congo (Pre´at et al. 2010), indicating hypersaline shallow water sub-supratidal conditions, in possible lagoonal setting with development of lithoherms and cyanobacterial mats. In SW Gabon, the Schisto-Calcaire Group is overlain by the Schisto-Gre´seux Group (Chevalier et al. 2002), which is correlated with the Mpioka Subgroup of the Bas-Congo Basin (Tack et al. 2001; Frimmel et al. 2006). The series is thinner in Gabon (up to 100 m thick) and of Late Neoproterozoic age (Thie´blemont et al. 2009). The Schisto-Gre´seux Group is divided into two subgroups: the lower subgroup is predominantly composed of siltstones and argillites with, in places, a 6– 15-m-thick basal conglomerate, and the upper subgroup contains feldspathic sandstones and argillites.

Stratigraphy of the Lindi/Ubangi Supergroup Neoproterozoic sedimentary rocks exposed on the northern Congo Craton are recognized in the Bangui Basin of southwestern CAR, the Fouroumbala-Bakouma Basin of south central CAR and the Lindian Basin of northeastern DRC (Fig. 13.1). Trompette (1994) considered all these sub-basins to be part of the Lindian Basin. Correlation is problematic due to the limited number of studies that have been carried out, poor or non-existent age constraints, poor outcrop conditions and problems concerning accessibility. Diamictite units have been recognized in all three basins (Fig. 13.4). The best studied is the Lindian Basin of northeastern DRC, which was described in detail by Verbeek (1970), and the stratigraphy of which is summarized below.

The Lindian Basin The Lindian Basin is an intracratonic basin that developed on the northern Congo Craton in the Neoproterozoic. The rocks of the Lindi Supergroup, which are up to 2500 m thick, rest unconformably upon crystalline Archaean and/or Early Proterozoic basement rocks and have undergone little or no metamorphism. To the west of the Lindian Basin, the continuation of this sequence in the DRC has been termed the Ubangi Supergroup. As summarized by Verbeek (1970) and Trompette (1994), the Lindi Supergroup is divided into three groups: the basal Ituri, the Lokoma and the Aruwimi groups, the last of which is subdivided into the Galamboge, Alolo and Banalia formations. However, in Figure 13.3 the Aruwimi Group as originally defined (Verbeek 1970) is here split into two units, with the uppermost Banalia Formation ranked as a separate, younger ‘Group’ (see below). Only one diamictite horizon, occurring at the base of the Lokoma Group and termed the Akwokwo Tillite, is interpreted as being glacial in origin (Fig. 13.3). The basal clast-supported conglomerate with coarse-grained arkoses and quartzarenites (30 –50-m-thick Penge Formation, Ituri Group) is overlain by thick oolitic and stromatolitic limestone with dolomite layers (the Lenda Formation). Poidevin (2007)

THE WEST CONGO AND LINDI/UBANGI SUPERGROUPS

West-Congo SW Gabon

C R Y O G E N I A N

Panafrican Tectonic

Lindi, DRC

BANALIA GROUP

†566 Ma

SCHISTOGRESEUX GROUP

MPIOKA SUBGROUP

Post-Marinoan

600

Lindi Ubangi/Lindi Fouroumbala- Bangui, CAR Bakouma CAR

INKISI SUBGROUP

POST-PRECAMBRIAN E D I A C A R I A N

West-Congo Bas-Congo, DRC

189

Ma

Marinoan Glaciation

SCHISTOCALCAIRE GROUP NIARI TILLITE

700 Sturtian Glaciation

BOUENZA FORMATION

BAKOUMA SUBGROUP

††575 Ma

SCHISTOCALCAIRE SUBGROUP UPPER DIAMICTITE ††645 Ma

Post-Sturtian

ARUWIMI GROUP

BANGUI

LOKOMA GROUP

BIMBO BONDO TILLITE KEMBE NAKANDO SUBGROUP

CONGLO.

ITURI GROUP

†650 Ma

HAUTSHILOANGO SUBGROUP

BOUGBOULOU SUBGROUP

LOWER DIAMICTITE

LOWER TILLITE

AKWOKWO TILLITE

††730-750 Ma

?

†923 ± 43 Ma

800 T O N I A N

LEGEND Pre-Sturtian

Diamictites/Mixtites/Tillites SANSIKWA SUBGROUP

Sandstones/Quartzites Cap carbonates/Dolomites

900 †††910 Ma

Limestones/Dolomites Shales/Argillites/Pelites † Frimmel et al. (2006) †† Poidevin (2007) ††† Tack et al. (2001)

1000

reported the presence of early Neoproterozoic microfaunas in these horizons. The top of the Ituri Group is marked by c. 50 m of micaceous shales, carbonates and sandstones of the Asoso Formation. The entire Ituri Group represents lagoonal or shallow marine platform type sedimentary rocks (Daly et al. 1992) and is restricted to the eastern part of the Lindian Basin. It has been tentatively correlated with both the Haut-Shiloango (Trompette 1994; Daly et al. 1992) and the Sansikwa Subgroup (Poidevin 2007) of the West Congolian Group. The Lokoma Group (.500 m thick) overlies the Ituri Group with an angular unconformity, and begins with the Akwokwo Tillite, which has only limited exposure in the eastern parts of the basin where it sits in erosional contact on the underlying Ituri Group. It is up to 40 m thick, with diverse and poorly sorted clasts (Verbeek 1970; Cahen 1978) in a grey-green clay matrix. The rest of the Lokoma Group consists of arkoses, conglomerates and argillites. In the original definition of Verbeek (1970), the Aruwimi Group is up to 1500 m thick and subdivided into three formations. The Galamboge Formation (c. 100 m thick) consists of cross-bedded

Fig. 13.4. Stratigraphic correlation of Neoproterozoic sequences of the western and northern Congo Craton. Correlations based on Sr isotopic data, radiometric age constraints and revised lithological relationships. Modified after Poidevin (2007).

medium-grained quartzoarenites interbebbed with argillites. The quartzoarenites are marked by a rapid transition to the Alolo Formation, and comprise 400-m-thick clayey to calcareous argillites, often stratified, interbedded with dark carbonaceous argillite layers, fine-grained sandstones, zoned pink and oolitic limestones and dolomitic lenses. The uppermost Banalia sequences are composed of .1000 m of massive or finely stratified and frequently cross-bedded fine-grained reddish arkoses interbedded with reddish micaceous and clayey argillites. The contact between the Banalia and the Alolo Formations is apparently conformable. In terms of thickness and depositional facies, the deltaic Banalia Redbeds have been correlated with the similar Inkisi ‘Subgroup’ of the Bas-Congo Basin and interpreted as evidence of postGondwana amalgamation (550 Ma) and pre-Karoo break-up in central Africa (Tack et al. 2008, 2009a). For this reason, they are ranked here, in both Figures 13.3 and 13.4, as (new) separate groups. On the other hand, new geophysical data in and around the DRC Cuvette Centrale suggest that not only the Banalia Redbeds but the complete Aruwimi Group (thus including also the lowermost Galamboge and overlying Alolo Formations) are

190

J. TAIT ET AL.

post-Pan African in age (Kadima et al. 2011). The precise age of the Galamboge and Alolo Formations (i.e. upper Neoproterozoic versus Phanerozoic) clearly needs to be determined using appropriate radiometric dating.

The Fouroumbala-Bakouma Basin The Fouroumbala-Bakouma Basin is located in the south central region of the CAR (Fig. 13.1). Three main sedimentary groups have been recognized (Fig. 13.3): the Bougboulou, Kembe´Nakando and Bakouma Subgroups (or Series). The stratigraphy of the Fouroumbala-Bakouma Basin was described in detail by Bigotte & Bonifas (1968), Poidevin et al. (1981) and Poidevin (1985) with recognition of two diamictite horizons. The base of the Bougboulou Subgroup rests with an angular unconformity on the Archaean Bangui-Kette´ basement. A basal diamictite, termed the Lower Tillite, has been identified in the extreme SW of the basin. The Bougboulou Subgroup comprises predominantly interbedded shales and sandstones, although in the south of the Fouroumbala Basin, silicified stromatolitic carbonates (the Kassa Formation) have been identified (Mestraud 1952, 1953). The Bougboulou Subgroup is unconformably overlain by the deltaic to neritic Kembe´-Nakando Subgroup comprising sandstones and black and red shales of the Mbiana Formation. The base of the overlying Bakouma Subgroup is marked by the Bondo Tillite, which is thought to be glacial in origin, containing varves and dropstones. The Bondo Tillite is overlain by the red and grey Bakouma Formation dolomites, which are similar to other postglacial (cap) carbonates of the Congo Craton (Poidevin 2007). This association, however, is only known from drill cores with no surficial exposure. The Bakouma Formation is overlain by the Bili and Dialanga Formations (Fig. 13.3). The Bili Formation comprises basal shales overlain by white and grey limestones. The clastic Dialanga Formation is the youngest member of the Bakouma Subgroup and comprises black shales and quartzoarenites. Poidevin (2007) correlated the Bakouma Group with the Aruwimi Group (original definition of Verbeek, 1970) of the Lindi Supergroup. If this is correct, the upper tillites (Bondo Tillites) are missing in the Aruwimi Group of the Lindi Supergroup.

The Bangui Basin The Bangui Basin was described lithostratigraphically by Babet (1935), Legoux & Hourcq (1943), Gerard & Gerard (1952), Bessoles & Trompette (1980) and Poidevin (1976, 1979a, 1979b, 1985, 2007). Neoproterozoic sedimentary rocks are exposed in SW CAR around the capital Bangui, and extend southwards to the Congo River and overlie the Nola Group quartzopelitic metasedimentary rocks, which are thought to be Palaeoproterozoic in age. The Nola Group sequences have been overthrust in the north by the Pan-African Oubanguides belt (i.e. Central African Belt), and intruded by the Nola dolerite dykes. These dykes are undated but yield a pan-African resetting Ar/Ar age of 571 Ma (Moloto-A-Kenguemba et al. 2008). Correlation with 950 Ma dolerite dykes in Gabon has been suggested (Vicat et al. 1996) but remains to be demonstrated. The Neoproterozoic sedimentary sequences of the Bangui Basin are subdivided into the Kembe conglomerates, the Bimbo sandstones and the Bangui carbonates (Fig. 13.4). The basal Kembe conglomerates (up to 200 m thick and exposed near the village of Bimbo along the Ubangui River), are tectonized and comprise fine-grained polymict conglomerates interbedded with mylonitized quartzitic sandstones and schistose shales. The overlying Bimbo sandstones (130 m thick) are a thick fluviatile sequence of massive dark shales interstratified with thick beds of conglomerate, sandstones and quartzarenites with crossstratification, which, according to Alvarez (1995) pass eastwards

into the fluvio-lacustrine Kembe´-Nakando sandstones of the Fouroumbala-Bakouma Basin in eastern CAR, and may be glacial in origin (Alvarez 2000). The Bimbo sandstones are unconformably overlain by the Bangui Series (Fig. 13.4), a thick carbonate sequence exposed mainly along the Ubangui River to the south of Bangui. The Bangui Series carbonates have been affected by Pan-African related deformation and, therefore, are considered to be Late Neoproterozoic in age. They have been correlated with the SchistoCalcaire Subgroup of the West Congolian Supergroup (Alvarez 1995). They have been subdivided into the distal Lesse´ Formation, the proximal outer shelf Bobassa Formation, the offshore barrier facies Mboma Formation and the lagoonal Fatima Formation. These carbonates comprise laminated microbial limestones, rhythmically bedded carbonates with clay and silt-rich interval interpreted as shallow subtidal or intertidal deposits, more massively bedded limestones, and are capped by grey microbial limestones intercalated with black dolomites indicating deposition in a lagoonal environment (Alvarez 1992). Alvarez (1995) interpreted these sedimentary rocks as characterizing an extensive carbonate ramp that developed along the northern edge of the Central Africa belt. However, it is now thought that they developed in local intracratonic basins and they have been correlated with the Bakouma Subgroup in more easterly regions of CAR.

Glaciogenic character of the WCB diamictites Until 1970, the diamictites of the West Congolian Group were generally considered to be glacial in origin, and in 1986 the Upper Diamictite of the Congo was described as a marine tillite associated with an ice cap on the Chaillu massif (Trompette & Boudzoumou 1988). In the Congo, however, Cahen & Lepersonne (1981) considered the Lower Diamictite Formation to be a tectonically related gravity flow deposit. Subsequently, Trompette (1994) considered the Upper Diamictite Formation to be a gravity-flow deposit originating from reworking of glacial material along the margins of the basin by small mountain glaciers developed on peripheral highlands. In Angola, the discussion was more acute. Both diamictites were interpreted as being tilloids, that is, non-glacial and deposited as debris flows (Schermerhorn & Stanton 1963), although the possibility of reworking of mountain glacier deposits (i.e. Alpine glacial deposits) was suggested for the lower diamictites (Schermerhorn 1974). Kro¨ner & Correia (1973) reinterpreted both diamictite levels in Angola as being true glaciomarine diamictites with local intercalations of continental tillites. Schermerhorn (1981) continued to ascribe the deposits to a deep-water submarine turbiditic setting, arguing against a glacial origin. Similarly, Vellutini & Vicat (1983) argued against any glacial influence and considered both diamictites of the West Congo Belt as basal conglomerates. In his synthesis of available data and taking all the arguments into account, Trompette (1994) surmises that there was most likely some glacial input, even if they do not represent true glaciomarine deposits. More recently, Tack et al. (2006) critically reappraised the origin of Neoproterozoic diamictites from central Afria. As a result, detailed micro-structural and -textural analyses (based on techniques applied to recent unconsolidated tills by several authors such as Evans & Benn 2004; Benn & Evans 1996; Menzies 2000a, b, 2006; Phillips 2006; Phillips et al. 2007; Van der Meer 1993, 1997, 2003) were carried out on samples of the Lower and Upper Diamictite of the Bas-Congo Basin, the Akwokwo Tillite of the Lindian Basin, and the Niari Formation of the Nyanga Basin (Delpomdor 2007c, Delpomdor et al. 2008, Pre´at 2008, Pre´at et al. 2008). These observations provide important new constraints on the various sedimentological and depositional environments. The diamictites from the SW Gabon, Bas-Congo and the Lindian Basin are principally composed of massive and stratified mud-supported matrix with rare intercalation of varved mudstones

THE WEST CONGO AND LINDI/UBANGI SUPERGROUPS

and sandstones. Clasts within the diamictites are only rarely striated, do not display crescent or conchoidal fractures, and no glacial pavements have been observed. Thus, a directly subglacial environment cannot be identified. The combination of observed microstructures demonstrates that all these diamictites were deposited under high strain rates with moderate to high stress conditions, influenced by water pressure or clays within the beds. The cumulative evidence of overprinting microstructures suggests that all these diamictites were deposited in fluvioglacial or glaciomarine conditions in proximal subglacial and/or more distal subaqueous environments.

Chemostratigraphy The only chemostratigraphic data available for the WCB are from a handful of recent studies (Frimmel et al. 2006; Poidevin 2007; Straathof et al. 2008; Pre´at et al. 2010, 2011). Strontium isotope ratios measured in carbonate rocks of the West Congo Supergroup range from 0.7066 to 0.7109 (Poidevin et al. 2007), although the highest ratios are surely overprints. The least-altered and most Sr-rich samples in the Haut Shiloango Subgroup have 87Sr/86Sr ratios of c. 0.7068 –0.7072, compared to ratios of 0.7074– 0.7075 in the lower Schisto-Calcaire Subgroup (Frimmel et al. 2006; Poidevin 2007). Ten d13Ccarbonate values in the Haut Shiloango Subgroup range from 3.2 to 8‰. The Schisto-Calcaire Subgroup defines a larger range in d13Ccarbonate, from – 5.4 to 9‰. Values in the lowermost part of the subgroup are consistently negative (Frimmel et al. 2006; Straathof et al. 2008; Pre´at et al. 2011), whereas those in the upper part of the subgroup are erratic, probably due to temporarily elevated evaporation rates (Frimmel 2009). The near-shore depositional environment for the Bas-Congo carbonates is also demonstrated geochemically by Frimmel (2009) who concluded that carbon isotopes obtained from successions developed in such environments, particularly when elevated evaporation has occurred, should be treated with great care and are not suitable for chemostratigraphic correlation.

Palaeolatitude and palaeogeography The Precambrian palaeomagnetic database for the Congo Craton is extremely sparse, with very few of the available data meeting modern reliability criteria. Nevertheless, the latest Mesoproterozoic position of the Congo can be constrained by poles obtained from the Sa˜o Francisco Craton (D’Agrella-Filho et al. 1990). These data place the western margin of the craton (present day co-ordinates) at intermediate palaeolatitudes (45 –508) for the time period 1.1 –1.0 Ga (De Waele et al. 2008). Neoproterozoic palaeogeography is constrained by two poles from the 795 + 7 Ma Gagwe lavas of Tanzania (Piper 1972; Meert et al. 1995, Ar/Ar age after Deblond et al. 2001) and the 743 + 30 Ma Mbozi intrusive rocks (Meert et al. 1995, K –Ar biotite age after Brock 1968). According to these data, by the early Cryogenian the Congo craton had moved into tropical latitudes, positioning the present-day western margin close to the equator, and by 755 Ma, had rotated some 908 while remaining in the tropics. Collisional deformation related to consolidation of western Gondwana was initiated c. 600 Ma in the western Congo and 580 –530 Ma in the south and southeast. Palaeomagnetic data from the Sinyai dolerites of Kenya (Meert & Van der Voo 1996) and geological evidence indicate that proto-Gondwana had assembled by the Early Cambrian. The implications of the available palaeomagnetic data are that the Neoproterozoic sedimentary rocks of the West Congo and Lindi Supergroups were deposited in tropical latitudes. In terms of more detailed palaeogeography, it is important to point out that the paucity of Precambrian data for the Congo Craton means the polarity of the palaeopoles is uncertain.

191

Geochronological constraints Radiometric data from the West Congo and Lindi/Ubangi Supergroups are sparse and geochronological constraints are based largely on bulk detrital zircon analyses and correlation with bettedated successions elsewhere on the Congo Craton. The Lower Diamictite Unit is constrained only to be younger than c. 910 Ma based on the youngest U –Pb ages from the Mayumbian Group volcanic rocks, which underlie the Sansikwa Subgroup (Tack et al. 2001). The youngest detrital zircon ages obtained from the overlying Haut-Shiloango Subgroup yield ages of 547 + 45, 671 + 20 and 709 + 20 Ma (1s errors, 10% discordance) from which, given the errors, a possible age of 650 Ma is interpreted by Frimmel et al. (2006). Clearly, more reliable age data are required to more accurately constrain maximum and minimum depositional ages for both the Lower and Upper Diamictite Formations. The 40Ar/39Ar age of 566 + 42 Ma obtained from a sill intruding the Lower Diamictite Unit and dating greenschist metamorphism during west Gondwana amalgamation (Frimmel et al. 2006) provides a loose minimum age constraint for the entire sedimentary package.

Correlation between the West Congolian Group and the Lindi Supergroup Updated lithostratigraphic correlations are summarized in Figure 13.4 and are based on Sr isotope ratios obtained from carbonates, SHRIMP data (Tack et al. 2001; Frimmel et al. 2006; Poidevin 2007) and revised lithological relationships. The Neoproterozoic West Congolian Group sedimentary rocks are relatively well constrained in comparison to those of the Lindian Supergroup in the northern part of the Congo Craton. In the BasCongo Basin, the West Congolian Group starts with siliciclastic deposits of the Sansikwa Subgroup. The lower age of the Sansikwa Subgroup is constrained by the underlying 912 + 7 Ma rhyolites (Tack et al. 2001). The upper age of the Sansikwa Subgroup is less well known, but has been inferred by Frimmel et al. (2006) to be c. 750 Ma based on correlation of the overlying Lower Diamictite unit with the Sturtian glacial event (assuming the Sturtian to have occurred at 720– 750 Ma, although this age is controversial). No carbonate rocks are observed in the Sansikwa Subgroup; the oldest carbonate rocks occur at the top of the post-Sturtian Haut-Shiloango Subgroup (Fig. 13.3). However, Poidevin (2007) proposed that the oldest limestones in the Lindi Supergroup (the Lenda Formation, Ituri Group) are pre-Sturtian and 730– 755 Ma in age on the basis 87Sr/86Sr ratios and correlation with the chemostratigraphic compilation of Halverson et al. (2007). However, the isotopic ratios reported by Poidevin (2007) are extremely variable (0.70663 –0.71090), making correlation difficult. Given the lack of any carbonate rocks prior to deposition of the lower diamictite unit (considered to be Sturtian in age in the BasCongo), the limestones of the Lenda Formation (Ituri Group) are considered to be post-Sturtian in age, as traditionally presented in the literature (Trompette 1994), and correlated with the HautShiloango Subgroup of the Bas-Congo Basin and the Bougboulou Series in the Fouroumbala-Bakouma Basin (Fig. 13.4). In Gabon, the Bouenza Group is a correlative of Haut-Shiloango Subgroup (Chevallier 2002). The Ituri Group is overlain by the Akwokwo Tillite which, as mentioned above, Poidevin (2007) correlated with the Lower Diamictite Formation of the West Congolian Group. However, recent field observations of Neoproterozoic diamictites in SW Gabon and/or in the DRC Bas-Congo and Katanga regions show that carbonate rocks comprising massive zoned pink to greyish dolomites immediately cap the upper diamictites in these regions, but are not found in association with the lower diamictites. These ‘cap carbonates’ are also recorded in the Fouroumbala-Bakouma Basin (the Bakouma Formation capping the Bondo Tillite) (Fig. 13.3). In

192

J. TAIT ET AL.

the Lindian Basin, large lenses of zoned pink dolomites are embedded in the Mamungi Formation overlying the Akwokwo Tillites. The close association of the Akwokwo Tillite with massive zoned pink to greyish dolomites is also confirmed from the RMCA (Tervuren) sample collection (Tack, pers. comm.). Hence, the occurrence in DRC and adjacent CAR of the cap carbonate is thought to be indicative of the Marinoan glacial event (Fig. 13.4). No carbonate is observed capping the diamictites of the Bangui Basin, but the Kembe´ Formation conglomerates are commonly correlated with the Bondo Tillite of the FouroumbalaBakouma Basin (Poidevin 1976, 1985, 2007; Bessoles & Trompette 1980; Cornacchia & Giorgi 1986). Finally, oolitic carbonate rocks of the Lindi Supergroup occur in the Alolo Formation (Aruwimi Group overlying the Lokoma Group; Fig. 13.3). Comparable oolitic carbonate rocks of post-Marinoan age are also known up sequence in the West Congo and Katanga regions. In the Bili and Dialinga Formations (upper part of the Bakouma Series) of the Fouroumbala-Bakouma Basin, some metabasalt layers are intercalated. In the Aruwimi Group of the Lindian Basin, Verbeek (1970) considers the Kaparata Breccia to be a conglomerate, possibly including some clasts of a volcanic rock. In fact, the RMCA (Tervuren) Kaparata samples are a hyaloclastic breccia, and are reminiscent of rocks observed in the Bas-Congo Basin, suggesting the occurrence in the Lindian Basin of pillow lavas and an up to now radiometrically undated tholeiitic basaltic event. In conclusion, unlike Poidevin (2007), we correlate the Akwokwo Tillite (Lindian Basin) with the Bondo Tillite (FouroumbalaBakouma Basin), the Upper Diamictite (Bas Congo) and Niari Tillite (SW Gabon), and conclude a Marinoan age. However, as already suggested by Poidevin (2007), the Upper Bakouma Series are correlated with the Aruwimi Group, as redefined in this paper (Fig. 13.3). This correlation is tentatively extended to the Mpioka Subgroup of the WCB (in SW Gabon and DRC BasCongo). Finally, the Redbeds of the Inkisi ‘Group’ of the WCB, with a maximum depositional age of c. 550 Ma (i.e. post-Pan African) are correlated with the Banalia ‘Group’ of NE DRC (formerly considered as the uppermost unit of the Neoproterozoic Lindi Supergroup). Without new geochronological and palaeomagnetic constraints for these Neoproterozoic basins it will be difficult to refine these correlations. In Gabon, the fieldwork was carried out under the terms of the ‘Programme Sysmin, 8e`me Fonds Europe´en du De´veloppement au groupement BRGMCGS-SANDER-MRAC’. J.T. and G.S. gratefully acknowledge financial support from the European Commission FP6 Programme (Marie Curie Excellence Grant). H. Frimmel and U. Zimmermann are gratefully thanked for their helpful and constructive reviews. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Abdelsalam, M. G., Liegeois, J. P. & Stern, R. J. 2002. The Saharan Metacraton. Journal of African Earth Sciences, 34, 119– 136. Alvarez, P. 1992. Re´partition de la se´dimentation dans le golfe Prote´rozoı¨que supe´rieur du Schisto-calcaire au Congo et Gabon. Implications en Afrique centrale. Palaeogeography, Palaeoclimatology, Palaeoecolology, 96, 281– 297. Alvarez, P. 1995. Evidence for a Neoproterozoic carbonate ramp on the northern edge of the Central African Craton: relations with Late Proterozoic intracratonic troughs. Geologische Rundschau, 84, 636– 648. Alvarez, P. 2000. A quantitative method for the study of nonfossiliferous clastic formations: pre-Pan-African sandstones from central Africa and the northern Democratic Republic of Congo [ex-Zaire]. Journal of African Earth Sciences, 31, 263– 284.

Babet, V. 1935. Esquisse ge´ologique provisoire de la re´gion comprise entre Bangui et la frontie`re du Cameroun. Chronique Mineralogique Col. Paris, 38, 160– 164. Benn, D. I. & Evans, D. J. A. 1996. The interpretation and classification of subglacially-deformed materials. Quaternary Science Reviews, 15, 23 – 52. Bessoles, B. & Trompette, R. 1980. Ge´ologie de l’Afrique: la chaıˆne panafricaine ‘zone mobile d’Afrique Centrale (partie sud) et zone mobile soudanaise’. Me´moire du Bureau de la Recherche Geleolgique et Minie`re, 92, 398. Bigotte, G. & Bonifas, G. 1968. Faits nouveaux sur la ge´ologie de la re´gion Bakouma. Chronique de Mines et Recherche Minerale, 36, 43 – 46. Bueno, J. F., Oliveira, E. P., McNaughton, N. J. & Laux, J. H. 2009. U –Pb dating of granites in the Neoproterozoic Sergipano Belt, NE-Brazil: implications for the timing and duration of continental collision and extrusion tectonics in the Borborema Province. Gondwana Research, 15, 86– 97. Cahen, L. 1978. La stratigraphie et la tectonique du Supergroupe OuestCongolien dans les zones me´diane et externe de l’oroge´ne`se OuestCongolien (Pan-African) au Bas-Zaı¨re et dans les re´gions voisines. Annales de la Muse´e Royale de l’Afrique Centrale, Tervuren, in 88, Science Geologique, 83, 150. Cahen, L. & Lepersonne, J. 1981. Proterozoic diamictites of Lower Zaire. In: Hambrey, M. A. & Harland, H. W. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 153–157. Chevallier, L., Makanga, J. F. & Thomas, R. J. 2002. Carte ge´ologique de la re´publique gabonaise, 1:1,000,000. Notice explicative. Council for Geoscience, South Africa. Cornacchia, M. & Giorgi, L. 1986. Les se´ries Pre´cambriennes d’origine se´dimentaire et volcano-se´dimentaire de la Re´publique Centrafiquaine. Annales de la Muse´e Royale de l’Afrique Centrale, Tervuren, in 88, Science Geologique, 93, 51. Dadet, P. 1969. Notice explicative de la carte ge´ologique de la re´publique du Congo-Brazzaville au 1:500 000. Me´moire du Bureau de la Recherche Geleolgique et Minie`re, 40, 103. D’Agrello-Filho, M. S., Pacca, I. G., Renne, P. R., Onstott, T. R. C. & Teixeira, W. A. 1990. Paleomagnetism of middle Proterozoic (1.01– 1.08 Ga) mafic dykes in Southeasterrn Bahia State –Sa˜o Francisco Craton, Brazil. Earth and Planetary Science Letters, 101, 332– 348. Daly, M. C., Lawrence, S. R., Diemu-Tshiband, K. & Matouana, B. 1992. Tectonic evolution of the Cuvette Centrale, Zaire. Journal of the Geological Society, London, 149, 539–546. Deblond, A., Punzalan, L. E., Boven, A. & Tack, L. 2001. The Malagarazi supergroup of southeast Burundi and its correlative Bukoba supergroup of northwest Tanzania: Neo- and Meso-proterozoic constraints from Ar – Ar ages of mafic intrusive rocks. Journal of African Earth Science, 32, 435– 449. Delhaye, F. & Sluys, M. 1923– 1924 and 1928– 1929. Observations ayant servi a l’elaboration de l’ ‘Esquisse geologique du Congo occidental’. Etude du systeme Schisto-Calcaire. 1er– 3e memoire. Annales de la Socie´te´ Belge de Ge´ologie, 1923– 1924, C. 50– 91 et 1928– 1929, C. 69– 114. Delor, C., Theveniaut, H. et al. 2008. New insights into the Precambrian Geology of Angola: basis for an updated lithochronological framework at 1:2,000,000 scale. 22nd Colloquium African Geology, Hammamet, Tunisia, Abstracts volume, 52 –53. Delpomdor, F. 2007a. Lithostratigraphie et se´dimentologie du faisceau de Sekelolo Sh8 (Sous-Groupe du haut-Shiloango), de la Formation de la Diamictite supe´rieure et de la Dolomie rose C1 du fiasceau du Kwilu CI (Sous-Groupe du Schisto-Calcaire), anticlinal de Congo Dia Kati, Bas-Congo, Re´publique De´mocratique du Congo (RDC), unpublished Internal report of Royal Museum of Central Africa, Tervuren (Belgium). Delpomdor, F. 2007b. Etude des de´pots diamictitiques du Ne´oprote´rozoı¨que supe´rieure en Re´publique De´mocratique du Congo et au sud-ouest du Rwanda, unpublished internal report of Royal Museum of the Central Africa, Tervuren (Belgium). Delpomdor, F. 2007c. Lithostratigraphie et se´dimentologie de la chaıˆne Ouest Congolienne du Ne´oprote´rozoı¨que supe´rieur (Formation de

THE WEST CONGO AND LINDI/UBANGI SUPERGROUPS

la Diamictite supe´rieure et Sous-groupe du Schisto-Calcaire) BasCongo, Re´publique De´mocratique du Congo, unpublished MSc thesis, Free University of Brussels. Delpomdor, F., Tack, L. & Pre´at, A. 2008. Microstructures in the Neoproterozoic tillites around the Congo River Basin (CRB), Democratic Republic of the Congo (DRC)– Comparison with the Karoo tillites from the Dekese borehole in the CRB. 22nd Colloquium African Geology, Hammamet, Tunisia, Abstracts volume. Delpomdor, F., Tait, J., Tack, L. & Pre´at, A. 2011. Neoproterozoic lithostratigraphy in the Democratic Republic of Congo (DRC). 23rd Colloquiom of African Geology, Johannesburg. De Paepe, P., Hertogen, J. & Tack, L. 1975. Mise en e´vidence de laves en coussin dans les facie`s volcaniques basiques du massif de Kibungu (Bas-Zaı¨re) et implications pour le magmatisme ouest-congolien. Annales de la Socie´te´ Belge de Ge´ologie, 98, 251– 270. De Waele, B., Lie´geois, J.-P., Nemchin, A. A. & Tembo, F. 2006. Isotopic and geochemical evidence of Proterozoic episodic crustal reworking within the Irumide Belt of south-central Africa, the southern metacratonic boundary of an Archaean Bangweulu Craton. Precambrian Research, 148, 225–256. De Waele, B., Johnson, S. P. & Pisarevsky, S. A. 2008. Palaeoproterozoic to Neoproterozoic growth and evolution of the eastern Congo Craton: its role in the Rodinia puzzle. Precambrian Research, 160, 127– 141. Evans, D. A. D. 2000. Stratigraphic, geochronological, and paleomagnetic constraints upon the Neoproterozoic climatic paradox. American Journal of Science, 300, 347–433. Evans, D. J. A. & Benn, D. I. 2004. A Practical Guide to the Study of Glacial Sediments. Arnold Publishers, London. Frimmel, H. 2009. Trace element distribution in Neoproterozoic carbonates as palaeoenvironmental indicator. Chemical Geology, 258, 338– 353. Frimmel, H. E., Tack, L., Basei, M. S., Nutman, A. P. & Boven, A. 2006. Provenance and chemostratigraphy of the Neoproterozoic West Congolian Group in the Democratic Republic of Congo. Journal of African Earth Sciences, 46, 221– 239. Ge´rard, G. 1958. Carte ge´ologique de l’Afrique Equatoriale Franc¸aise au 1/2 000 000 avec notice explicative. Brazzaville Directorate des Mines et Ge´ologie A.E.F., 198, 4 feuilles. Ge´rard, G. & Ge´rard, J. 1952. Stratigraphie du Pre´cambrien d’Oubangui-Chari occidental (A.E.F.) et essai de corre´lations avec les territoires voisins. 19th International Geological Congress (Algiers), A.S.G.A., 145– 153. Halverson, G. P., Dudas, F. O., Maloof, A. C. & Bowring, S. A. 2007. Evolution of the 87Sr/86Sr composition of Neoproterozoic seawater. Palaeogeography Palaeoclimatology Palaeoecology, 256, 103– 129. Hoffman, P. F. 2011. Glacigenic and associated strata of the Otavi carbonate platform and foreslope, northern Namibia: evidence for extreme glacioeustatic fluctuation. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 195– 210. Hudeley, H. & Belmonte, Y. 1966. Carte ge´ologique de la Re´publique Gabonaise, e´chelle 1/1 000 000. Bureau de Recherches Ge´ologiques et Minie`res (B.R.G.M.). Kadima, K. E., Mwene, N. S. S. & Francis, L. 2011. A Proterozoic-rift origin for the structure and the evolution of the cratonic Congo basin. Earth and Planetary Science Letters, 304, 240–250. Kampunzu, A. B., Kapenda, D. & Manteka, B. 1991. Basic magmatism and geotectonic evolution of the Pan African belt in central Africa: evidence from the Katangan and West Congolian segments. Tectonophysics, 190, 363–371. Kroner, A. & Correia, H. 1973. Further evidence for glaciogenic origin of Late Precambrian mixtites in Angola. Nature –Physical Science, 246, 115– 117. Legoux, P. & Hourcq, V. 1943. Esquisse ge´ologique de l’Afrique Occidental Franc¸aise, Bulletin du Service des Mines de l’Afrique Occidentale Franc¸aise. Lepersonne, J. 1951. Donne´es nouvelles sur la stratigraphie des terrains anciens du Bas-Congo. Bulletin de la Socie´te´ Belge de Ge´ologie, 60, 169– 189.

193

Master, S. & Wendorff, M. 2011. Neoproterozoic glaciogenic diamictites of the Katanga Supergroup, Central Africa. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. A. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society of London, Memoir, 36, 173– 184. Meert, J. G. & Van der Voo, R. 1996. Paleomagnetic and 40Ar/39Ar study of the Sinyai Dolerite, Kenya: implications for Gondwana assembly. Journal of Geology, 104, 131– 142. Meert, J. G., Van der Voo, R. & Ayub, S. 1995. Paleomagnetic investigation of the Neoproterozoic Gagwe lavas and Mbozi complex, Tanzania and the assembly of Gondwana. Precambrian Research, 74, 225– 244. Menzies, J. 2000a. Micromorphological analyses of microfabrics and microstructures, indicative of deformation processes, in glacial sediments. In: Maltman, A. J., Hubbard, B. & Hambrey, M. J. (eds), Deformation of Glacial Materials. Geological Society, London, 176, 245–248. Menzies, J. 2000b. Microstructures in diamictites of the lower Gondwana Formation (Huronian), near Elliot Lake Ontario: evidence for deforming-bed conditions at the grounding line? Journal of Sedimentary Research, 70, 210–216. Menzies, J., van der Meer, J. J. M. & Rose, J. 2006. Till as a glacial ‘tectomict’, its internal architecture, and the development of a ‘typing’ method for till differentiation. Geomorphology, 75, 172– 200. Mestraud, J.-L. 1952. Formation du socle en Oubangui-Chari central. 19th International Geological Congress (Algiers) A.S.G.A., 155– 162. Mestraud, J.-L. 1953. Notice explicative sur la feuille Bangassou Ouest. Carte ge´ologique de reconnaissance au 1/500.000. Direction Mines et Ge´ologie, l’Afrique Equatoriale Franc¸aise, Brazzaville. Moloto-A-Kenguemba, G. R., Trindade, R. I. F., Monie, P., Nedelec, A. & Siqueira, R. 2008. A late Neoproterozoic paleomagnetic pole for the Congo Craton: tectonic setting, paleomagnetism and geochronology of the Nola dike swarm (Central African Republic). Precambrian Research, 164, 214– 226. Noce, C. M., Pedrosa-Soares, A. C., da Silva, L. C., Armstrong, R. & Piuzana, D. 2007. Evolution of polycyclic basement complexes in the Arac¸uai Orogen, based on U–Pb SHRIMP data: implications for Brazil– Africa links in Paleoproterozoic time. Precambrian Research, 159, 60– 78. Oliveira, E. P., Toteu, S. F. et al. 2006. Geologic correlation between the Neoproterozoic Sergipano belt (NE Brazil) and the Yaounde belt (Cameroon, Africa). Journal of African Earth Sciences, 44, 470– 478. Pedrosa-Soares, A. C., Alkmin, F. F., Tack, L., Noce, C. M., Babinski, M., Silva, L. C. & Martins-Neto, M. A. 2008. Similarities and differences between the Brazilian and African counterparts of the Neoproterozoic Arac¸uai– West Congo Orogen. In: Pankhurst, R. J., Trouw, R. A. J., Brito Neves, B. B. & De Wit, M. J. (eds) West Gondwana: Pre-Cenozoic Correlations Across the South Atlantic Region. Geological Society, London, Special Publications, 294, 153– 172. Phillips, E. 2006. Micromorphology of a debris flow deposit: evidence of basal shearing, hydrofacturing, liquefaction and rotational deformation during emplacement. Quaternary Science Reviews, 25, 720– 738. Phillips, E., Merrit, J., Auton, C. & Golledge, N. 2007. Microstructures in subglacial and proglacial sediements: understanding faults, folds and fabrics, and the influence of water on the style of deformation. Quaternary Science Reviews, 26, 1499–1528. Pin, C. & Poidevin, J. L. 1987. U– Pb zircon evidence for a Pan-African granulite facies metamorphism in the Central-African-Republic – a new interpretation of the high-grade series of the northern border of the Congo Craton. Precambrian Research, 36, 303– 312. Pinna, P., Cocherie, A., Thieblemont, D., Feybesse, J.-L. & Lagny, P. 1996. Evolution ge´odynamique du Craton Est-Africain et de´terminisme gıˆtologique. Geodynamic evolution and metallogenic controls in the East-African Craton (Tanzania, Kenya, Uganda). BRGM, Chronique Recherche Miniere, 525, 33 – 43. Piper, J. D. A. 1972. Paleomagnetic study of Bukoban System, Tanzania. Geophysical Journal of the Royal Astronomical Society, 28, 111–127.

194

J. TAIT ET AL.

Poidevin, J.-L. 1976. Les formations du Pre´cambrien supe´rieur de la re´gion de Bangui (R.C.A.). Bulletin de la Societe´ Ge´ologique de France, 18, 999– 1003. Poidevin, J.-L. 1979a. Echelle stratigraphie des formations pre´cambriennes de Centrafrique (R.C.A.). Re´sume´ 10e`me Colloque de Ge´ologie Africaine, Montpellier. Poidevin, J.-L. 1979b. Les basaltes et dole´rites Pre´cambrien supe´rieur de la re´gion de Bakouma (Empire centraficain). 7e`me Re´uion Annuelle des Sciences de la Terre, Lyon, 374. Poidevin, J.-L. 1985. Le Prote´rozoı¨que supe´rieur de la Re´publique centrafricaine. Annales de Muse´um Royal d’Afrique Centrale Tervuren, Belgique, Se´rie in 8e`me, Sciences Ge´ologiques, 91, 75. Poidevin, J.-L. 2007. Stratigraphie isotopique du strontium et datations des formations carbonate´es et glacioge´niques ne´oprote´rozoı¨ques du Nord et de l’Ouest du Craton du Congo. Comptes Rendus Geoscience, 339, 259– 273. Poidevin, J.-L., Alabert, J. & Miauton, J. D. 1981. Ge´ologie des se´ries du Pre´cambrien supe´rieur de la re´gion de Bakouma (Re´publique centrafricaine). Bulletin du Bureau de Recherches Geologiques et Minieres, Section IV, 4, 313–320. Pre´at, A. 2008. Etude au microscope polarisant de 165 lames minces de roches carbonate´es du synclinal de la Nyanga (feuilles a` 1:200 000 de Fougamou et de N’Dende´). Rapport Universite´ Libre de Bruxelles, Tervuren, Belgique. Pre´at, A., Bouton, P., Kolo, K., Prian, J.-P., Simo Ndounze, S. & Thie´blemont, D. 2008. Se´dimentologie et isotopes (carbone et oxyge`ne) des carbonates Pre´cambriens du Gabon: rapport au mode de fonctionnement des bassins Ne´o- et Pale´oprote´rozoı¨ques. Communication Spec. Session Geologica Belgica, Namur, Belgique. Pre´at, A., Delpomdor, F., Kolo, K., Gillan, D. & Prian, J.-P. 2011. Stromatolites and cyanobacterial mats in peritidal evaporitive environments in the Neoproterozoic of Bas-Congo (Democratic Republic of Congo) and South of Gabon. In: Tewari, V. C. & Seckback, J. (eds) Stromatolites: Interactions of Microbes with Sediment, Cellular Origin and Life in Extreme Habitats and Astrobiology. Springer Verlag, 43 –63. Pre´at, A., Kolo, K., Prian, J.-P. & Delpomdor, F. 2010. A peritidal evaporite environment in the Neoproterozoic of South Gabon (SchistoCalcaire Subgroup, Nyanga Basin). Precambrian Research, 177, 253– 265. Prian, J.-P. 2008. Notice ge´ologique et ressources mine´rales de la carte de N’Dende´ a` 1/200 000. Document provisoire, carte ge´ologique du Gabon. Schermerhorn, L. J. G. 1974. No evidence for glacial origin of Late Precambrian tilloids in Angola. Nature, 252, 114–115. Schermerhorn, L. J. G. 1981. Late Precambrian Tilloids of northwest Angola. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 158– 161. Schermerhorn, L. J. G. & Stanton, W. I. 1963. Tilloids in the West Congo geosyncline. Quarterly Journal of the Geological Society of London, 119, 201–234. Stanton, W. I., Schermerhorn, L. J. G. & Korpershoek, H. R. 1963. The West Congo System. Boletim dos Servic¸os de Geologia e Minas de Angola, 8, 69 –78. Straathof, G. B., Tait, J., Cibambula, E., Kanda Nkula, V. & Zimmermann, U. 2008. Constraints on the glacial events from the Neoproterozoic West Congolian Group, 33rd International Geological Congress, Oslo. Tack, L., Wingate, M. T. D., Lie´geois, J.-P., Fernandez-Alonzo, M. & Deblond, A. 2001. Early Neoproterozoic magmatism

(1000 –910 Ma) of the Zadinian and Mayumbian Groups (BasCongo), onset of Rodinia rifting at the western edge of the Congo Craton. Precambrian Research, 110, 277–306. Tack, L., Fernandez-Alonso, M., Kanda Nkula, V., Mpoyi, J., Delvaux, D., Trefois, Ph. & Baudet, D. 2006. Neoproterozoic Diamictites around the Congo River Basin: a critical reappraisal of their origin. 21st Colloquium African Geology (CAG21), 3– 5 July 2006, Maputo, Mozambique, Abstract Book, 152– 153. Tack, L., Delvaux, D. et al. 2008. The 1000 m thick Redbeds sequence of the Congo River Basin (CRB): a generally overlooked testimony in Central Africa of post-Gondwana amalgamation (550 Ma) and preKaroo break-up (320 Ma). 22nd Colloquium on African Geology, Hammamet, Tunisia, Abstracts volume, 86 –88. Tack, L., Delvaux, D. et al. 2009a. The 1000 m thick Redbeds sequence of the Congo River Basin (CRB): a generally overlooked testimony in Central Africa of post-Gondwana amalgamation (550 Ma) and preKaroo break-up (320 Ma). Geological Society of London Fermor Meeting, Edinburgh. Tack, L., Wingate, M. T. D. et al. 2009b. The Mesoproterozoic ‘Kibaran Event’ in Central Africa: a 1375 Ma intracratonic emplacement of a Large Igneous Province (LIP). Geological Society of London Fermor Meeting, Edinburgh. Thie´blemont, D., Castaing, C., Billa, M., Bouton, P. & Pre´at, A. et collaborateurs. 2009. Notice explicative de la carte ge´ologique et des Ressources mine´rales de la Re´publique gabonaise a` 1:1 000 000. Editions DGMG, Ministe`re des Mines, du Pe´trole, des Hydrocarbures. Libreville. Toteu, S. F., Fouateu, R. Y. et al. 2006. U– Pb dating of plutonic rocks involved in the nappe tectonic in southern Cameroon: consequence for the Pan-African orogenic evolution of the Central African fold belt. Journal of African Earth Sciences, 44, 479– 493. Trompette, R. 1994. Geology of Western Gondwana (2000– 500 Ma). Pan-African-Brasiliano aggregation of South America and Africa. Balkema, Rotterdam. Trompette, R. & Boudzoumou, F. 1988. Paleogeographic significance of stromatolitic buildups on Late Proterozoic Platforms– the example of the West Congo Basin. Palaeogeography, Palaeoclimatology, Palaeoecology, 66, 101–112. Van der Meer, J. J. M. 1993. Microscopic evidence of subglacial deformation. Quaternary Science Reviews, 12, 553– 587. Van der Meer, J. J. M. 1997. Particle and aggregate mobility in till: microscopic evidence of subglacial processes. Quaternary Science Reviews, 16, 827–831. Van der Meer, J. J. M. 1999. Particle and aggregate mobility in till: microscopic evidence of subglacial processes. Quaternary Science Reviews, 22, 1659–1685. Van der Meer, J. J. M., Menzies, J. & Rose, J. 2003. Subglacial till: the deforming glacier bed. Quaternary Science Reviews, 22, 1659– 1685. Vellutini, P. & Vicat, J. P. 1983. On the origin of basal conglomeratic formations of the West-Congolan Geosyncline (Gabon, Congo, Zaire, Angola). Precambrian Research, 23, 87– 101. Verbeek, T. 1970. Ge´ologie et lithostratigraphie du Lindien (Pre´cambrien supe´rieur du nord de la Re´publique De´mocratique du Congo). Annale de la Muse´e Royale de l’Afrique Centreale, se´rie n88, 66, 309. Vicat, J.-P., Le´ger, J. M., Nsifa, E., Piguet, P., Nzenti, J. P., Tchameni, R. & Pouclet, A. 1996. Distinction au sein du craton congolais du Sud-Ouest du Cameroun, de deux e´pisodes doleitiques initiant les cycles orogeniques eburnee´ en (Pale´oprote´rozoique) et panafricain (Ne´oprote´rozoique). Comptes Rendus de l’Acade´mie des Sciences. Se´rie 2. Sciences de la terre et des plane`tes, 323, 575– 582.

Chapter 14 Glaciogenic and associated strata of the Otavi carbonate platform and foreslope, northern Namibia: evidence for large base-level and glacioeustatic changes PAUL F. HOFFMAN1,2 1

Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA 02138, USA 2

School of Earth and Ocean Sciences, University of Victoria, Victoria, BC V8W 2Y2, Canada (e-mail: [email protected])

Abstract: Two discrete, mappable, glaciogenic formations occur within the Otavi Group, a 3 + 1-km-thick carbonate-dominated platform of late Neoproterozoic age, developed on the SW promontory of the Congo craton in northern Namibia and exposed in bordering late Ediacaran fold belts. Each is overlain abruptly by an expanded postglacial carbonate sequence, the younger of which begins with a globally-correlative transgressive dolopelarenite. The older Chuos glaciation (,746 Ma) occurred during a time of north-south crustal stretching. Debris derived from upturned older rocks collected in structural depressions. The younger Ghaub glaciation (635 Ma) occurred, after stretching ceased, on a thermally-subsiding marine platform and its distally-tapered foreslope. A continuous ice grounding-zone wedge (GZW) occurs on the distal foreslope, while the upper foreslope and outer platform are devoid of glacial debris and only small pockets of lodgement facies exist on the inner platform. Debris in the GZW is derived from a distinctive falling-stand wedge that is unique to the foreslope and from immediately older strata mined preferentially from the inner platform. The GZW rests on a smooth surface that includes a transverse steep-walled trough presumably cut by an ice-stream, within which is a towering doubly-crested moraine composed of composite, massive, carbonate diamictite. The surface suggests that the ice-sheet was grounded on the distal foreslope, implying a large fall in base level at a glacial maximum that predates the GZW. The glacial record ends with Fe-stained beds, rich in ice-rafted debris, that are notably absent from the moraine, upper foreslope and platform, which were apparently above sea-level at that time.

The existence of glaciogenic diamictite within the Neoproterozoic carbonate succession of the Otavi Mountains in north-central Namibia (Fig. 14.1a) was first recognized by le Roex (1941), who cited as evidence of glacial action the presence of faceted and striated clasts, absence of sorting or clast gradation, uniformity of gritty fine-grained matrix, and import of granitic and arenaceous clasts into a conformable carbonate succession. Because its floor is not morphologically exposed, le Roex (1941) forthrightly admitted that the diamictite could have been deposited by grounded ice or floating ice. Martin (1965a, b) argued that the regional extent of this unit and the existence of similar diamictites of roughly the same age throughout southern Africa and on other continents favoured a glacial origin. He interpreted the Otavi Tillite (le Roex 1941) as a glaciomarine unit on account of its stratigraphic interposition between marine carbonate formations. Kro¨ner & Rankama (1972) and Hedberg (1979) also favoured a glaciomarine interpretation. Martin (1965b) noted the close association of banded Fe-formation with diamictite in the Otavi Group and attributed ‘this peculiar combination of sediments’ to ‘oxygen deficiency in stagnating bottom waters caused by an ice cover’. A glacial origin for the carbonate-clast diamictite (now Ghaub Formation) on the Fransfontein foreslope (Fig. 14.1b) was rejected by Frets (1969), who favoured submarine landsliding in response to a fall in base level. This view was taken up by Martin et al. (1985), who at age 75 (and 20 years removed from Africa) recanted his earlier conclusions and hypothesized that a major eustatic event, possibly caused by glaciation in other parts of the globe, had triggered submarine landslides simultaneously throughout Namibia. The discovery of large submarine landslides of Quaternary age on the modern African margin (Dingle 1977) may have prompted this reinterpretation. Most present-day workers, however, favour the glaciomarine interpretation (Hoffmann & Prave 1996; Hoffman et al. 1998a, 2007; Evans 2000; Condon et al. 2002; Hoffman 2005, 2011; Domack & Hoffman 2011). Although structureless carbonate-clast diamictite (with rare basement-derived clasts) is the predominant lithology, thin intervals of well-stratified detrital carbonate with variable amounts of rafted debris occur at the top and bottom of the formation and

discontinuously within it. Condon et al. (2002) argue that this evidence for ice-line instability and iceberg-rafting is incompatible with an ice-covered ocean at the time of Ghaub deposition. In rejecting a glacial origin altogether, Eyles & Januszczak (2004a, b) revived the ideas of Frets (1969) and Martin et al. (1985). In their opinion, the Ghaub Formation consists ‘entirely of a wide variety of mass flow facies, . . . derived from faulted underlying carbonate platform strata’. This interpretation is difficult to reconcile with the regionally consistent stratigraphic position of the Ghaub Formation with respect to the overlying Keilberg Member (Hoffmann & Prave 1996; Hoffman et al. 1998a, 2007). It turns a blind eye to prior stratigraphic mapping, sedimentology and chemostratigraphy showing that faulting ended well before the Ghaub Formation was deposited (Soffer 1998; Hoffman 1999; Halverson et al. 2002). During Ghaub time, no fault-related structural rotations occurred on the platform or on its margin (Hoffman & Halverson 2008).

Structural framework The Otavi Group is exposed in an arcuate fold-and-thrust belt around the junction of the Late Ediacaran – Early Cambrian Kaoko and Damara orogens (Fig. 14.1a). At the apex of the fold belt is the Kamanjab inlier, a large basement-cored structural culmination. For most of the twentieth century, studies focused on the Otavi Mountains near the eastern terminus of the fold belt (Miller 2008; Bechsta¨dt et al. 2009), where major carbonate-hosted Cu – Pb– Zn –V mineralization occurs. Since Namibian independence in 1990, scientifically motivated research has favoured the better exposed sections around the western Kamanjab inlier. Weakly folded Otavi Group strata extend across the Owambo Basin in the subsurface, making the total area of the ancient carbonate platform in Namibia .110 000 km2 (Miller 1997). The western margin of the carbonate platform is complicated by a major, sinistral-oblique thrust zone (Sesfontein thrust) bounding the Kaoko Belt (Goscombe et al. 2003). The southern margin is preserved in the east around the Otavi syncline (Smit 1962;

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 195– 209. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.14

196

P. F. HOFFMAN

Fig. 14.1. (a) Geological map of the Otavi Group fold belt and its relation to the Owambo Basin (Congo craton) and to the Kaoko and Damara orogenic belts (Hoffman 2005). Dashed white lines indicate the southern and western edges of the Otavi Platform. The black rectangle shows the area of Figure 14.1b. (b) Geological map of the Fransfontein Ridge and its relation to the Otavi Platform of the Upper Huab outlier and Danube´ (after Hoffman et al. 2007). Also shown is the division of the foreslope into proximal and distal zones, the location of the Duurwater trough and moraine, and district road C35.

Hedberg 1979) and to the west along the autochthonous southern flank of the Kamanjab inlier (Fig. 14.1b). Distal stratigraphic equivalents of the Otavi Group occur on tectonically stretched and later foreshortened Congo crust within the central and eastern zones of the Kaoko Belt (Swart 1992; Stanistreet & Charlesworth 1999; Goscombe & Grey 2007; Paciullo et al. 2007) and in the central and northern zones of the Damara Belt (Hoffmann et al. 2004; Johnson et al. 2006; Clifford 2008; Hoffman & Halverson 2008).

Stratigraphy The Otavi Group is a 3 + 1-km-thick, carbonate-dominated succession of Cryogenian and early Ediacaran age, deposited between c. 770 and 580 Ma (Halverson et al. 2005). It is conformably underlain by fluviatile feldspathic quartzites and conglomerates of the Nosib Group (Fig. 14.2), and overlain by marine and non-marine, synorogenic clastics of the Mulden Group. The Otavi and Mulden groups are separated by a paraconformity with .1.0 km of karstic palaeotopographic relief (Frets 1969; Hoffman & Hartz 1999; Hoffman & Halverson 2008).

Subgroups of the Otavi Group The Otavi Group was originally divided into two subgroups, Abenab and Tsumeb, at a surface corresponding to the floor of what was then thought to be a single glaciogenic formation

(Hedberg 1979; SACS 1980), assumed correlative with the glaciogenic Chuos Formation of the central Damara Belt (Gevers 1931). Hoffmann & Prave (1996) first established the existence of two discrete and extensive glaciogenic formations in the Otavi Group, the older of which (Varianto Formation of SACS 1980) is equivalent to the type Chuos Formation and the younger of which (the Otavi Tillite of le Roex 1941) they renamed the Ghaub Formation. This critical distinction has been amply confirmed by subsequent studies (Hoffman et al. 1998a, b; Kennedy et al. 1998; Halverson et al. 2005; Hoffman & Halverson 2008). Hoffmann & Prave (1996) divided the Otavi Group into three subgroups at surfaces corresponding to the floors of the two glaciogenic formations. Thus the Ombombo (om-bomb-boh) Subgroup ends below the Chuos Formation, the Abenab (ah-ben-ab) Subgroup below the Ghaub Formation, and the Tsumeb (tsoo-meb) Subgroup below the Mulden Group. South of the Huab palaeoridge (Fig. 14.2), the Ugab (oo-khab) Subgroup occupies a stratigraphic position analogous to the Ombombo Subgroup, but U –Pb zircon ages show it to be significantly younger (,746 + 2 Ma) than the Devede (de-ved-ee) Formation of the Ombombo Subgroup, which is .760 + 1 Ma (Halverson et al. 2005).

Crustal stretching and thermal subsidence in relation to glaciation Subsidence accommodating the Otavi Group was driven by north – south crustal stretching, which lasted at least episodically from

THE OTAVI CARBONATE PLATFORM AND FORESLOPE

South

Otavi Platform

50

-0.5

200

δ13C (per mil) –10 –5 0 5

197

10

Tsumeb Subgroup

" Huttenberg

-1.0

Elandshoek Foreslope -2.0 km

Chuos Ugab

Maieberg Ombaatjie

.

Huab Ridge

Abenab Subgroup

Gruis

Rasthof . . . . . . . . Devede . . . 746 ± 2 Ma 759 ± 1 Ma . . . . Naauw Palaeoproterozoic metamorphic. Nosib -poort . . plutonic complex (Congo craton) Karibib

Ghaub

Keilberg

Makalani Ridge

Ugab Sbgp Ombombo Subgroup

0 5 10 –10 –5 δ 13C (per mil V-PDB) Congo Angola

cannibalized clastics (syn-rift)

shallow-water carbonate deep-water carbonate glaciogenic diamictite

bimodal alkaline volcanics

. .

feldspathic sandstone (palaeoflow)

c. 770 Ma (Halverson et al. 2005) until a rift-sag transition at the base of the Ombaatjie (om-bye-gee) Formation (Fig. 14.2) of the Abenab Subgroup (Halverson et al. 2002). Accordingly, the Chuos glaciation occurred during a prolonged period of active faulting, which is reflected by the diversity of its debris and by a low-angle (1.58) structural unconformity at its base (Fig. 14.2) that truncates .2 km of Ombombo and Nosib, or Ugab and Naauwpoort (nauf-port) strata (Hoffman et al. 1998a). No causal connection between glaciation and crustal stretching (Eyles & Januszczak 2004a, b) is indicated: stretching began millions of years before the glaciation and continued for millions of years afterwards. In contrast to the Chuos, the Ghaub glaciation occurred at a time of broad regional subsidence, which is reflected by platform-wide structural conformity between depositional cycles of the upper Ombaatjie Formation and the Keilberg cap dolostone (Halverson et al. 2002; Hoffman et al. 2007). The Ombaatjie Formation was regionally flat-lying at the time of the Ghaub glaciation. The Tsumeb Subgroup is a structurally conformable succession on the regional scale and thermal subsidence is the inferred accommodation mechanism for the Ombaatjie through Hu¨ttenberg formations (Halverson et al. 2002).

Palaeogeography of the platform and the southern foreslope The unfaulted southern margin of the platform (Figs 14.1a & 14.2) is well defined lithologically in units above and below the Ghaub Formation (Tsumeb Subgroup and Ombaatjie Formation, respectively). To the north are stacked peritidal parasequences, commonly bounded by subaerial exposure surfaces marked by tepee structures and tepee breccias (Kendall & Warren 1987). To the south is a distally tapered foreslope wedge of progressively deeper-water carbonate facies (Hoffman & Halverson 2008). More proximal parts of the foreslope (i.e. 0– 5 km seaward of the slope break) are dominated by ribbony lutites and arenites lacking subaerial exposure surfaces. More distally (i.e. .5 km seaward of the slope break), parallel-sided rhythmites, rhythmite slump breccias and carbonate turbidites prevail. Overall, the Tsumeb Subgroup

Tanz.

Zam. Zim.

Nam. Bots. Swaz.

South Africa

Mal.

Moz.

Fig. 14.2. Stratigraphic restoration of the Otavi Group in a north–south cross-section based on measured sections and mapped stratigraphic cutoffs around the western Kamanjab inlier (after Hoffman et al. 2007). Note the existing U –Pb zircon age constraints from volcanic lavas and tuffs (Hoffman et al. 1996; Halverson et al. 2005), the composite d13C profile (Halverson et al. 2005) and the location map showing the Kalahari and Congo cratons (shaded).

tapers from c. 1400 m on the platform to 320 –540 m for isotopically correlative strata on the distal foreslope (Halverson et al. 2005). The Abenab Subgroup tapers from 500–800 m on the platform to 70– 200 m on the distal foreslope (Hoffman 2005). Only the Keilberg cap dolostone exhibits shallow-water bedforms on the distal foreslope, reflecting large-amplitude glacioeustatic fluctuation (Hoffman et al. 2007). The foreslope wedge is best exposed on the 60-km-long Fransfontein Ridge (Fig. 14.1b), a present-day topographic feature underlain by a simple, south- to SE-dipping, structural homocline of Otavi Group carbonates. The strata dip 40–608 on average, with near-vertical cleavage in limestones and fine clastics, and local north-vergent thrust imbrication of major rheological interfaces (e.g. dolomite against fine clastics at the Otavi-Mulden group contact, and dolostone against marly limestone within the Maieberg Formation). The foreslope is also preserved in structural outliers 60 and 80 km west of Fransfontein (on the farms Bethanis and Toekoms, and Vrede and Opdraend, respectively; see Hoffman & Halverson 2008). It also occurs for 250 km east of Fransfontein along the southern flank of the Kamanjab inlier and reportedly on the southern limb of the Otavi Syncline (Smit 1962; Hedberg 1979). Thus, the foreslope wedge is discontinuously exposed for 330 km along strike. The foreslope divides naturally on the basis of lithofacies (see above) and stratigraphic architecture (Fig. 14.3b) into proximal and distal zones, projected to lie 0–5 and .5 km seaward of the slope break, respectively (Fig. 14.1b). The western two-thirds of the Ridge strike east –west, subparallel to the inferred palaeoslope contours. It exposes a longitudinal section of the distal slope. The eastern third of the ridge strikes NE and exposes a section that angles obliquely up the proximal slope to the edge of the platform (or slope break) near its northernmost point on the farm Danube´ 59. Comparison with recent fine-grained carbonate slopes like the western (leeward) margin of the Great Bahama Bank (Eberli & Ginsburg 1987; Adams & Schlager 2000) suggests that the distal foreslope (where the glaciogenic Ghaub Formation occurs) lay at palaeodepths .0.5 km below the rim of the platform (Hoffman 2005, 2011).

198

P. F. HOFFMAN

1986), but the paucity of bed-scale stratification suggests that debris flows were of secondary importance in the Chuos Formation of the Otavi Group. The associated strata, in decreasing order of abundance, are conglomerates and pebbly sandstones, laminated siltstones with rafted debris, and banded Fe-formation. The coarse clastics are co-derived with the diamictites but are grain-supported, clast-imbricated and crudely stratified or cross-stratified. They were clearly deposited by running water, but whether the flow was proglacial or subglacial is not known. The conglomerates and pebbly sandstones generally occur at the top and/or bottom of the Chuos Formation, and are richer in dolomite clasts and/or felspathic basement debris compared with similar lithofacies in the underlying, non-glaciogenic, Okakuyu (oh-ka-koo-you) Formation (Ombombo Subgroup). Relatively thin units of finely laminated siltstone or silty mudstone with rafted lonestones occur locally at the base or within the Chuos Formation. Discontinuous beds and lenses of hematitic Fe-formation are intimately intercalated with structureless or stratified diamictites in different areas (Martin 1965a). Discharges of oxygenated meltwater at the grounding lines of tidewater glaciers could have triggered the precipitation of Fe-formation precursor in the presence of Fe(II)laden seawater.

Glaciogenic and associated strata Chuos Formation The Chuos Formation is highly variable in thickness up to 1.0 km and the predominant lithology, especially in the thicker sections, is structureless to weakly stratified, matrix-supported diamictite. The diamictite forms composite bodies derived in varying proportions from basement rocks (metavolcanic, metaplutonic and metasedimentary rocks, notably orthoquartzite), sub-feldspathic arenites of the Nosib Group, and dolomites, cherts and cannibalized clastics of the Ombombo and Ugab subgroups. Clasts of strongly porphyritic, mafic and felsic lava, derived from the Naauwpoort Formation (747 + 2 Ma), are locally abundant. The matrices of the diamictites are foliated to structureless, mud-rich wackestones. Some of the foliated fabrics are pre-tectonic. Striated clasts are probably of glacial origin. Fe- and/or Mn-rich cements coat the detrital grains, giving the matrix a strong colour – brick red, olive green or black. Fabric analysis to determine emplacement mechanisms are compromised by tectonic strains, but the diamictites appear to represent both ice-contact (foliated) and proximal rain-out deposits. Diamictites in the central zone of the Damara orogen were emplaced as debris flows (Henry et al.

SOUTH –20

Distance inward from the glacial-age slope break (km) 0 20 40 60 80

–10

OUTER PLATFORM

400

FR AN SF ON TE IN FO RE SL OP E

0 –200

iso

che

m

aeolianite

Trezona anomaly δ13C = 0 o/oo datum

GREAT OTAVI CARBONATE BANK

–400 –600

INNER PLATFORM

subaerial surface

sea-floor cements

200

NORTH 100

400 m

Depth wrt glacial-age platform rim (m)

(a)

Maieberg Fm: shoalwater dolostone grainstone

POSTGLACIAL

Ghaub prism

SYN-DEGLACIAL

Ghaub Fm: diamictite prism (ice grounding zone)

SYN-GLACIAL

–800

Maieberg Fm: deepwater limestone rhythmite Keilberg Mb: transgressive 'cap' dolostone

ADVANCE GLACIAL

Franni-aus Mb: oolite debris flows (lowstand wedge) Ombaatjie Fm: shelf-slope carbonate (post-Trezona 0 o/oo)

PRE-GLACIAL

Ombaatjie Fm: shelf-slope carbonate (pre-Trezona 0 o/oo)

–0.2

Keilberg cap dolostone

–0.5 –0.6 –0.7 –0.8 –0.9 –1.0

Franni-aus Member low-stand wedge

SL OP

–0.2

AL

Ghaub diamictite prism

M

–0.1

OX I

–0.4

PR

–0.3

E

0.0 Palaeodepth wrt top of platform (km)

Palaeodepth wrt permanent post-glacial flood

(b)

–0.3

DISTAL SLOPE

–0.4 –0.5

Ombaatjie Fm

–0.6

Narachaams Member siltstone of cycle b8

–0.7 –0.8 10

9

South

8

5 7 6 4 3 2 Horizontal distance from slope break (km)

1

0

–1

North

Fig. 14.3. (a) Stratigraphic – palaeobathymetric cross-section of the Otavi carbonate platform and foreslope, showing showing the Ghaub diamictite wedge and its relations to the Frannis-aus falling-stand wedge and the Maieberg cap-carbonate sequence. The 0‰ cross-over of the Trezona d13C anomaly serves as the datum for the platform. Foreslope palaeodepths are estimated assuming a profile (below – 200 m) like the western margin of the present Great Bahama Bank (see text). Ghaub diamictite 20 km inboard of the slope break is the ‘Otavi Tillite’ (le Roex 1941; Bechsta¨dt et al. 2009). Note parallelism between the Trezona datum and the top of the Maieberg sequence, disproving structural rotations due to faulting associated with the Ghaub diamictites. (b) Proximal and distal differentiation of the Fransfontein foreslope based on stratigraphic relations between the Franni-aus falling-stand wedge, the Ghaub diamictite wedge and the Keilberg cap dolostone. The palaeodepth scale with respect to permanent (i.e. after isostatic adjustment) post-glacial flood assumes a water depth on the outer platform of 140 m for the Maieberg Formation limestone rhythmite, which was deposited below storm wave and accommodated c. 300 m of additional cap-carbonate strata.

THE OTAVI CARBONATE PLATFORM AND FORESLOPE

Ghaub Formation The palaeogeographic distribution of the Ghaub Formation is critical to its interpretation. As mapped, roughly 99% of its mass is contained within a linear wedge that tracks the distal foreslope, .5 km downslope from the outer edge of the platform (Fig. 14.1b). It pinches out slightly upslope from the zero-isopach of the Franni-aus Member (Fig. 14.3), a glacioeustatic fallingstand wedge composed of carbonate turbidites and debris-flows of non-glacial origin (see below). Abundant Franni-aus-derived debris within the Ghaub Formation indicates that the falling-stand wedge originally extended farther upslope, where it was subsequently eroded and redeposited downslope during the glaciation. In the Summas Mountains, 30 km south of the platform, the Ghaub Formation is c. 4.0 m thick and its glacial affinity is scarcely recognizable (Clifford 2008; Hoffman & Halverson 2008). The geometry of the Ghaub Formation is a wedge, continuous along strike, pinching-out upslope and severely tapered downslope. The Ghaub Formation is totally absent from the proximal foreslope and the raised outer zone of the platform (Fig. 14.3). There, the Keilberg Member (cap dolostone) rests disconformably on the upper Ombaatjie Formation (Figs 14.2 & 14.3). Pockets of carbonate diamictite reappear beneath the Keilberg Member on the inner zone of the platform, which was preferentially eroded by up to 80 m relative to the outer zone during the glaciation (Hoffman et al. 2007). The type section of the Otavi Tillite (le Roex 1941; Bechsta¨dt et al. 2009) is an exceptional 60-m-thick body of diamictite, situated c. 20 km north of the edge of the platform. On the whole, however, the platform and proximal foreslope were areas of erosion during the Ghaub glaciation, and the main zone of sedimentation was on the distal foreslope. In order to document the internal facies architecture and lateral variability of the Ghaub Formation, 72 stratigraphic sections were measured on the Fransfontein Ridge as a whole, including 40 sections in a 13-km-long stretch east of the town of Fransfontein (Fig. 14.4). The sections examined by Condon et al. (2002) and Eyles & Januszczak (2007) are among those measured, as are the sections recently logged in greater detail (Domack & Hoffman 2011). Facies associations and stratal architecture. The Ghaub Formation is composed of three facies associations. The first consists of weakly stratified to structureless diamictites, individually ranging from 1 to 75 m in thickness. Pebble- to boulder-sized, rounded to angular clasts of limestone and/or dolostone (rarely basementderived granitoids) ‘float’ in a structureless to faintly laminated matrix of unsorted detrital carbonate mud, silt, sand and granules. Because of middle greenschist-grade metamorphism, the clasts are thoroughly welded to their carbonate matrix, making it impossible to observe clast surface morphology or micromorphology (e.g. striae). Isotopic data indicate that the clasts are derived from the upper part of the Ombaatjie Formation (Fig. 14.2) and from the Frannis-aus Member (see below), which contributed the very coarse-grained oolite clasts and grains conspicuous in many of the diamictites. Some diamictites gradationally overlie wellstratified units, whereas others have erosive bases and internal reactivation surfaces. Collectively, weakly stratified and structureless diamictites make up 68% of the Ghaub Formation west of the Duurwater moraine (see below) and 82% overall. The second facies association, amounting to 17% of the formation as a whole, consists of well-stratified detritus and comprises suspension-plume fallout, turbidites, debrites, contourites and rafted debris, in varying proportions. All the detritus is carbonate, except to the west of Fransfontein town (particularly on Bethanis farm), where the basal stratified interval is composed of siliciclastic siltstone with rare outsize quartz granules. This terrigenous sediment is likely reworked from the Narachaams (nahra-khams) Member (Figs 14.3b & 14.5), a fine-grained siliciclastic

199

incursion in the upper Ombaatjie Formation. Plume fallout forms laterally continuous drapes of laminated mud and fine silt with both normal and reverse grain-size gradation. Plume fallout hosts rafted debris of variable size and concentration. Turbidites and debris flows consist of poorly sorted, sandy to pebbly material in discrete tabular or channelized sedimentation units that are commonly grain-size graded towards the top and less commonly reverse-graded near the base. Contourites are starved (isolated) ripples with foreset laminae consistently indicating current flow from ENE to WSW, subparallel to depositional strike. Rafted debris are lonestones not carried by debris flows or bedload sedimentation units. They are most abundant in slowly deposited sediments (plume fallout) and least abundant in rapidly deposited material (turbidites and debris flows). Subjacent laminae are pinched or punctured, adjacent ones are laterally ejected, and superjacent ones onlap or drape the rafted debris. The third facies association consists of current-rippled or crossbedded, sand- or granule-sized, detrital carbonate with sparse lonestones. Distal facies include fine sand with trains of climbing ripples, implying combined bedload and suspension fallout. Palaeoflow directions are southwestward with large apparent dispersion. Proximal facies involve gravelly sand organized in multi-metric clinoforms that coarsen up-dip. Grain-supported pebble conglomerate is the most proximal (and least abundant) component of this facies association. The sandy facies occurs in units up to 5.5 m thick that can rarely be traced laterally for more than a few 100 m. They amount to less than 2% of the formation as a whole. West of the Duurwater moraine (Fig. 14.4), the facies associations are organized as follows. Each vertical section contains 2 to 7 (average 4.6) diamictite units, usually separated by thinner intervals of well-stratified and/or sandy detritus, but occasionally amalgamated diamictite-on-diamictite with erosive contacts. The stratified intervals commonly coarsen upwards, and their contacts with overlying diamictites are gradational or erosively truncated. Compositionally, they normally herald the overlying diamictite. These characteristics are all well illustrated by the stratified interval at the base of the formation. The tops of diamictites are commonly abrupt, in some cases associated with coarse-grained debris flows and in others by a direct transition into fine-grained, relatively distal sandy facies. In the 5-km stretch east of Fransfontein town, correlation of diamictite bodies from section to section suggests an overall imbricate structure, younging westward in the line of section (Hoffman 2005). The 2D nature of the homoclinal exposure makes it impossible to determine directly if the imbrication represents lateral accretion, subparallel to the inferred slope contours, or basinward accretion in a highly oblique section. The herein-defined Bethanis Member (Fig. 14.4) forms a continuous blanket, 5–15 m thick, of well-stratified detrital carbonate choked with rafted debris of all sizes. It transgresses the imbricate structure of the rest of the formation. In contrast to the older wellstratified intervals, which coarsen upwards and are followed by diamictite, the Bethanis Member becomes finer grained and more distal upwards in all its major constituents (i.e. debrites, turbidites and plume fallout) except for rafted debris. The latter includes boulders up to 5 m in diameter composed of laterally linked, hemispheroidal, stromatolitic dolostone. This clast type is not observed in the underlying diamictites or rafted debris. Boulder-sized rafted debris continues to the top of the Bethanis Member, where the detrital components give way gradationally, over a stratigraphic interval of only 2– 5 cm, to peloidalmicropeloidal dolostone of the Keilberg Member (Hoffmann & Prave 1996; Hoffman et al. 2007). The isotopic systematics of the Keilberg Member (Hoffman et al. 2007) preclude a detrital origin for the dolomite as postulated by Eyles & Januszczak (2007). The Bethanis Member is dark brown in colour, apparently due to Fe and Mn oxides. Starved ripples generated by westerly directed contour currents occur in some sections. The Bethanis

200

P. F. HOFFMAN

D5 (a)

Distance in kilometres east of district road C35 9.0 10.0 11.0 D7 D8

8.0 D6

12.0

20

13.0 D10

D9

P7010

7.0

Keilberg

P6506

P8011

0.0 D0

(b)

Member

P6514

P7028

P7027

P6508

P7026

P7025

P7021 P7022

Narachaams

P7023

P7014

Franni-aus Member

-40

P7011

P7009

P7013 P4027

Ghaub -20

P7012

0

P7024 P6510

Bethanis

Bethanis

Distance in kilometres east of district road C35 Datum base Bethanis Mb 2.0 3.0 4.0 5.0 6.0 D 2 D 4 D1 D3 Keilberg Member

1.0

20

Bethanis Member 0

-20

-40

P5003 P7003

P7002

P8010

Ghaub Formation

marly dolostone turbidity flows

weakly- and non-stratified limestone-dolomite diamictite

peloidal dolostone, giant wave ripples

weakly- and non-stratified limestone diamictite

dolostone, isopachous sheet-crack cements

weakly- and non-stratified dolomite-chert diamictite

D2 - drainage number east of C35

P8012

P8008

P6536

P8007

P7007

P7006 P7008

Franni-aus Member

Keilberg Member cap dolostone

P7001

-140

E0803

-120

P6538

P7018

-100

E0801

-80

P6537

P6535

P7004

-60

P2093

Ghaub Formation

current-rippled carbonate sandstone, minor ice-rafted debris stratified diamictite (debris flows, contourites, ice-rafted debris)

Fig. 14.4. Facies associations of the Ghaub Formation on the distal foreslope of the Fransfontein Ridge: (a) between the western edge of Duurwater trough (at 7.5 km east) and the western edge of Duurwater moraine (at 13.0 km east) and (b) west of the Duurwater trough.

Member is lithologically distinct from earlier stratified units of the Ghaub Formation and represents the terminal collapse of the ice sheet on the Otavi platform. Duurwater trough and moraine. In the central part of the Fransfontein Ridge (Fig. 14.1b), the erosion surface on which the Ghaub Formation rests cuts out a steep-walled trough, 18 km wide and

0.1 km deep, as defined with respect to the underlying stratigraphy (Fig. 14.5). On the floor of the trough, carbonates of the fallingstand wedge (Franni-aus Member) and terrigenous siltstone of the underlying Narachaams Member are missing, and the Ghaub Formation rests directly on older units of the Abenab Subgroup. The trough is roughly coaxial with a major complex of deepwater carbonate (oolitic) sand bodies (submarine channel and levee

THE OTAVI CARBONATE PLATFORM AND FORESLOPE

0

Palaeodepth in metres below top of platform

–100

Duurwater Trough

WEST N

Otavi

DT Platform

Damara

DM

EAST

crystal fans above cap dolostone

–200 –300

201

Belt

–400 –500 –600 –700

Karibib Formation (includes Keilberg cap dolostone)

Ghaub Formation falling-stand o o o o o o o o o wedge o o Narachaams Member

o

moraine ridge

–800

grainstone prism

–900

crystalline basement

–1000 –1100 –1200 –1300

Chuos diamictite

crystalline basement

Abenab Subgroup slope carbonates –5

0

5

10

15

20 km

Abenab Subgroup

Distance east of district road C35

deposits) in the lower Abenab Subgroup. It straddles a palaeovalley on the crystalline erosion surface lined by basement-derived boulder diamictite and minor iron-formation of the Chuos Formation (Fig. 14.5). Hoffman (2005) attributes the Ghaub-age trough to a transverse, southward-flowing ice stream. It occupied a preexisting drainage system developed long before the Ghaub glaciation. Compared with Quaternary ice-stream troughs, the Duurwater trough is quite modest in size, similar to the Isfjorden trough on the European continental shelf west of Svalbard (Ottesen et al. 2005). Like the Ghaub Formation itself, what makes the Duurwater trough unusual is its location on the distal slope, well below the depth of any known Quaternary ice-stream trough. The Ghaub wedge thins on both sides of the Duurwater trough to about half its average thickness (Fig. 14.5), without change in facies. In the middle of the trough, it swells to a pair of thickness maxima of 288 m (east) and 602 m (west). The maxima are 3 km apart and separated by a 182-m-thick saddle. As the wedge thickens, it loses its well-stratified facies and at both crests consists exclusively of amalgamated, weakly stratified to structureless, boulder diamictites. For want of a datum in the upper Abenab Subgroup beneath the trough (see above), the true palaeo-elevation of the morainal crest above the rims of the trough is uncertain. That it was a major bathymetric feature is reflected by the anomalous nature of the directly overlying units. The Keilberg cap dolostone is less than half its thickness off the moraine, and the normal finegrained micropeloids are winnowed away, leaving lags of unusually coarse-grained peloids (Hoffman et al. 2007, fig. 5b). The directly overlying Maieberg limestone hosts spectacular crystal fans (Hoffman 2005, fig. 11a), pseudomorphous after aragonitic sea-floor cements. These are elsewhere absent from the distal foreslope, but are modestly well-developed on the proximal foreslope (e.g. Hoffman et al. 2007, section S5). The orthogonal aspect ratio of 0.08 for the morainal buildup (7.5 km wide at the base) in the line of exposure implies a transverse orientation assuming a ridge-like form. It is termed a medial moraine (Hoffman 2005) with respect to its location within the Duurwater trough, and its double crest could reflect the fact that true medial moraines consist of paired lateral moraines back-to-back. On the other hand, morainal ridges associated with

Fig. 14.5. Stratigraphic relations in the Abenab Subgroup and Ghaub Formation on the Fransfontein Ridge, reconstructed to show the inferred palaeobathymetry of the Duurwater trough and moraine (see Fig. 14.1b for location) at the end of the Ghaub glaciation (simplified after Hoffman 2005). Inset shows a postulated palaeogeography (see text) in which the Duurwater trough (DT) marks a tributary ice streamlet to a trunk ice stream flowing westward along the northern zone of the Damara Belt. Duurwater moraine (DM) is interpreted as one of a set of terminal moraines (black dots) formed by the retreating trunk ice stream, contemporaneous with deposition of the Bethanis Member.

Quaternary ice streams are mostly lateral moraines (e.g. Isfjorden Svalbard, Ottesen et al. 2005) or terminal moraines oriented at high angles to the direction of ice-stream flow (Ottesen et al. 2005; McMullen et al. 2007).

Boundary relations with overlying and underlying non-glacial units Chuos Formation The Chuos Formation truncates the basement complex and all older Neoproterozoic strata at a low-angle (,1.58) unconformity (Fig. 14.2). Carbonates directly beneath the unconformity are typically shattered and silicified. In contrast, the top of the Chuos Formation is a sharp but conformable contact lacking evidence for substantial hiatus and only locally of significant reworking (Hoffman & Halverson 2008). The top of the Chuos Formation is strongly enriched in hematite, but concentrations along tectonic cleavage planes suggests secondary Fe migration. The basal Rasthof facies is generally a dark grey, parallel-laminated dolomicrite with variable proportions of calcitic, allodapic, centimetric density-flows (Yoshioka et al. 2003; Pruss et al. 2010). It appears to be the deepest-water facies within the Rasthof Formation, meaning that the Rasthof lacks a transgressive facies tract. Hoffman & Schrag (2002) suggested that seawater failed to reach critical oversaturation with respect to CaCO3 or [CaMg]CO3 until after the glacioeustatic rise accompanying Chuos deglaciation. This failure could be related to insufficient availability of exposed carbonate globally, as only carbonate weathering could have supplied sufficient alkalinity on the timescale of glacioeustatic recovery (Higgins & Schrag 2003).

Ghaub Formation Relation to the Franni-aus Member. The Ghaub Formation paraconformably overlies the Franni-aus Member (Fig. 14.3) on the distal foreslope (Hoffman 2005). The contact is a smooth, planar, erosion surface despite the coarse fragmental nature of the underlying

202

P. F. HOFFMAN

debris flows. The Franni-aus Member (Hoffman & Halverson 2008) is a coarsening-upward stack of limestone and dolomite turbidites and debris flows, characterized in the upper part by variably disaggregated slabs and grains of very coarse-grained (,3 mm) oolite, typically heavily silicified. The contact with the Ghaub Formation is marked by an abrupt decrease in grain size, a change in colour from pale grey to medium brown, and the gradual appearance of increasingly numerous and sizeable dropstones, which grade upwards into the first, non-stratified, rain-out diamictite. The ooids are inferred to have originally formed in highly agitated waters at the surf zone and were redeposited gravitationally before complete lithification. As there is no known source for these ooids in the Ombaatjie or any other formation on the platform, it is hypothesized that they originally formed on the proximal foreslope during glacioeustatic regression (Hoffman 2005; Hoffman et al. 2007). Continued base-level fall hastened their redeposition more distally as debris flows. The only strata of equivalent age on the platform are ,30 m of aeolian(?) dolarenite sandwiched between the youngest marine cycle of the Ombaatjie Formation and the Keilberg Member close to the outer edge of the platform in the Upper Huab area (Fig. 14.1b) (Soffer 1998; Halverson et al. 2002). The Franni-aus Member is cut out by the Duurwater trough, where the Ghaub Formation rests directly on siliciclastic siltstone of the Narachaams Member or underlying slope carbonates of the lower Abenab Subgroup (Fig. 14.5). Reworked siltstone at the base of the Ghaub Formation outside the trough is most likely derived from the Narachaams Member, implying that the Duurwater trough or others like it had been eroded before the Ghaub Formation was deposited (Hoffman 2005). If the Duurwater trough was cut by an ice stream, glacial maximum conditions are not represented in the Ghaub Formation. Relation to the Ombaatjie Formation on the platform. Lenticles of Ghaub Formation overlie a continuous paraconformity on the platform having ,80 m of stratigraphic relief with respect to the Ombaatjie Formation (Halverson et al. 2002; Hoffman et al. 2007). The erosion surface must be the combined product of karstification during the initial glacioeustatic fall and subsequent erosion by the ice sheet after it developed. It can be reconstructed in detail using the 0-per-mil intersection of the Trezona carbon-isotope anomaly as a datum (Halverson et al. 2002; Hoffman et al. 2007). At the end of the glaciation, the platform had a raised rim and a bowl-shaped interior with local highs. The Keilberg cap dolostone consistently swells over positive-relief features on the platform (Hoffman et al. 2007) and sea-floor cements (pseudomorphosed aragonite crystal fans) are localized above them in the overlying Maieberg limestone rhythmite. Relation to the Keilberg Member. The Ghaub Formation is everywhere overlain conformably and without significant hiatus by the Keilberg Member (Hoffman et al. 2007). Well-stratified proglacial deposits choked with dropstones (Bethanis Member) precede the Keilberg except on the Duurwater moraine, where the Bethanis Member is absent and the Keilberg Member rests directly on non-stratified diamictite. The base of the Keilberg is generally a laminated peloidal-micropeloidal dolostone with lowangle cross-stratification. In distal-slope sections, however, the cross-stratified facies is commonly preceded by ,1.0 m of deeperwater dolomitic turbidites (Hoffman et al. 2007; Hoffman & Macdonald 2010).

Chemostratigraphy The Otavi Group is ideally suited for chemostratigraphy on account of its carbonate-dominated lithology, ocean margin setting and estimated duration of c. 190 Ma (Halverson et al. 2005).

Carbon isotopes Pre-glacial negative anomalies. A composite d13Ccarb curve for the

Otavi platform is given in Figure 14.2. The major features of the curve have been replicated in numerous sections and compare favourably with correlative sections outside Namibia (Halverson et al. 2005; Halverson 2006; Nogueira et al. 2007). Both glaciations were preceded by precipitous declines in d13C, following persistent (.500 m) intervals of strong 13C enrichment (d13C ¼ 4 –8‰ PDB). The decline before the Ghaub glaciation is deeper, from þ7.5‰ to –5.0‰ PDB; it represents the ‘Trezona anomaly’ (Halverson et al. 2005; Halverson 2006) observed below the Elatina, Stelfox (Ice Brook) and Smalfjord glaciations in Australia, Canada and Norway, respectively. The older decline is observed at the top of the Ugab Subgroup, south of the Huab palaeoridge, and its nadir of only 0‰ is probably truncated by sub-Chuos erosion. It possibly represents the here-named ‘Islay anomaly’, observed to precede the Port Askaig, Petrovbreen and Julius River glaciations of Scotland, East Svalbard and Tasmania, respectively (Calver 1998; Brasier & Shields 2000; Halverson et al. 2005; Halverson 2006; McCay et al. 2006; Prave et al. 2009). Post-glacial negative anomalies. Negative d13C anomalies are

associated with the post-glacial Rasthof and Maieberg formations (Fig. 14.2), and differences between them are common to the pairs of Cryogenian glaciations in Australia and Canada (Kennedy et al. 1998; Hoffman & Schrag 2002). The anomaly in the Rasthof Formation has been documented in fine detail by Yoshioka et al. (2003) and their findings are valid over most of the Otavi platform (Pruss et al. 2010). The negative anomaly is limited to the lower (abiotic rhythmite) member, which is 2–70 m thick regionally and 14 m in its detailed section. Its top and bottom are dolomitic but the middle 10 m is calcitic. The basal dolomite rhythmite rises asymptotically in d13C from –4.0‰ towards –2.0‰, the middle calcite rhythmite rises gradually by an additional 1‰, and the upper dolomite rhythmite rises from –1.0‰ asymptotically to þ2‰. The middle (microbialaminite) and upper (grainstone) members of the Rasthof Formation have d13C values between þ4 and þ6‰. The negative d13C anomaly following the Ghaub glaciation extends through the full ,400-m thickness of the Maieberg Formation (Fig. 14.2). Isotopic profiles of the Keilberg cap dolostone differ significantly according to palaeogeography because of diachroneity associated with glacioeustatic transgression: shallowwater facies in distal slope sections are entirely older than the same facies on the platform (Hoffman et al. 2007). Distal slope sections begin at –0.5‰ and fall asymptotically towards –3.0‰; platform sections begin near –3.0‰ and rise imperceptibly to –2.5‰ over most of their thickness, before falling steeply to –4.5‰ at the dolostone-limestone transition. Proximal slope sections occupy the middle part of the overall sigmoidal trend. If the timescale of the global meltdown was on the order of 104 years (Peltier et al. 2004), severe constraints are placed on the origin of the secular d13C change of 4‰ (Hoffman et al. 2007). Independent of the secular changes, the absolute d13C values vary systematically across the platform. Sections at the edge of the platform and on the proximal slope are 1.0‰ and 1.5‰ enriched in 13C, respectively, compared with sections in the platform interior. It is implausible, due to rapid mixing, that this lateral isotopic gradient could be maintained in ocean surface waters but, if the dissolved inorganic carbon (DIC) pool was dominated by CO2 (Snowball Earth hypothesis), the gradient in d13Ccarb could reflect a lateral temperature gradient resulting from strong surface warming over the platform and upwelling of colder waters at the slope (Hoffman et al. 2007). The isotopic gradient would result from temperature-dependent isotopic fractionation of CO23 – ions with respect to aqueous and atmospheric CO2. Thus, the lateral gradient in d13C reflects spatial variation in surface temperature, and the secular changes in d13C in all areas

THE OTAVI CARBONATE PLATFORM AND FORESLOPE

reflect progressive warming over time as global ice cover recedes. Importantly, this interpretation is contingent upon low-pH (,7.2) conditions, unique to a snowball aftermath, in which CO2 is the dominant carbon-bearing species. The observed consistent spatial and temporal variations in d13C (Hoffman et al. 2007) demonstrate that the Keilberg cap dolostone does not consist of detrital dolomite (Eyles & Januszczak 2007). Early diagenetic breccias and cements like those in presumed correlative cap dolostones in South China and West Africa (Jiang et al. 2003, 2006; Shields et al. 2007) are largely absent from the Keilberg Member (Hoffman et al. 2007). The lower limestone rhythmites of the Maieberg Formation have rather uniform d13C values of –6.0 to –6.5‰ (Fig. 14.2). The drop of c. 0.75‰ (and 2– 3‰ in d18O) at the dolostone – limestone transition is a step function apparently related to mineralogy. The combined d13C and d18O changes are consistent with low-temperature equilibrium fractionation between coexisting dolomite and calcite (Friedman & O’Neil 1977; Vasconcelos et al. 2005). Reverse steps occur at the limestone-dolostone transition in the upper Maieberg Formation. The Maieberg is the only formation in the Otavi Group (including the Rasthof Formation) exhibiting a systematic isotope fractionation between stratigraphically associated calcite and dolomite (Hoffman & Halverson 2008). The rise in d13C with stratigraphic height through the middle and upper Maieberg Formation varies laterally in its trajectory, rising less near the edge of the platform and more in the platform interior. This may reflect faster sedimentation rates on the outer platform and diachronous shoaling from the edge of the platform toward the interior. The ‘isochems’ in Figure 14.3a denote schematically the surfaces of uniform d13C composition.

Strontium isotopes 87

Sr/86Sr ratios have been measured in limestones of the Otavi Group deposited before, between and after the Cryogenian glaciations. Because Sr is a mobile trace element in carbonate rocks, samples with low Sr concentrations (,800 ppm), high Rb –Sr ratios, low d18O values and high 87Sr/86Sr ratios are screened out. No data from dolostones survive screening because of their low Sr concentrations. According to these criteria, the best limestones from the Ombombo Subgroup have 87Sr/86Sr ratios c. 0.7065 (P. F. Hoffman & A. J. Kaufman, unpublished data), the Rasthof cap carbonate c. 0.7068 (Yoshioka et al. 2003), the lower Ombaatjie Formation c. 0.7073 (Halverson et al. 2007) and the Maieberg cap carbonate c. 0.7072 (Halverson et al. 2007). These data are broadly consistent with results from both western and eastern Laurentia (Brasier & Shields 2000; Halverson et al. 2007) and support the correlation of the pairs of glaciogenic formations in each region.

Sulphur isotopes (CAS) Secular variations in the concentration and sulphur isotopic composition (d34S) of carbonate-associated sulphate (CAS) have been profiled in sections of the Abenab Subgroup and the Maieberg cap-carbonate sequence (Hurtgen et al. 2002, 2006). CAS concentrations are mostly ,200 ppm, lower than Phanerozoic carbonates, and this combined with large (.20‰) and apparently rapid isotopic fluctuations suggests that seawater sulphate concentrations were lower than in Phanerozoic oceans (Hurtgen et al. 2002). Large, complex, but broadly similar d34S anomalies are associated with the Rasthof and Maieberg cap-carbonate sequences. Their lower and upper parts are isotopically enriched to 40 and 50‰ (CDT), respectively, while their middle part, corresponding closely to the deepest-water facies on the platform, are isotopically depleted to c. 15‰. In the Maieberg Formation, the steepest

203

isotopic change straddles the dolostone – limestone transition, which also coincides with the disappearance of wave-generated bedforms and is interpreted as the incursion due to flooding of anoxic sulphidic waters onto the platform, accompanied by the oxidation of sulphide to sulphate at the oxic – anoxic interface (Hurtgen et al. 2006).

Reactive Fe and Mn concentrations Hurtgen et al. (2006) report reactive Fe and Mn concentrations for three, widely spaced, closely sampled sections of the Maieberg cap-carbonate sequence. Fe and Mn concentrations from acidsoluble (mainly Fe2þ and Mn2þ carbonates) and diothioniteextractable (mainly oxides and oxyhydroxides) sources were measured, but not sulphide-bound Fe. There is excellent agreement between the sections. All concentrations rise from very low values in the lower half of the Keilberg cap dolostone towards high values in the middle Maieberg limestone member, before falling in the upper Maieberg dolomitic grainstone. A sharp peak in acid-soluble Fe (and Ba) concentration occurs at the cap dolostone – limestone transition. Combined with d13C, d18O, d34Spyrite and d34SCAS data from the same samples, Hurtgen et al. (2006) propose a model involving progressive mixing between cold, euxinic (i.e. anoxicsulphidic), deep water and a lid of warm, oxic, brackish water, during and after deglaciation (see also Shields 2005).

Oxygen isotopes Oxygen isotopes in carbonate rocks are susceptible to alteration by fluid flows under diagenetic and metamorphic conditions. d18O values for relatively unaltered Otavi Group carbonates fall mostly in the range of – 4 to –6‰ VPDB (Hurtgen et al. 2006, unpublished data). Of the three formations containing both limestone and dolostone, including fabric-retentive dolostone, one (Maieberg Formation) displays consistent isotopic fractionations of both d18O and d13C between the two phases, whereas the others (Rasthof and Ombaatjie formations) show no pattern whatsoever on a cross-plot of d13C v. d18O (Hoffman & Halverson 2008). In the Maieberg Formation (including its Keilberg Member), limestone is isotopically depleted by c. 1.7‰ in d13C and c. 3.8‰ in d18O, relative to dolostone. This fractionation remains locked despite the large secular change (5‰ in d13C) the formation records. This analysis includes the upper dolostone grainstone member as well as the Keilberg Member. The Maieberg data are consistent with low-temperature (c. 50 8C) isotopic equilibrium fractionation between dolomite and calcite (Friedman & O’Neil 1977), and by implication that the dolomite formed in contact with seawater. If one ‘corrects’ the limestone values to bring them in line with adjacent dolomites (e.g. add 3.8‰ in d18O), a well-defined negative d18O anomaly of 1 to 3‰ is evident in the first 130 m of the Maieberg Formation (limestone from 13 to 330 m). This is possibly an isotopic record of the meltwater plume attending deglaciation, before it was mixed with salty glacial deepwater (Hoffman 1999; Shields 2005).

Boron and calcium isotopes Kasemann et al. (2005, 2010) report d11B and d44Ca measurements for carbonates of the Ombaatjie and Maieberg formations as proxies for pre- and post-glacial seawater. There is a large (c. 20‰) isotopic fractionation between borate ion (which is incorporated into carbonate) and boric acid in seawater, and the relative proportions of these species change from nearly 0 to almost 1.0 over the pH range of 7.5– 9.5. Changes in seawater d11B over geological time (i.e. millions of years) may also reflect changes in

204

P. F. HOFFMAN

ocean inputs and outputs of boron. Because continental weathering tends to lower seawater d11B and mid-ocean ridge hydrothermal exchange to raise it, we may suppose that seawater should become relatively more enriched if continental weathering was held down for millions of years under snowball conditions. Instead, post-glacial d11B values are more depleted, reaching a nadir at the top of the Keilberg cap dolostone and in the directly overlying Maieberg limestones. Using their d44Ca data to correct for changes in Ca (and indirectly boron) ocean input and output, Kasemann et al. (2005) calculate changes in the maximum levels of atmospheric pCO2 over time. Their results allow for a rise in pCO2 of up to 0.1 atm (100 000 ppm) in the glacial aftermath, broadly consistent with the Snowball Earth hypothesis.

Palaeolatitudes and palaeogeography Neoproterozoic sedimentary rocks in Namibia were extensively and thoroughly remagnetized during the orogenic amalgamation of Gondwanaland in the Cambrian. Palaeopoles from mafic members of the post-tectonic Mbozi igneous complex in Tanzania (Meert et al. 1995), felsic components of which give K –Ar cooling ages of 755 + 25 Ma (recalculated from Cahen & Snelling 1966; Evans 2000), suggest a palaeolatitude of 10 + 58 for the Otavi platform c. 40 Ma before the Chuos glaciation, assuming a rigid Congo craton (Evans 2000). Dual-polarity palaeopoles from the Nola dyke swarm (Central African Republic) dating from its metamorphism and remagnetization at 571 + 9 Ma (Moloto-AKenguemba et al. 2008) place the Otavi platform near the palaeoequator (04 + 58) c. 65 Ma after the Ghaub glaciation. Preliminary palaeomagnetic data from the basal part of the Keilberg cap dolostone place the Otavi platform in the (southern) subtropics with its present southern margin facing the palaeoequator (R.I.P. Trindade, pers. comm.). These fragmentary results are at least consistent with more robust data sets indicating that other Neoproterozoic carbonate-dominated successions formed at palaeolatitudes ,358 (e.g. Trindade et al. 2003; Macouin et al. 2004; Kilner et al. 2005; Maloof et al. 2006), and therefore that the meridional climatic gradient was not reversed as proposed in the large orbital obliquity hypothesis for low-latitude glaciation (Williams 1993; see also Evans 2006).

Geochronological constraints Zircons were extracted from tuffaceous layers within the Chuos Formation, from strata directly preceding the Ghaub glaciation on the platform and slope, and from both cap carbonates. No primary volcanogenic zircons were recovered. The glacial history of the Otavi platform is currently constrained by just three U –Pb zircon ages. The first provides a maximum constraint for the Chuos Formation on the platform. An age of 760 + 1 Ma (Halverson et al. 2005) was obtained for a tuff near the top of the Devede Formation of the Ombombo Subgroup (Fig. 14.2), nearly 350 m stratigraphically below the Chuos at that locality. The age is close to 756 + 2 Ma for the Oas quartz-syenite pluton (Hoffman et al. 1996), representing a suite of peralkaline igneous rocks associated with crustal stretching. The younger Naauwpoort bimodal (basalt-rhyolite) volcanics provide a tighter maximum constraint on Chuos glaciation in the Summas Mountains area (Fig. 14.1a). There, rhyolite lava that is unconformably overlain by (tectonically overturned) Chuos diamictite has an age of 747 + 2 Ma (Hoffman et al. 1996). An ash-flow tuff with an indistinguishable age of 746 + 2 Ma in the same area is separated from the Chuos by 720 m of shallow-water mixed carbonates and clastics of the Ugab Subgroup (Hoffman et al. 1996; Hoffman & Halverson 2008). Accordingly, the onset of the Chuos glaciation must be substantially younger than 746 Ma.

No constraints on the Ghaub glaciation have been obtained from the Otavi platform, but a tuff layer within carbonates of the central Damara Belt carrying ice-rafted debris, believed correlative with the Ghaub Formation, gives an age of 635.5 + 0.5 Ma (Hoffmann et al. 2004; Condon et al. 2005). This is indistinguishable from the age of 635.2 + 0.6 Ma for a tuff lying upon the postglacial (Nantuo diamictite) cap dolostone in South China (Condon et al. 2005). The Ghaub glaciation is therefore inferred to have ended in 635 Ma.

Discussion Tectonic/palaeogeographical setting and origin of the Ghaub Formation Eyles & Januszczak (2007; see also Eyles 2004; Eyles & Januszczak 2004a, b) examined the Ghaub Formation in two sections, near the western terminus of the Fransfontein Ridge (Narachaams) and at the Fransfontein drainage gap (Fig. 14.1b). They conclude that it consists of subaqueous mass-flow deposits of non-glacial origin, coincident with rifting and breakup of the Congo craton. Accordingly, the Ghaub Formation has no bearing on the Neoproterozoic glacial record and provides no support for the Snowball Earth hypothesis (Eyles & Januszczak 2007). Their argument boils down to four interconnected points. First, the Ghaub Formation lacks key glacial indicators such as faceted, striated and bullet-shaped clasts. Second, the pre-, syn- and postGhaub facies are indistinguishable, meaning that the identification of a glaciogenic interval is entirely arbitrary. Third, the Ghaub diamictites occur strictly in base-of-slope palaeoenvironments, consistent with their origin as subaqueous mass-flows. Fourth, they were deposited contemporaneously with rift faulting and continental break-up at the edge of the Congo craton, suggesting that active faulting was the trigger for mass wasting. These conclusions are so at variance with those favoured by other authors (e.g. Condon et al. 2002; Domack & Hoffman 2011) that a point-by-point discussion is warranted. Diagnostic glacial indicators. In the type area of the Ghaub diamic-

tite, le Roex (1941) reported that ‘a high percentage of the pebbles show typical glacial faceting, particularly the hard quartzitic types’. Some also show striations, ‘which are much better preserved on the quartzitic than the calcareous types’ (le Roex 1941). Quartzitic pebbles do not occur in the Ghaub Formation on the Fransfontein foreslope, only carbonate clasts (along with rare granitic detritus at Bethanis). The problem is that the carbonate clasts and the carbonate matrix were thoroughly welded together during middle greenschist facies metamorphism. Facets and striae might exist on every clast, or on none: it would be impossible to tell the difference. In the absence of clast morphology, dropstones are the most diagnostic glacial feature (along with the overall facies associations and stratal architecture). Classic attributes of dropstones are well-developed in the Ghaub Formation, including pinched and punctured substrata, thickened and ejected side-strata, draped or onlapping superstrata, and independence from bottom sediment transport events. Ejection folds are directly analogous to the ‘overturned flap’ on the rim of Meteor Crater, Arizona (Shoemaker 1963). Eyles & Januszczak (2004a, 2007) emphasize that dropstones also occur in non-glacial settings, but the frequency of their occurrence in the Ghaub Formation is unmatched by any nonglacial deposit. With a conservative average of one dropstone per 1000 ppm3, the Bethanis Member alone held 50 trillion (5  1013) dropstones given its original dimensions of 0.1  5  100 km3. Pre-, syn- and post-glacial facies. The lower and upper contacts of

the Ghaub Formation can be unambiguously mapped on the basis of lithology with an accuracy of centimetres along the

THE OTAVI CARBONATE PLATFORM AND FORESLOPE

length of the Fransfontein foreslope. The same is true of the outliers to the west (Bethanis-Toekoms and Vrede-Opdraend). The carbonate turbidite-hosted debris flows of the underlying Franni-aus Member and overlying Maieberg Formation are oligomictic and intraformational; those of the Ghaub Formation are polymictic and extraformational. Weakly stratified to structureless diamictites make up 82% of the Ghaub Formation (based on 72 measured sections) but are uncommon or absent in the adjacent formations. Similarly, dropstones are profligate in the Ghaub but rare or non-existent in the Franni-aus and Maieberg. In short, the Ghaub is a true formation, lithologically mappable on the regional scale. How could Eyles & Januszczak (2007) conclude that the same essential features occur within and below the Ghaub Formation? The answer is the following. At Narachaams (see their figs 4 and 5), they place the base of the Ghaub Formation at the base of the Bethanis Member, which constitutes only the top 2 m of the 54-m-thick Ghaub Formation in that section. At Fransfontein (their figs 3 and 5), everything they call Ghaub Formation lies below the Bethanis Member. In other words, their ‘Ghaub Formation’ at Fransfontein is stratigraphically equivalent to their ‘Abenab Subgroup’ at Narachaams. Small wonder their ‘pre-Ghaub’ and ‘syn-Ghaub’ facies are similar; they are one and the same. Palaeoenvironmental setting of the Ghaub Formation. Eyles &

Januszczak (2007) concur with the interpretation of the distal Fransfontein foreslope (Fig. 14.1b) as a deep-water marine environment (Henry et al. 1990; Halverson et al. 2002, 2005; Hoffman 2005). Their interpretation of the Ghaub Formation as a stack of submarine mass flows is consistent with this setting, but it does not account for sedimentary structures (low-angle crosslamination and giant wave ripples) in the overlying Keilberg cap dolostone, which demonstrate that the distal foreslope was above storm wave-base at the glacial – deglacial transition (Hoffman et al. 2007). The diachronous nature and extent of the cap dolostone (same-depth deposits on the platform and distal foreslope) support a large-amplitude glacioeustatic fluctuation, as a result of which the ice grounding-line advanced to the distal foreslope before retreating at the glacial termination (Hoffman et al. 2007). The facies associations and stratal architecture of the Ghaub Formation have much in common with ice grounding-line wedges of Quaternary age on the continental shelves and upper slopes of polar seas (e.g. Alley et al. 1989; Boulton 1990; Powell 1990; King et al. 1991; King 1993; Powell & Domack 1995). The wellstratified facies association represents proglacial marine deposits. The sandy facies association resembles ice grounding-line fans (Powell 1990). Weakly-stratified to structureless diamictites represent ice-proximal and ice-contact deposits. Diamictites that grade downwards into stratified proglacial deposits probably accumulated as rain-out deposits immediately seaward of the ice grounding line. Ice-contact deposits include diamictites with erosive bases, internal reactivation surfaces and discontinuous stringers of finely laminated siltstone. The silt-stringers are interpreted as ice-bed separation cracks beneath stagnant grounded ice. They were susceptible to deformation associated with recurrent ice flowage. Eyles & Januszczak (2007) fail to mention the widespread relics of Ghaub diamictite on the Otavi platform, including the 60 m of diamictite with granitic and quartzitic clasts in the Otavi Mountains type section (le Roex 1941; Hoffmann & Prave 1996). As these diamictites are underlain and overlain by shallowwater carbonates (Ombaatjie Formation and Keilberg Member, respectively), it is improbable that they originated as submarine mass-flows. Role of rift faulting. Le Roex (1941) argued that the quartzitic and

granitic clasts in his diamictite could not be locally derived (e.g. through faulting) because the diamictite is separated from quartzite

205

and granitic basement in the Otavi Mountains by hundreds of metres of shallow-water (stromatolitic and oolitic) carbonate strata forming a conformable stratigraphic succession. He concluded that the clasts must have been transported from distant sources, where quartzite and granitic basement were exposed in Ghaub time. Halverson et al. (2002) greatly strengthened le Roex’s (1941) argument with detailed isotopic chemostratigraphy of the upper Ombaatjie Formation. They documented a steep decline in d13C from þ5‰ or higher down to – 5‰ in the final pair of depositional cycles (parasequences) at 17 locations distributed from the edge of the platform far into its interior. This isotopic shift has been correlated globally and is named the Trezona anomaly after the formation in South Australia where it was first encountered (Halverson et al. 2002, 2005). Plotting all the sections using the interpolated 0‰ intersection as the datum provides a quantitative basis for reconstructing the palaeotopography of the erosion surface on the platform beneath the Ghaub diamictite, or the Keilberg Member where the Ghaub is absent (Halverson et al. 2002; Hoffman et al. 2007). There is up to 80 m of local relief on the erosion surface but no evidence of uplift or back-rotation of the outer platform. If a basinward-dipping normal fault was active at the edge of the platform (Eyles & Januszczak 2007), uplift and back-rotation would occur as an isostatic response to tectonic unloading of the footwall. The resulting unconformities as well as cannibalistic clastic deposits on the back-rotated dip-slopes are precisely the criteria successfully used to determine the location and timing of rift faulting in the lower Otavi Group, prior to the Ombaatjie Formation (Soffer 1998; Hoffman 1999; Halverson et al. 2002; Hoffman & Halverson 2008). However, there is no evidence of such activity at the time of the Ghaub Formation. The edge of the platform preserves more upper Ombaatjie strata than does the interior, not less, and over the platform as a whole the upper Ombaatjie cycles and the Keilberg Member are parallel. This is a very sensitive test because even a small angular rotation of a rift ‘shoulder’ of modest dimensions would result in significant stratigraphic truncation in a shallow marine environment. None is observed. Before Eyles & Januszczak (2004a, 2007) began their study, the history of rift faulting on the Otavi platform was known to have ended before the Ombaatjie Formation was deposited, millions of years before the Ghaub glaciation (Hoffman et al. 1998a; Halverson et al. 2002; Hoffman 2002).

Significance of ice-rafted dropstones for maximum ice-shelf extent The presence of ice-rafted dropstones in the Ghaub Formation on the Fransfontein Ridge was taken as evidence for the existence of free-floating icebergs (Condon et al. 2002), which would be inconsistent with a globe-encircling ice shelf as predicted by the Snowball Earth hypothesis (Hoffman & Schrag 2002; Warren et al. 2002; Goodman & Pierrehumbert 2003). Two issues are raised by this interpretation. Were the dropstones carried by icebergs or were they released by basal melting of ice-shelf ice carrying englacial debris across the grounding line? Free-floating icebergs transport debris farther from the grounding line than ice-shelf ice. Iceberg transport is favoured for the Bethanis Member, but the high proportion (88%) of weakly stratified and structureless diamicitites in the rest of the Ghaub Formation supports deposition close to the grounding line. Thus, in the absence of iceberg ‘dumps’ (Condon et al. 2002), the origin of dropstones below the Bethanis Member is ambiguous with respect to the continuity of ocean ice cover. The second issue is whether the Ghaub Formation represents the maximum glacial conditions, or only the waning stages of glaciation. If the smooth, planar and laterally continuous surface at the base of the Ghaub Formation on the Fransfontein Ridge represents

206

P. F. HOFFMAN

a subglacial erosion surface, then the overlying glacial deposits do not represent the first, or even necessarily the maximum, glacial stage. Likewise the Duurwater trough, if it was cut by an ice stream, must represent a glacial maximum that is unrepresented by deposits in the Ghaub Formation itself (Hoffman 2005). Accordingly, any evidence for open water within the Ghaub Formation on the Fransfontein Ridge is ambiguous with regards to the maximum extent of the ice. A complete glacial record will only be found seaward of the maximum ice grounding line.

Significance of grounding-line oscillations Condon et al. (2002) also suggested that the evidence in the Ghaub Formation for grounding-line oscillations was inconsistent with an ice-covered ocean. The logic behind this assertion is unclear: changes in the mass-balance of an ice sheet (driven, for example, by orbital forcing) or, alternatively, episodicity in ice-sheet dynamics (e.g. the binge – purge cycles of MacAyeal 1993) might just as well occur if the hydrological cycle was driven by the sublimation of sea ice as by the evaporation of seawater. Relative to grounding-line oscillations on Quaternary continental shelves, however, the oscillations recorded in the Ghaub Formation appear to have been more limited in scale, possibly because of the inclination of the Otavi foreslope (Pollard & deConto 2007; but see also Alley et al. 2007).

Magnitude of base-level and glacioeustatic changes The existence of grounded ice on the distal foreslope at an inferred palaeodepth of c. 0.5 km below the rim of the platform, combined with shallow-water structures (low-angle cross-strata and giant wave ripples) in the directly overlying cap dolostone and post-glacial flooding of the platform below storm wave-base in the succeeding limestone (Fig. 14.3a) imply a base-level change of .0.5 km at the glacial termination. The glacioeustatic change must have been far greater than this because of the isostatic rebound of the platform due to the removal of the ice sheet, the hydroisostatic sinking of the sea floor (and therefore of sea level) due to the addition of glacial meltwater, and the additional lowering of sea level in the vicinity of a retreating ice sheet due to the weakening of its gravitational attraction on the adjacent ocean (Clark 1976; Farrell & Clark 1976). The time scale over which the platform was flooded, given the thickness of the deepwater deposits (Fig. 14.3a), must have far exceeded the timescale of the isostatic adjustments. If the thickness of the ice sheet near the edge of the Otavi platform was c. 1.0 km, comparable to the edge of the East Antarctic ice sheet today (Lythe et al. 2001), then the isostatic rebound would have been c. 0.27 km (rice/rmantle). If the global average thickness of ice sheets on all continents was .2.5 km, as suggested by climate models of a Snowball Earth (Donnadieu et al. 2003; Pollard & Kasting 2004), then .1.25 km of meltwater would have been added to the ocean because its area is roughly twice that of the continents. The resulting hydroisostatic sinking of the seafloor would have been .0.38 km (1.25/rmantle). The additional lowering of sea level in the vicinity of the former ice sheet could be as great as c. 0.1 km (Farrell & Clark 1976). Accordingly, a glacioeustatic change of .1.25 km is fully compatible with the inferred base-level change of .0.5 km (i.e. 1.25–0.27–0.38 –0.1 ¼ 0.5 km).

Significance of the associated carbonates The ‘reverse’ solubility of calcium carbonate accounts for the preferential accumulation of carbonate sediments in the warmer parts of the surface ocean. Carbonate dominates shallow-water

sediments forming today only in latitudes ,358 (Ko¨ppen & Wegener 1924, p. 19; Rodgers 1957). The zonal distribution of carbonate sediments did not change between the warm (non-glacial) and cool (glacial – interglacial) periods of the Phanerozoic (Blackett 1961; Briden & Irving 1964; Briden 1970; Ziegler et al. 1984; Opdyke et al. 1990; Witzke 1990; Kiessling 2001). This is because carbonate production and preservation reflect relative, not absolute temperatures. Moreover, it implies that the presence or absence of polar ice sheets caused no perceptible change in the flux of alkalinity entering the ocean. It is therefore a reasonable inference that the Otavi carbonate platform occupied one of the warmer parts of the global ocean before, between and after the Cryogenian glaciations. If one accepts that an ice sheet on the Otavi platform had a continuous tidewater grounding line in one of the warmer parts of the Cryogenian ocean, then contemporaneous ice sheets must have existed at higher elevations and palaeolatitudes, wherever there was net meteoric precipitation or glacial influx. Research on the Otavi Group was initially supported by the Geological Survey of Namibia and the Canadian National Science and Engineering Research Council (NSERC). Since 1994, support was generously provided by the United States National Science Foundation (NSF), NASA Astrobiology Institute, Harvard University Center for the Environment (HUCE), and the Canadian Institute for Advanced Research (CIFAR). Among the many who have contributed to this work, I am particularly indebted to S. A. Bowring, E. W. Domack, G. P. Halverson, J. A. Higgins, M. T. Hurtgen, A. J. Kaufman, F. A. Macdonald, A. C. Maloof, D. P. Schrag, G. Soffer and R. Trindade. Roy McG. Miller gave valuable advice on stratigraphic names. G. Jiang is thanked for his perceptive comments on the manuscript. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Adams, E. W. & Schlager, W. 2000. Basic types of submarine slope curvature. Journal of Sedimentary Research, 70, 814– 828. Alley, R. B., Blankenship, D. D., Rooney, S. T. & Bentley, C. R. 1989. Sedimentation beneath ice shelves– the view from ice stream B. Marine Geology, 85, 101– 120. Alley, R. B., Anandakrishnan, S., Dupont, T. K., Parizek, B. R. & Pollard, D. 2007. Effect of sedimentation on ice-sheet groundingline stability. Science, 315, 1838–1841. Bechsta¨dt, T., Ja¨ger, H., Spence, G. & Werner, G. 2009. Late Cryogenian (Neoproterozoic) glacial and post-glacial successions at the southern margin of the Congo Craton, northern Namibia: facies, paleogeography and hydrocarbon perspective. In: Craig, J., Thurow, J., Thusu, B., Whitham, A. & Abutarruma, (eds) Global Neoproterozoic Petroleum Systems: The Emerging Potential in North Africa. Geological Society, London, Special Publications, 326, 255– 287. Blackett, P. M. S. 1961. Comparison of ancient climates with the ancient latitudes deduced from rock magnetic measurements. Proceedings of the Royal Society of London, Series A, 263, 1 –30. Boulton, G. S. 1990. Sedimentary and sea-level changes during glacial cycles and their control on glacimarine facies architecture. In: Dowdeswell, J. A. & Scourse, J. D. (eds) Glacimarine Environments: Processes, Sediments. Geological Society of London, Special Publications, 53, 15 –52. Brasier, M. D. & Shields, G. 2000. Neoproterozoic chemostratigraphy and correlation of the Port Askaig glaciation, Dalradian Supergroup of Scotland. Journal of the Geological Society, London, 157, 909–914. Briden, J. C. 1970. Palaeolatitude distribution of precipitated sediments. In: Runcorn, S. K. (ed.) Palaeogeophysics. Academic Press, London, 437– 444. Briden, J. C. & Irving, A. 1964. Palaeolatitude spectra of sedimentary palaeoclimatic indicators. In: Nairn, A. E. M. (ed.) Problems in Palaeoclimatology. John Wiley & Sons, New York, 199–224. Cahen, L. & Snelling, N. J. 1966. The Geochronology of Equatorial Africa. North-Holland Publishing Co., Amsterdam.

THE OTAVI CARBONATE PLATFORM AND FORESLOPE

Calver, C. R. 1998. Isotope stratigraphy of the Neoproterozoic Togari Group, Tasmania. Australian Journal of Earth Sciences, 45, 865– 874. Clark, J. A. 1976. Greenland’s rapid postglacial emergence: A result of ice-water gravitational attraction. Geology, 4, 310– 312. Clifford, T. N. 2008. The geology of the Neoproterozoic Swakob-Otavi transition zone in the Outjo District, northern Damara Orogen, Namibia. South African Journal of Geology, 111, 117– 140, 3 maps. Condon, D. J., Prave, A. R. & Benn, D. I. 2002 Neoproterozoic glacial rainout intervals: observations and implications. Geology, 30, 35 – 38. Condon, D., Zhu, M., Bowring, S. A., Wang, W., Yang, A. & Jin, Y. 2005. U– Pb ages from the Neoproterozoic Doushantuo Formation, China. Science, 308, 95– 98. Dingle, R. V. 1977. The anatomy of a large submarine slump on a sheared continental margin (SE Africa). Journal of the Geological Society (London), 134, 293 –310. Domack, E. W. & Hoffman, P. F. 2011. An ice grounding-line wedge from the Ghaub glaciation (635 Ma) on the distal foreslope of the Otavi platform, Namibia, and its bearing on the Snowball Earth hypothesis. Geological Society of America Bulletin, 123, 1448– 1477. Donnadieu, Y., Fluteau, F., Ramstein, G., Ritz, C. & Besse, J. 2003. Is there a conflict between the Neoproterozoic glacial deposits and the snowball Earth interpretation: an improved understanding with numerical modeling. Earth and Planetary Science Letters, 208, 101– 112. Eberli, G. P. & Ginsburg, R. N. 1987. Segmentation and coalescence of Cenozoic carbonate platforms, northwestern Great Bahama Bank. Geology, 15, 75 –79. Evans, D. A. D. 2000. Stratigraphic, geochronological, and paleomagnetic constraints upon the Neoproterozoic climatic paradox. American Journal of Science, 300, 347–433. Evans, D. A. D. 2006. Proterozoic low orbital obliquity and axial-dipolar geomagnetic field from evaporite palaeolatitudes. Nature, 444, 51 – 55. Eyles, N. 2004. Frozen in time: concepts of ‘global glaciation’ from 1837 (die Eiszeit) to 1998 (the Snowball Earth). Geoscience Canada, 31, 157– 166. Eyles, N. & Januszczak, N. 2004a. ‘Zipper-rift’: a tectonic model for Neoproterozoic glaciations during the breakup of Rodinia after 750 Ma. Earth-Science Reviews, 65, 1– 73. Eyles, N. & Januszczak, N. 2004b. Interpreting the Neoproterozoic glacial record: the importance of tectonics. In: Jenkins, G. S., McMenamin, M. A. S., McKey, C. P. & Sohl, L. (eds) The Extreme Proterozoic: Geology, Geochemistry, and Climate. Geophysical Monograph 146. American Geophysical Union, Washington, DC, 125–144. Eyles, N. & Januszczak, N. 2007. Syntectonic subaqueous mass flows of the Neoproterozoic Otavi Group, Namibia: where is the evidence of global glaciation? Basin Research, 19, 179–198, doi: 10.1111/ j.1365-2117.00319.x Farrell, W. E. & Clark, J. A. 1976. On postglacial sea level. Geophysical Journal of the Royal Astronomical Society, 46, 647–667. Frets, D. C. 1969. Geology and structure of the Huab-Welwitschia area, South West Africa. Precambrian Research Unit Bulletin, 5, 235, 2 maps, University of Cape Town, South Africa. Friedman, I. & O’Neil, J. R. 1977. Compilation of stable-isotope fractionation factors of geochemical interest. In: Fleischer, M. (ed.) Data of Geochemistry, 6th edn. United States Geological Survey, Washington, DC, Professional Paper 44-KK. Gevers, T. W. 1931. An ancient tillite in South-West Africa. Transactions of the Geological Society of South Africa, 34, 1 –17. Goodman, J. & Pierrehumbert, R. T. 2003. Glacial flow of floating marine ice in ‘Snowball Earth’. Journal of Geophysical Research, 108, doi: 10.1029/2002JC001471. Goscombe, B. & Gray, D. R. 2007. The Coastal Terrane of the Kaoko Belt, Namibia: outboard arc-terrane and tectonic significance. Precambrian Research, 155, 139–158. Goscombe, B., Hand, M. & Gray, D. 2003. Structure of the Kaoko Belt, Namibia: progressive evolution of a classic transpressional orogen. Journal of Structural Geology, 25, 1049– 1081.

207

Halverson, G. P. 2006. A Neoproterozoic chronology. In: Xiao, S. & Kaufman, A. J. (eds) Neoproterozoic Geobiology and Paleobiology. Springer, Dordrecht, 231– 271. Halverson, G. P., Hoffman, P. F., Schrag, D. P. & Kaufman, J. A. 2002. A major perturbation of the carbon cycle before the Ghaub glaciation (Neoproterozoic) in Namibia: prelude to snowball Earth? Geophysics, Geochemistry, Geosystems, 3, doi: 10.1029/ 2001GC000244. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Toward a Neoproterozoic composite carbon-isotope record. Geological Society of America Bulletin, 117, 1181–1207, doi: 10.1130/B25630.1 ¨ ., Maloof, A. C. & Bowring, S. A. Halverson, G. P., Duda´s, F. O 2007. Evolution of the 87Sr/86Sr composition of Neoproterozoic seawater. Palaeogeography, Palaeoclimatology, Palaeoecology, 256, 103–129. Hedberg, R. M. 1979. Stratigraphy of the Owamboland Basin, South West Africa. Precambrian Research Unit Bulletin, 24, 325, 6 maps, University of Cape Town, South Africa. Henry, G., Stanistreet, I. G. & Maiden, K. J. 1986. Preliminary results of a sedimentological study of the Chuos Formation in the central zone of the Damara Orogen: evidence for mass flow processes and glacial activity. Communications of the Geological Survey of South West Africa/Namibia, 2, 75 –92. Henry, G., Clendenin, C. W., Stanistreet, I. G. & Malden, K. J. 1990. A multiple detachment model for the early rifting stage of the Late Proterozoic Damara Orogen in Namibia. Geology, 18, 67– 71. Higgins, J. A. & Schrag, D. P. 2003. Aftermath of a snowball Earth. Geophysics, Geochemistry, Geosystems, 4, doi: 10.1029/2002GC000403. Hoffman, P. F. 1999. The break-up of Rodinia, birth of Gondwana, true polar wander, and the snowball Earth. Journal of African Earth Sciences, 28, 17 –33. Hoffman, P. F. 2002. Carbonates bounding glacial deposits: evidence for Snowball Earth episodes and greenhouse aftermaths in the Neoproterozoic Otavi Group of northern Namibia. Excursion Guide, 16th International Sedimentological Conference, Auckland Park, South Africa, 39. Hoffman, P. F. 2005. 28th DeBeers Alex Du Toit Memorial Lecture: On Cryogenian (Neoproterozoic) ice-sheet dynamics and the limitations of the glacial sedimentary record. South African Journal of Geology, 108, 557–576. Hoffman, P. F. 2011. Strange bedfellows: glacial diamictite and cap carbonate from the Marinoan (635 Ma) glaciation in Namibia. Sedimentology, 58, 57– 119. Hoffman, P. F. & Halverson, G. P. 2008. The Otavi Group of the Northern Platform and the Northern Margin Zone. In: Miller, R. McG. (ed.) The Geology of Namibia, vol. 2. Geological Survey of Namibia, Windhoek, 13.69– 13.136. Hoffman, P. F. & Hartz, E. H. 1999. Large, coherent, submarine landslide associated with Pan-African foreland flexure. Geology, 27, 687– 690. Hoffman, P. F. & Macdonald, F. A. 2010. Sheet-crack cements and early regression in Marinoan (635 Ma) cap dolostones: regional benchmarks of vanishing ice-sheets? Earth and Planetary Science Letters, 300, 374– 384. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129– 155. Hoffman, P. F., Hawkins, D. P., Isachsen, C. E. & Bowring, S. A. 1996. Precise U– Pb zircon ages for early Damaran magmatism in the Summas Mountains and Welwitschia Inlier, northern Damara belt, Namibia. Communications of the Geological Survey of Namibia, 11, 47– 52. Hoffman, P. F., Kaufman, J. A. & Halverson, G. P. 1998a. Comings and goings of global glaciations on a Neoproterozoic carbonate platform in Namibia. GSA Today, 8, 1 – 9. Hoffman, P. F., Kaufman, A. J., Halverson, G. P. & Schrag, D. P. 1998b. A Neoproterozoic snowball Earth. Science, 281, 1342– 1346. Hoffman, P. F., Halverson, G. P., Domack, E. W., Husson, J. M., Higgins, J. A. & Schrag, D. P. 2007. Are basal Ediacaran (635 Ma) post-glacial ‘cap dolostones’ diachronous? Earth and Planetary Science Letters, 258, 114–131.

208

P. F. HOFFMAN

Hoffmann, K.-H. & Prave, A. R. 1996. A preliminary note on a revised subdivision and regional correlation of the Otavi Group based on glaciogenic diamictites and associated cap dolomites. Communications of the Geological Survey of Namibia, 11, 77 –82. Hoffmann, K.-H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. U– Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817– 820. Hurtgen, M. T., Arthur, M. A., Suits, N. S. & Kaufman, A. J. 2002. The sulfur isotopic composition of Neoproterozoic seawater sulfate: implications for a snowball Earth? Earth and Planetary Science Letters, 203, 413–429. Hurtgen, M. T., Halverson, G. P., Arthur, M. A. & Hoffman, P. F. 2006. Sulfur cycling in the aftermath of a 635 Ma snowball glaciation: evidence for a syn-glacial sulfidic deep ocean. Earth and Planetary Science Letters, 245, 551– 570. Jiang, G., Kennedy, M. J. & Christie-Blick, N. 2003. Stable isotopic evidence for methane seeps in Neoproterozoic postglacial cap carbonates. Nature, 426, 822– 826. Jiang, G., Kennedy, M. J., Christie-Blick, N., Wu, H. & Zhang, S. 2006. Stratigraphy, sedimentary structures, and textures of the late Neoproterozoic Doushantuo cap carbonate in South China. Journal of Sedimentary Research, 76, 978–995. Johnson, S. D., Poujol, M. & Kisters, A. F. M. 2006. Constraining the timing and migration of collisional tectonics in the Damara Belt, Namibia: U– Pb zircon ages for the syntectonic Salem-type Stinkbank granite. South African Journal of Geology, 109, 611– 624. Kasemann, S. A., Hawkesworth, C. J., Prave, A. R., Fallick, A. E. & Pearson, P. N. 2005. Boron and calcium isotope composition in Neoproterozoic carbonate rocks from Namibia: evidence for extreme environmental change. Earth and Planetary Science Letters, 231, 73 –86. Kasemann, S. A., Prave, A. R., Fallick, A. E., Hawkesworth, C. J. & Hoffmann, K.-H. 2010. Neoproterozoic ice ages, boron isotopes, and ocean acidification: implications for a snowball Earth. Geology, 38, 775– 778. Kendall, C. G. St. C. & Warren, J. 1987. A review of the origin and setting of tepees and their associated facies. Sedimentology, 34, 1007– 1028. Kennedy, M. J., Runnegar, B., Prave, A. R., Hoffmann, K.-H. & Arthur, M. A. 1998. Two or four Neoproterozoic glaciations? Geology, 26, 1059–1063. Kiessling, W. 2001. Paleoclimatic significance of Phanerozoic reefs. Geology, 29, 751–754. Kilner, B., Mac Niocaill, C. & Brasier, M. 2005. Low-latitude glaciation in the Neoproterozoic of Oman. Geology, 33, 413– 416. King, L. H. 1993. Till in the marine environment. Journal of Quaternary Science, 8, 347–358. King, L. H., Kokoengen, K., Fader, G. B. J. & Gunleiksrud, T. 1991. Till-tongue stratigraphy. Geological Society of America Bulletin, 103, 637– 659. Ko¨ppen, W. & Wegener, A. 1924. Die Klimate der geologischen Vorzeit. Gebru¨der Borntraeger, Berlin. Kro¨ner, A. & Rankama, K. 1972. Late Precambrian glaciogenic sedimentary rocks in southern Africa: a compilation with definitions and correlations. Precambrian Research Unit Bulletin, 11, 37, University of Cape Town, South Africa. Lythe, M. B. & Vaughan, D. G.the BEDMAP Consortium. 2001. BEDMAP: a new ice thickness and subglacial topographic model of Antarctica. Journal of Geophysical Research, 106, 11335–11351. MacAyeal, D. R. 1993. Binge-purge oscillations of the Laurentide ice sheet as a cause of the North Atlantic’s Heinrich events. Paleoceanography, 8, 775–784. Macouin, M., Besse, J., Ader, M., Gilder, S., Yang, Z., Sun, Z. & Agrinier, P. 2004. Combined paleomagnetic and isotopic data from the Doushantuo carbonates, South China: implications for the ‘snowball Earth’ hypothesis. Earth and Planetary Science Letters, 224, 387– 398. Maloof, A. C., Halverson, G. P., Kirschvink, J. L., Schrag, D. P., Weiss, B. P. & Hoffman, P. F. 2006. Combined paleomagnetic, isotopic and stratigraphic evidence for true polar wander from the Neoproterozoic Akademikerbreen Group, Svalbard. Geological

Society of America Bulletin, 118, 1099–1124, doi: 10.1130/ B25892.1. Martin, H. 1965a. The Precambrian geology of South West Africa and Namaqualand. Precambrian Research Unit Bulletin, 1, 159, 11 plates, 2 maps, University of Cape Town, South Africa. Martin, H. 1965b. Beobachtungen zum Problem der jung-pra¨kambrischen Glazialen Ablagerungen in Su¨dwestafrika (Observations concerning the problem of the late Precambrian glacial deposits in South West Africa). Geologische Rundschau, 54, 115– 127. Martin, H., Porada, H. & Walliser, O. H. 1985. Mixtite deposits of the Damara Sequence, Namibia, problems of interpretation. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 159–196. McCay, G. A., Prave, A. R., Alsop, G. I. & Fallick, A. E. 2006. Glacial trinity: Neoproterozoic Earth history within the British-Irish Caledonides. Geology, 34, 909–912, doi: 10.1130/G22694A.1 McMullen, K., Domack, E., Leventer, A., Olson, C., Dunbar, R. & Brachfeld, S. 2006. Glacial morphology and sediment formation in the Metz Trough, East Antarctica. Palaeogeography, Palaeoclimatology, Palaeoecology, 231, 169–180. Meert, J. G., Van der Voo, R. & Ayub, S. 1995. Paleomagnetic investigation of the Neoproterozoic Gagwe lavas and Mbozi complex, Tanzania, and the assembly of Gondwana. Precambrian Research, 74, 225– 244. Miller, R. McG. 1997. The Owambo Basin of northern Namibia. In: Selley, R. C. (ed.) African Basins. Sedimentary Basins of the World, 3. Elsevier, Amsterdam, 237– 268. Miller, R. McG. 2008. Otavi Group of the Otavi Mountainland (OML), the region west of the OML, and the eastern Kaokoveld. In: Miller, R. McG. (ed.) The Geology of Namibia, 2. Geological Survey of Namibia, Windhoek, 13.52– 13.69. Moloto-A-Kenguemba, G., Trindade, R. I. F., Monie´, P., Ne´de´lec, A. & Siqueira, R. 2008. A late Neoproterozoic paleomagnetic pole for the Congo craton: tectonic setting, paleomagnetism and geochronology of the Nola dike swarm (Central African Republic). Precambrian Research, 164, 214– 226. Nogueira, A. C. R., Riccomini, C., Sial, A. N., Moura, C. A. V., Trindade, R. I. F. & Fairchild, T. R. 2007. Carbon and strontium isotope fluctuations and paleoceanographic changes in the late Neoproterozoic Araras carbonate platform, southern Amazon craton, Brazil. Chemical Geology, 237, 168– 190. Opdyke, B. N. & Wilkinson, B. H. 1990. Palaeolatitude distribution of Phanerozoic marine ooids and cements. Palaeogeography, Palaeoclimatology, Palaeoecology, 78, 135–148. Ottesen, D., Dowdeswell, J. A. & Rise, L. 2005. Submarine landforms and the reconstruction of fast-flowing ice streams within a large Quaternary ice sheet: the 2500-km-long Norwegian– Svalbard margin (578 – 808N). Geological Society of America Bulletin, 117, 1033– 1050. Paciullo, F. V. P., Ribeiro, A., Trouw, R. A. J. & Passchier, C. W. 2007. Facies and facies association of the siliciclastic Brak River and carbonate Gemsbok formations in the Lower Ugab River valley, Namibia, W. Africa. Journal of African Earth Sciences, 47, 121– 134. Peltier, W. R., Tarasov, L., Vettoretti, G. & Solheim, L. P. 2004. Climate dynamics in deep time: modeling the ‘snowball bifurcation’ and assessing the plausibility of its occurrence. In: Jenkins, G. S., McMenamin, M. A. S., McKey, C. P. & Sohl, L. (eds) The Extreme Proterozoic: Geology, Geochemistry, and Climate. Geophysical Monograph 146, American Geophysical Union, Washington, DC, 107–124. Pollard, D. & DeConto, R. M. 2007. A coupled ice-sheet/ice-shelf/ sediment model applied to a marine margin flowline: forced and unforced variations. In: Hambrey, M. J., Christoffersen, P., Glasser, N. F. & Hubbard, B. (eds) Glacial Marine Processes and Products. International Association of Sedimentologists Special Publication No. 39, Blackwell Publishing, Malden, MA, 37– 52. Pollard, D. & Kasting, J. F. 2004. Climate-ice sheet simulations of Neoproterozoic glaciation before and after collapse to Snowball Earth. In: Jenkins, G. S., McMenamin, M. A. S., McKey, C. P. & Sohl, L. (eds) The Extreme Proterozoic: Geology, Geochemistry, and Climate. Geophysical Monograph 146, American Geophysical Union, Washington, DC, 91 –105.

THE OTAVI CARBONATE PLATFORM AND FORESLOPE

Powell, R. D. 1990. Glacimarine processes at grounding-line fans and their growth to ice-contact deltas. In: Dowdeswell, J. A. & Scourse, J. D. (eds) Glacimarine Environments: Processes and Sediments. Geological Society Special Publication No. 53, London, 53 – 73. Powell, R. D. & Domack, E. 1995. Modern glacimarine environments. In: Menzies, J. (ed.) Glacial Environments, vol. 1: Modern Glacial Environments: Processes, Dynamics and Sediments. ButterworthHeinemann, Oxford, 445– 486. Prave, A. R., Strachan, R. A. & Fallick, A. E. 2009. Global C cycle perturbations recorded in marbles: a record of Neoproterozoic Earth history within the Dalradian succession of the Shetland Islands, Scotland. Journal of the Geological Society, London, 166, 129– 135. Pruss, S. B., Bosak, T., Macdonald, F. A., McLane, M. & Hoffman, P. F. 2010. Microbial facies in a Sturtian cap carbonate, the Rasthof Formation, Otavi Group, northern Namibia. Precambrian Research, 181, 187– 108. Rodgers, J. 1957. The distribution of marine carbonate sediments: a review. In: Le Blanc, R. J. & Breeding, J. G. (eds) Regional Aspects of Carbonate Deposition. Society of Economic Paleontologists and Mineralogists (SEPM) Special Publication No. 5, Tulsa, Oklahoma, 2– 14. le Roex, H. D. 1941. A tillite in the Otavi Mountains, S.W.A. Transactions of the Geological Society of South Africa, 44, 207–218, 2 plates, 1 map. SACS (South African Committee for Stratigraphy) 1980. Stratigraphy of South Africa. Part 1 (comp. L.E. Kent): Lithostratigraphy of the Republic of South Africa, South West Africa/Namibia and the republics of Bophuthatswana, Transkei and Venda. Handbook, Geological Survey of South Africa, 8, 415– 437. Shields, G. A. 2005. Neoproterozoic cap carbonates: a critical appraisal of existing models and the plumeworld hypothesis. Terra Nova, 17, 299– 310. Shields, G. A., Deynoux, M., Strauss, H., Paquet, H. & Nahon, D. 2007. Barite-bearing cap dolostone of the Taoude´ni Basin, northwest Africa: sedimentary and isotopic evidence for methane seepage after a Neoproterozoic glaciation. Precambrian Research, 154, 209– 235. Shoemaker, E. M. 1963. Impact mechanics at Meteor Crater, Arizona. In: Middlehurst, B. M. & Kuiper, G. P. (eds.) The Moon Meteorites Comets 4. University of Chicago Press, Chicago, 301– 336.

209

Smit, J. M. 1962. Stratigraphy and metamorphism of the Otavi Series southeast of Otavi, South West Africa. Transactions of the Geological Society of South Africa, 65, 63 –78. Soffer, G. 1998. Evolution of a Neoproterozoic continental margin subject to tropical glaciation. BA thesis, Harvard College, Cambridge, USA. Stanistreet, I. G. & Charlesworth, E. G. 1999. Damaran basementcored fold nappes incorporating pre-collisional basins, Kaoko Belt, Namibia, and controls on Mesozoic supercontinental breakup. Economic Geology Research Unit, Information Circular, 332, University of the Witwatersrand, Johannesburg, 14. Swart, R. 1992. Facies analysis of late Proterozoic carbonate turbidites in the Zerrissene Basin, Damara Orogen, Namibia. Journal of African Earth Sciences, 14, 283–294. Trindade, R. I. F., Font, E., D’Agrella-Filho, M. S., Nogueira, A. C. R. & Riccomini, C. 2003. Low-latitude and multiple geomagnetic reversals in the Neoproterozoic Puga cap carbonate, Amazon craton. Terra Nova, 15, 441– 446, doi: 10.1046/j.1365-3121.2003. 00510.x. Vasconcelos, C., McKenzie, J. A., Warthmann, R. & Bernasconi, S. M. 2005. Calibration of the d18O paleothermometer for dolomite precipitated in microbial cultures and natural environments. Geology, 33, 317– 320. doi: 10.1130/G20992.1. Warren, S. G., Brandt, R. E., Grenfell, T. C. & McKay, C. P. 2002. Snowball Earth: ice thickness on the tropical ocean. Journal of Geophysical Research, 107, doi: 10.1029/2001JC001123. Williams, G. E. 1993. History of the Earth’s obliquity. Earth-Science Reviews, 34, 1 –45. Witzke, B. J. 1990. Palaeoclimatic constraints for Palaeozoic palaeolatitudes of Laurentia and Euramerica. In: McKerrow, W. S. & Scotese, C. R. (eds) Palaeozoic Palaeogeography and Biogeography. Geological Society of London Memoir, 12, 57 –73. Yoshioka, H., Asahara, Y., Tojo, B. & Kawakami, S. 2003. Systematic variations in C, O, and Sr isotopes and elemental concentrations in Neoproterozoic carbonates in Namibia: implications for a glacial to interglacial transition. Precambrian Research, 124, 69– 85. Ziegler, A. M., Hulver, M. L., Lottes, A. L. & Schmachtenberg, W. F. 1984. Uniformitarianism and paleoclimates: inferences from the distribution of carbonate rocks. In: Brenchley, P. J. (ed.) Fossils and Climate. John Wiley & Sons, New York, 3 –25.

Chapter 15 The Witvlei Group of East-Central Namibia A. R. PRAVE1*, K.-H. HOFFMANN2, W. HEGENBERGER2 & A. E. FALLICK3 1

Department of Earth Sciences, University of St Andrews, St Andrews KY16 9AL, UK 2

Geological Survey of Namibia, 1 Aviation Road, Windhoek, Namibia

3

Scottish Universities Environmental Research Centre, East Kilbride G75 0QF, UK *Corresponding author (e-mail: [email protected])

Abstract: The Witvlei Group is preserved in two regional synclinoria in the Gobabis-Witvlei area of east-central Namibia and as isolated outcrops 90 km SW of Rehoboth, itself some 200 km south of that area. It consists of mixed, coarse- to fine-grained siliciclastic and carbonate strata deposited in deep- to shallow-marine, and locally non-marine, settings along the post-rift continental margin of the Kalahari Craton prior to the onset of foreland basin sedimentation recorded by the overlying terminal Neoproterozoic–Cambrian Nama Group. No direct age constrains exist for the Witvlei Group, but it post-dates c. 800 Ma rift-related rocks and pre-dates the c. 548 Ma base of the Nama Group, thereby placing it as Cryogenian to Ediacaran in age. The Witvlei Group consists of three main units, from oldest to youngest, the Blaubeker, Court and Buschmannsklippe Formations. The Blaubeker Formation is highly variable in thickness and can be as much as 1000 m thick. It consists mostly of massive, polymict diamictite and, in the area of the type locality, contains conglomerate and pebbly sandstone beds. The diamictic strata combined with the presence of numerous faceted and striated clasts provide the evidence for glaciogenic influences on sedimentation. The highly variable thickness pattern likely reflects the infill of palaeo-valleys formed by the deep erosion and scouring of bedrock by ice, and the conglomerates and pebbly sandstones record glacial outwash processes. The Tahiti Formation is a locally developed, fine-grained sandstone above the Blaubeker Formation. It is poorly exposed and its exact stratigraphic relationship to the Blaubeker rocks and overlying Court Formation remains to be determined. The Blaubeker rocks are overlain sharply by the basal unit of the Court Formation, the Gobabis Member. This Member is from 20 to 60 m thick and consists mostly of dark and light grey laminated dolostones that display a d13Ccarbonate profile that rises from values of 24‰ in the lowermost beds to values of 5‰ in the topmost. The Gobabis Member is conformably overlain by the shales, marls and thin limestones of the Constance Member followed by quartzites of the uppermost unit of the Court Formation, the Simmenau Member. The basal unit of the Buschmannsklippe Formation is the light to tan and pink grey dolostone of the Bildah Member. Its basal contact is sharp everywhere, and it is gradationally overlain by a coarsening (shoaling) upward succession from shales, thin limestones (some exhibiting formerly aragonitic fans) and fine sandstones of the La Fraque Member, to interbedded quartzites and stromatolitic and cherty dolostones of the Okambara Member. The d13Ccarbonate profile for the Buschmannsklippe rocks shows that the basal beds of the Bildah Member begin at –4‰, followed by a decline to –6‰ in the lower La Fraque limestones and then a rise to –3‰ in the dolostones of the Okambara Member before being truncated by the base of the regionally unconformably overlying basal Weissberg Quartzite Member of the lower Nama Group. Although no glacial sediments have been recognized below the Bildah Member, its lithofacies character, stratigraphic position and C-isotopic profile are compatible with and strikingly similar to younger Cryogenian cap carbonates. Thus, the Witvlei Group arguably contains both the older and younger cap carbonates of Neoproterozoic time, but only the older Cryogenian glacial deposit.

Glaciogenic rocks of the Witvlei Group form part of a mixed siliciclastic and carbonate sedimentary succession of Late Neoproterozoic (Cryogenian – early Ediacaran) age that is preserved along the NW margin (present-day coordinates) of the Kalahari Craton (Fig. 15.1). It averages 200 –400 m in thickness (but has a composite thickness of more than 1000 m) and rests unconformably on a varied substrate comprised of thick sequences of pre-glacial, early – middle Neoproterozoic (Tonian – early Cryogenian) rift-related arkosic sandstones, conglomerates and argillites of the Eskadron and Doornpoort Formations (Tsumis Group) and the Kamtsas Formation (Nosib Group) and, locally, older Palaeo/Mesoproterozoic metasedimentary and igneous rocks. The Witvlei Group records sedimentation in a range of shallow- to deep-marine and locally non-marine settings that developed following rifting and antecedent to the onset of foreland basin development represented by the unconformably overlying terminal Neoproterozoic (late Ediacaran) – Cambrian Nama Group. Hegenberger (1993) provides the most complete description and synthesis of the Witvlei Group rocks, while Hoffmann (1989) presents lithostratigraphic correlations of the Witvlei Group succession within a regional stratigraphic framework encompassing the Naukluft Nappe Complex, the southern and northern Damara Belt (including the Otavi Group) and the Gariep Belt of southern Namibia. Only one glaciogenic unit has been recognized in the Witvlei Group, the diamictic Blaubeker Formation, but two ‘cap

carbonates’ are present (Fig. 15.2). The older of the two is the Gobabis Member of the Court Formation; it caps the Blaubeker diamictite. The younger one is the basal unit of the Buschmannsklippe Formation, the Bildah Member; it sits sharply on and transgresses a variable substrate of the Court Formation and, where absent, older Kamtsas and Eskadron-Doornpoort Formation rocks, but no glaciogenic rocks have been recognized.

Structural framework The Witvlei Group is preserved in two NE– SW trending regional synclinoria located within the southern foreland zone of the Damara Belt in east-central Namibia (Fig. 15.1; Hegenberger 1993). The western one, the Witvlei Synclinorium, averages about 20 km in width and is 180 km in length. It comprises several open to tight second- and numerous third-order folds, and exposures are good along many of the moderately to steeply dipping limbs. The eastern structure is termed the Gobabis Synclinorium and covers an area of c. 80 km by 100 km; dips are relatively shallow and exposure is poor. Minor faulting disrupts the stratigraphy in both synclinoria, but displacements are small and typically no more than a few tens of metres. Several larger steep faults are associated with the Witvlei Synclinorium but they do not compromise the stratigraphic coherence of the succession.

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 211– 216. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.15

212

A. R. PRAVE ET AL.

Fig. 15.1. Generalized geological map of the southern Damara foreland showing the distribution of Witvlei Group rocks in the Witvlei-Gobabis area (after Hegenberger 1993). Inset map shows location of main figure and the major tectonic regions and sedimentary basins. B, Blaubeker; GB, Gariep Belt; NNC, Naukluft Nappe Complex; R, Rehoboth.

Folding was associated with very low-grade metamorphism, decreasing from NW to SE, and is attributed to the late Precambrian –Cambrian Damaran orogeny (Miller 1983). The original extent of the Witvlei Group is not known but was likely much greater than the currently preserved area of outcrop. To the NW, the Witvlei Group outcrop belt is truncated by the

Damaran structural front, defined by thrust faults involving preWitvlei Group rocks including crystalline basement. To the east, south and SW, the Witvlei Group rocks are buried beneath Phanerozoic strata (Karoo and Kalahari Groups) and how far they extend under that cover sequence is not known, although drill holes some 40 km south of the southwestern tip of the Witvlei Synclinorium have intersected Witvlei Group rocks (Hegenberger 1993). Further SW, within the western part of the southern Damara foreland, glacial sediments of the Blaubeker Formation are preserved as local erosional remnants on Blaubeker farm, the designated type locality (Schalk 1970) some 90 km SW of Rehoboth, with thick strata of the Nama Group of the northern Nama (Zaris) basin resting directly on the older Neoproterozoic Tsumis and Nosib Group rift-related sediments, and, still further west, Palaeo- and Mesoproterozoic basement. Absence of rocks equivalent to the Court and Buschmannsklippe Formation is either due to non-deposition or erosional removal prior to deposition of the Nama. In the Gobabis-Witvlei area, the lower units of the Witvlei Group, the Blaubeker and Court Formation, display lateral thickness and facies variations and a major intraformational unconformity at the base of the Simmenau Member that has been interpreted as evidence for influence by synsedimentary tectonism (Hegenberger 1993). No unequivocal basin-bounding faults and/ or growth structures have been identified, but the Nina Anticline (which separates the Gobabis and Witvlei Synclinoria) was actively up-warped during this time, resulting in erosion of lower units of the Court Formation across the crest of that structure. The upper unit of the Witvlei Group, the Buschmannsklippe Formation, exhibits systematic thickness variations and similar facies characteristics throughout the region, and displays an overall westward thickening (Hegenberger 1993). This suggests that sedimentation occurred during a phase of relatively uniform subsidence.

Fig. 15.2. Stratigraphy of the Witvlei Group (after Hoffmann 1989; Hegenberger 1993). Glaciogenic rocks (Blaubeker Formation) and cap carbonates (Gobabis and Bildah Members) are highlighted in bold and with shaded ornament. See text for discussion.

Stratigraphy The rocks that make up the Witvlei Group, including the locally preserved diamictite beds in the Gobabis-Witvlei area, were first

THE WITVLEI GROUP

described collectively as the Buschmannsklippe Formation by Martin (1965) and regarded for many years as a local, northeastern facies of the Nama Group (e.g. Hegenberger & Seeger 1980). Since then, lithostratigraphic subdivision and nomenclature of the succession has undergone several revisions, the most important of which is the recognition that the lower and middle parts of the original Buschmannsklippe Formation are older than the Nama Group, which led to the establishment of the Witvlei Group as a major new division for these units (Hoffmann 1989; see Hegenberger 1993 for a review). Here we adopt the subdivision by Hegenberger (1993), but include the Blaubeker Formation at the base as recommended by Hoffmann (1989). The Witvlei Group thus consists of three main units, the Blaubeker, Court and Buschmannsklippe Formations (Fig. 15.2). The Blaubeker Formation is patchily preserved, the Court Formation is most complete in the Gobabis Synclinorium, and the Buschmannsklippe Formation is best developed and exposed in the Witvlei Synclinorium. The Blaubeker Formation is characterized by polymict diamictite. At the type locality on Blaubeker farm and surrounding areas, it overlies the Kamtsas Formation with an angular unconformity and consists of a basal unit of interbedded conglomerate and quartzite followed by massive polymict diamictite, which is unconformably overlain by the Nama Group; both the Court and Buschmannsklippe Formations are absent. Total thickness varies between 250 and 300 m (Schalk 1970). In the Gobabais-Witvlei area, the Blaubeker Formation is present in two main outcrop areas where it unconformably overlies the Kamtsas Formation. The thickest and best-developed exposures are in the south-central portions of the Gobabis Synclinorium (particularly in the valleys of the White and Black Nossob Rivers) and it is also present along the easternmost edge of the south-central flanks of the Witvlei Synclinorium (Hegenberger 1993). In these areas the Blaubeker Formation is composed entirely of massive polymict diamictite and is highly variable in thickness, ranging from as little as 10 m to as much as 1000 m (Hegenberger & Seeger 1980). The Court Formation consists of four units (Hegenberger 1993). The oldest is the Tahiti Member, a 1– 10-m-thick finegrained feldspathic arenite. This member is poorly exposed and few sedimentological details exist. It is restricted in its occurrence to a few outcrops along the extreme east-southeastern edge of the Gobabis Synclinorium (and may well be a facies of the Blaubeker Formation, but this remains to be proven). The first, welldeveloped and laterally extensive unit of the Court Formation is the Gobabis Member. It mostly consists of grey laminated dolostones and local stromatolitic limestone, and varies between 20 and 60 m in thickness. The youngest units of the Court Formation are the Constance Member, a recessive, poorly exposed 100– 200-m-thick succession of pale-coloured shale, marl and thinbedded limestone, and the Simmenau Member, a 100 –150-mthick fine- to medium-grained quartzite to feldspathic arenite. The Buschmannsklippe Formation is divided into three units, from oldest to youngest, the Bildah, La Fraque and Okambara Members (Hegenberger 1993). The Formation as a whole defines a west-SW-thickening wedge, from about 50 m thick in the easternmost sections of the Gobabis Synclinorium to more than 300 m thick in the Witvlei Synclinorium. The Bildah Member is typically 40– 60 m thick and consists of light grey to pale tan and pink dolostone. Its base defines a regional unconformity that oversteps the underlying Court Formation to rest unconformably on older Neoproterozoic rocks. Overlying the Bildah is the La Fraque Member, a succession of interbedded shale, thin-bedded limestone and fine sandstone. It thickens to the west and reaches a maximum thickness of c. 150 m in the central portions of the Witvlei Synclinorium. The Okambara Member, like the La Fraque, similarly thickens westward attaining a maximum thickness of about 125 m. It is characterized by cream to pink-grey coloured commonly stromatolitic dolostone and cherty dolostone and subordinate fine, white quartzite. In the thicker sections in the central portions of the Witlvlei Synclinorium, Hegenberger

213

(1993) subdivided the Okambara into an upper and lower dolostone-dominated unit separated by a quartzite-dominated unit. The top of the Witvlei Group is marked by the regional unconformity at the base of the Weissberg Member (Dabis Formation) of the lower Nama Group.

Glaciogenic deposits and associated strata The Blaubeker Formation No detailed modern sedimentological investigations of the Blaubeker Formation have been carried out, and the following summary is based on descriptions by Martin (1965), Schalk (1970), Kro¨ner & Rankama (1972), Hegenberger & Seeger (1980) and field observation by the authors. In the type locality and adjacent areas west of Rehoboth, exposures are patchy, but show that the Blaubeker Formation consists of a lower unit of pebbly quartzite with rare shale interbeds. This unit is several tens of metres thick and is sharply overlain by massive polymict diamictite between 200 and 300 m thick. In the Gobabis-Witvlei area, the Blaubeker consists entirely of massive, non-graded, non-stratified, polymict diamictite; the best exposures are at Farm Tahiti in the Black Nossob River valley. In both synclinoria, clasts are dispersed in a greygreen to reddish-grey, fine siliciclastic matrix; the matrix is massive and no sandstone lenticules or laminae have been reported. Clasts typically range between 0.05 and 0.5 m in the longest dimension, but can be as large as 1 m. They consist mainly of quartzite, lesser felsic and mafic volcanics, granitoids, gneiss and schist. Shapes vary from angular to sub-rounded and many of the clasts display variably orientated striations and faceting. This is the primary observational evidence to infer a glacially influenced origin for the Blaubeker diamictite. At Gobabis, a 0.1 – 0.5-m-thick, crumbly weathering, poorly sorted, massive brown sandstone containing dispersed quartzite clasts and cobbles occurs between the underlying Kamtsas Formation and the sharp base of the Gobabis Member. This sandstone may be a facies of the Blaubeker Formation, but this remains to be proven.

The Doornpoort/Eskadron and Kamtsas Formation These rocks are pre-glacial in age but are briefly described here because many of the clasts in the Blaubeker diamictites were reworked from these older units. All three formations can reach thicknesses measured in many thousands of metres and are dominated by feldspathic quartzite, pebbly quartzite and arenite and local conglomerate. The conglomeratic portions contain clasts ranging in size from pebbles to boulders that consist of quartz, quartzite, felsic and mafic volcanic rocks, granites and granitic gneiss (Hegenberger & Seeger 1980).

The Court Formation Resting sharply on the Blaubeker Formation is the Gobabis Member of the Court Formation (locally, the Tahiti Formation sandstones occur on top of the Blaubeker rocks). In the vicinity of the type locality, the Gobabis Member is 50–60 m thick and consists of two main units. The lower unit is of variable thickness (10 –30 m) and characterized by thin-bedded, planar- to wavyparallel laminated dolomicrite. The centimetre-thick laminae alternate between dark grey and lighter grey in colour, giving the Gobabis Member a distinctive striped pattern. In places, synsedimentary (or early lithification) disruption of bedding has resulted in locally formed intraformational breccias; these, however, are rare. Much of the remainder of the Member is composed of similarly dark –light coloured dolomicrite, but the lamination is crinkly-wrinkly in character and exhibits microbial

214

A. R. PRAVE ET AL.

roll-up structures. Columnar stromatolites are developed locally, and some of these are in limestone rather than dolostone.

The Buschmannsklippe Formation At its type locality, the Bildah Member rests with sharp contact on quartzites of the Kamtsas Formation and consists of four units, from the base upward these are: (i) light grey to tan coloured, flat to wavy-parallel laminated dolomicrite containing locally developed soft-sediment slumping and irregularly shaped calcite-spar-filled vugs; (ii) microbially laminated to stromatolitic light grey dolomicrite with quartz- and calcite-spar-filled vugs and tubes (developed in variable densities); (iii) fine-grained dolostone, light grey in colour, exhibiting cross-bedding and flat lamination; and (iv) thin-bedded to laminated, fine-grained, grey dolostone with mudstone partings. The Bildah cap dolostone varies in thickness from 60 to 100 m, and much of this variability is due to the lesser or greater development of the stromatolitic unit. Gradationally overlying the Bildah dolostones is the reddishbrown shale and thin-bedded limestone of the lower part of the La Fraque Member; some of the limestones exhibit centimetrescale aragonite fan structures. These beds pass transitionally upward into tan-brown coloured, cross-bedded and hummocky cross-stratified, commonly calcareous-cemented, sandstone marking the upper part of the La Fraque Member. Gradationally overlying the La Fraque rocks are the light-coloured quartzites and cream-coloured dolostones of the Okambara Member. The Okambara rocks contain a variety of sedimentary features including crinkly-wrinkly laminated dolomicrite, columnar, branching stromatolites, as well as laterally linked heads, imbricated (edgewise)-, wave-rosette- and flat-pebble intraclastic limestones and dolostones, cross-bedded dolograinstones, rare oolitic dolostone, chicken-wire-fabric quartz nodules, cauliflower chert and enterolithic bedding. The tops of some beds exhibit desiccation cracks.

Boundary relations with overlying and underlying non-glacial units The Blaubeker Formation everywhere rests unconformably on the Kamtsas Formation. The highly varied thickness pattern of the Blaubeker (from 10 m or less to more than 1000 m) is evidence that its base must be deeply erosive. In the area of the eponymous type locality, strata of the lower Nama Group (Dabis Formation) rest unconformably on the Blaubeker rocks. In the GobabisWitvlei area, with the exception of where the areally restricted Tahiti Member is present, the diamictic strata of the Blaubeker Member are everywhere sharply capped by the basal laminated dolostone of the Gobabis Member. The Gobabis Member is overlain by the shales and thin limestones of the Constance Member, but the contact between these two Members is poorly exposed and commonly obscured by calcrete, so we are uncertain if it is sharp or gradational. The base of the Bildah Member rests unconformably on a variable substrate consisting of the Court and/or Kamtsas or EskadronDoortpoort Formations and is everywhere sharp. The contact between the Bildah Member and the overlying La Fraque Member is gradational over a few metres; the topmost several metres of the Bildah Member exhibit an upward increase in the proportion and thickness of mudstone partings concomitant with a thinning upward and proportional decrease in dolostone beds.

Chemostratigraphy Kaufman et al. (1991, 1997) Saylor et al. (1998) and Kennedy et al. (1998) did reconnaissance C-isotopic studies on the Witvlei Group rocks and Gorjan et al. (2003) performed S-isotopic analyses. We

have expanded upon the C-isotopic work and the following is a preliminary description of new, unpublished, data from the carbonate-bearing units of the Witvlei Group by the authors. A total of 193 stable isotopic analyses have been carried out, but additional work is required before more robust trends can be constructed, especially for the less well exposed interglacial units The C-isotopic profile of the older cap carbonate, the Gobabis Member, displays a progressively rising trend: the basal layer has d13Ccarbonate values of –4‰ to – 3‰ and these rise to –2‰ to –1‰ in the overlying thin-bedded and laminated dolomicrites. The contact between the lower planar- to wavy-laminated unit and the upper crinkly-wrinkly laminated unit with roll-up structures is everywhere covered and/or mantled with calcrete, but the lowest exposed beds of the upper unit begin with d13Ccarbonate values of 5‰, and these stay between 3‰ and 5‰ to the top of the Gobabis Member. Because the contact is covered, we are uncertain if the rise to positive values is gradational or abrupt. The inter-glacial succession consists of the Constance and Simmenau Members of the Court Formation. The former is poorly exposed and the latter does not contain carbonate rocks, so we have few data on these units and more work needs to be carried out before anything substantial can be stated regarding their chemostratigraphy. The basal laminated unit of the Bildah cap dolostone consistently has d13Ccarbonate values around –4‰; these rise slightly to values around –3‰ in the middle and upper units of the cap dolostone. At the contact with, and through the transition into, the La Fraque Member, C-isotopic values decline to –5‰ to –6‰ through the limestone rhythmite and thin-bedded limestone units (it is noteworthy that at this stratigraphic level, fans considered to be formerly aragonitic in composition are present). Above this and through the Okambara Member, the d13Ccarbonate profile rises to values between –3‰ and –2‰ before being truncated by the basal unconformity of the overlying Weissberg Member of the Nama Group.

Other characteristics (e.g. economic deposits, biomarkers) No biomarker data have been obtained on the Witvlei Group rocks, nor are these rocks known to contain any major deposits or mineralizaton of economic importance. The Bildah Member has been quarried locally for dimension stone.

Palaeolatitude and palaeogeography Kro¨ner et al. (1980) obtained palaeomagnetic data from three sites of the Blaubeker Formation at its type locality. The data indicated that the magnetization post-dated folding and was attributed to thermal events associated with Damaran orogenesis. No other palaeomagnetic data exist on the Witvlei Group rocks, but it is generally assumed that they, as part of the Kalahari craton, were deposited in low-latitudinal positions during most of the Cryogenian (e.g. Li et al. 2008).

Geochronological constraints No direct geochronological data exist on the Witvlei Group rocks. The Witvlei Group rests unconformably on rocks that are correlative with those elsewhere in central and southern Namibia considered as c. 1000– 800 Ma in age; it is overlain unconformably by the late Ediacaran – Cambrian Nama Group (Hegenberger 1993). Age constraints are good for the lower part of the Nama Group and an ash bed within the Zaris Formation in the main Nama basin to the SW has yielded a U – Pb zircon age of 548 Ma (Grotzinger et al. 1995). These broad constraints show that the Witvlei Group is entirely Neoproterozoic (Cryogenian to Ediacaran) in age.

THE WITVLEI GROUP

The lithological and chemostratigraphic similarity of the Bildah Member to the Keilberg cap dolostone (Hoffmann 1989; Kennedy et al. 1998) can be used to infer an age of 635 Ma, that is, the timing of meltback associated with the Marinoan-equivalent glacial and cap-carbonate rocks in northern Namibia and southern China (Hoffmann et al. 2004; Condon et al. 2005). The postMarinoan d13C crossover to positive values (e.g. Halverson et al. 2005) is not recorded in the Buschmannsklippe C-isotopic data, most likely due to erosional truncation along the unconformity surface defining the base of the Nama Group. If so, then the time gap between the base of the Nama Group and top of the Witvlei Group could be considerable and on the order of many tens of millions of years. Direct age constraints are also lacking for the base of the Witvlei Group, but can be inferred from the proposed correlation of the Gobabis Member and Blaubeker Formation with the Rasthof cap carbonate and Chuos Formation of northern Namibia (Hoffmann 1989; Kennedy et al. 1998), which overlies felsic volcanic rocks with a U –Pb zircon age of 746 Ma (Hoffman et al. 1996). A still younger age has been suggested as c. 710 Ma (Halverson et al. 2005). It is reasonable, then, to infer a similar age for the Blaubeker-Gobabis glacial –cap couplet. If this is correct, then the Witvlei Group likely spans some 100 Ma in time, from the Cryogenian through to at least the early –middle Ediacaran.

Discussion The Blaubeker Formation is the only unit in the Witvlei Group succession that can be attributed to recording glacial influences on sedimentation in the form of diamictic rocks containing striated and faceted clasts. It occurs in narrow outcrop belts displaying highly variable thicknesses, and these features most likely are the result of glacial scouring and carving of valleys and channels. The pebbly sandstones and conglomerates associated with the diamictites represent glacial outwash. The Blaubeker is sharply capped by the Gobabis Member dolostones. The overall facies development of this Member is identical to that documented and described from the Rasthof Formation, the older of the two cap carbonates in the Otavi succession of northern Namibia (Hoffmann 1989; Hoffmann & Prave 1996; Kennedy et al. 1998; Hoffman & Schrag 2002). This has been used in combination with C-isotopic data to correlate the Gobabis dolostones and the Blaubeker diamictite with the Rasthof Formation and Chuos Formation diamictite (Hoffmann 1989; Kennedy et al. 1998; Hoffman & Schrag 2002). The overlying Constance and Simmenau Members form the inter-glacial package and are typified by fine-grain sizes, flat lamination, low-angle crossstratification and current ripples. Rare desiccation cracks are preserved in the Constance Member and the Simmenau Member contains mud-chip conglomerates in the base of broad, shallowchannelled beds. These features have been used to infer that the Constance Member was deposited in a low-energy intermittently exposed, shallow-water setting and that the Simmenau Member records a distal, fine-grained braidplain (Hegenberger 1993). Combined, the Court Formation records a post-glacial transgression from deeper marine (Gobabis Member) through shoaling into shallower marine (Constance Member) settings. Nowhere along the outcrop belt of the Witvlei Group have rocks been recognized as glacial in origin beneath the Bildah Member. Based on striking similarities in lithology and facies, it was proposed to be equivalent to the characteristic pink-grey dolostones that cap diamictites in the Naukluft Nappe Complex and the Gariep Belt (Hoffmann 1989). This established, for the first time, that the Bildah, even given the absence of an underlying diamictite, was the same (cap) carbonate unit overlying a regional glacial unconformity surface. This is similar to what is now recognized from many places worldwide where glacial rocks are only patchily preserved and it is the overlying cap carbonates that are

215

the most extensively and consistently developed (e.g. Hoffman & Schrag 2002; Fairchild & Kennedy 2007). Since then, the Cisotopic profile and overall facies of the Bildah Member has been shown to be identical to those known to cap the younger Cryogenian glaciation elsewhere and is now interpreted as being equivalent to the Keilberg cap dolostone of the Otavi Group in northern Namibia (Hoffmann & Prave 1996; Kennedy et al. 1998; Hoffman & Schrag 2002). The overlying La Fraque – Okambara succession records the post-glacial transgression through shoaling sequence: from the deeper-marine settings of the lower La Fraque (interbedded shale and originally aragonitefan-bearing limestone rhythmites), to storm-dominated inner shelf settings (hummocky cross-stratified sandstones) of the upper La Fraque, to shallow subtidal and intertidal (quartzreplacement evaporite fabrics and desiccated microbial laminites) depositional settings of the Okambara rocks (Hegenberger 1993). K.-H. H. and W. H. thank the Geological Survey of Namibia for support. A. R. P. and A. E. F. acknowledge NERC and The Carnegie Trust for The Universities of Scotland for support. This represents a contribution of the IUGS- and UNESCOfunded IGCP (International Geoscience Programme) Project #512.

References Condon, D., Zhu, M., Bowring, S., Wang, W., Yang, A. & Jin, Y. 2005. U–Pb ages from the Neoproterozoic Doushantuo Formation, China. Science, 308, 95– 98. Fairchild, I. & Kennedy, M. J. 2007. Neoproterozoic glaciation in the Earth System. Journal of the Geological Society, London, 164, 895– 921. Gorjan, P., Walter, M. R. & Swart, R. 2003. Global Neoproterozoic (Sturtian) post-glacial sulfide – sulfur isotope anomaly recognized in Namibia. Journal of African Earth Sciences, 36, 89 –98. Grotzinger, J. P., Bowring, S., Saylor, B. Z. & Kaufman, A. J. 1995. Biostratigraphic and geochronologic constraints on early animal evolution. Science, 270, 598– 604. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Towards a Neoproterozoic composite carbon isotope record. Geological Society of America Bulletin, 117, 1181–1207. Hegenberger, W. 1993. Stratigraphy and sedimentology of the Late Precambrian Witvlei and Nama Groups, East of Windhoek. Geological Survey of Namibia, Memoir 17, 82. Hegenberger, W. & Seeger, K. G. 1980. The geology of the Gobabis area. Explanation of Sheet 2218, scale 1:250,000. Geological Survey of South West Africa/Namibia Windhoek, 11. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129– 155. Hoffman, P. F., Hawkins, D. P., Isachsen, C. E. & Bowring, S. A. 1996. Precise U– Pb zircon ages for early Damaran magmatism in the Summas Mountains and Welwitschia Inlier, northern Damara belt, Namibia. Communications of the Geological Survey of Namibia, 11, 47– 52. Hoffmann, K. H. 1989. New aspects of lithostratigraphic subdivision and correlation of late Proterozoic to early Canbrianrocks of the southern Damara Belt and their correlation with the central and northern Damara Belt and the Gariep Belt. Communications of the Geological Survey of Namibia, 5, 59 –67. Hoffmann, K. H. & Prave, A. R. 1996. A preliminary note on a revised subdivision and regional correlation of the Otavi Group based on glaciogenic diamictites and associated cap dolostones. Communications of the Geological Society of Namibia, 11, 81 –86. Hoffmann, K. H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. A U– Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817– 820. Kaufman, A. J., Hayes, J. M., Knoll, A. H. & Germs, G. J. B. 1991. Isotopic compositions of carbonates and organic carbon from upper Proterozoic successions in Namibia: stratigraphic variation and the effects of diagenesis and metamorphism. Precambrian Research, 49, 301– 327.

216

A. R. PRAVE ET AL.

Kaufman, A. J., Hayes, J. M., Knoll, A. H. & Narbonne, G. M. 1997. Isotopes, ice ages, and terminal Proterozoic earth history. Proceedings of the National Academy of Science, 94, 6600– 6605. Kennedy, M. J., Runnegar, B., Prave, A. R., Hoffmann, K. H. & Arthur, M. 1998. Two or four Neoproterozoic glaciations? Geology, 26, 1059–1063. Kro¨ner, A., McWilliams, M. O., Germs, G. J. B., Reid, A. B. & Schalk, K. E. L. 1980. Paleomagnetism of late Precambrian to early Paleozoic mixtite-bearing formations in Namibia (South West Africa): the Nama Group and Blaubeker Formation. American Journal of Science, 280, 942–968. Kro¨ner, A. & Rankama, K. 1972. Late Precambrain glaciogenic sedimentary rocks in southern Africa: a compilation with definitions and correlations. Bulletin of the Precambrian Research Unit, University of Cape Town, 11, 37.

Li, Z. X., Bogdanova, S. V. et al. 2008. Assembly, configuration, and break-up history of Rodinia: a synthesis. Precambrian Research, 160, 179– 210. Martin, H. 1965. The Precambrian Geology of South West Africa and Namaqualand. Precambrian Research Unit, University of Cape Town, Cape Town. Miller, R. McG. 1983. The pan-African Damara Orogen of South West Africa/Namibia. In: Miller, R. McG. (ed.) Evolution of the Damara Orogen of Sout West Africa/Namibia. Special Publication of the Geological Survey of South Africa, 11, 431– 515. Saylor, B. Z., Kaufman, A. J., Grotzinger, J. P. & Urban, F. 1998. A composite reference section for terminal Proterozoic strata of southern Namibia. Journal of Sedimentary Research, 68, 1223– 1235. Schalk, K. E. L. 1970. Some late Precambrian formations in central South West Africa. Annals of the Geological Survey of South Africa, 8, 29 – 47.

Chapter 16 The Chameis Gate Member, Chameis Group, Marmora Terrane, Namibia HARTWIG E. FRIMMEL1,2 1

Department of Geological Sciences, University of Cape Town, Rondebosch 7701, South Africa

2

Present address: Geodynamics & Geomaterials Research Division, Institute of Geography and Geology, University of Wuerzburg, Am Hubland, D-97074 Wuerzburg, Germany (e-mail: [email protected]) Abstract: The Chameis Gate Member is a poorly exposed and poorly investigated diamictite in the Chameis Subterrane of the Marmora Terrane, which forms the western, completely allochthonous part of the Pan-African Gariep Belt (southwestern Namibia). Its significance lies in its position in an entirely oceanic unit, the Dernburg Formation, which is dominated by mafic volcanic rocks. The diamictite contains exotic dropstones in a mafic volcaniclastic matrix, thus providing evidence for transport by ice away from the continental margin into an oceanic environment. No direct age data are available and stratigraphic relationships are obscured by limited outcrop and intense syn-orogogenic deformation. Preliminary chemostratigraphic data obtained on carbonate rocks below and above the diamictite, imprecise Pb–Pb age data on the largely volcaniclastic silicate fraction within associated stromatolitic reef carbonates, and imprecise Ar–Ar data on early hornblende related to sea-floor metamorphism of the associated volcanic rocks all point to an age loosely constrained between 640 and 580 Ma. Based on a comparison between the tectono-stratigraphic units of the Marmora Terrane with the continental Port Nolloth Group on the one side and the Rocha Group of the Punta del Este Terrane in Uruguay on the other side of the terrane, it is suggested that the diamictite was deposited in a back-arc basin that developed in response to the 640 –590 Ma volcanic arc of the Dom Feliciano Belt in southeastern Brazil and eastern Uruguay.

The Neoproterozoic Marmora Terrane is an internal, entirely allochthonous tectonic unit in the western part of the Gariep Belt (Fig. 16.1). Although the Marmora Terrane stretches along the Atlantic coast of Namibia from 27.468S southwards into South Africa to 28.768S, just south of the Orange River mouth, exposures of the Chameis Gate Member are restricted to a narrow coastal strip in the so-called Sperrgebiet (‘restricted area’) of southwestern Namibia. As all outcrops of this member are effectively located within the mining lease area of the alluvial diamond operations of Namdeb Pty Ltd. along the coast NW of Oranjemund, access to these outcrops is restricted. Moreover, the outcrops are in a predominantly sandy desert devoid of infrastructure, and much of the rocks of the Chameis Gate Member are covered by shifting sand dunes. As a consequence, this unit is only very poorly investigated. The area was first mapped by Kaiser (1926), but he did not distinguish the diamictite that characterizes the Chameis Gate Member. The largest exposures of the member can be found on wind-blown surfaces to the west and south of a security gate, named Chameis Gate, at 27.8378S, 15.7278E. Minor outcrops can be found at Bakers Bay at 27.6558S, 15.5368E and further NW at 27.4538S, 15.4258E (Fig. 16.2). The Chameis Gate Member is so far the only glaciogenic unit that has been recognized in that subterrane, and it was first described by Frimmel (2000b). Preliminary lithogeochemical and chemostratigraphic data for the bounding strata have been presented by Frimmel & Jiang (2001). No further follow-up studies have been carried out, and much of its stratigraphic interpretation hinges on circumstantial evidence. The main significance of the Chameis Gate Member lies in its palaeodepositional environment. In contrast to most, if not all, other diamictite units described so far, this diamictite occurs associated with oceanic rocks with very little continental influence except for pre-Gariep basement-derived clasts and dropstones in the diamictite (Frimmel & Jiang 2001).

Structural framework and tectonic evolution The Chameis Gate Member forms part of the largely oceanic Marmora Terrane, which represents an allochthonous nappe complex that was thrust in an easterly to southeasterly direction

onto the para-autochthonous continental Port Nolloth Zone along the Schakalsberge Thrust (Fig. 16.1). The Marmora Terrane has been subdivided into three tectono-stratigraphic units, the Schakalsberge, Oranjemund and Chameis complexes (Hartnady & Von Veh 1990) or subterranes (Frimmel 2000b). The tectonically lowest of these units, the Schakalsberge Subterrane, consists mainly of mafic metavolcanic rocks of the Grootderm Formation with a dolomitic, in places stromatolitic to oolitic carbonate on top, interpreted as a former guyot that evolved from an oceanic seamount (Frimmel et al. 1996). The tectonically intervening Oranjemund Subterrane consists of predominantly siliciclastic, commonly turbiditic, metasedimentary rocks, with only very minor carbonate and ferruginous chert, originally referred to as the Oranjemund Formation (Frimmel 2000b) but since raised to the rank of group (Basei et al. 2005). The Chameis Subterrane, which lithostratigraphically comprises the Chameis Group and contains the Chameis Gate Member, is the tectonically highest unit. It is intensely deformed, resembling a tectonic melange zone, which makes a stratigraphic subdivision difficult. The lower part consists of predominantly mafic metavolcanic rocks, similar to those of the Grootderm Formation, whereas the upper part consists of carbonates and siliciclastic metasedimentary rocks that resemble those of the Holgat Formation in the Port Nolloth Zone and those of the upper Oranjemund Group. No basement has been identified in the Chameis Subterrane. Available geochemical and petrological data for the mafic rocks (both extrusive and intrusive) across the Marmora Terrane are most compatible with an oceanic intra-plate setting (Frimmel et al. 1996). For some metagabbros in the Chameis Subterrane, geochemical characteristics of mid-ocean ridge basalt are indicated. Although exact ages for the time of magmatism are lacking, some crude constraints on the timing of sea-floor metamorphism (Frimmel & Frank 1998) and of oceanic volcanism (Frimmel & Fo¨lling 2004), detailed below, suggest an interval sometime between 640 and 580 Ma for the time of oceanic crust formation. Based on zircon provenance data from the Oranjemund Group siliciclastic rocks and comparable rocks from the Rocha Group in the Punta del Este Terrane in eastern Uruguay, a back-arc basin to the 640–590 Ma volcanic arc of the Dom Feliciano Belt in southeastern Brazil and eastern Uruguay has been suggested as the most likely depositional environment for the lower

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 217– 221. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.16

218

H. E. FRIMMEL

Fig. 16.1. Tectonic sub-division of the Marmora Terrane in the western Gariep Belt (modified after Hartnady & von Veh 1990).

stratigraphic units of the Marmora Terrane (Basei et al. 2005). During amalgamation of SW Gondwana, the basin was closed, resulting in syn-orogenic sedimentation in a foredeep in front and on top of already stacked thrust sheets of the Marmora Terrane. This model explains the remarkable similarity in the younger sedimentary deposits across tectonic boundaries in both the Port Nolloth Zone and the Marmora Terrane (Frimmel & Fo¨lling 2004). A penetrative foliation (s1) dips to the west and is axial –planar relative to tight to isoclinal, east-vergent F1 folds. Locally, s1 was refoliated around open to tight, NE-trending F2 kink bands. The contact with the underlying Port Nolloth Zone is a major thrust fault with top-to-SE transport (Schakalsberge Thrust). The intensity of folding increases towards the Schakalsberge Thrust. At the northern boundary of the Marmora Terrane, F1 folds within the Chameis Subterrane are truncated by the Schakalsberge Thrust, highlighting its syn-D2 timing. The contact between the Oranjemund and the Schakalsberge Subterranes is a west- to NW-dipping fault plane. The contact between the Chameis and Oranjemund Subterranes is a zone of intense F2 folding with top-to-SE transport and subsequent back-folding that resulted in locally steeply SE-dipping crenulation. The preferred kinematic model for the evolution of the main orogenic structures involves two stages. The first is eastward transport of the oceanic rocks of the Marmora Terrane, internal stacking of these, and eventually obduction onto the continental margin of the Port Nolloth Zone. The variable angle between principal direction of compression and the pre-existing arcuate north- to NW-trending continental margin led to strain partitioning

Fig. 16.2. Map showing the distribution of the Chameis Gate Member within the Chameis Subterrane, Marmora Terrane (Gariep Belt), in southwestern Namibia.

with NE-, east- and SE-directed transport along the northeastern, eastern and southeastern basin margin, respectively (Frimmel 2000a). This was followed by the second stage, sinistral transpression with overall top-to-SE transport. The resulting folds and thrusts are east-verging along the northeastern lateral ramps and SE-verging along the southern frontal ramps (in South Africa) within the outer Port Nolloth Zone (Gresse 1994). The Marmora Terrane differs from the Port Nolloth Zone not only with regard to stratigraphy and deformation, but also in its metamorphic history. In the Port Nolloth Zone, only one stage of regional metamorphism is discerned, the peak of which (c. 520 8C and 3.0 –3.5 kbar) was reached after D2 but prior to D3; it is ascribed to crustal thickening following overthrusting of the Marmora Terrane onto the Port Nolloth Zone. The

THE CHAMEIS GATE MEMBER

metamorphic evolution of the Marmora Terrane is more complex. There, the mafic and ultramafic rocks experienced polyphase metamorphism that is not recorded by their sedimentary cover. As many as three metamorphic amphibole generations therein reflect different metamorphic conditions of formation (Frimmel & Hartnady 1992) and yield distinctly different Ar –Ar age spectra (Frimmel & Frank 1998). They reflect three metamorphic stages: (i) sea floor metamorphism, (ii) a relatively high-pressure metamorphic event, probably related to first accretion of mafic crust and (iii) regional metamorphism as a result of continental collision.

Stratigraphy The Chameis Gate Member forms part of the Chameis Group, the rocks of which constitute the Chameis Subterrane of the Marmora Terrane. The stratigraphic subdivision of this group given here (Fig. 16.3) follows a preliminary subdivision presented by Frimmel (2000b). It should be noted that the intensity of synorogenic tectonic overprint of the entire subterrane makes it virtually impossible to provide reliable constraints on stratigraphic thickness and, in places, even on stratigraphic boundaries. A thick pile of mafic metavolcanic rocks at the base of the succession (Dernburg Formation) with minor, local intercalations of carbonates (Sholtzberg Member) and diamictite (Chameis Gate Member) is overlain by a carbonate sequence (Dreimaster Member) and a siliciclastic, mainly arenitic succession with turbiditic affinities. The latter two units constitute the Bogenfels Formation. The Dernburg Formation was intruded by a series of tectonically dismembered metagabbro bodies (Bakers Bay Suite).

219

Glaciogenic deposits and associated strata The Chameis Gate Member in the Dernburg Formation The Dernburg Formation, which contains in its upper part the glaciogenic unit of interest here, consists predominantly of thinly laminated greenschist (metatuff), mafic hyaloclastites, metabasalt and serpentinized metapicrite. Compositional layering on a millimetre- to centimetre-scale reflects variations in magma composition from mafic to intermediate. Acidic layers are rare. The metabasalt displays pillow structures in strain-protected domains. In places, mica-rich tuffitic to pelitic beds are intercalated. Locally, the mafic rocks show evidence of intense metasomatism, leading to epidosite, albitite, hornblendite and ophicarbonate. The Chameis Gate Member is distinguished from the remainder of the formation by its diamictite lithology. No stratification is discernable. The matrix of the diamictite is similar to the laminated greenschist (metatuff and -tuffite) and intercalated metapelite above, but the clast composition is distinguishing. It contains exotic lonestones and dropstones, up to 1.5 m in diameter. They comprise various granitoids, gneisses, quartzite and dolomite, and display a variable degree of rounding. In less intensely deformed places, the dark grey pelitic or green chloritic beds are deflected only on the lower side of a given larger clast, thus attesting to its emplacement as a dropstone. Apart from the exotic clast content, no other glacial indicators, such as glacial striations, have been observed. Associated with the metavolcanic and volcaniclastic rocks of the Dernburg Formation are also distinct, in places Fe-rich, dolostone outcrops, distinguished as Sholtzberg Member (Frimmel 2000b).

Fig. 16.3. Generalized lithostratigraphy of the Marmora Terrane and correlation with that of the Port Nolloth Zone, all of which constitutes the Gariep Supergroup (modified after Frimmel & Fo¨lling 2004).

220

H. E. FRIMMEL

Their exact stratigraphic position is unclear because of tectonic contacts, but they appear in the upper part of the formation, close to the diamictite of the Chameis Gate Member. At one locality (27.6078S, 15.0408E), relics of circular stromatolite mounds are preserved. Elsewhere, the dolostone is associated with magnesian metapelite that contains layers or lenses of tourmalinite and albitite. Sodic metasomatism is widespread, resulting in the crystallization of abundant magnesioriebeckite in all lithotypes in and around the Sholtzberg Member. An evaporitic origin of this member is indicated by mineralogical and fluid inclusion data as well as B isotopic evidence (Frimmel & Jiang 2001). For the lack of any continental influence in these deposits (Rb, Zr, Nb, Th and U concentrations are an order of magnitude lower than in normal marine carbonates) an atoll setting on an oceanic seamount with temporary evaporation of seawater in the central basin is envisaged for the Sholtzberg Member depositional environment (Frimmel 2000b).

The Bogenfels Formation The term ‘Bogenfels Formation’ was introduced by Martin (1965) for those rocks that had previously been described as ‘Folded Nama’ (Kaiser 1926). At that time, no distinction was made between the Marmora Terrane and the Port Nolloth Zone. To conform to the current tectonic nomenclature, it has been suggested that the Bogenfels Formation be redefined as the unit of carbonate and siliciclastic rocks above the Dernburg Formation within the Chameis Subterrane (Frimmel 2000b). It thus represents the upper part of the Chameis Group. Its base is a regionally extensive, laterally continuous, laminated, medium-grey limestone, followed by a massive, thick-bedded, fine-grained, light creamywhite to grey dolostone and finally a dolomitic breccia in which the immediately underlying carbonates appear re-worked. This carbonate succession of the Dreimaster Member is overlain by a siliciclastic succession of alternating feldspathic and quartz arenite, chlorite schist and calcpelite. The feldspathic arenite, a greywacke to arkose, displays rare graded bedding of turbiditic appearance. In contrast to the Dernburg Formation, which is entirely different to any unit in the Port Nolloth Zone, the Bogenfels Formation is lithologically comparable to the Holgat Formation of the Port Nolloth Zone (Fig. 16.3).

Boundary relations with overlying and underlying non-glacial units Upper and lower boundaries of the Chameis Gate Member are either tectonic or covered by sand. The spatial distribution of the member is confined to areas made up of metavolcanic and -volcaniclastic rocks of the Dernburg Formation. The lower contact of that formation is not exposed. The upper contact of the Dernburg Formation with the Dreimaster Member (Bogenfels Formation) is sharp and conformable, defined by the first appearance of limestone. The stratigraphic relationship of the Chameis Gate Member with the Sholtzberg Member within the Dernburg Formation is unclear and only inferred because of a lack of primary contacts.

Chemostratigraphy Limited trace element concentrations as well as C, O and Sr isotope data have been reported for carbonates of the Sholtzberg and Dreimaster Members in the stratigraphic vicinity of the glaciogenic Chameis Gate Member (Frimmel 2000b; Frimmel & Jiang 2001; Frimmel & Fo¨lling 2004). 87Sr/86Sr ratios of 0.7075– 0.7078 were obtained on evaporitic dolostone of the Sholtzberg Member

with very low Rb –Sr ratios (,0.0005). The presence of perfectly preserved dolomitic ooids in the succession and the lack of geochemical evidence of diagenetic alteration suggest that the above Sr isotope ratios are very close to primary. That dolostone has positive d13CCarb ratios (as much as 2.82‰ relative to the PDB standard) and d18OCarb as high as –4.3‰ (relative to PDB), whereas the limestone of the younger Dreimaster Member is characterized by consistently negative d13CCarb ratios slightly below zero, with the lowest ratios ( –2.6‰) determined within the first few metres above the bottom contact. A further distinctive feature of this limestone is its high Sr content (as much as 1412 ppm).

Other characteristics No economic ore deposits are known from the Chameis Subterrane, but base metal anomalies have been noted, particularly in association with Na-metasomatism that affected most of the area (H. E. Frimmel, unpublished data). The fluids involved were highly oxidizing, which is evident from the widespread occurrence of hematite veins. The metasedimentary rocks of the Marmora Terrane still await a micropalaeontological investigation, which might help in further constraining their age.

Palaeolatitude and palaeogeography No reliable palaeomagnetic data are available for the entire Gariep Belt and thus also for the Marmora Terrane. The low-grade regional metamorphic overprint most likely led to the remagnetization of the rocks during the Gariepian orogeny. Thus, any assumption about palaeolatitude and palaeogeography is based only on circumstantial evidence and comparison with continental-scale models (e.g. Collins & Pisarevsky 2005). The presence of former evaporite deposits in the Marmora Terrane speak against a high-latitude position at the time of sedimentation.

Geochronological constraints The minimum age is provided by the age of metamorphism related to the accretion of the Marmora Terrane, which has been constrained at 575 + 2 Ma by Ar –Ar data on a second, relatively high-pressure metamorphic amphibole generation in metagabbro of the Bakers Bay Suite (Frimmel & Frank 1998). Slightly older Ar –Ar ages of 600–610 Ma, obtained in the same study on an early, very low-pressure metamorphic amphibole generation related to sea floor metamorphism, might set a minimum constraint on the timing of oceanic crust formation, but are subject to a large uncertainty. An attempt to date the dolostone of the Sholtzberg Member by the double-spike Pb– Pb technique did not yield a reliable result for the carbonate fraction. Its silicate fraction, controlled by minute, presumably volcanogenic zircon grains, gave an imprecise age of 598 + 60 Ma (Frimmel & Fo¨lling 2004), which is in agreement with the mentioned Ar –Ar data.

Discussion The presence of exotic dropstones in marine, volcaniclastic rocks, the massive, unbedded nature of the diamictite and lack of graded bedding support a glaciomarine origin of the Chameis Gate Member, but no definitive glaciogenic features have been observed. Deposition distal from a glaciated continental margin by processes of settling from suspension and ice rafting is suggested. The Dreimaster Member carbonates are interpreted as a cap-carbonate sequence based on the stratigraphic position and C isotopic compositions equivalent to those of the post-Numees Bloeddrif Member in the Port Nolloth Zone.

THE CHAMEIS GATE MEMBER

The lack of precise radiometric age control for the Chameis Gate Member precludes a reliable stratigraphic correlation with glaciogenic deposits elsewhere. Although the available data base is rather limited, tentative correlation with other glaciogenic deposits in the region can be attempted on the basis of Sr and C isotope data on over- and underlying carbonates and the limited radiometric age data. The C isotopic evidence for the Sholtzberg Member is ambiguous. The noted enrichment in 13C therein could be merely a reflection of elevated evaporation rates. However, its Sr isotopic composition is lower than in late Ediacaran carbonates, but compares well with that reported regionally for the Pickelhaube Formation (Fo¨lling & Frimmel 2002) and globally for early to middle Ediacaran carbonates (Halverson et al. 2007). Fieldwork in the Sperrgebiet would have been impossible without the logistic support and permissions to enter restricted areas granted by Namdeb Pty Ltd. Funding by the South African National Research Foundation is gratefully acknowledged. Comments from G. Halverson in particular helped in improving the manuscript. This work represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Projects #512 and #478.

References Basei, M. A. S., Frimmel, H. E., Nutman, A. P., Preciozzi, F. & Jacob, J. 2005. A connection between the Neoproterozoic Dom Feliciano (Brazil/Uruguay) and Gariep (Namibia/South Africa) orogenic belts — evidence from a reconnaissance provenance study. Precambrian Research, 139, 195–221. Collins, A. S. & Pisarevsky, S. A. 2005. Amalgamating eastern Gondwana: the evolution of the circum-Indian orogens. Earth-Science Reviews, 71, 229–270. Fo¨lling, P. G. & Frimmel, H. E. 2002. Chemostratigraphic correlation of carbonate successions in the Gariep and Saldania Belts, Namibia and South Africa. Basin Research, 14, 69– 88.

221

Frimmel, H. E. 2000a. The Pan-African Gariep Belt in southwestern Namibia and western South Africa. Communications of the Geological Survey of Namibia, 12, 197–209. Frimmel, H. E. 2000b. The stratigraphy of the Chameis Sub-terrane in the Gariep Belt in southwestern Namibia. Communications of the Geological Survey of Namibia, 12, 179–186. Frimmel, H. E. & Fo¨lling, P. G. 2004. Late Vendian closure of the Adamastor Ocean: timing of tectonic inversion and syn-orogenic sedimentation in the Gariep Basin. Gondwana Research, 7, 685– 699. Frimmel, H. E. & Frank, W. 1998. Neoproterozoic tectono-thermal evolution of the Gariep Belt and its basement, Namibia/South Africa. Precambrian Research, 90, 1 – 28. Frimmel, H. E. & Hartnady, C. J. H. 1992. Blue amphiboles and their significance for the metamorphic history of the Pan-African Gariep belt, Namibia. Journal of Metamorphic Geology, 10, 651– 669. Frimmel, H. E. & Jiang, S.-Y. 2001. Marine evaporites from an oceanic island in the Neoproterozoic Adamastor ocean. Precambrian Research, 105, 57– 71. Frimmel, H. E., Hartnady, C. J. H. & Koller, F. 1996. Geochemistry and tectonic setting of magmatic units in the Pan-African Gariep Belt, Namibia. Chemical Geology, 130, 101–121. Gresse, P. G. 1994. Strain partitioning in the southern Gariep Arc as reflected by sheath folds and stretching lineations. South African Journal of Geology, 97, 52– 61. ¨ ., Maloof, A. C. & Bowring, S. A. Halverson, G. P., Dudas, F. O 2007. Evolution of the 87Sr/86Sr composition of Neoproterozoic seawater. Palaeogeography, Palaeoclimate, Palaeoecology, 256, 103– 129. Hartnady, C. J. H. & von Veh, M. W. 1990. Tectonostratigraphic and structural history of the Late Proterozoic –Early Palaeozoic Gariep Belt, Cape Province, South Africa. Geological Society of South Africa, Cape Town. Kaiser, E. 1926. Die Diamanten Wu¨ste. Dietrich Reimer, Berlin. Martin, H. 1965. The Precambrian geology of South West Africa and Namaqualand. Precambrian Research Unit, University of Cape Town, Bulletin, 4, 1– 177.

Chapter 17 The Kaigas and Numees formations, Port Nolloth Group, in South Africa and Namibia HARTWIG E. FRIMMEL1,2 1

Department of Geological Sciences, University of Cape Town, Rondebosch 7701, South Africa

2

Present address: Geodynamics & Geomaterials Research Division, Institute of Geography and Geology, University of Wuerzburg, Am Hubland, D-97074 Wuerzburg, Germany (e-mail: [email protected]) Abstract: The Port Nolloth Group makes up the eastern, external part of the Pan-African Gariep Belt (Port Nolloth Zone) in southern Namibia and western South Africa. It contains two glaciogenic diamictite units, the older Kaigas Formation and the younger Numees Formation, with intercalated and overlying carbonate-dominated units. Available chemostratigraphic information include O, C and Sr isotope data. Micropalaeontological and geochronological data point to an early Cryogenian age (c. 750 Ma) of the Kaigas Formation and possibly a middle Ediacaran age (c. 580 Ma) for the Numees Formation. The former was deposited in an evolving, but eventually failed, continental rift on the western flank of the Kalahari Craton, probably at low latitude. The Numees Formation is a laterally continuous, up to 600-m-thick glaciomarine deposit for which a passive continental margin setting has been suggested. Alternatively, based on more recent data, the depositional setting might have been a back-arc basin. The eroded remnants of the corresponding arc are present in the Dom Feliciano Belt.

The Kaigas and Numees Formations represent the older and younger glaciogenic diamictite-dominated units, respectively, within the para-authochthonous Port Nolloth Zone of the PanAfrican Gariep Belt (Fig. 17.1). Owing to intense transpressional overprinting of large parts of this zone, the distinction between the two formations has been unclear for decades. This has led to considerable confusion in the stratigraphic terminology used in the older literature, with some workers even having argued for the existence of only a single diamictite unit in the entire belt (Ha¨lbich & Alchin 1995; Jasper et al. 2000). Regional mapping (von Veh 1993; Frimmel 2008), as well as the distinctly different litho-, chemo-, chrono- and biostratigraphic characteristics in the respective overlying carbonate-rich successions (Fo¨lling et al. 2000; Fo¨lling & Frimmel 2002; Frimmel & Fo¨lling 2004; Gaucher et al. 2005), however, have demonstrated the existence of the two separate diamictite units at different stratigraphic positions. Here, the stratigraphic nomenclature as accepted by the South African Commission on Stratigraphy (Frimmel 2003a) is followed but with the understanding that from a sequence stratigraphic perspective, it is in need of revision. Rocks of both the Kaigas and Numees Formations are exposed at several sites in southernmost Namibia and in the adjoining part of South Africa (Fig. 17.2). The spatial distribution of outcrops follows that of the outer zone of the Gariep Belt, which occurs in an arcuate, south-trending, 20–60-km-wide zone stretching from Lu¨deritz in Namibia, across the Orange River to the western Richtersveld region of the Northern Cape Province in South Africa, and striking out to sea south of Port Nolloth (Fig. 17.1). The Numees Formation, including outcrops that are now assigned to the Kaigas Formation, was the first Neoproterozoic unit in southern Africa to have been recognized as being of glacial origin (Rogers 1916; Beetz 1926), an interpretation that has never been seriously challenged by subsequent workers (Martin 1965a). The original Kaigas Series of Rogers (1916) included most of the strata now assigned to the Hilda Subgroup. The Kaigas Formation as redefined by De Villiers & So¨hnge (1959) was limited to beds along the Kaigas (also known as Gaigas) River, where excellent, almost continuous exposures along a ridge to the east of the river bed at 28.6318S, 17.0988E (Fig. 17.2) constitute the holostratotype (Fig. 17.3). Diamictite in the type area for the Numees Series of Rogers (1916) was later included by von Veh (1993) in the Kaigas Formation because of lithological similarities. Likewise, the diamictite exposures west and NW of Numees Peak

at the so-called Numees Prospect (28.29168S, 16.97018E) were initially regarded as part of the Numees Formation (De Villiers & So¨hnge 1959) but since have been recognized to be older and part of the Kaigas Formation (Kro¨ner 1974; von Veh 1993). Although the name Numees Formation is derived from a place (Numees Prospect) that is now assigned to the Kaigas Formation, the formation name is retained because the main occurrence of the Numees Formation is only about 5 km to the west and the name is firmly entrenched in the literature. The type area for the Numees Formation is south of the Orange River at 28.18 –28.338S, 16.87 –16.938E. Following the initial description of the South African portion of the Port Nolloth Zone by Rogers (1916), the area was remapped in greater detail by De Villiers & So¨hnge (1959). Subsequent structural and lithostratigraphic studies by Kro¨ner (1974) and von Veh (1993) laid the foundation for the current lithostratigraphic subdivision of the entire zone (Frimmel 2003b). The Namibian sector was mapped in greater detail by McMillan (1968). Davies & Coward (1982) provided a first interpretation of the Gariep Belt’s structural evolution. This structural work was followed up in the South African sector of the Port Nolloth Zone by Gresse (1994) and in the Namibian sector by Ha¨lbich & Alchin (1995) and Jasper et al. (2000). An attempt to resolve the sedimentology of parts of the stratigraphic succession was made by Jasper et al. (1992 –1993). Metamorphism (Frimmel 1995), geochemistry and tectonic setting of the various igneous bodies (Frimmel et al. 1996a), timing of tectonic events and the thermal evolution have also been studied (Reid et al. 1991; Grotzinger et al. 1995; Frimmel et al. 1996b; Frimmel & Frank 1998). Attempts to place the rock record of the Port Nolloth Zone, including the Kaigas and Numees Formations, into a regional tectonic framework and to correlate units on a subcontinental scale have been made by Frimmel et al. (2002). Our current understanding of the tectonic settings for the deposition of the glaciogenic deposits is based on work by Basei et al. (2005), and the lithostratigraphic correlation has been updated by Gaucher et al. (2005).

Structural framework and tectonic evolution In the past, the entire Port Nolloth Group was interpreted as having been deposited in a basin (Gariep Basin) that continuously

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 223– 231. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.17

224

H. W. FRIMMEL

Fig. 17.1. Position of the Port Nolloth Zone within the Gariep Belt (modified after Hartnady & von Veh 1990).

evolved from a continental rift to a passive margin and eventually became a foredeep prior to continental collision at c. 545 Ma (Frimmel et al. 1996a; Frimmel & Frank 1998). An alternative to the traditional Wilson Cycle model for the Gariep Belt has emerged from a comparison with the tectono-thermal evolution of the South American counterparts in southeastern Brazil and eastern Uruguay (Basei et al. 2005; Frimmel & Basei 2006). According to these more recent studies, the Gariep Basin might have been a successor basin in which sediments had been deposited in two different basins and were subsequently stacked on top of each other. The Kaigas Formation belongs to an older sequence that reflects progressive continental rifting. Predominantly siliciclastic sediments of the Stinkfontein Subgroup were first deposited into a widening rift graben, bounded by listric faults. The presence of debris flow and turbidity fan deposits as well as large slump masses indicates widespread syn-depositional deformation related to continental rifting. Large, but only locally developed dolomitic olistostromes near the northeastern basin margin are explained by seismic activity along active, basin-bounding growth faults. These normal faults were subsequently inverted during the Gariepian orogeny, but the position of some of the major syn-rift faults is still recognizable. A major growth fault must have existed along today’s eastern to northeastern margin of the belt, separating basement rocks to the east and NE from the rift graben fill. This fault is now present as an inverted NW –SE trending thrust fault system, with the Rosh Pinah Fault being a good example. Along those faults, an elongated, roughly north –south trending basement horst (Aurus Horst) developed, separating the so-called Rosh

Fig. 17.2. Map showing the distribution of the Kaigas and Numees Formations within the Port Nolloth Zone (made up of rocks of the Port Nolloth Group), Gariep Belt, in southern Namibia and westernmost South Africa (western Richtersveld). Also shown are the localities of the type sections for the Kaigas and Numees Formations.

Pinah Graben in the east, with a number of smaller, separate basins, from a half-graben to the west. At an advanced rifting stage, eustatic sea-level drop was accompanied by the entry of glaciers into the basin along its eastern margin to the mainland. Debris released from those glaciers makes up the bulk of the Kaigas Formation. This stage was accompanied by bimodal, predominantly felsic, rift volcanism (Rosh Pinah Formation) on a growth fault along the eastern basin margin. Although still in an overall extensional regime, the eustatic drop in sea level led to the starvation of the Rosh Pinah Basin and consequently to anoxic bottom waters, which provided a suitable environment for sedimentary – exhalative base-metal mineralization (Rosh Pinah mine and protore of the secondary Skorpion deposits). Renewed transgression led to the drowning of parts of the basement horst

KAIGAS AND NUMEES FORMATIONS

225

the emergence of the Dom Feliciano volcanic arc towards the (north)west. Consequently, the upper Hilda Subgroup and the overlying Numees Formation would have been deposited in a back-arc basin behind the Dom Feliciano arc. Inversion from extension to compression took place during post-Numees sedimentation. A foredeep developed in which the youngest sediments of the Port Nolloth Group (Holgat Formation) were deposited (Frimmel & Fo¨lling 2004). At the same time, first stacking and accretion of the advancing oceanic rock pile in the shrinking basin took place. This was followed by continental collision in a SE-directed transpressional regime, which was accompanied by regional low-grade metamorphism, east-vergent folding, thrusting and back-thrusting. Many of the original normal faults were reactivated at that stage (von Veh 1993; Ha¨lbich & Alchin 1995). In places, syn-tectonic foliation occurred at a high angle to bedding, but primary sedimentary contacts remained well preserved in strain-protected domains.

Stratigraphy Both the Kaigas and Numees Formations form part of the Port Nolloth Group (Fig. 17.4), which, in turn, is the part of the Gariep Supergroup that makes up the bulk of the Port Nolloth Zone. The Port Nolloth Group unconformably overlies rocks of the Mesoproterozoic Namaqua Metamorphic Province and, in places, alkaline intrusive rocks of the 833– 771 Ma Richtersveld Suite (Frimmel et al. 2001). It is, in turn, unconformably overlain by the Kuibis Subgroup of the Nama Group in the east, and towards the west it is overlain tectonically by the Marmora Terrane, the allochthonous internal sector of the Gariep Belt. The stratigraphy of the Port Nolloth Group begins with a predominantly siliciclastic, minor volcanic to volcaniclastic continental rift graben fill, the Stinkfontein Subgroup. This is followed by the stand-alone Kaigas Formation, which, in turn, is overlain by the Hilda Subgroup. Within the Hilda Subgroup, four formations are distinguished, some of which are time-equivalents of each other.

Fig. 17.3. Holostratotype of the Kaigas Formation (for location see Fig. 17.2; modified from Frimmel et al. 2003a).

as well as the Kaigas Formation and the deposition of a varied, carbonate-dominated succession (Pickelhaube Formation). One of the great enigmas of the Gariepian stratigraphy has been the lack of sediment in the post-rift successions. If a successful rift is assumed for the tectonic evolution of the Gariep Basin (Adamastor Ocean), as postulated by most workers in the past (Hartnady et al. 1985; Germs 1995; Frimmel et al. 1996a), the post-Kaigas period should have been marked by further transgression and the development of a thick sedimentary passive margin sequence. First regional drowning of the rift shoulders is indeed indicated by the laterally continuous post-Kaigas carbonates, and the various carbonate-dominated facies of the Hilda Subgroup provide good evidence of the development of carbonate platforms. Yet the total thickness of the Hilda Subgroup (except for the volcanically controlled Rosh Pinah area) is not more than 650 m. This lack of sediment is explained by a major hiatus and/or slope failure as the Hilda Subgroup is largely composed of turbidites, olistostromes and potentially cannibalizing debris flows. Fieldwork by the author has shown that the erosive surface beneath the siliciclastic Wallekraal Formation within the Hilda Subgroup extends across the entire Port Nolloth Zone and is a far more widespread feature than previously recognized. It is a first-order sequence boundary. The temporal extent of the pre-Wallekraal hiatus remains unresolved. If the above back-arc model applies, the original basinbounding faults would have become re-activated in response to

Fig. 17.4. Generalized lithostratigraphy of the Port Nolloth Group (Gariep Supergroup); unit thickness in metres given in parentheses.

226

H. W. FRIMMEL

In the volcanically influenced basin margin, the Kaigas Formation diamictite is followed by the volcanic to volcaniclastic Rosh Pinah Formation, whereas elsewhere it is followed by the carbonatedominated Pickelhaube Formation. The former comprises rhyolitic lava flows, subaerial and submarine pyroclastic deposits of highly variable grain size (from agglomerate to ash tuff and ignimbrite), whereas the latter consists predominantly of carbonate rocks (laterally discontinuous, massive dolomitic marble and laminated to medium-bedded siliceous and graphite-bearing dolomitic, partly allodapic limestone, with slump breccias) and subordinate feldspathic arenite and graphitic metapelite alternating locally with calcpelite. These are overlain paraconformably, or with an erosional unconformity, by coarse-grained clastic sedimentary rocks of the Wallekraal Formation in which parts of the older Pickelhaube carbonate platform appear re-worked. The Wallekraal Formation consists predominantly of very coarse-grained feldspathic arenite and less abundant well-rounded polymict conglomerate, phyllitic to calcareous, locally graphitic, metapelite, isolated lenses of dolostone, typically re-worked into olistostromes, and thin-bedded allodapic limestone. The youngest unit of the Hilda Subgroup is the Dabie River Formation, which consists essentially of stromatolitic reef facies, including dolomite breccia and oolite. The Hilda Subgroup is followed by diamicite and minor banded iron-formation of the Numees Formation, which, in turn, is overlain by the Holgat Formation. The latter commences with a carbonate sequence above the Numees Formation diamictite (Bloeddrif Member), and in places a thick package of turbiditic, siliciclastic metasedimentary rocks. Because glaciogenic deposits are the emphasis of this paper, the stratigraphic description of the two glaciogenic units and those that are associated with them follow in more detail below.

Glaciogenic deposits and associated strata Descriptions of the two principal glaciogenic units in the Port Nolloth Group, the Kaigas and Numees Formations, have been provided by von Veh (1993), Frimmel (2000, 2008) and Frimmel et al. 2003a, b). To also obtain some information on the rock record that provides possible information on the events that led up to the glacial intervals and those that followed, the bounding stratigraphic units below and above the glaciogenic units will be also described briefly.

The Vredefontein Formation Strata underlying the Kaigas Formation are known as the Vredefontein Formation (Stinkfontein Subgroup). The Vredefontein Formation is up to 300 m thick. Internal sedimentary structures, accentuated by heavy mineral concentrations along laminations, are well preserved in the dominantly medium-bedded feldspathic arenite. These include various forms of cross-bedding, asymmetrical ripple marks indicating palaeocurrent directions from the east or SE, as well as various soft sediment deformation features (von Veh 1993). In the upper part of the formation, minor felsic, intermediate and mafic metavolcanic rocks are intercalated (Middlemost 1963). These rocks are only known from the Richtersveld in South Africa, but they provide an important reference for the genesis of more abundant volcanic rocks at a higher stratigraphic position in Namibia (Rosh Pinah Formation). Sandy to gritty breccia that occurs towards the top of the Vredefontein Formation contains fragments of angular to sub-rounded vein quartz, granitoid, microgranite, gneiss, orthoquartzite, calcareous grit and felsic volcanic rocks in a massive or indistinctly cross-bedded feldspathic grit matrix. The distribution of the Stinkfontein Subgroup in Namibia remains enigmatic, because of the occurrence of similar

siliciclastic successions higher up in the stratigraphy, lateral facies changes that make long-distance correlation with the type sections in South Africa problematic, and the commonly tectonic nature of its upper and lower contacts. Occurrences of the subgroup are confined to the northeastern basin margin. Lithologically, the Stinkfontein Subgroup in Namibia is similar to the Vredefontein Formation in the Richtersveld, whereas the relatively clean quartz arenite, typical of the Lekkersing Formation in the lower part of the Stinkfontein Subgroup, is only poorly developed in Namibia. The overall higher feldspar content in the siliciclastic rift deposits in Namibia might reflect an overall more proximal position of the depositional environments relative to the area further south in South Africa. This finds support from the observation that along the NW –SE trending branch of the basin north of the Aurus Horst, the entire Stinkfontein Formation is missing and rocks of the next younger stratigraphic units, the Kaigas Formation, where present, or the Hilda Subgroup rest directly on the basement.

The Kaigas Formation The Kaigas Formation is only locally developed along the eastern and northeastern margin of the basin. It is characterized by its marked variations in thickness along strike, with individual occurrences pinching out over only several hundred metres. At its type locality (Fig. 17.2) in the southern Richtersveld (South Africa), it reaches a thickness of 115 m. The formation consists predominantly of medium- to thick-bedded diamictite and subordinate massive, locally cross-bedded or graded bedded, feldspathic arenite and argillite. Ripple cross laminations, indicating palaeocurrent directions from the east, rip-up of mudstone clasts, load and flute casts, and sinuous ripple marks are locally present. The diamictite beds display sharp upper and lower contacts and lack internal structures except for crudely graded bedding. The clast size distribution is strongly bimodal with generally larger clasts present in thicker diamictite beds. In places, outsized erratic blocks can also be found in pelitic beds. An overall decrease in clast size away from the former basin margin is noted. Individual clasts and blocks commonly appear sandwiched (cold deformation by ice). Faceted clasts can be found in less deformed regions. Unequivocal glacial striations are difficult to identify because of the tectonic overprint under greenschist-facies temperatures, which modified most clast surfaces. Locally, dolostone is a common clast lithology in the southern part of the Port Nolloth Zone, but elsewhere the Kaigas Formation diamictite is notably free of carbonate clasts. The clast population is generally dominated by basement-derived rock types, such as granite, gneisses, amphibolite, schists and re-worked arenite from the underlying Stinkfontein Subgroup. The general lack of carbonate clasts (with the exception of the southern Port Nolloth Zone) is a major feature by which the Kaigas Formation differs from the younger Numees Formation. Further features that make possible a distinction of the Kaigas from the Numees Formation are the absence of a near-basal banded iron-formation or ferruginous arenite zone and the relative abundance of interbedded arenite and pelite. In Namibia, the formation reaches its maximum thickness in an area east of the Aurus Mountains around Borehole 5 at 16.5278E, 27.6798S (Fig. 17.2), where it rests directly on basement. From there it pinches out rapidly towards the SE, but re-emerges in several small occurrences along the basement edge to the north and SE of Rosh Pinah (Alchin et al. 2005; Frimmel 2008). On the farm Spitzkop 111, the Kaigas Formation diamictite grades laterally into a coarse-grained, proximal pyroclastic deposit (at 16.7068E, 27.8278S) that is assigned to the Rosh Pinah Formation, indicating sedimentation that overlapped in time with volcanism. Elsewhere, the background sedimentation towards the top of the

KAIGAS AND NUMEES FORMATIONS

Kaigas Formation gradually changes from siliciclastic to calcareous. An example of that can be found east of the Aurus Mountain at 16.4868E, 27.6518S, where the matrix of the diamictite and intercalated arenite becomes progressively dolomitic towards the top of the formation, thus heralding a fundamental change in the style of sedimentation that led to the next stratigraphic unit.

The Pickelhaube and Rosh Pinah Formations Where the Kaigas Formation is developed, the Pickelhaube Formation rests on it conformably. Elsewhere, the Pickelhaube Formation occurs with an erosional unconformity on basement. It starts with laminated, variably dolomitized, medium to dark grey limestone, followed by a varied sequence of predominantly argillite and marl with minor feldspathic quartz arenite in which small-scale channels and asymmetric ripple marks indicate a palaeocurrent direction from the SE. Intercalated are a number of limestone and dolostone beds, which become more prominent towards the top of the formation. The absence of conglomerate, scarcity of cross-bedding, presence of thin bedding, and the increasing carbonate content of the Pickelhaube Formation arenites indicate quieter, submerged conditions compared to the underlying Stinkfontein Subgroup and Kaigas Formation, as can be expected in the distal parts of a fan complex. Some of the clastic beds may represent turbidites and minor debris flow deposits. The original limestone deposits are largely dolomitized in the more proximal positions but retained their syn-sedimentary mineralogy farther away from the basin margin. Laminated allodapic limestone (calcarenite) is a ubiquitous feature of this formation, which reaches a maximum thickness of about 280 m. Where the formation onlaps directly onto basement, basementderived clasts are embedded in allodapic limestone of the lower Pickelhaube Formation. The carbonate beds in the middle of the formation show extensive syn-sedimentary brecciation that is reminiscent of tempestite. This is in contrast to slump breccias and carbonate debris flow deposits, some of them containing extrabasinal clasts, which are common in the thicker carbonate beds. The Rosh Pinah Formation (Kapok Formation of Martin 1965b) is distinguished from the Pickelhaube Formation by the presence of felsic volcanic and volcaniclastic rocks in the former. The greatest accumulation of felsic volcanic rocks, also referred to as Spitzkop Formation by some authors (Siegfried & Moore 1990), occurs some 15 –20 km north of Rosh Pinah (on farm Spitzkop 111) and is considered to represent the main volcanic centre. The Rosh Pinah Formation displays repetitive sedimentary cycles that reflect quiescent periods intervening with rapid deposition related to reactivation on basin bounding faults with subsequent thermal and mechanical subsidence (Alchin et al. 2005). Although deposition of the volcanosedimentary succession was sustained in the more active, proximal parts of the basin, distal background sedimentation ranged from deposition of planar laminated allodapic limestone to mudstone, followed by platform carbonates (Pickelhaube Formation). In proximal positions, tilting of the rift shoulders led to partial erosion of the Rosh Pinah Formation and thus to intraformational breccias and olistostromes. Volcanic lithotypes in the Rosh Pinah Formation comprise massive to flow banded quartz-alkali feldspar rhyolite to rhyodacite, hyaloclastic breccias, often locally reworked (lapilli tuff breccias), and a variety of volcaniclastic units, which reflect different energy and distance from the eruptive centre. They range from proximal agglomerate, to lapillistone, and distal coarse- and finegrained tuffs. Inversely graded lapillistone/tuff units are present and indicate water-lain pumice ash flow deposits. A special variety is an alkali feldspar-rich crystal tuff to tuffite. A bimodal character of the Rosh Pinah volcanism is indicated by the local presence of metabasalt and metagabbro and the lack of igneous rocks of intermediate composition (Frimmel et al. 1996a). Sedimentary rocks directly associated with the volcanic facies

227

comprise intercalated ripple-marked quartzite and dolostone, which is, in places, highly ferruginous due to fumarolic activity (Frimmel & Lane 2005).

The Numees Formation In places where the stromatolitic reef facies of the Dabie River Formation is not developed, the siliciclastic Wallekraal Formation grades conformably into the Numees Formation. The stratotype section of the formation as accepted by the South African Committee for Stratigraphy (Frimmel et al. 2003b) is located in the Richtersveld region in westernmost South Africa (Fig. 17.5). That area suffers, however, from considerable tectonic complication and less tectonized sections of the formation occur in the Namibian part of the belt. The total thickness of the Numees Formation is estimated to be not more than 600 m, but decreases towards the east where the formation onlaps directly onto preGariep basement. The dominant lithology is a generally massive, very thick bedded (bedding poorly defined) diamictite with small to very large (2 mm –10 m) subrounded clasts. The lithology of the clasts comprises pegmatite, leucogranite, granitic-gneiss, quartzite, schist, dolostone, minor limestone of the immediately underlying Dabie River Formation and re-worked diamictite. Carbonate clasts are particularly abundant in the lower parts of the formation but become rare in the upper parts, where basementderived clasts dominate. Clast and matrix grain sizes decrease upwards. Intercalated minor ferrugineous, feldspathic arenite is coarse-grained, medium-bedded and forms the matrix of those parts of diamictite that are dark grey to blue-green in colour. The latter becomes dark brown on weathered surfaces. In addition, thin-bedded, varve-like siltstone beds with intercalated coarsergrained, graded sandstone beds occur. Some of these beds contain dropstones with soft-sediment deformation around their bottom contact. A highly ferruginous unit, consisting of thinly laminated chlorite schist, iron-formation and diamictite, is distinguished as the Jakkalsberg Member near the base of the formation (Frimmel et al. 2003b). This member is best developed in the southern portion of the belt in South Africa (Fig. 17.5), but is largely missing in Namibia (except for a few outcrops just north of the Orange River). In places, exotic, mainly basement-derived dropstones are found within the iron-formation.

The Bloeddrif Member (Holgat Formation) On top of the Numees Formation follows a distinct carbonate unit that is distinguished as the Bloeddrif Member within the otherwise siliciclastic Holgat Formation. The true thickness of the formation is difficult to constrain because of intense folding and thrust duplication. It appears, however, that stratigraphic onlap led to a decrease in thickness from several hundred metres in the west to less than 100 m in the east. The formation is best developed in the western Richtersveld in South Africa, where the type sections are located. In less deformed regions near the basin margin, such as some 17 km to the NE of Rosh Pinah and on farm Nord Witputs 22 as well as on farm Namuskluft east of Rosh Pinah, internal, vertical tube-like structures of infilled micritic sediment and cement form a conspicuous feature of the Bloeddrif Member. They resemble structures that have been described from many other post-glacial carbonates (Cloud 1974), and have been variably interpreted as microbial in origin (Hegenberger 1993; Hoffman et al. 1998; Corsetti & Grotzinger 2005) or by gas escape following the destabilization of gas hydrate during warming of terrestrial permafrost (Kennedy 2001). Some microbial activity in the Bloeddrif Member is clearly indicated by the local presence of stromatolites. More distal, relatively pure limestone deposits that are

228

H. W. FRIMMEL

Fig. 17.5. Location of the holostratotypes of the Numees Formation (a) and the Jakkalsberg Member (b). The type section of the Numees Formation is 380 m thick, rests in fault contact on feldspathic arenite of the Wallekraal Formation, and begins with 30 m of thickly bedded diamictite. Banded iron-formation, thinly laminated Fe-rich metapelite and quartzite of the Jakkalsberg Member, displaying graded bedding, follow conformably and attain a thickness of 5 m. They are overlain by 95 m of Fe-rich, green, thickly bedded diamictite and feldspathic arenite intercalations, locally displaying horizontal lamination. The diamictite is rich in dolomite clasts. This is followed by 250 m of thickly bedded, yellow-brown diamictite with only few dolomite clasts but predominantly basement-derived granite and gneiss clasts. The upper contact of the formation is tectonic, with a repetition of the Jakkalsberg Member (from Frimmel et al. 2003b).

characteristically devoid of continental detritus can be distinguished from more proximal deposits that experienced more intense, early diagenetic dolomitization and have intercalations of thin arenite beds (Frimmel & Fo¨lling 2004). A particularly good example of a very shallow, proximal facies of this member occurs to the north of Aurus Mountain at 16.2168E, 27.4538S, where a thinly laminated dark grey to cream-coloured limestone with numerous erosion surfaces and mud cracks grades upwards into a chaotic dolomitic breccia that is unconformably overlain by white, locally cross-bedded dolomitic arenite that became disrupted and filled with sand dykes (Frimmel 2008). In most places, the laminated limestone is overlain by a massive, dark grey dolostone that is followed by a thickly bedded, distinctly pink dolostone. It should be noted, however, that the distal sections through the Holgat Formation appear continuous without breaks in sedimentation.

towards the top of the formation. The upper contact with the Pickelhaube Formation is in many places conformable, but locally (mostly in South Africa) unconformable. The lower contact of the Numees Formation to the underlying Hilda Subgroup is locally conformable (Kro¨ner 1974), although in most places it is tectonic. Wherever the Dabie River Formation carbonates are not developed, the siliciclastic Wallekraal Formation grades conformably into the Numees Formation. The upper boundary of the Numees Formation with the carbonates of the Bloeddrif Member is typically sharp, conformable or paraconformable. The stratigraphic relationships of the Jakkalsberg Member within the Numees Formation with the bounding units are quite clear from a regional perspective. The iron-formations are always, without a single exception, near the bottom of the Numees Formation diamictite, and with conformable contacts, either sharp or gradational. To take those out and place them into a different stratigraphic position altogether is not supported by the field evidence.

Boundary relations with overlying and underlying non-glacial units Chemostratigraphy The boundary relations of the Kaigas Formation with the underlying strata are marked by the laterally discontinuous distribution of the formation. At the type locality in the southern part of the Port Nolloth Zone, the contact of the Kaigas Formation with the underlying feldspathic arenite of the Vredefontein Formation is conformable and gradational, and is defined by the first appearance of diamictite. Further north, the contact with the underlying Stinkfontein Subgroup is, in many places, an angular discordance, or the Kaigas Formation rests directly on the pre-Gariep basement, but a conformable contact exists NE of Rosh Pinah (Fig. 17.4). On the farm Spitzkop 111 north of Rosh Pinah, the Kaigas Formation diamictite grades laterally into a coarse-grained, proximal pyroclastic deposit of the Rosh Pinah Formation, indicating sedimentation that overlapped in time with volcanism. Elsewhere, the background sedimentation towards the top of the Kaigas Formation gradually changes from siliciclastic to calcareous; the matrix of the diamictite and intercalated arenite becomes progressively dolomitic

The first stable isotopic compositional data for carbonates from the Port Nolloth Zone were reported by Kaufman et al. (1991), but those data suffered from poor stratigraphic control. Since then, a number of chemostratigraphic profiles with isotope ratios carefully screened for the effects of post-depositional alteration, through various sections of carbonate-dominated successions in the Port Nolloth Zone (Fo¨lling & Frimmel 2002), have helped in distinguishing between post-Kaigas and post-Numees carbonates because of distinctly different isotopic trends. The former are characterized by an increase in d13CCarb from negative ratios (,4‰ relative to PDB standard) above the Kaigas Formation diamictite to values as high as þ8.7‰ further up-section. This trend is accompanied by a decrease in 87Sr/86SrCarb from 0.7076 to 0.7071. The Sr isotope ratios are considered near primary because of high Sr (1828 –2916 ppm) and generally low Rb concentrations (,13 ppm, Rb/Sr 0.007, Fo¨lling & Frimmel 2002). In contrast,

KAIGAS AND NUMEES FORMATIONS

the post-Numees carbonates of the Bloeddrif Member have significantly higher, but consistent 87Sr/86SrCarb ratios of 0.7085 in the limestone immediately above the diamictite and 0.7082 in the remainder of the member. Considering Rb/Sr ratios of much less than 0.001 (Sr and Rb concentrations of 1078 –2483 and ,5.3 ppm, respectively), the measured 87Sr/86SrCarb ratios are again regarded as effectively primary (Fo¨lling & Frimmel 2002). The bottom of the Bloeddrif Member limestone is markedly depleted in 13C (d13CCarb ¼ 24.6‰), whereas the rest has d13CCarb consistently clustering within the narrow range between 21.0 and þ1.0‰. In a follow-up study (Frimmel & Fo¨lling 2004) it could be shown, however, that the increase in d13CCarb is strongly dependent on palaeoenvironment, with much higher isotope ratios (þ4.0‰) achieved in the most proximal positions for the same stratigraphic interval. Alternatively, deposition of the cap carbonate could be diachronous from the slope up to the shelf (e.g. Hoffman et al. 2007). Although O isotopes are more susceptible to diagenetic alteration, a distinct negative d18OCarb anomaly (decrease by as much as 4‰) has been noted in the first 25 m of the post-Numees cap carbonates in various outcrops and has been interpreted as reflecting freshwater mixing with seawater during large-scale melting of glaciers, in agreement with the slightly higher 87Sr/86SrCarb at the same position (Fo¨lling & Frimmel 2002). By and large, the post-Numees carbonates are more enriched in Sr than the post-Kaigas carbonates. Strontium concentrations of several thousand ppm have been reported for the distal portions (characterized by a lack of Rb, Y and Zr) of the Bloeddrif Member limestone, whereas the near-shore equivalents contain an order of magnitude less Sr (Frimmel & Fo¨lling 2004).

Other characteristics The Rosh Pinah Formation is economically significant for its base-metal enrichment. Stratabound and largely stratiform sulphidic Pb and Zn ore bodies are being mined at Rosh Pinah (past production and resources are c. 50 Mt at an average grade of 7 wt% Zn, 2 wt% Pb, 0.1 wt% Cu and 11 ppm Ag). The ore is hosted by a variably silicified carbonaceous mudstone that reaches as much as 50 m in thickness, with intercalated carbonate lenses, within an overall arkosic succession of the middle Rosh Pinah Formation (Alchin et al. 2005). A protore in a similar stratigraphic position is regarded as the source for the secondary oxidized Zn deposit at Skorpion (25 Mt, average grade of 10.6 wt% Zn; Borg et al. 2003), located c. 22 km NW of Rosh Pinah. In spite of a low-grade metamorphic overprint, it has been possible to recover a number of organic-walled microfossils that help in setting some biostratigraphic constraints for units above and below the Numees Formation (Gaucher et al. 2005). Poorly preserved, highly carbonized acritarchs characterize units beneath the Numees Formation. Of biostratigraphic significance is the occurrence of Bavlinella faveolata in the Pickelhaube and upper Wallekraal Formations. Acritarchs in the Bloeddrif Member, Holgat Formation, are characterized by a low-diversity assemblage consisting of Soldadophycus, Myxococcoides, Coniunctiophycus and Leiosphaeridia, with a dominance of Soldadophycus bossii, and the absence of acanthomorphs and large sphaeromorphs. Associated with this acritarch assemblage is the agglutinated foraminifer Titanotheca sp.

Palaeolatitude and palaeogeography No reliable palaeomagnetic data are available for the Gariep Belt or for the location of the Kalahari Craton in the Neoproterozoic in general. Today, the Gariep Belt is positioned between the Kalahari and the Rio de la Plata Cratons. Both Pb and Nd isotopic evidence (Frimmel et al. 2004; Frimmel & Basei 2006) point to a

229

provenance of the entire Port Nolloth Group on the Kalahari Craton. This is supported by detrital zircon age spectra that conform to the ages known from the immediate basement on the Kalahari Craton but differ from those typical for the Rio de la Plata Craton (Basei et al. 2005). Both the Kaigas and the Numees Formation diamictites would have been deposited in a marginal cratonic position for which, in most palaeogeographic reconstructions, a low palaeolatitude is assumed, though on circumstantial evidence, throughout the Neoproterozoic time span of interest here (Collins & Pisarevsky 2005).

Geochronological constraints A maximum age constraint for the onset of sedimentation in the Gariep Basin is given by the youngest age obtained on pre-Gariep basement, that is, 771 + 6 Ma for the Lekkersing Granite (Frimmel et al. 2001). The Kaigas Formation must be younger than that age but cannot be younger than the volcanism evident in the Rosh Pinah Formation. Felsic volcanic rocks above the Kaigas Formation yielded Pb–Pb and U – Pb single zircon ages of 741 + 6 Ma and 752 + 6 Ma, respectively (Frimmel et al. 1996b; Borg et al. 2003). This is supported by a double-spike Pb– Pb carbonate age of 728 + 32 Ma obtained for the overlying cap carbonates of the Pickelhaube Formation (Fo¨lling et al. 2000), which has been interpreted as dating the time of diagenesis. More problematic is the absolute age of the Numees Formation, as no radiometric data are available for this unit. Minimum age constraints are given by a double-spike Pb–Pb carbonate age of 555 + 28 Ma for the overlying limestone of the Bloeddrif Member of the Holgat Formation (Fo¨lling et al. 2000) and the age of regional, syn-collisional metamorphism that is constrained at 545 + 2 Ma by Ar –Ar data on syn-tectonic hornblende and micas (Frimmel & Frank 1998).

Discussion The sedimentary features of the Kaigas Formation diamictite, such as sharp upper and lower bedding contacts, lack of internal structures (except for crudely graded bedding), correlation between clast size and bedding thickness, rapid decrease in clast size towards the basin centre, and the limited geographical distribution suggest rapid deposition as debris flows. Transport by turbidity currents in a subaqueous fan environment is indicated by common upwards fining beds in the intercalated feldspathic arenite. The presence of westward-directed gravity flow deposits stepping back onto the basement foreland suggests a phase of marine transgression along an active fault scarp. This is also indicated by local dolostone clasts in the diamictite, which implies the presence of shallow-water, pre-Kaigas carbonate platform deposits, at least in the south. A component of melt-out glacial debris in the Kaigas Formation diamictite, such as dropstones, bimodal clast distribution, sandwiched and faceted clasts, and outsized erratic blocks in shaly beds, suggest a glaciomarine or fluvioglacial origin of parts of the Kaigas Formation. Uplifted horsts along the margin of the basin may have been the sites of mountain glaciers that contributed some glacial material. Although there is little doubt of a glacial influence during Kaigas times, most of the Kaigas Formation reflects sediment gravity flows in consequence of subaqueous slumping, probably triggered by tectonic activity along the growing rift shoulders. In contrast, the Numees Formation represents a truly glacial deposit. Deposition adjacent to a glaciated continental margin by processes of settling from suspension and ice rafting is indicated by the extensive lateral distribution of the diamictite, its textural homogeneity and largely unbedded nature, and the presence of out-sized extrabasinal lonestones. The associated sediments, in

230

H. W. FRIMMEL

particular the banded iron-formation of the Jakkalsberg Member, record marine rather than glaciofluvial depositional conditions. Huge blocks, more than 10 m in length, of completely different lithology (dolomite, gneiss) occurring next to each other within the same diamictite bed (commonly found near the Orange River) can only be explained reasonably by melt-out from drifting icebergs. The Bloeddrif Member carbonates are interpreted as a typical cap carbonate sequence. Although the Kaigas and Numees Formations have been tentatively correlated with the global Sturtian and Marinoan glaciations (Frimmel et al. 2002), this interpretation has been questioned. The age of the Kaigas Formation is c. 750 Ma, older than the Sturtian glaciation and, purely based on evidence from the Gariep Belt, there is insufficient evidence from the Gariep Belt to postulate a global glacial event for the Kaigas Formation. On a regional scale, the Kaigas glaciogenic deposits may be correlated with the Chuos Formation of northern Namibia, based on similar radiometric ages for associated felsic volcanic rocks there (Hoffman et al. 1996; Hoffman & Prave 1996). As evidence for glacial conditions at that time exists also from other continents (Vieira et al. 2007; Zheng et al. 2007), it may well represent a global glacial event. The lack of precise radiometric age control for the Numees Formation precludes a reliable stratigraphic correlation with glaciogenic deposits elsewhere. A precise age of 636 + 1 Ma has been obtained on an ash bed in metamorphosed glaciomarine diamictite from the Central Zone of the Damara Belt (Hoffmann et al. 2004). These authors consider the dated diamictite unit as equivalent to the Ghaub Formation in the northern platform and consequently postulate a correlation with the Marinoan glaciation. A correlation of the Numees Formation with the Ghaub Formation appears intuitively probable (as suggested by Frimmel et al. 2002), but is not supported by more recent biostratigraphic evidence (Gaucher et al. 2005). The microfossil assemblage in the post-Numees rocks is markedly different from those in the preNumees strata, and the former compares well with Ediacaran assemblages elsewhere. An Ediacaran age of the Holgat Formation is also supported by the Pb–Pb age of 555 + 28 Ma obtained for the Bloeddrif cap carbonate (Fo¨lling et al. 2000) and the comparatively high initial 87Sr/86SrCarb (Fo¨lling & Frimmel 2002). A correlation of the Numees Formation with the c. 580 Ma Gaskiers glaciation is therefore preferred. Thus, if the successor basin model is accepted for the Gariep Basin, the apparent absence of any syn-Marinoan glacial deposits in the Port Nolloth Zone may be explained by the hiatus between the continental rift stage (770– 740 Ma) and the back-arc basin stage (c. 610– 580 Ma) in the overall basin evolution. The South African National Research Foundation is thanked for funding the research on the Gariep Belt for many years. Reviews by F. MacDonald, R. McMiller and G. Halverson significantly improved the original manuscript. This work represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Projects #512 and #478.

References Alchin, D. J., Frimmel, H. E. & Jacobs, L. E. 2005. Stratigraphic setting of the metalliferous Rosh Pinah Formation and the Spitzkop and Koivib Suites in the Pan-African Gariep Belt, southwestern Namibia. South African Journal of Geology, 108, 19 –34. Basei, M. A. S., Frimmel, H. E., Nutman, A. P., Preciozzi, F. & Jacob, J. 2005. A connection between the Neoproterozoic Dom Feliciano (Brazil/Uruguay) and Gariep (Namibia/South Africa) orogenic belts – evidence from a reconnaissance provenance study. Precambrian Research, 139, 195–221. ¨ ber Glazialschichten an der Basis der Nama- und Beetz, W. 1926. U Konkipformation in der Namib Su¨dwestafrikas. Neues Jahrbuch f. Mineralogie, Geologie und Pala¨ontologie, Abteilung B, 56, 437– 481.

Borg, G., Ka¨rner, K., Buxton, M., Armstrong, R. & Van der Merwe, S. W. 2003. Geology of the Skorpion zinc deposit, southern Namibia. Economic Geology, 98, 749–771. Cloud, P. E. 1974. Giant stromatolites and associated vertical tubes from the Upper Proterozoic Noonday Dolomite, Death Valley region, eastern California. Geological Society of America, Bulletin, 85, 1869– 1882. Collins, A. S. & Pisarevsky, S. A. 2005. Amalgamating eastern Gondwana: the evolution of the circum-Indian orogens. Earth-Science Reviews, 71, 229–270. Corsetti, F. A. & Grotzinger, J. P. 2005. Origin of tube structures in Neoproterozoic post-glacial cap carbonates: example from Noonday Dolomite, Death Valley, United Sates. Palaios, 20, 348– 362. Davies, C. & Coward, M. P. 1982. The structural evolution of the Gariep Arc in southern Namibia. Precambrian Research, 17, 173– 198. De Villiers, J. & So¨hnge, P. G. 1959. The geology of the Richtersveld. Memoirs of the Geological Survey of South Africa, 48, 1– 295. Fo¨lling, P. G. & Frimmel, H. E. 2002. Chemostratigraphic correlation of carbonate successions in the Gariep and Saldania Belts, Namibia and South Africa. Basin Research, 14, 69– 88. Fo¨lling, P. G., Zartman, R. E. & Frimmel, H. E. 2000. A novel approach to double-spike Pb –Pb dating of carbonate rocks: examples from Neoproterozoic sequences in southern Africa. Chemical Geology, 171, 97 –122. Frimmel, H. E. 1995. Metamorphic evolution of the Gariep Belt. South African Journal of Geology, 98, 176–190. Frimmel, H. E. 2000. The Pan-African Gariep Belt in southwestern Namibia and western South Africa. Communications of the Geological Survey of Namibia, 12, 197–209. Frimmel, H. E. 2003a. Gariep Supergroup. In: Johnson, M. R. (ed.) Catalogue of South African Lithostratigraphic Units. Council for Geoscience, Pretoria, 5 –7. Frimmel, H. E. 2003b. Port Nolloth Group. In: Johnson, M. R. (ed.) Catalogue of South African Lithostratigraphic Units. Council for Geoscience, Pretoria, 35 – 37. Frimmel, H. E. 2008. The Gariep Belt. In: Miller, R. M. (ed.) The Geology of Namibia. Geological Survey of Namibia, Windhoek, 2, 14-1 –14-39. Frimmel, H. E. & Basei, M. A. S. 2006. Tracking down the Neoproterozoic connection between southern Africa and South America – a revised geodynamic model for SW-Gondwana amalgamation. In: Gaucher, C. & Bossi, J. (eds) V. South American Symposium on Isotope Geology. 24 –27 August, Punta del Este, 94– 97. Frimmel, H. E. & Fo¨lling, P. G. 2004. Late Vendian closure of the Adamastor Ocean: timing of tectonic inversion and syn-orogenic sedimentation in the Gariep Basin. Gondwana Research, 7, 685– 699. Frimmel, H. E. & Frank, W. 1998. Neoproterozoic tectono-thermal evolution of the Gariep Belt and its basement, Namibia/South Africa. Precambrian Research, 90, 1 –28. Frimmel, H. E. & Lane, K. 2005. Geochemistry of carbonate beds in the Neoproterozoic Rosh Pinah Formation, Namibia: implications on depositional setting and hydrothermal ore formation. South African Journal of Geology, 108, 5 – 18. Frimmel, H. E., Hartnady, C. J. H. & Koller, F. 1996a. Geochemistry and tectonic setting of magmatic units in the Pan-African Gariep Belt, Namibia. Chemical Geology, 130, 101– 121. Frimmel, H. E., Klo¨tzli, U. & Siegfried, P. 1996b. New Pb– Pb single zircon age constraints on the timing of Neoproterozoic glaciation and continental break-up in Namibia. The Journal of Geology, 104, 459– 469. Frimmel, H. E., Zartman, R. E. & Spa¨th, A. 2001. Dating Neoproterozoic continental break-up in the Richtersveld Igneous Complex, South Africa. The Journal of Geology, 109, 493– 508. Frimmel, H. E., Fo¨lling, P. G. & Eriksson, P. 2002. Neoproterozoic tectonic and climatic evolution recorded in the Gariep Belt, Namibia and South Africa. Basin Research, 14, 55 –67. Frimmel, H. E., von Veh, M. W. & Fo¨lling, P. G. 2003a. Kaigas Formation. In: Johnson, M. R. (ed.) Catalogue of South African Lithostratigraphic Units. Council for Geoscience, Pretoria, 17– 19.

KAIGAS AND NUMEES FORMATIONS

Frimmel, H. E., von Veh, M. W. & Fo¨lling, P. G. 2003b. Numees Formation. In: Johnson, M. R. (ed.) Catalogue of South African Lithostratigraphic Units. Council for Geoscience, Pretoria, 25– 28. Frimmel, H. E., Jonasson, I. & Mubita, P. 2004. An Eburnean base metal source for sediment-hosted zinc-lead deposits in Neoproterozoic units of Namibia: lead isotopic and geochemical evidence. Mineralium Deposita, 39, 328–343. Gaucher, C., Frimmel, H. E. & Germs, G. J. B. 2005. Organic-walled microfossils and biostratigraphy of the upper Port Nolloth Group (Namibia): implications for the latest Neoproterozoic glaciations. Geological Magazine, 142, 539– 559. Germs, G. J. B. 1995. The Neoproterozoic of southwestern Africa, with emphasis on platform stratigraphy and paleontology. Precambrian Research, 73, 137– 151. Gresse, P. G. 1994. Strain partitioning in the southern Gariep Arc as reflected by sheath folds and stretching lineations. South African Journal of Geology, 97, 52 – 61. Grotzinger, J. P., Bowring, S. A., Saylor, B. Z. & Kaufman, A. J. 1995. Biostratigraphic and geochronologic constraints on early animal evolution. Science, 270, 598–604. Ha¨lbich, I. W. & Alchin, D. J. 1995. The Gariep belt: stratigraphicstructural evidence for obliquely transformed grabens and backfolded thrust stacks in a combined thick-skin thin-skin structural setting. Journal of African Earth Sciences, 21, 9 –33. Hartnady, C., Joubert, P. & Stowe, C. 1985. Proterozoic crustal evolution in southwestern Africa. Episodes, 8, 236– 244. Hartnady, C. J. H. & von Veh, M. W. 1990. Tectonostratigraphic and structural history of the Late Proterozoic – Early Palaeozoic Gariep Belt, Cape Province, South Africa. Geological Society of South Africa, Cape Town. Hegenberger, W. 1993. Stratigraphy and sedimentology of the Late Precambrian Witvlei and Nama Groups, east of Windhoek. Memoirs of the Geological Survey of Namibia, 17, 1– 82. Hoffman, P. F., Hawkins, D. P., Isachsen, C. E. & Bowring, S. A. 1996. Precise U–Pb zircon ages for early Damaran magmatism in the Summas Mountains and Welwitschia inlier, northern Damara belt, Namibia. Communications of the Geological Survey of Namibia, 11, 47 – 52. Hoffman, P. F., Kaufman, A. J. & Halverson, G. P. 1998. Comings and goings of global glaciations on a Neoproterozoic tropical platform in Namibia. GSA Today, 8, 1– 9. Hoffman, P. F., Halverson, G. P., Domack, E. W., Husson, J. M., Higgins, J. A. & Schrag, D. P. 2007. Are basal Ediacaran (635 Ma) post-glacial ‘cap dolostones’ diachronous? Earth and Planetary Science Letters, 258, 114–131. Hoffmann, K.-H. & Prave, A. R. 1996. A preliminary note on a revised subdivision and regional correlation of the Otavi Group based on glaciogenic diamictites and associated cap dolostones. Communications of the Geological Survey of Namibia, 11, 77 – 82. Hoffmann, K.-H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. A U– Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817– 820. Jasper, M. U., Stanistreet, I. G. & Charlesworth, E. G. 1992– 1993. Report: preliminary results of a study of the structural and

231

sedimentological evolution of the late Proterozoic/early Palaeozoic Gariep Belt, southern Namibia. Communications of the Geological Survey of Namibia, 8, 99 –118. Jasper, M. J. U., Stanistreet, I. G. & Charlesworth, E. G. 2000. Neoproterozoic inversion tectonics, half-graben depositories and glacial controversies, Gariep fold-thrust belt, southern Namibia. In: Miller, R. M. (ed.) Henno Martin Commemorative Volume. Communications of the Geological Survey of Namibia, Windhoek, 1887–1896. Kaufman, A. J., Hayes, J. M., Knoll, A. H. & Germs, G. J. B. 1991. Isotopic composition of carbonates and organic carbon from upper Proterozoic successions in Namibia: stratigraphic variation and the effects of diagenesis and metamorphism. Precambrian Research, 49, 301– 327. Kennedy, M. J. 2001. Are Proterozoic cap carbonates and isotopic excursions a record of gas hydrate destabilization following Earth’s coldest intervals? Geology, 29, 443–446. Kro¨ner, A. 1974. The Gariep Group, Part I: Late Precambrian formations in the western Richtersveld, northern Cape Province. Precambrian Research Unit, University of Cape Town, Bulletin, 13, 1– 115. Martin, H. 1965a. Beobachtungen zum Problem der jung-pra¨kambrischen glazialen Ablagerungen in Su¨dwestafrika. Geologische Rundschau, 54, 115– 127. Martin, H. 1965b. The Precambrian geology of South West Africa and Namaqualand. Precambrian Research Unit, University of Cape Town, Bulletin, 4, 1– 177. McMillan, M. D. 1968. The geology of the Witputs-Sendelingsdrif area. Precambrian Research Unit, University of Cape Town, Bulletin, 4, 1– 177. Middlemost, A. E. K. 1963. Geology of the southeastern Richtersveld. Unpublished PhD thesis, University of Cape Town, Cape Town. Reid, D. L., Ransome, I. G. D., Onstott, T. C. & Adams, C. J. 1991. Time of emplacement and metamorphism of Late Precambrian mafic dykes associated with the Pan-African Gariep orogeny, Southern Africa: implications for the age of the Nama Group. Journal of African Earth Sciences, 13, 531–541. Rogers, A. W. 1916. The geology of part of Namaqualand. Transactions of the Geological Society of South Africa, 18, 72 – 101. Siegfried, P. R. & Moore, J. M. 1990. The Rosh Pinah Zn– Pb– Cu– Ag massive sulphide deposit–a product of early rift-related volcanism? In: Geocongress 90. Geological Society of South Africa, Cape Town, 512– 513. Vieira, L. C., Trindade, R. I. F., Nogueira, A. C. R. & Ader, M. 2007. Identification of a Sturtian cap carbonate in the Neoproterozoic Sete Lagoas carbonate platform, Bambuı´ Group, Brazil. Comptes Rendus Geoscience, 339, 240–258. von Veh, M. W. 1993. The stratigraphy and structural evolution of the Late Proterozoic Gariep Belt in the Sendelingsdrif-Annisfontein area, northwestern Cape Province. Precambrian Research Unit, University of Cape Town, Bulletin, 38, 1– 174. Zheng, Y.-F., Wu, Y.-B., Gong, B., Chen, R.-X., Tang, J. & Zhao, Z.-F. 2007. Tectonic driving of Neoproterozoic glaciations: evidence from extreme oxygen isotope signature of meteoric water in granite. Earth and Planetary Science Letters, 256, 196– 210.

Chapter 18 The Karoetjes Kop and Bloupoort formations, Gifberg Group, South Africa HARTWIG E. FRIMMEL1,2 1

Department of Geological Sciences, University of Cape Town, Rondebosch 7701, South Africa

2

Present address: Geodynamics & Geomaterials Research Division, Institute of Geography and Geology, University of Wuerzburg, Am Hubland, D-97074 Wuerzburg, Germany (e-mail: [email protected]) Abstract: The Vredendal Outlier near the South African west coast lies in an intermediate position between the late Neoproterozoic Gariep Belt further north and the Cambrian Saldania Belt further south. It consists of a sedimentary succession of siliciclastic and carbonate rocks, unified as the Gifberg Group, which contains two diamictite-bearing units, the Karoetjes Kop Formation at the base and the Bloupoort Formation near the top of the group. The diamictite of the Karoetjes Kop Formation represents mainly debris flow deposits in a continental rift setting, with some contribution from retreating glaciers. In contrast, the younger diamticite in the Bloupoort Formation is glaciomarine, is associated with banded iron-formation, is underlain by stromatolitic reef carbonates and is overlain by carbonates. Most of the Gifberg Group is poorly exposed and poorly investigated. In the absence of radiometric age data, stratigraphic interpretation and correlation is based on lithological and chemostratigraphic evidence. The entire Gifberg Group is considered to be equivalent to the much better investigated Port Nolloth Group in the Gariep Belt (sensu stricto). Whereas the Karoetjes Kop Formation is correlated with the c. 750 Ma Kaigas Formation of the Gariep Belt, the diamictite and associated banded iron-formation of the Bloupoort Formation are regarded as correlatives of the Numees Formation of the Gariep Belt. The entire Gifberg Group was subjected to transpressional deformation and accompanying low-grade metamorphism during continental collision between the Rio de la Plata and Kalahari plates at the end of the Neoproterozoic and again during the Cambrian accretionary orogeny along the southwestern margin of Gondwana, which led to the development of the Saldania Belt further south.

The Karoetjes Kop and Bloupoort Formations both contain diamictite. They occur in a succession of metasedimentary Neoproterozoic rocks that are relatively poorly exposed along the South African west coast between latitudes 308S and 328S and in the plains that stretch from the coast up to 80 km inland in the area around Vanrhynsdorp (Figs 18.1 and 18.2). This area was first mapped by Rogers (Rogers & Schwarz 1904; Rogers 1911). Subsequent studies (Brink 1950; Jansen 1960; Kro¨ner 1968; Lamont 1947) led to rather contrasting stratigraphic interpretations. The basis for the current stratigraphic subdivision is the work of Gresse (1992), who re-mapped large parts of the area and studied the sedimentology, structure and metamorphism. Originally, Gresse included this metasedimentary succession as the Gifberg Subgroup in the Vanrhynsdorp Group. The younger part of that group has since been recognized as a correlative of the lower to middle Nama Group in Namibia and represents the fill of a southern subbasin of the larger Nama foreland basin (Gresse & Germs 1993). The Gifberg rocks occur unconformably below these foreland basin deposits and bear strong lithological similarities to sedimentary successions of the Neoproterozoic Port Nolloth Group in the Gariep Belt further north. Subsequently, they were, therefore, described as a separate group, the Gifberg Group (de Beer et al. 2002). Their area of distribution is referred to as the Vredendal Outlier, which occupies an intermediate position between the Gariep Belt in the north and the Saldania Belt in the south. A glacial influence during deposition of the Karoetjes Kop Formation was indirectly inferred by Gresse (1992), who was the first to compare the diamicite of that formation with the glaciogenic Kaigas Formation diamictite in the Port Nolloth Zone. Less ambiguous is the glacial origin in the case of the diamictite within the younger Bloupoort Formation. Although it was described originally as agglomerate (Rogers & Schwarz 1904), general agreement exists now on its glaciogenic character following a more detailed description of the rock association by Buehrmann (1981). The correlation of individual formations between the Port Nolloth and Gifberg Groups has remained speculative, and uncertainty exists particularly with regard to the age of the Bloupoort Formation diamictite (de Beer et al. 2002). Recent lithogeochemical and isotopic data on carbonates that are positioned

between the two diamictite-bearing formations of the Gifberg Group (Frimmel 2008) helped in better constraining the stratigraphic position of that group.

Structural framework and tectonic evolution Although a polyphase deformational history is evident from field observations and structural data (Gresse 1992), the overall structure appears to be relatively simple with a gentle dip of primary stratigraphic contacts towards the SW. The local distribution of rock types can be largely explained by the interference between two generations of folds (F2 and F3). Earlier F1 folds are rarely preserved. As all fold axes follow a similar trend that is subparallel to the orogenic belt and colinear with the stretching direction, a transpressive regime is inferred. The style of deformation is identical to that in the Gariep Belt further north, where it has been explained by sinistral transpression (Von Veh 1993; Gresse 1994). Similarly, the deformation in the Vredendal Outlier can be explained by sinistral transpression that reflects compression towards the east with SSE-verging folds and thrusts (F1), west-verging back-folds and back-thrusts (F2), ENE-verging F3 folds and conjugate kinks. By analogy with findings from the Gariep Belt further north (Frimmel & Frank 1998), an age of c. 545 Ma is assigned to F1 and F2. The F3 fold- and thrust-event also affected the siliciclastic sedimentary rocks of the Vanrhynsdorp Group (Gresse 1992) and is thus younger. Ar – Ar age data on syn-F3 micas are in the range 500– 480 Ma (Gresse et al. 1988; Frimmel & Frank 1998). Syn-orogenic deformation was accompanied by regional metamorphism. The grade of metamorphism increases from the foreland in the east towards the west. Although the Vanrhynsdorp Group rocks in the east have experienced only sub-greenschist facies conditions, those near the thrust contact with the rocks of the Gifberg Group reached greenschist-facies conditions. Consequently, this phase of dynamic metamorphism must have taken place after the deposition of the Vanrhynsdorp foreland basin fill and it was syn-kinematic with respect to F3. An increase in the metamorphic grade from lowest greenschist facies near Vanrhynsdorp to lowermost amphibolite facies at the Atlantic coast is

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 233– 237. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.18

234

H. E. FRIMMEL

starts with a white to blue-grey basal conglomerate in which the basement lithotypes appear reworked. This is followed by a grey, fine-grained quartzitic greywacke in which thin, well-rounded quartz (or quartzite) pebble beds are intercalated. In places, this greywacke rests directly on the basement. A massive, poorly bedded, laterally discontinuous diamictite follows gradationally above the quartzitic greywacke. It reaches a maximum thickness of a few tens of metres and contains poorly sorted pebbles, cobbles and boulders of basement-derived gneiss, quartzite, vein quartz and biotite schist, all of which are set in a metamorphosed biotite-rich matrix. In places, the diamictite rests directly on basement and grades laterally into the quartzitic greywacke. Above the diamictite follows a reddish weathering, laterally continuous feldspathic quartzite with intercalated metapelite (sericite schist). Large trough cross-bedded sets are common, reaching several metres in thickness. This lithotype grades upwards into thick-bedded, coarse-grained, white to light blue-grey quartzite. No complete section through the entire formation is exposed. Consequently, no information on the true stratigraphic thickness is available.

The Widouw Formation

Fig. 18.1. Map showing the location of the Vredendal Outlier within the network of Pan-African orogenic belts of southwestern Africa.

evident from metapelites of the Gifberg Group, with kyanitestaurolite-garnet-micaschist developed in the extreme NW of the Vredendal Outlier.

Stratigraphy The oldest unit of the Gifberg Group (Fig. 18.3) is the mixed siliciclastic, diamictite-bearing Karoetjes Kop Formation. Variably recrystallized limestone and dolostone, as well as minor metagreywacke, quartzite and phyllite of the Widouw Formation follow above the Karoetjes Kop Formation. The overlying Aties Formation is dominated by carbonaceous schist, with subordinate phyllite to biotite-schist (in the westernmost exposures kyanitebearing), meta-arenite, and impure, thin carbonate beds. The youngest formation is the Bloupoort Formation, which comprises a variety of carbonate rocks, calcarenite, diamictite, pyritic and graphitic phyllite, ferruginous chert and iron-formation.

Glaciogenic deposits and associated strata The Karoetjes Kop Formation This basal formation overlies high-grade metamorphic gneisses of the Mesoproterozoic Namaqualand Metamorphic Complex. It

Carbonate intersections of more than 200 m were recorded in boreholes through the central part of the Vredendal Outlier (Gresse 1992), but the true thickness may be less due to tectonic thickening by isoclinal folding. The carbonate beds thin out towards the north where a basement high probably existed throughout Neoproterozoic to early Cambrian times (Kamieskroon Ridge, Fig. 18.1). Variably coloured (from white, pink, ochre to black), flat laminated, medium- to thick-bedded limestone dominates the lower part of the formation. In several places, individual beds are laterally discontinuous, irregular and reflect former dissolution surfaces (Frimmel 2008). Although the limestone unit is generally very uniform, intra-formational syn-sedimentary limestone breccia beds as well as stratiform black chert boudins occur in the upper parts. Contorted dolomitized beds reach 20–30 cm in thickness and contain white blades of calcite, several centimetres long, which have been compared with herringbone calcite and interpreted as pseudomorphs after enterolithic gypsum (Frimmel 2008). Impure limestone contains as much as 40 vol% coarsegrained, angular quartz clasts, which indicate an overall proximal setting for the depositional environment of these carbonates. Marked differences in grain size reflect different degrees of metamorphic recrystallization. Several types of dolostone, ranging in relative age from syn-sedimentary/early diagenetic to metamorphic, can be distinguished. The former includes stromatolitic dolostone (Gresse 1992).

The Aties Formation Although the lower contact of the Aties Formation with the underlying carbonates of the Widouw Formation is typically sheared, an originally conformable contact is inferred. The Aties Formation is made up of predominantly black carbonaceous, pyritic schist with interbedded phyllite to biotite schist, meta-arenite and thin, impure limestone and dolostone beds. The arenitic beds are quartzsandstone and greywacke in composition, but some of them are highly feldspathic (arkose), coarse-grained and grade in places into intraformational feldspathic conglomerate beds. Many of the well-bedded greywacke beds display normal grading and trough cross-bedding on a metre scale. Locally, monomictic quartz pebble conglomerate occurs within white quartzite. The presence of, inter alia, conglomerate and diamictite, associated with limestone and dolostone has been reported as ‘interbedded’ with phyllite of the Aties Formation (de Beer et al. 2002). It is suspected, however, that this diamictite-bearing succession is part of the overlying Bloupoort Formation.

THE KAROETJES KOP AND BLOUPOORT FORMATIONS

235

Fig. 18.2. Areal distribution of the individual formations that constitute the Gifberg Group.

The Bloupoort Formation The distinction of this formation from the underlying Aties Formatin has been problematic and prompted Gresse (1992) to include all the Bloupoort rocks as a separate member within the Aties Formation. Subsequently, it was recognized that they are younger and represent a relatively complex unit, and their stratigraphic rank was raised to that of formation (de Beer et al. 2002). The main distribution of this formation is to the SE of Vanrhynsdorp (Fig. 18.2), where a mixed carbonate unit grades into a diamictite-dominated unit. The former is almost entirely dolomitic and comprises laminated, stromatolitic (Conophyton-like), oolitic, intraclastic, conglomeratic, and brecciated and siliceous dolostone. Sandy to gritty dolostone is rich in variably rounded quartz clasts and contains up to 15 vol% feldspar (Gresse 1992). Oolites, pelletoids, pellets and lumps are common and were deposited in carbonate mud (micrite). Dolomitic breccia consists of micritic fragments that are cemented by sparry dolomite or micrite. This dolomitic unit grades upwards into dolomitic arenite and ferruginous diamictite, which constitute the Swartleikrans Bed of Buehrmann (1981). The diamictite consists of very poorly sorted, angular fragments and subrounded pebbles, cobbles and boulders of basement-derived orthogneiss and granulite, as well as basin-derived dolostone and arenite, and minor ferruginous chert, jasper, vein quartz, limestone, oolitic dolostone and feldspar grains, all of which are set in a ferruginous, hematite-rich argillitic matrix. Bedding is only very poorly developed in the overall massive diamictite, but in places very crude graded bedding from diamictite into ferruginous sandstone and siltstone to argillite can be observed on a metre scale. Overall, the proportion of interbedded ferruginous argillite increases towards the top of the unit. Laterally, the diamictite grades into banded iron-formation with isolated dropstones and interbedded chert and jasper (Rooihoogte Bed of Buehrmann 1981). Locally, the diamictite is overlain by thinly laminated pink dolostone and dolomitic calcarenite. Intense thrust-faulting and folding in all of these units makes it virtually impossible to constrain their true sedimentary thickness.

Boundary relations with overlying and underlying non-glacial units The lower contact of the diamictite in the Karoetjes Kop Formation with the underlying quartzitic greywacke is gradational. Where the diamictite rests directly on basement gneiss, the contact is an erosional unconformity. The upper contact of this diamictite with the overlying quartzite is sharp and conformable. The contact between the Karoetjes Kop Formation and the Widouw Formation is not exposed, and the relative age relationship between the two formations is only inferred from their spatial distribution. The lower contact of the diamictite-hosting Bloupoort Formation is typically tectonic. The upper contact of the formation is either a thrust-fault, or it is covered by younger sedimentary rocks of the Cape Supergroup or by aeolian sand. Within the formation, the lower boundary of the diamictite with the underlying dolostone is gradational, with an intervening dolomitic arenite, whereas the upper boundary against the overlying dolostone is sharp and appears conformable. The entire succession is, however, strongly deformed. A fivefold rhythmical repetition of an ‘interbedded sequence of dolomite, ironstone and ferruginous mixtite’, each about 5 –15 m in thickness, described by Gresse (1992), could well reflect tectonic repetition.

Chemostratigraphy Limited trace element analyses as well as C, O and Sr isotope data have been reported for carbonates of the Widouw Formation (Frimmel 2008). Limestone of this formation is very rich in Sr (as much as 4728 ppm) and has extremely low Rb/Sr ratios. Their 87Sr/86Sr ratios can be taken as near-primary and they range consistently throughout the succession between 0.70844 and 0.70847. The C and O isotope ratios in the limestone cover a wide range of d13C between – 4.2 and þ4.8‰ (relative to V-PDB) and d18O between –9.4 and –3.1 (relative to V-PDB). Syn-sedimentary or early diagenetic dolostone is enriched in 18O

236

H. E. FRIMMEL

Vanrhynsdorp Group, undifferentiated (c. 550–530 Ma)

Gifberg Group

Bloupoort Formation

(e.g. Collins & Pisarevsky 2005). The depositional basin for the Gifberg Group was most likely the southern continuation of the Gariep Basin, specifically the basins in which the sediments of the Port Nolloth Group were deposited. Today the Gariep Belt is positioned between the Kalahari and the Rio de la Plata Cratons. Both Pb and Nd isotopic evidence (Frimmel et al. 2004; Frimmel & Basei 2006) point to a provenance of the entire Gariep Supergroup, including the Gifberg Group, on the Kalahari Craton. For parts of the Widouw Formation, an evaporitic influence has been suggested based on petrographic observations, trace element distribution and isotope data (Frimmel 2008), which speaks against a high-latitude position at the time of sedimentation.

Geochronological constraints

Aties Formation Widouw Formation

Karoetjes Kop Formation

Basement (1.0 Ga) Limestone, dolostone

Arenite, rudite

Iron Formation

Argillite

Diamictite

Gneiss, schist

Fig. 18.3. Generalized lithostratigraphy of the Gifberg Group (modified from de Beer et al. 2002 and Frimmel & Fo¨lling 2004).

relative to the limestone and it shows a similar, but narrower variation in d13C.

Other characteristics Calcitic marble of the Widouw Formation is mined at several quarries in the vicinity of Vredendal. The pyritic schists of the Aties Formation are strongly weathered with numerous quartz veins that carry Cu, Fe and Mn staining. Numerous surface Fe-oxide deposits (‘gossans’) are derived from the palaeoweathering of the pyrite-rich beds on a Paleocene palaeosurface, the so-called African Surface (Partridge & Maud 1987). No biostratigraphic data have been published so far for rocks of the Gifberg Group.

Palaeolatitude and palaeogeography No reliable palaeomagnetic data are available for the Vredendal Outlier. The low-grade regional metamorphic overprint most likely led to the remagnetization of the rocks during the Pan-African/Gariepian orogeny. Thus, any assumption about palaeolatitude and palaeogeography is based only on circumstantial evidence and comparison with continental-scale models

No radiometric age data that might help in constraining the absolute age of the diamictite units in the Gifberg Group are available. By analogy with the Port Nolloth Group in the eastern Gariep Belt, the maximum age of sedimentation is given by the youngest age obtained on the pre-Gariep basement. This is a U –Pb single zircon age of 771 + 6 Ma for the Lekkersing granite (Frimmel et al. 2001). The minimum age for sedimentation is given by the age of metamorphism. Ar – Ar age data obtained on syn-tectonic muscovite and biotite from coastal outcrops in the west of the Vredendal Outlier (Frimmel & Frank 1998) range from 513 to 491 Ma. This range includes a Ar – Ar whole-rock age of 496 + 2 Ma obtained on a phyllite from the Vanrhynsdorp Group (Gresse et al. 1988). These ages have been interpreted to date a late Pan-African tectonic pulse that led to NE-directed thrusting of both the Gifberg Group and the younger Vanrhynsdorp Group. It is therefore likely that the first regional metamorphism that affected the Gifberg Group rocks took place earlier during the continental collision phase of the Gariepian orogeny. By analogy with results obtained on the Port Nolloth Group (Frimmel & Frank 1998) and the Nama Group (Grotzinger et al. 1995) further north, an age of c. 545 Ma is assumed for that stage of metamorphism and tectonism that caused the unconformity between the Gifberg and Vanrhynsdorp Groups.

Discussion No evidence exists for any marine influence in the Karoetjes Kop Formation for which a narrow continental rift, dominated by alluvial fans, plains and fan deltas, is indicated. The depositional environment envisaged for this formation is therefore similar to that inferred for the Stinkfontein Subgroup and the Kaigas Formation in the Port Nolloth Zone (Chapter 17), except for a total lack of any evidence for marine influence in the Karoetjes Kop Formation. A number of very small outcrops of the Karoetjes Kop Formation also exist along the coast further to the NW of the Vredendal Outlier, thus suggesting a continuation of the former basin margin from the Port Nolloth Zone southwards into the area around Vredendal. Consequently, the current eastern boundary of these Neoproterozoic rocks would be a good approximation of the original eastern basin margin. As with the Kaigas Formation, the diamictite in the Karoetjes Kop Formation is interpreted as mainly a debris-flow deposit because of its lack of internal structure and very limited geographical distribution (lateral discontinuity) along the basin edge. By analogy with the Kaigas Formation, a fluvio-glacial origin for parts of the formation is suggested. The isotopic record of the Widouw Formation carbonates has been interpreted as reflecting an overall 13C-depleted sea, with the higher d13C ratios resulting from elevated evaporation rates rather than changes in the global seawater composition (Frimmel 2008). These shallow-water marine deposits are therefore correlated with the post-Kaigas transgression.

THE KAROETJES KOP AND BLOUPOORT FORMATIONS

The regional and global correlation of the diamictite in the Bloupoort Formation remains problematic. Little doubt exists about the glacial nature of the diamictite. The association with, and lateral transition into, banded iron-formation with occasional lonestones and dropstones clearly indicate a glaciomarine origin. The appearance of the underlying carbonates, which bear a remarkable resemblance to those of the Dabie River Formation in the Port Nolloth Zone, the ferruginous matrix of the diamictite, its massive structure and its association with iron-formation are all features that are identical to the Numees Formation in the Port Nolloth Zone. In the absence of radiometric and biostratigraphic evidence, the lower dolomitic unit of the Bloupoort Formation is correlated with the Dabie River Formation, the diamictite and iron-formation with the Numees Formation, and the upper carbonates and siliciclastic rocks with the Holgat Formation. Fieldwork in the Vredendal area was supported by Kumba Resources. Funding by the South African National Research Foundation is gratefully acknowledged. F. MacDonald and G. Halverson are thanked for helpful comments on the original manuscript. This work represents a contribution of the IUGS- and UNESCOfunded IGCP (International Geoscience Programme) Projects #512 and #478.

References Brink, W. C. 1950. The geology, structure and petrology of the Nuwerus area. Annals of the University of Stellenbosch, 16, 97 –221. Buehrmann, H. T. 1981. The geology of the Lower Kobe Valley, Vanrhynsdorp District. Annals of the University of Stellenbosch, 3, Series A1, 67 –144. Collins, A. S. & Pisarevsky, S. A. 2005. Amalgamating eastern Gondwana: the evolution of the circum-Indian orogens. Earth-Science Reviews, 71, 229–270. de Beer, C. H., Gresse, P. G., Theron, J. N. & Almond, J. E. 2002. The Geology of the Calvinia Area, Explanation Sheet 3118 Calvinia 1:250000 scale. Council for Geoscience, Pretoria. Frimmel, H. E. 2008. An evaporitic facies in Neoproterozoic post-glacial carbonates: the Gifberg Group, South Africa. Gondwana Research, 13, 453– 468. Frimmel, H. E. & Basei, M. A. S. 2006. Tracking down the Neoproterozoic connection between southern Africa and South America – a revised geodynamic model for SW-Gondwana amalgamation. In: Gaucher, C. & Bossi, J. (eds) V. South American Symposium on Isotope Geology. 24 –27 August, Punta del Este, 94 –97. Frimmel, H. E. & Fo¨lling, P. G. 2004. Late Vendian closure of the Adamastor Ocean: timing of tectonic inversion and syn-orogenic sedimentation in the Gariep Basin. Gondwana Research, 7, 685– 699.

237

Frimmel, H. E. & Frank, W. 1998. Neoproterozoic tectono-thermal evolution of the Gariep Belt and its basement, Namibia/South Africa. Precambrian Research, 90, 1 – 28. Frimmel, H. E., Jonasson, I. & Mubita, P. 2004. An Eburnean base metal source for sediment-hosted zinc-lead deposits in Neoproterozoic units of Namibia: lead isotopic and geochemical evidence. Mineralium Deposita, 39, 328–343. Frimmel, H. E., Zartman, R. E. & Spa¨th, A. 2001. Dating Neoproterozoic continental break-up in the Richtersveld Igneous Complex, South Africa. The Journal of Geology, 109, 493 – 508. Gresse, P. G. 1992. The tectono-sedimentary history of the Vanrhynsdorp Group. Memoirs of the Geological Survey of South Africa, 79, 1– 163. Gresse, P. G. 1994. Strain partitioning in the southern Gariep Arc as reflected by sheath folds and stretching lineations. South African Journal of Geology, 97, 52– 61. Gresse, P. G., Fitch, F. J. & Miller, J. A. 1988. 40Ar/39Ar dating of the Cambro-Ordovician Vanrhynsdorp tectonite in southern Namaqualand. South African Journal of Geology, 91, 257–263. Gresse, P. G. & Germs, G. J. B. 1993. The Nama foreland basin: sedimentation, major unconformity bounded sequences and multisided active margin advance. Precambrian Research, 63, 247– 272. Grotzinger, J. P., Bowring, S. A., Saylor, B. Z. & Kaufman, A. J. 1995. Biostratigraphic and geochronologic constraints on early animal evolution. Science, 270, 598–604. Jansen, H. 1960. The geology of the area around Nieuwoudtville. In: Von Backstro¨m, J. W. (ed.) Explanation Sheet 241 (Nieuwoudthville). Geological Survey of South Africa, Pretoria, 49. Kro¨ner, A. 1968. The gneiss-sediment relationship northwest of Vanrhynsdorp, Cape Province. Bulletin, Precambrian Research Unit, University of Cape Town, 3, 1– 233. Lamont, G. T. 1947. The geology of part of the Van Rhynsdorp Division, Cape Province. Unpublished PhD thesis, University of Cape Town, Rondebosch. Partridge, T. C. & Maud, R. R. 1987. Geomorphic evolution of southern Africa since the Mesozoic. South African Journal of Geology, 90, 179– 208. Rogers, A. W. 1911. Report on the geological survey of parts of the divisions of Van Rhyn’s Dorp, and Namaqualand. Annual Report of the Geological Commission, Cape of Good Hope, 1911, 7– 84. Rogers, A. W. & Schwarz, E. H. L. 1904. Geological survey of the northwestern part of Van Rhyn’s Dorp. Annual Report of the Geological Commission, Cape of Good Hope, 1904, 9– 46. Von Veh, M. W. 1993. The stratigraphy and structural evolution of the Late Proterozoic Gariep Belt in the Sendelingsdrif-Annisfontein area, northwestern Cape Province. Precambrian Research Unit, University of Cape Town, Bulletin, 38, 1– 174.

Chapter 19 The Ayn Formation of the Mirbat Group, Dhofar, Oman PHILIP A. ALLEN1*, RUBEN RIEU2, JAMES L. ETIENNE3, ALBERT MATTER4 & ANDREA COZZI5 1

Department of Earth Science and Engineering, Imperial College London, South Kensington Campus, London SW7 2AZ, UK 2

TOTAL E&P Nederland, Bordewijklaan 18, Den Haag, The Netherlands

3

Neftex Petroleum Consultants Ltd, 97 Milton Park, Abingdon, Oxfordshire OX14 4RY, UK

4

Institute of Geological Sciences, University of Bern, Baltzerstrasse 1, CH-3012 Bern, Switzerland 5

ENI Angola, Rua Nicola Gomes Spencer, 140, PO Box 1289, Luanda, Angola *Corresponding author (e-mail: [email protected])

Abstract: Glacial deposits are found in the Ayn Formation and Shareef Formation of the Mirbat Group close to Mirbat in Dhofar, southern Oman. The Mirbat Group is most likely a correlative of the Abu Mahara Group of the Huqf Supergroup of northern Oman. The Ayn Formation, the main subject of this chapter, comprises ,400 m of mainly coarse-grained glaciogenic deposits, ponded in 2to .8-km-wide N- to NW-oriented palaeovalleys eroded into crystalline basement, with few or no deposits preserved on intervening palaeohighs. The Shareef Formation occurs as thin, lenticular, erosional remnants beneath the unconformably overlying Cretaceous. The Ayn Formation is overlain by a thin (,3 m), discontinuous cap carbonate that passes from carbonate-cemented talus on the basin margin to stromatolitic carbonate on palaeohighs and resedimented gravity flows on palaeovalley flanks. The Ayn Formation is younger than its youngest detrital zircons and the youngest late plutons in crystalline basement, constraining it to ,c. 720 Ma, but its exact age is unknown. The detrital zircon population comprises exclusively Neoproterozoic sources, suggesting derivation from the juvenile Neoproterozoic crust of the Arabian area. The composition of fine-grained matrix in glaciogenic diamictite units and of non-glacial mudstones, plotted using the chemical index of alteration (CIA), suggests strong variations in the intensity of palaeoweathering on contemporary land surfaces between the mechanical weathering-dominated Ayn Formation, and the chemical weathering-dominated overlying Arkahawl Formation, which supports the notion of major glaciation followed by rapid climatic transit as basin margins were flooded and buried with sediment during post-glacial transgression. The carbon isotopic ratio (d13C) of the post-glacial carbonate is strongly variable from 23.5‰ to þ5.8‰, whereas carbonate fissures in the underlying basement range between þ4.1‰ and þ5.7‰. Two independent palaeomagnetic studies have yielded low palaeomagnetic latitudes for the Mirbat Group.

Precambrian sedimentary rocks have been known in Oman since early surveys of the Dhofar area of southern Oman (Carter 1852; Lees 1928) when the Mirbat Sandstone was recognized. Subsequently, a fourfold division of the sedimentary rocks in the Huqf area of east-central Oman was devised (Henson & Elliot 1958; Morton 1959; Beydoun 1960, 1964), and the Neoproterozoic – Early Cambrian succession referred to as the Huqf Group (Glennie 1977; Gorin et al. 1982). The Huqf Group was later raised in status to a Supergroup, comprising Abu Mahara, Nafun and Ara Groups (Glennie et al. 1974; Gorin et al. 1982; Hughes-Clarke 1988; Wright et al. 1990), the former containing glaciomarine deposits in the Jabal Akhdar region of northern Oman. Rieu et al. (2006, 2007a) and Allen (2007) recommended the use of the term Mirbat Group, to include the Ayn, Arkahawl, Marsham and Shareef Formations, in Dhofar. The Mirbat Group is unfossiliferous and comprises ,400-m-thick glacially influenced deposits of the Ayn Formation, overlain by c. 1 km of non-glacial marine deposits of the Arkahawl and Marsham Formations, followed by a ,40-m-thick unit of diamictite of the Shareef Formation. The precise correlation of the Mirbat Group with the Huqf Supergroup of east-central Oman and the Jabal Akhdar of northern Oman has been a subject of ongoing debate, largely because of the inadequacy of the geochronological database. Recent publications (Rieu et al. 2006, 2007a; Allen 2007) correlate the Mirbat Group of Dhofar with the Abu Mahara Group of northern Oman, whereas Kilner et al. (2005) correlated it with the Nafun Group. The glaciogenic origin of the Ayn Formation was first recognized by Qidwai et al. (1988) and the central part of its outcrop area was studied sedimentologically by Kellerhals (1993) and Kellerhals & Matter (2003), who reported a wide range of facies including subglacial, glaciomarine, fluvial, deltaic, turbiditic and

lacustrine deposits. Rieu et al. (2006) focused on a number of previously undescribed outcrops and interpreted the Ayn Formation in the context of the whole outcrop area north and west of Mirbat. Many Neoproterozoic sedimentary successions are preserved in marine basins strongly influenced by subaqueous gravity flows, leading to an active debate on the glacial v. non-glacial origin of the poorly sorted deposits labelled diamictites (Schermerhorn 1974; Hambrey & Harland 1981; Young & Gostin 1991; Young 1992; Ross et al. 1995; Arnaud & Eyles 2002b; Allen et al. 2004; Eyles & Januszczak 2004, 2007). In contrast, the Ayn Formation records a terrestrial to marginal marine succession, and therefore provides a valuable example of a glaciated basin margin where glaciofluvial and glaciodeltaic deposits are well developed. The Ayn Formation is overlain by a discontinuous cap carbonate that exhibits large and abrupt lateral changes in sedimentary facies. The presence of glaciogenic diamictites in thin (,40 m) isolated remnants beneath the erosional sub-Cretaceous unconformity in Jabal Samhan was recognized by Kellerhals (1993) and confirmed by Rieu (2006), who termed them the Shareef Formation. Their presence demonstrates a second glacial epoch preserved in the Mirbat Group, but the paucity of outcrop limits further studies.

Structural framework and basement geology The Mirbat Group crops out superbly along a 20 km NE –SW striking escarpment near the town of Mirbat (Fig. 19.1). It comprises up to 1.5 km of weakly deformed, almost entirely siliciclastic rocks that generally dip c. 108 to the NW. In the northeastern part of the outcrop belt, a sub-vertical fault juxtaposes the lower part of

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 239– 249. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.19

240

P. A. ALLEN ET AL.

Fig. 19.1. (a) Outcrop areas of Neoproterozoic basement and sedimentary rocks in Oman. (b) Geological map of the Mirbat area showing the location of palaeovalleys discussed in the text. (c) Summary stratigraphic column of the Mirbat Group. Based on Rieu et al. (2006) and Rieu et al. (2007a, b).

the Mirbat Group and basement with the sub-Cretaceous unconformity (see map in fig. 3 of Rieu & Allen 2008), and in the southern extremity of the outcrop belt small thrust faults cutting the Mirbat granodiorite fold the overstepping carbonate and terminate upwards in the Arkahawl Formation. In general, however, the Mirbat area escaped the deformation associated with the ‘Western Deformation Front’ to the NW (Allen 2007). It has been suggested that the Mirbat Group was formed on the southeastern flank of a c. 1000-km-long NE –SW striking rift basin imaged from the subsurface (Loosveld et al. 1996). The general

sediment transport direction in the Mirbat Group derived from palaeocurrent data was to the NW (Kellerhals 1993; Kellerhals & Matter 2003; Rieu et al. 2006, 2007b; Rieu & Allen 2008). This direction would therefore correspond to a transverse supply from a margin situated to the SE. The Mirbat Group unconformably overlies a complex crystalline and metamorphic basement (Platel et al. 1992a, b; Hauser & Zurbriggen 1992, 1994; Wu¨rsten 1994; Worthing 2005; Mercolli et al. 2006). Four major units are recognized in the basement at Mirbat (Mercolli et al. 2006; Bowring et al. 2007): (i) the Juffa

THE AYN FORMATION

Group, composed of metasedimentary mica-gneisses and amphibolites with a .1300 Ma (Mesoproterozoic) siliciclastic sedimentary protolith; (ii) the Sadh Group, comprising highly deformed and metamorphosed banded gneisses intruded by slightly deformed diorites and tonalites (Mahall Complex); (iii) the Tonalite Group, composed of three large calc-alkaline plutons dated 780– 800 Ma; and (iv) the Granite Group, comprising different types of dykes and small intrusive bodies including the Pegmatite Complex (770 –750 Ma), and the Mirbat Granodiorite (706 + 40 Ma), Leger Granite (726 + 0.4 Ma) and Shaat Dyke Swarm (c. 750 –700 Ma). A phase of accretion of juvenile Neoproterozoic crust (.900 Ma) was followed by extensive calc-alkaline magmatism, amphibolite facies metamorphism and intense deformation culminating at c. 800 Ma. Worthing (2005) believed the NW –SE orientated calc-alkaline Shaat Dyke Swarm to originate from subduction-modified lithospheric mantle, mobilized by crustal extension. The dykes appear to be similar in orientation and geochemistry to a dyke swarm (c. 700 Ma) in the Al Bayda island-arc terrane of Yemen (Windley et al. 1996). Late stocks such as the Mirbat Granodiorite (and probably the Leger Granite) and the Shaat dykes were most likely exhumed by the time of deposition of the oldest sedimentary rocks of the Mirbat Group (Qidwai et al. 1988; Platel et al. 1992a, b; Kellerhals & Matter 2003; Rieu et al. 2006, 2007a).

Stratigraphy The ,1.5-km-thick Mirbat Group comprises a lowermost glacial interval (,400-m-thick Ayn Formation) occupying deep palaeovalleys incised in crystalline basement (Fig. 19.1). The overlying Arkahawl Formation records at its base a major post-glacial transgression over the previous basin margin, followed by a 300– 400-m-thick turbidite complex consisting of 1–5-km-wide lobes embedded in fine-grained distal fan deposits (Rieu & Allen 2008). The turbidite complex gradationally passes up into c. 500 m of distal marine mudstone and siltstone. The overlying c. 100m-thick Marsham Formation records the progradation of shallow marine and fluviatile deposits, before a second glaciation represented by the Shareef Formation, barely preserved as thin remnants beneath the sub-Cretaceous unconformity.

Glaciogenic deposits and associated strata In the Jabal Akhdar of the Oman Mountains, Neoproterozoic glaciomarine deposits are found in the Abu Mahara Group (Ghubrah and Fiq Formations), occupying the tectonically deformed core of a major east –west oriented anticline, with no basement rocks exposed (Allen et al. 2011). In contrast, the Abu Maharaequivalent glaciogenic sedimentary rocks of the Ayn and Shareef Formations of the Mirbat Group of Dhofar are relatively undeformed and occur above an irregular crystalline basement (Kellerhals & Matter 2003; Rieu et al. 2006). Rieu et al. (2007a) proposed that the Arkahawl and Marsham Formations represent the missing stratigraphy between the Ghubrah and Fiq Formations in the Jabal Akhdar.

Ayn Formation The Ayn Formation comprises ,400 m of mainly coarse-grained glaciogenic deposits, ponded in 2- to .8-km-wide N- to NWoriented palaeovalleys eroded into basement, with few or no deposits preserved on intervening palaeohighs. Sedimentological analysis of near-continuous sections through the three westernmost palaeovalley-fills of the Ayn Formation (Fig. 19.2) allow four facies associations to be distinguished: (i) subaerial,

241

(ii) fluviodeltaic, (iii) proximal glaciomarine and (iv) distal glaciomarine. Details can be found in Rieu et al. (2006). Subaerial facies association. Units of the subaerial facies association, generally ,1 m thick, unconformably overlie crystalline and metamorphic basement at many localities, and comprise in situ brecciated bedrock interpreted as regolith, and poorly sorted, clast-supported breccia and crudely stratified diamictite thought to represent slope deposits. Clast type commonly reflects the directly underlying bedrock lithology. There is no evidence such as striated clasts or an underlying striated pavement that supports a glacial origin of the diamictites in this facies association. Fluviodeltaic facies association. The fluviodeltaic facies association comprises mainly coarse-grained sandstones and conglomerates that make up the lower part of the stratigraphy of palaeovalleys 1, 2 and 3. In palaeovalley 3, this facies association is also found higher up in the stratigraphy in two 20–30-m-thick units, alternated with glaciomarine deposits. Characteristic architectural elements include ,25-m-wide and ,2-m-thick channel fills, and 10– 15-m-thick, stacked tabular foreset beds with a c. 258 apparent dip. Massive, graded and cross-stratified sandstones and conglomerates are thought to represent low-sinuosity, bedload-dominated fluvial deposits, whereas the stacked tabular units are Gilbert-type deltas that identify the delta front. Interbedded siltstones were deposited by hemipelagic fallout in pro-delta bottomsets, and dropstone-bearing silty laminites demonstrate rainout from floating ice (Ovenshine 1970; Smith 2000) in a background glaciomarine or glaciolacustrine setting characterized by dilute underflows or interflows (e.g. Mackiewicz et al. 1984; Powell & Molnia 1989). In the central outcrop belt (palaeovalley 4), Kellerhals & Matter (2003) describe glacial striations preserved within the fluviodeltaic facies association, demonstrating dynamic glacier advance over fluvioglacial environments. Proximal glaciomarine facies association. Coarse-grained proximal

glaciomarine deposits constitute the bulk of the upper Ayn stratigraphy throughout the area. Deposits of this facies association overstep palaeovalley margins and may comprise single units that reach up to c. 150 m in thickness in palaeovalley 4 (Kellerhals & Matter 2003). Massive and weakly stratified diamictites with abundant striated clasts are the most common lithology in this facies association. Stratified diamictites contain common outsized clasts (,1 m) deflecting underlying lamination, and many striated and/or faceted clasts. Interbedded sandstones and conglomerates locally fill channels or occur as pebble lags. The presence of many large outsized, striated and faceted clasts is suggestive of rainout from floating ice (e.g. Ovenshine 1970; Syvitski et al. 1996; Smith 2000) of debris carried by subglacial transport prior to deposition (Boulton 1978; Benn & Evans 1998). Conglomerates within stratified diamictite units represent either local concentrations of clasts derived from iceberg overturn (Ovenshine 1970), lag deposits resulting from intensive wave or current reworking of ice-rafted diamictite in relatively shallow marine conditions (Eyles et al. 1985; Eyles 1988; Moncrieff & Hambrey 1990) or debris flows of remobilized sediment (Wright & Anderson 1982). Channel bodies comprising graded and crossstratified sandstones ‘embedded’ in diamictites may have been deposited by currents issuing from meltwater tunnel outlets (Rust & Romanelli 1975; Banerjee & MacDonald 1975; Powell 1981). The entire facies association suggests a proximal glaciomarine environment, close to the grounded ice-margin (Eyles et al. 1985; Moncrieff & Hambrey 1990; Syvitski et al. 1996), possibly in a grounding-line fan (Powell 1990; Lønne 1995). Distal glaciomarine facies association. The distal glaciomarine facies association consists mainly of laminated mudstones with subordinate sandstone laminae. Outsized, commonly striated, clasts

242 P. A. ALLEN ET AL.

Fig. 19.2. Summary sedimentological logs of the Ayn Formation from palaeovalleys 1 to 4 and interpretive correlation panel (after Rieu et al. 2006). (a) Logged sections through palaeovalleys and possible correlation (section PV-4 modified after Kellerhals & Matter, 2003). Logs; m, mudstone; s, sandstone; g, gravel/conglomerate. Facies notation follows Rieu et al. (2006). (b) Lateral correlation of palaeovalley fills and elevation of basement, based on logged sections and field mapping. Field map in Rieu (2006) and Rieu & Allen (2008).

THE AYN FORMATION

deflecting laminae occur throughout this facies association, but are much less common (c. 1%) than in the stratified diamictites of the proximal glaciomarine facies association. Distal glaciomarine facies also include graded sandstones and rare gravel and thin pebble conglomerates. The predominance of fine-grained laminated deposits points to a relatively quiet depositional environment dominated by hemipelagic sedimentation. Sediment gravity flow deposits are more sheet-like and thinner than those found in the proximal glaciomarine facies association, which together with the low abundance of outsized clasts, interpreted as ice-rafted debris, supports a relatively distal glaciomarine depositional environment (e.g. Mackiewicz et al. 1984; Eyles et al. 1985; Syvitski et al. 1996). In the central outcrop area (palaeovalley 4), Kellerhals & Matter (2003, p. 60) describe graded, turbiditic sandstones with dropstones at the tops of beds, interleaved with diamictite units, demonstrating contemporaneous sediment gravity flow processes and rainout from melting icebergs.

Post-glacial carbonate above Ayn Formation The post-glacial Arkahawl Formation, in contrast to the Ayn Formation, is laterally extensive and covers all remaining basement relief left after deposition of the Ayn Formation (Rieu 2006). At the base of the Arkahawl Formation is a ,3-m-thick transgressive cap carbonate that sharply overlies the underlying glaciomarine or fluviodeltaic deposits of the Ayn Formation or, where these are absent, directly overlies crystalline basement (Rieu et al. 2006). This carbonate is locally overlain by marine sandstones but more commonly by a 30 –40-m-thick succession of distal marine shales, indicating rapid deepening of water depths. Carbonate facies. The carbonate is laterally discontinuous and varies considerably in thickness and sedimentary facies. Carbonates are locally well developed on palaeohighs, whereas they are absent or represented by discontinuous, mixed carbonate – siliciclastic mass flow deposits in the deeper parts of the basin (i.e. in palaeovalleys). The carbonate comprises shallow-water, massflow and fissure-filling facies.

† Shallow-water facies association. In the western part of the outcrop area, onlapping onto crystalline basement or thin breccia, the carbonate is mainly a pink, recrystallized micritic dolomite, passing upward into wave-rippled calcareous sandstones and then a .1.5-m-thick, generally massive dolomite. In the centre of the outcrop area a locally well-developed carbonate with a basal conglomeratic lag overlies fluviodeltaic deposits. The carbonate is a dark grey, locally organic-rich, recrystallized micritic limestone showing mainly millimetrethin, undulating laminations, small-scale symmetrical wave ripples (eastern part of palaeovalley 3) and ,1-m-high, elongate domal stromatolites (centre of palaeovalley 3), which together indicate shallow-water deposition. † Carbonate mass-flow facies association. In palaeovalleys 1, 2 and 4, the carbonate is either absent or represented by carbonate-rich mass flow deposits containing a significant (,c. 75%) amount of siliciclastic material. Mass-flow deposits include centimetre-thick graded beds, interpreted as turbidites, decimetre-thick beds with soft-sediment deformation, interpreted as slump deposits, and poorly sorted conglomerates interpreted as debris flows. A ,15-m-thick erosive unit of boulder and cobble conglomerates overlies basement and glacial deposits in palaeovalley 1, containing a matrix of siliciclastic sand and detrital carbonate. The presence of bedding-parallel veneers of carbonate beneath and above this unit indicates that the conglomerates, which contain rare carbonate clasts, were deposited while carbonate was being produced in the basin. † Carbonate-filled fissures. In the western part of the area, a slightly foliated granodiorite (Mirbat Granodiorite) is host to

243

several populations of cross-cutting, millimetre-to-centimetrewide and ,10 m deep, fissures filled with detrital carbonate (dolomite) mixed with minor amounts of sand-sized siliciclastic particles, and void-filling cements. The fills of the fissures originate from the overlying carbonate (dolomite), which is further supported by their similarity in C-isotopic values (see below). In one location a small-scale (,2 cm high) set of foresets of a ‘mini delta’ developed over a centimetre-high negative step in the wall of the fracture indicates pumping of sediment and fluids upwards through a connected fracture system (Rieu et al. 2006). Fissures were evidently opened up at the same time as carbonate deposition during post-glacial transgression.

Boundary relations with overlying and underlying glacial strata The Mirbat Group is sandwiched between a basal unconformity represented by a series of deep palaeovalleys cut into crystalline basement, exposed in the bevelled coastal plain close to Mirbat town, and a regional sub-Cretaceous planar unconformity exposed high in the cliffs below Jabal Samhan. A map of the Group and its boundary relations is found in Rieu & Allen (2008, fig. 3).

Chemostratigraphy Carbon-isotopic composition of the post-glacial carbonate Rieu et al. (2006) measured the carbon isotopic composition of inorganic carbonate (d13C) in the carbonate at five different sites across the area and in the carbonate-filled fissures in basement (Fig. 19.3). Tests for diagenetic effects are discussed in Rieu et al. (2006). The most negative d13C values of the carbonate are found in the central part of the field area where they increase from 23.5‰ at the base to þ5.1‰ at the top in laminated and stromatolitic carbonate (sections C4 and C5). Further to the west, laminated carbonate overlying basement shows an increasing trend from þ0.9‰ at the base to þ2.7‰ at the top (section C3) and nearby carbonate associated with conglomerate shows upward-increasing values from þ3.5‰ to þ4.3‰ (section C3). The most positive values are found in the westernmost part of the area in carbonatefilled fissures in crystalline basement (between þ4.1‰ and þ5.7‰) and in the overlying laminated and wave-rippled carbonate, which shows an upward decreasing trend from þ5.8‰ at the base to þ2.0‰ at the top (sections C1 and C2). The lateral variation in d13C in the carbonate is interpreted to represent a secular variation of the C-isotopic composition of ocean water. Owing to progressive flooding of the palaeotopographically highest parts of the basin (i.e. in the west), carbonate deposition in these locations would have been delayed with respect to the basin centre. The resulting d13C signal of the carbonate on the basement high is therefore ‘base-truncated’. This interpretation allows the construction of a composite isotope profile through the carbonate, showing an increasing trend from 23‰ to þ5.8‰ and then decreasing to þ2‰ at the top (Fig. 19.3). In the central area (sections C4, C5; palaeovalley 3) values reach a maximum of þ5.1‰, but the decreasing arm is missing (‘top-truncated’), which is in accord with the observation that carbonate deposition was restricted mainly to shallower environments.

Neoproterozoic weathering: Chemical Index of Alteration Rieu et al. (2007b) studied the compositional and mineralogical changes in the Mirbat Group in order to test the possible effect of severe climate swings in the Neoproterozoic on the intensity of chemical weathering in source areas (Nesbitt & Young 1982;

244

P. A. ALLEN ET AL.

Fig. 19.3. Summary logs, d13C data and major element data of the post-glacial carbonate (after Rieu et al. 2006). See Figure 19.2 for location of C1 –C5. (a) Isotopic profiles and summary logs are shown in their relative vertical positions, high in the SW where the cap carbonate onlaps basement, and low in the centre of the outcrop area where the cap overlies fluvial deposits. (b) Composite isotopic profile, with sections C1 and C4 compressed in order to account for different sedimentation rates and to match sections C2 and C5, respectively. Dashed arrows indicate the relation between cap carbonate and carbonate fissures. (c) Element concentrations for selected samples from composite section.

THE AYN FORMATION

McLennan et al. 1993; Fedo et al. 1995; Nesbitt et al. 1996; Scheffler et al. 2003; Bahlburg & Dobrzinski 2011). A similar study of Neoproterozoic sedimentary rocks using the chemical index of alteration (CIA) is found in Dobrzinski et al. (2004). To ensure a well-mixed provenance and to minimize the effects of hydrodynamic sorting, Rieu et al.’s (2007b, c) study was limited to mudstone beds and the mudstone matrices of diamictites. Throughout the Mirbat Group there are significant compositional and mineralogical variations (Fig. 19.4). When plotted in A –CN –K and Qtz – Pl– Kfs space, the data define trends parallel to A – CN and Qtz –Pl boundaries, thus suggesting that compositional variations result from variability in the extent of chemical weathering in the regolith in source areas. Glacial deposits of the Ayn Formation (AY2 –AY8) plot close to the feldspar join and close to the composition of unaltered granodiorite, indicating that their composition has been little affected by chemical alteration, as reflected by their low average CIA value of 54. Pre-glacial deposits of the Ayn and Marsham Formation, and non-glacial deposits of the Arkahawl Formation in particular, are enriched in Al2O3. This enrichment suggests the incorporation of weathered material, in agreement with the relatively high abundance of clay minerals, and is reflected by average CIA values of 60 and 74 for the pre-glacial and non-glacial intervals, respectively. In Qtz – Pl– Kfs space, a similar trend is revealed, with samples from the glacial, pre-glacial and non-glacial intervals, respectively, being progressively enriched in quartz with respect to plagioclase, resulting in progressively higher average MIA (mineralogical index of alteration ¼ Qtz/(Qtz þ Kfs þ Pl)) values (18, 36 and 46, respectively). Rieu et al. (2007b) believed that these

trends were best explained by variations in the intensity of palaeoweathering of contemporary land surfaces.

Palaeolatitude and palaeogeography Kilner et al. (2005) reported palaeomagnetic data that passed fold and reversal tests from Oman, which gave the Muscat region a palaeolatitude of 138 in the late Neoproterozoic. They recognized a remnant magnetization that was acquired between deposition and a period of deformation in the Cambrian – Ordovician (Blendinger et al. 1990; Gass et al. 1990), and argued (p. 415) that the magnetization was acquired at deposition since a similar pattern of reversals occurs at the same stratigraphic level in widely separated locations in the north, centre and south of Oman. Subsequent work, however, points strongly to different ages for these stratigraphic sections (Allen & Leather 2006; Rieu et al. 2006; Rieu et al. 2007a). The difference between the Oman palaeopoles and those of Gondwanan terranes (Meert 2003) suggested to Kilner et al. (2005) that Oman did not accrete with the ‘African’ terranes comprising greater Gondwana until after c. 600 Ma, the approximate age of the youngest stratigraphy sampled for palaeomagnetism. Kempf et al. (2000) recovered a Precambrian pole position from the south of Oman (implying a palaeolatitude of 98) but noted a similarity with published poles from Gondwana (Congo, India, Australia) at c. 550 Ma, which was thought to be the depositional age. The rocks in south Oman providing the low palaeolatitude are c. 700 Ma in age, not 550 Ma (Rieu et al. 2006, 2007a), which suggests that either the south of Oman was in low

MIA

CIA

1200

50

60

70

0

20

40

60

80

glacial epoch II

Mhm Shf

unconformity

245

decreasing temperatures

800

Arkahawl Fm

Non-glacial interlude or ‘interglacial’

cap carbonate

rapidly increasing temperatures

Ayn Fm

400

glacial epoch I

0m 50

60

70

CIA

0

20

40

MIA

60

80

Fig. 19.4. Variations in weathering indices (CIA and MIA) in the Mirbat Group, displayed as a function of their stratigraphic height (from Rieu et al. 2007b). Low CIA and MIA values are associated with glacial and pre-glacial conditions inferred from sedimentological facies, and relatively high CIA and MIA values are associated with non-glacial marine deposits. Errors on CIA are ,2%.

246

P. A. ALLEN ET AL.

palaeolatitudes at c. 700 Ma and experienced similar palaeolatitudes at 550 Ma, or that the magnetization was acquired during a later event close to 550 Ma, when Oman and the adjacent parts of Gondwana were indeed situated in the tropics on the basis of their common evaporitic sedimentary facies (Allen 2007). Geological evidence summarized in Allen (2007) indicates an earlier accretion of Oman into the megasuture forming the eastern border (in present-day coordinates) of Gondwana. Currently, therefore, the palaeomagnetic data described by Kempf et al. (2000) and Kilner et al. (2005) lack a rigorous stratigraphic and geochronological context.

870

810

Shareef Fm (glaciation)

Geochronological constraints

n = 177 (69)

810 840

Mirbat Group

The age of the Mirbat Group is uncertain. The maximum age must be younger than the age of the youngest intrusions in the basement complex truncated by the basal unconformity. The carbonate stratigraphically above the Ayn Formation is therefore younger than the age of the underlying Mirbat Granodiorite (706 + 40 Ma, Rb –Sr date, Gass et al. 1990). The Leger Granite (726 + 0.4 Ma U –Pb zircon date, Bowring et al. 2007) cannot be proven to pre-date the Mirbat Group because it is truncated by the sub-Cretaceous unconformity to the NE of the outcrop belt of the Mirbat Group, but it is nevertheless highly likely (see below). These late intrusions were most likely exhumed by the time of deposition of the oldest sedimentary rocks of the Mirbat Group (Qidwai et al. 1988; Platel et al. 1992a, b; Kellerhals & Matter 2003; Rieu et al. 2006, 2007a). The dykes at Mirbat do not cut the unconformably overlying sedimentary Mirbat Group, as recognized by Qidwai et al. (1988), so the Mirbat Group sedimentary rocks must be younger than the 696 –744 Ma age of the dykes derived from Sm– Nd and Rb – Sr dating (Worthing 2005). In order to further constrain the age of the Mirbat Group and the contributing source areas, use has been made of detrital zircon geochronology. A total of 1057 new U– Pb detrital zircon ages were produced by laser ablation ICP-MS from the Huqf Supergroup (Rieu et al. 2007a; Allen 2007). The sampled stratigraphic intervals in the Mirbat Group include the Ayn, Arkahawl, Marsham and Shareef Formations. Samples of the Ayn Formation come from three different diamictite units (D1, D2 and D3 of Kellerhals & Matter 2003) in Wadi Autunt. The sample of the Arkahawl Formation was obtained from turbiditic sandstones in Wadi Hinuna, c. 75 m above the top of the Ayn Formation. From the Marsham Formation two samples were collected near Jabal Shareef from shoreface sandstones. Diamictites of the Shareef Formation were sampled in the same section, as well as in a section 3 km to the NE. The stratigraphic units of the Mirbat Group contain remarkably similar detrital zircon populations, comprising zircons almost exclusively of Neoproterozoic age (Rieu et al. 2007a) (Fig. 19.5). When only the most concordant data and those with the lowest analytical errors are considered, main peaks in the Ayn Formation are revealed at c. 722, 810 and 840 Ma. In the Arkahawl and Marsham formations, only the c. 810 Ma and c. 840 Ma peaks are present, whereas in the Shareef Formation the zircon population is dominated by c. 870 Ma ages with only a small ‘shoulder’ at c. 810 Ma. The maximum age constraint for the depositional age of the Ayn Formation and the overlying part of the Mirbat Group is provided by the c. 722 Ma subpopulation in the Ayn Formation, represented by four grains with a mean age of 722 +19 Ma. This age is in agreement with the constraints provided by the 700–750 Ma age of the underlying basement based on Rb – Sr, K –Ar and Sm– Nd whole-rock and mineral analyses (Gass et al. 1990; Worthing 2005; Mercolli et al. 2006), but is considered to be more robust. The detrital zircon populations are dominated by Neoproterozoic ages that are in good agreement with the age range (0.7 – 1.0 Ga) of local basement rocks in the Mirbat area. Although there is no stratigraphic contact between the Mirbat Group and

Arkahawl + Marsham Fms n = 92 (59)

840 810

Ayn Fm (glaciation)

722

n = 180 (111)

0.5

1.0

1.5

Age (x109 years) Fig. 19.5. Frequency distribution of detrital zircons from the Mirbat Group (modified from Rieu et al. 2007a). Grey areas are data produced by filter 1 (,25% discordant; 2s errors ,20% for 235U– 207Pb and ,10% for 238 U – 206Pb). Black lines are data produced by filter 2 (,10% discordant; 2s errors ,10% for 235U – 207Pb and 238U– 206Pb). Number of analyses (n) produced with filters 1 and 2 are displayed outside and inside parentheses respectively. Numbers for peaks are in Ma.

the Leger Granite (726 + 0.4 Ma), the 722 +19 Ma subpopulation of zircons in the Ayn Formation may have been derived from this or associated intrusions, suggesting that the Leger Granite intruded prior to deposition of the Ayn Formation and

THE AYN FORMATION

was exhumed shortly after. This subpopulation is similar in age to the zircons found in tuffaceous ashes of the Ghubrah Formation in northern Oman (Brasier et al. 2000).

Discussion: the nature of Neoproterozoic glaciation The Ayn stratigraphy suggests that in basin-marginal settings, periods of glaciomarine deposition alternated with periods of fluvial and deltaic deposition during glacial recession. Thus, terrestrial sediment routing systems functioned strongly during these periods of recession, by which plentiful sediment could be transported by high-energy fluvial systems and delivered to more distal, hemipelagic-dominated areas. CIA analysis suggests a dominance of mechanical weathering at this time. Although the simplest interpretation is that the entire glacial epoch was characterized by dynamic glacier advance and retreat cycles, it is possible that preservation of stratigraphy took place preferentially during a long, pulsed deglacial phase. The stratigraphy of the Ayn Formation is most easily reconcilable with ‘soft snowball’ models allowing for ice margin fluctuations and the presence of equatorial oceans (Hyde et al. 2000; Crowley et al. 2001; Peltier et al. 2007). A central tenet of the Snowball Earth hypothesis is the occurrence of laterally extensive cap carbonates, which sharply overlie most Neoproterozoic glacial deposits, with typically negative C-isotopic ratios that have been used for global correlation (Kaufman & Knoll 1995; McKirdy et al. 2001; Halverson et al. 2005). Cap carbonates have long been considered paradoxical because they suggest an abrupt change from glacial to tropical conditions (Hoffman & Schrag 2002, and references therein). Models requiring less extreme environmental changes have been proposed to explain the presence of cap carbonates (Grotzinger & Knoll 1995; Kennedy 1996; Hoffman et al. 1998; Kennedy et al. 2001; Jiang et al. 2003). Eyles & Januszczak (2004) doubted that cap carbonates had any precise palaeoenvironmental significance, and claimed that many of them are made up of detrital carbonate derived by erosion rather than representing a distinct oceanographic condition. It is important to stress first that the Mirbat cap carbonate is clearly associated with a significant palaeoclimatic change, because it marks the termination of a glacial epoch. Second, a detrital origin is unlikely, because it occurs within a .1-km-thick succession devoid of any other primary or detrital carbonate. However, the cap carbonate described here contrasts with many other cap carbonates worldwide, as it is discontinuous and shows considerable variations in facies, thickness and C-isotopic signal over very short lateral distances (hundreds of metres). Rieu et al.’s (2006) data suggest that carbonate was produced in the shallower parts of the basin and redistributed downslope from these palaeohighs, rather than as a blanket of carbonate precipitated from an oversaturated ocean (Hoffman et al. 1998; Hoffman & Schrag 2002; Shields 2005). Carbonate production therefore appears to have been determined by the same depth-dependent effects as we see in the photosynthesis-mediated processes of Phanerozoic carbonate production. If carbonate precipitation took place beyond the shallow waters of the palaeohighs, steep submarine relief in the palaeovalleys may have prevented the primary deposited carbonate from being preserved in situ.

Conclusions The Ayn Formation of the Mirbat Group is a c. 400-m-thick succession of glaciofluvial, glaciodeltaic and glaciomarine sediments deposited in several deeply incised palaeovalleys cutting the metamorphic and igneous basement in the Mirbat area of Dhofar. The Formation is overlain by a discontinuous, thin (,3 m) and diachronous cap carbonate that marks an abrupt overstepping of the basin margin and deepening of palaeoenvironments. The Shareef

247

Formation is a second, little-known glaciogenic unit preserved patchily beneath the sub-Cretaceous unconformity. The sedimentary evolution of the Mirbat Group is paralleled by variations in the CIA derived from mudstones and the finegrained matrices of diamictites. The CIA suggests conditions with minimal chemical weathering during the deposition of the Ayn Formation, followed by intensive weathering in the overlying Arkahawl Formation above the cap carbonate. This geochemical trend strongly supports the Ayn Formation as glaciogenic, followed by climatic transit as basin margins were flooded during post-glacial transgression. The age of the Ayn Formation is constrained by (i) the age of the youngest detrital zircons (four zircons with a mean of 722 + 19 Ma), (ii) the age of the Shaat dyke swarm cutting basement and which does not penetrate the Ayn Formation (696 –744 Ma), (iii) the age of late plutons such as the Mirbat Granodiorite (706 + 40 Ma), which is definitely erosionally overlain by cap carbonate and the Leger Granite (726 + 0.4 Ma), which is most likely but not definitely older than the Ayn Formation. It is therefore highly likely that the Ayn Formation is younger than c. 720 Ma. The age of the Shareef Formation is unknown. The Mirbat Group contains detrital zircons of exclusively Neoproterozoic age, closely matching the range of ages of metamorphic and igneous rocks in the underlying basement, which formed during a period of major crustal growth associated with subduction at the southeastern periphery of the Arabian-Nubian Shield. The senior author is grateful for the financial and logistical support of Petroleum Development Oman (PDO) and Enterprise Ireland, supplemented in later years by the Research Committee of ETH-Zu¨rich. We are grateful for the constructive input on field visits of J. Grotzinger, S. Bowring, P. Hoffman, G. Halverson, PDO staff (particularly J. Amthor) and the participants of the field trip associated with the IAS Regional Meeting in Muscat in 2005. I am grateful for the reviews of N. Dobrinski, I. Fairchild and E. Arnaud. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Allen, P. A. 2007. The Huqf Supergroup of Oman: basin development and context for Neoproterozoic glaciation. Earth Science Reviews, 84, 139– 185. Allen, P. A. & Leather, J. 2006. Siliciclastic marine sedimentation in the aftermath of a Marinoan glacial epoch: the Masirah Bay Formation, Huqf Supergroup, of Oman. Precambrian Research, 144, 167–198, doi: 10.1016/j.precamres.2005.10.006. Allen, P. A., Leather, J. & Brasier, M. D. 2004. The Neoproterozoic Fiq glaciation and its aftermath, Huqf Supergroup of Oman. Basin Research, 16, 507– 534, doi: 10.1111/j.1365-2117.2004.00249.x. Allen, P. A., Leather, J. et al. 2011. The Abu Mahara Group (Ghubrah and Fiq Formations), Jabal Akhdar, Oman. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 251–262. Arnaud, E. & Eyles, C. H. 2002. Glacial influence on Neoproterozoic sedimentation: the Smalfjord Formation, northern Norway. Sedimentology, 49, 765– 788. Bahlburg, H. & Dobrzinski, N. 2011. A review of the Chemical Index of Alteration (CIA) and its application to the study of Neoproterozoic glacial deposits and climate transitions. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 81– 92. Banerjee, I. & MacDonald, B. C. 1975. Nature of esker sedimentation. In: Jopling, A. V. & McDonald, B. C. (eds) Glaciofluvial and Glaciolacustrine Sedimentation. Society of Economic Paleontologists and Mineralogists, Special Publication, 23, 132– 154. Benn, D. I. & Evans, D. J. A. 1998. Glaciers and Glaciation. Arnold, London.

248

P. A. ALLEN ET AL.

Beydoun, Z. R. 1960. Synopsis of the geology of East Aden Protectorate. Report of the 21st International Geological Congress, Copenhagen, Part 21, 131– 149. Beydoun, Z. R . 1964. The stratigraphy and structure of the East Aden Protectorate. Overseas geology and mineral resources supplement series. Bulletin supplement, 5, 1– 107. Blendinger, W., van Vliet, A. & Hughes Clarke, M. W. 1990. Updoming, rifting and continental margin development during the Late Palaeozoic in northern Oman. In: Robertson, A. H. F., Searle, M. P. & Ries, A. C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publication, 49, 27 –37. Boulton, G. S. 1978. Boulder shapes and grain-size distribution of debris as indicators of transport paths through a glacier and till genesis. Sedimentology, 25, 773– 799. Bowring, S. A., Grotzinger, J. P., Condon, D. J., Ramezani, J., Newall, M. J. & Allen, P. A. 2007. Geochronologic constraints on the chronostratigraphic framework of the Neoproterozoic Huqf Supergroup, Sultanate of Oman. American Journal Science, 307, 1097– 1145. Brasier, M. D., McCarron, G., Tucker, R., Leather, J., Allen, P. A. & Shields, G. 2000. New U– Pb zircon dates for the Neoproterozoic Ghubrah glaciation and for the age of the top of the Huqf Supergroup, Oman. Geology, 28, 175–178, doi: 10.1130/00917613(2000)028,0175:NUPZDF.2.2.CO;2. Carter, H. J. 1852. Memoir on the geology of the south east coast of Arabia. Journal of the Royal Asiatic Society, Bombay Br., 4, 21–96. Crowley, T. H., Hyde, W. T. & Peltier, W. R. 2001. CO2 levels required for glaciation of a ‘near-snowball’ Earth. Geophysical Research Letters, 28, 283–286, doi: 10.1029/2000GL011836. Dobrzinski, N., Bahlburg, H., Strauss, H. & Zhang, Q. 2004. Geochemical climate proxies applied to the Neoproterozoic glacial succession on the Yangtze Platform, South China. In: Jenkins, G. S., McMenamin, M. A. S., McKay, C. P. & Sohl, L. (eds) The Extreme Proterozoic: Geology, Geochemistry and Climate. American Geophysical Union Monograph, 146, 13– 32. Eyles, C. H. 1988. Glacially and tidally influenced shallow marine sedimentation of the Late Precambrian Port Askaig Formation. Palaeogeography, Palaeoclimatology, Palaeoecology, 68, 1 –25. Eyles, N. & Januszczak, N. 2004. Zipper-rift: a tectonic model for Neoproterozoic glaciations during the break-up of Rodinia after 750 Ma. Earth Science Reviews, 65, 1 – 73, doi: 10.1016/ S0012-8252(03)00080-1. Eyles, N. & Januszczak, N. 2007. Syntectonic subaqueous mass flows of the Neoproterozoic Otavi Group, Namibia: where is the evidence for global glaciation? Basin Research, 19, 179–198, doi: 10.1111/ j.1365-2117.2007.00319.x. Eyles, C. H., Eyles, N. & Miall, A. D. 1985. Models of glacimarine sedimentation and their application to the interpretation of ancient glacial sequences. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 15 –84. Fedo, C. M., Nesbitt, H. W. & Young, G. M. 1995. Unravelling the effects of potassium metasomatism in sedimentary rocks and paleosols, with implications for paleoweathering conditions and provenance. Geology, 23, 921– 924, doi: 10.1130/00917613(1995)023,0921:UTEOPM.2.3.CO;2. Gass, I. G., Ries, A. C., Shackleton, R. M. & Smewing, J. D. 1990. Tectonics, geochronology and geochemistry of the Precambrian rocks of Oman. In: Robertson, A. H. F., Searle, M. P. & Ries, A. C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 585– 599. Glennie, K. W. 1977. Outline of the geology of Oman. Me´moires de la Societe´ Ge´ologique de France, hors se´rie, 8, 25– 31. Glennie, K. W., Boeuf, M. G. A., Hughes Clark, M. W., MoodyStuart, M., Pilaar, W. F. H. & Reinhardt, B. M. 1974. Geology of the Oman Mountains. KSEPL, Rijswijk, The Netherlands. Gorin, G. E., Raacz, L. G. & Walter, M. R. 1982. Late Precambrian – Cambrian sediments of Huqf Group, Sultanate of Oman. American Association Petroleum Geologists Bulletin, 66, 2609– 2627. Grotzinger, J. P. & Knoll, A. H. 1995. Anomalous carbonate precipitates: is the Precambrian the key to the Permian? Palaios, 10, 578– 596.

Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Toward a Neoproterozoic composite carbonisotope record. Geological Society of America Bulletin, 117, 1181– 1207. Hambrey, M. J. & Harland, W. B. 1981. Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge. Hauser, A. & Zurbriggen, R. 1992. Geology of the Crystalline Basement of the Hadbin area (Salalah area, Dhofar, Sultanate of Oman). Unpublished MSc thesis, University of Berne, Switzerland, 6– 221. Hauser, A. & Zurbriggen, R. 1994. Geology of the crystalline basement of the Hadbin area (Salalah area, Dhofar, Sultanate of Oman). Schweizerische Mineralogisch Petrographische Mitteilungen, 74, 213– 226. Henson, F. R. S. & Elliot, G. F. 1958. Exhibition of specimens of ‘Collenia’ from pre-Permian sediments in south Arabia. Proceedings Geological Society London, 1561, 89– 90. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129–155, doi: 10.1046/j.1365-3121.2002.00408.x. Hoffman, P. F., Kaufmann, A. J., Halverson, G. P. & Schrag, D. P. 1998. A Neoproterozoic snowball Earth. Science, 281, 1342–1346, doi: 10.1126/science.281.5381.1342. Hughes Clarke, M. W. 1988. Stratigraphy and rock nomenclature in the oil-producing area of interior Oman. Journal Petroleum Geology, 11, 5 –60. Hyde, W. T, Crowley, T. J., Baum, S. K. & Peltier, W. R. 2000. Neoproterozoic ‘snowball Earth’ simulations with a coupled climate/ ice-sheet model. Nature, 405, 425– 429, doi: 10.1038/35013005. Jiang, G., Kennedy, M. J. & Christie-Blick, N. 2003. Stable isotopic evidence for methane seeps in Neoproterozoic post-glacial cap carbonates. Nature, 426, 822– 826. Kaufman, A. J. & Knoll, A. H. 1995. Neoproterozoic variations in the C-isotopic composition of seawater: stratigraphic and biogeochemical implications. Precambrian Research, 73, 27– 49. Kellerhals, P. 1993. Igneous Petrology of the Mirbat Complex and Facies Analysis of the Mirbat Sandstone Formation (Middle and Upper Member). Unpublished MSc thesis, University of Berne, Switzerland. Kellerhals, P. & Matter, A. 2003. Facies analysis of a glaciomarine sequence, the Neoproterozoic Mirbat Sandstone Formation, Sultanate of Oman. Eclogae Geologicae Helvetiae, 96, 49 – 70. Kennedy, M. J. 1996. Stratigraphy, sedimentology and isotope geochemistry of Australian Neoproterozoic post-glacial cap dolostones: deglaciation, d13C excursions and carbonate precipitation. Journal Sedimentary Research, 66, 1050– 1064. Kennedy, M. J., Christie-Blick, N. & Sohl, L. E. 2001. Are Proterozoic cap carbonates and isotopic excursions a record of gas hydrate destabilization following Earth’s coldest intervals? Geology, 29, 443– 446. Kempf, O., Kellerhals, P., Lowrie, W. & Matter, A. 2000. Paleomagnetic directions in late Precambrian glaciomarine sediments of the Mirbat Sandstone Formation, Oman. Earth & Planetary Science Letters, 175, 181– 190, doi: 10.1016/S0012-821X(99)00307-6. Kilner, B., MacNiocaill, C. & Brasier, M. D. 2005. Low-latitude glaciation in the Neoproterozoic of Oman. Geology, 33, 413– 416, doi: 10.1130/G21227.1. Lees, G. M. 1928. The geology and tectonics of Oman and parts of southeastern Arabia. Quarterly Journal Geological Society London, 84, 585– 670. Lønne, I. 1995. Sedimentary facies and depositional architecture of ice-contact glaciomarine systems. Sedimentary Geology, 98, 13– 43. Loosveld, R., Bell, A. & Terken, J. 1996. The tectonic evolution of interior Oman. GeoArabia, 1, 28 –50. Mackiewicz, N. E., Powell, R. D., Carlson, P. R. & Molnia, B. F. 1984. Interlaminated ice-proximal glacimarine sediments in Muir Inlet, Alaska. Marine Geology, 57, 113–147. McKirdy, D. M., Burgess, J. M. et al. 2001. A chemostratigraphic overview of the late Cryogenian interglacial sequence in the Adelaide Fold-Thrust Belt, South Australia. Precambrian Research, 106, 149– 186. McLennan, S. M., Hemming, S., McDaniel, D. K. & Hanson, G. N. 1993. Geochemical approaches to sedimentation, provenance and tectonics. In: Johnsson, M. J. & Basu, A. (eds) Processes

THE AYN FORMATION

Controlling the Composition of Clastic Sediment. Geological Society of America Special Paper, 284, 21 –40. Meert, J. G. 2003. A synopsis of events related to the assembly of eastern Gondwana. Tectonophysics, 362, 1 –40. Mercolli, I., Briner, A. P., Frei, R., Scho¨nberg, R., Nagler, T. F., Kramers, J. & Peters, T. 2006. Lithostratigraphy and geochronology of the Neoproterozoic crystalline basement of Salalah, Dhofar, Sultanate of Oman. Precambrian Research, 145, 182– 206. Moncrieff, A. C. M. & Hambrey, M. J. 1990. Marginal marine glacial sedimentation in the late Precambrian succession of east Greenland. In: Dowdeswell, J. A. & Scourse, J. D. (eds) Glacimarine Environments: Processes and Sediments. Geological Society, London, Special Publications, 53, 387– 410. Morton, D. M. 1959. The Geology of Oman. 5th World Petroleum Congress, 1959, Section 1, paper 14. Nesbitt, H. W. & Young, G. M. 1982. Early Proterozoic climates and plate motions inferred from major element chemistry of lutites. Nature, 299, 715– 717, doi: 10.1038/299715a0. Nesbitt, H. W., Young, G. M., McLennan, S. M. & Keays, R. R. 1996. Effects of chemical weathering and sorting on the petrogenesis of siliciclastic sediments, with implications for provenance studies. Journal of Geology, 104, 525– 542. Ovenshine, A. T. 1970. Observations of iceberg rafting in Glacier Bay, Alaska, and the identification of ancient ice-rafted deposits. Geological Society America Bulletin, 81, 891– 894. Peltier, W. R., Liu, Y. & Crowley, J. W. 2007. Snowball Earth prevention by dissolved organic carbon remineralization. Nature, 450, 813– 818. Platel, J. P., Roger, J., Peters, T., Mercolli, I., Kramers, J. D. & Le Me´tour, J. 1992a. Geological Map of Salalah, Sheet NE 40-09, 1: 250 000, with explanatory notes. Directorate General of Minerals, Oman Ministry of Petroleum and Minerals. Platel, J. P., Le Me´tour, J., Berthiaux, A., Buerrier, M. & Roger, J. 1992b. Geological map of Juzor Al Halaaniyat, sheet NE 40-10, scale 1:250,000. Directorate General of Minerals, Oman Ministry of Petroleum and Minerals. Powell, R. D. 1981. A model for sedimentation by tidewater glaciers. Annals Glaciology, 2, 129–134. Powell, R. D. 1990. Glacimarine processes at grounding-line fans and their growth to ice-contact deltas. In: Dowdeswell, J. A. & Scourse, J. D. (eds) Glacimarine Environments: Processes and Sediments. Geological Society, London, Special Publications, 53, 53 – 73. Powell, R. D. & Molnia, B. F. 1989. Glacimarine sedimentary processes, facies and morphology of the south-southeast Alaska Shelf and fjords. Marine Geology, 85, 359– 390. Qidwai, H. A., Khalifa, M. I. & Ba-Mkhalif, K. A. 1988. Evidence of Permo-Carboniferous glaciation in the basal Murbat Sandstone Formation, Southern Region, Sultanate of Oman. Journal Petroleum Geology, 11, 81 –88. Rieu, R. 2006. Sedimentology, Stratigraphy and Geochemistry of the Glacially Influenced Neoproterozoic Mirbat Group, Oman. PhD thesis, ETH-Zu¨rich. Rieu, R. & Allen, P. A. 2008. Siliciclastic sedimentation in the interlude between two Neoproterozoic glaciations, Mirbat area, southern Oman: a missing link in the Huqf Supergroup. GeoArabia, 13, 45 – 72. Rieu, R., Allen, P. A., Etienne, J. L., Cozzi, A. & Weichert, U. 2006. A Neoproterozoic glacially influenced basin margin succession and ‘atypical’ cap carbonate associated with bedrock paleovalleys, Mirbat area, southern Oman. Basin Research, 18, 471– 496. Rieu, R., Allen, P. A., Cozzi, A., Kosler, J. & Bussy, F. 2007a. A composite stratigraphy for the Neoproterozoic Huqf Supergroup of Oman: integrating new litho-, chemo- and chronostratigraphic data

249

of the Mirbat area, southern Oman. Journal of the Geological Society of London, 164, 997– 1009. Rieu, R., Allen, P. A., Plo¨tze, M. & Pettke, T. 2007b. Compositional and mineralogical variations in a Neoproterozoic glacially influenced succession, Mirbat area, south Oman: implications for paleoweathering conditions. Precambrian Research, 154, 248– 265, doi: 10.1016/ j.precamres.2007.01.003. Rieu, R., Allen, P. A., Plo¨tze, M. & Pettke, T. 2007c. Climatic cycles during a Neoproterozoic ‘snowball’ glacial epoch. Geology, 35, 299– 302, doi: 10.1130/G23400A.1. Ross, G. M., Bloch, J. D. & Krouse, H. R. 1995. Neoproterozoic strata of the southern Canadian Cordillera and the isotopic evolution of seawater sulfate. Precambrian Research, 73, 71 –99. Rust, B. R. & Romanelli, R. 1975. Late Quaternary subaqueous outwash deposits near Ottawa, Canada. In: Joplin, A. V. & McDonald, B. C. (eds) Glaciofluvial and Glaciolacustrine Sedimentation. Society of Economic Paleontologists and Mineralogists, Special Publication, 25, 177– 192. Scheffler, K., Hoernes, S. & Schwark, L. 2003. Global changes during Carboniferous-Permian glaciation of Gondwana: linking polar and equatorial climate evolution by geochemical proxies. Geology, 31, 605–608, doi: 10.1130/0091-7613(2003)031, 0605:GCDCGO.2.0.CO;2. Schermerhorn, L. J. G. 1974. Late Precambrian mixtites: glacial and/or nonglacial? American Journal Science, 274, 673– 824. Shields, G. A. 2005. Neoproterozoic cap carbonates: a critical appraisal of existing models and the plumeworld hypothesis. Terra Nova, 4, 299– 310. Smith, I. 2000. Diamictic sediments within high Arctic lake sediment cores: evidence for lake ice rafting along the lateral glacial margin. Sedimentology, 47, 1157– 1179. Syvitski, J. P. M., Andrews, J. T. & Dowdeswell, J. A. 1996. Sediment deposition in an iceberg-dominated glacimarine environment, East Greenland: basin-fill implications. In: Solheim, A., Riis, F., Elverhoi, A., Faleide, J. I., Jensen, L. N. & Cloetingh, S. (eds) Impact of Glaciations on Basin Evolution: Data and Models from the Norwegian Margin and Adjacent Areas. Global Planetary Change, 12, 251– 270. Windley, B. F., Whitehouse, M. J. & Ba-Bttat, M. A. O. 1996. Early Precambrian gneiss terranes and Pan-African island arcs in Yemen: crustal accretion of the eastern Arabian Shield. Geology, 24, 131–134. Worthing, M. A. 2005. Petrology and geochronology of a Neoproterozoic dyke swarm from Marbat, south Oman. Journal African Earth Sciences, 41, 248–265. Wright, R. & Anderson, J. B. 1982. The importance of sediment gravity flow sediment transport and sorting in a glacial marine environment: eastern Weddell Sea, Antarctica. Geological Society America Bulletin, 93, 951–963. Wright, V. P., Ries, A. C. & Munn, S. G. 1990. Intraplatformal basin-fill from the Infracambrian Huqf Group, east-central Oman. In: Robertson, A. H. F., Seale, M. P. & Ries, A. C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 601–616. Wu¨rsten, F. 1994. The Precambrian crystalline basement of Salalah. Unpublished PhD thesis, University of Berne, Switzerland. Young, G. M. 1992. Neoproterozoic glaciation in the Broken Hill area, New South Wales, Australia. Geological Society America Bulletin, 104, 840–850. Young, G. M. & Gostin, V. A. 1991. Late Proterozoic (Sturtian) succession of the North Flinders Basin, South Australia: an example of temperate glaciation in an active rift setting. In: Anderson, J. B. & Ashley, G. M. (eds) Glacial Marine Sedimentation: Paleoclimatic Significance. Geological Society America Special Paper, 261, 207– 223.

Chapter 20 The Abu Mahara Group (Ghubrah and Fiq formations), Jabal Akhdar, Oman PHILIP A. ALLEN1*, JONATHAN LEATHER2, MARTIN D. BRASIER3, RUBEN RIEU4, MARGARET MCCARRON3, ERWAN LE GUERROUE´5, JAMES L. ETIENNE6 & ANDREA COZZI7 1

Department of Earth Science and Engineering, Imperial College London, South Kensington Campus, London SW7 2AZ, UK 2

Tullow Oil, 5th Floor, Block C, Central Park Leopardstown, Dublin 18, Ireland

3

Department of Earth Sciences, Oxford University, Parks Road, Oxford OX1 3PR, UK 4

TOTAL E&P Nederland, Bordewijklaan 18, Den Haag, The Netherlands

5

Beicip-Franlab, 232, Avenue Napole´on Bonaparte, 92502 Rueil-Malmaison, PO Box 213, France 6

Neftex Petroleum Consultants Ltd, 97 Milton Park, Abingdon, Oxfordshire OX14 4RY, UK 7

ENI Angola, Rua Nicola Gomes Spencer, 140, PO Box 1289, Luanda, Angola *Corresponding author (e-mail: [email protected])

Abstract: The Abu Mahara Group (c. 725–,645 Ma) of the Huqf Supergroup in the Jabal Akhdar of northern Oman hosts two glacial successions in the Ghubrah and Fiq formations, separated by the ,50-m-thick volcanogenic Saqlah Member. The .400-m-thick Ghubrah Formation is dominated by distal glaciogenic rainout diamictites, laminites and turbiditic siltstones, whereas the ,1.5-kmthick Fiq Formation exhibits a cyclical stratigraphy of proximal and distal marine glaciogenic facies, and non-glacial sediment gravity flow and shallow marine facies. The Fiq Formation is overlain by a transgressive, isotopically light carbonate known as the Hadash Formation. A tuffaceous ash interbedded with glacial diamictites of the Ghubrah Formation in Wadi Mistal has yielded a U –Pb zircon age of 713.7 + 0.5 Ma. The Fiq Formation contains detrital zircons as young as 645 Ma. The use of the CIA (Chemical Index of Alteration) shows the Fiq Formation to be climatically cyclic, with alternations of high and low chemical weathering of contemporary land surfaces driven by phases of glaciation and deglaciation. The transgression into the post-glacial Masirah Bay Formation is marked by a major increase in chemical weathering.

The Huqf Supergroup crops out in northern Oman (Jabal Akhdar), east-central Oman (Huqf area) and southern Oman (Mirbat area of Dhofar) (Fig. 20.1) (Allen et al. 2011) and is penetrated by many boreholes in the salt basins of the Oman interior. The Huqf Group recognized by Glennie (1977) in the Huqf area of east-central Oman was subdivided into the Abu Mahara, Khufai, Shuram and Buah Formations by Gorin et al. (1982), but no glaciogenic deposits are found in this region. In the Jabal Akhdar of northern Oman, Kapp & Llewellyn (1965) defined the Mistal, Hajir, Mu’aydin and Kharus Formations, which were correlated (Tschopp 1967) with the Huqf succession of east-central Oman. Kapp & Llewellyn (1965), Glennie et al. (1974), and geologists of the Bureau de Recherches Ge´ologiques et Minie`res (Beurrier et al. 1986; Rabu et al. 1986) recognized and mapped the Mistal Formation (syn. Mistal Conglomerate Formation) as the oldest exposed Neoproterozoic sedimentary rocks in the Jabal Akhdar. Rabu et al. (1986) divided the Mistal Formation into four members, in ascending order, the Ghubrah, Saqla, Fiq and Amq Members. In an attempt to streamline stratigraphic nomenclature across the surface outcrops and subsurface well penetrations of the salt basins, PDO geoscientists (PDO internal reports, Loosveld et al. 1996) redefined the Abu Mahara Group as comprising a lower diamictite-rich Ghadir Manqil Formation and an upper Masirah Bay Formation, and raised the status of the succession to the Huqf Supergroup, comprising Abu Mahara, Nafun and Ara Groups (Glennie et al. 1974; Gorin et al. 1982; Hughes-Clarke 1988; Wright et al. 1990). McCarron (2000) referred to the distinctive carbonate capping the diamictites of the Fiq Member as the Hadash Formation, based on the name of the village close to its type locality above Wadi Mistal in the Jabal Akhdar. When the basin evolution of the Huqf Supergroup was better understood, the term Abu Mahara Group was restricted to the stratigraphy below the transgressive Hadash Formation carbonate (McCarron 2000; Leather et al. 2002; Allen et al. 2004; Allen & Leather 2006), and the

Masirah Bay Formation assigned to the Nafun Group. The stratigraphic nomenclature currently in use divides the Abu Mahara Group into a lower Ghubrah Formation and an upper Fiq Formation, separated locally by a volcanogenic unit. This unit is transitional upwards into Fiq Formation, so is included in the Fiq as the Saqlah Member. Neoproterozoic glacial deposits are found in the Abu Mahara Group in the Jabal Akhdar of the Oman Mountains, occupying the core of a major east – west oriented anticline (Fig. 20.2). Partly coeval rocks are found in Dhofar (southern Oman) (Allen et al. 2011), but the Abu Mahara Group is absent from the Huqf area of east-central Oman. The Group has been penetrated by a limited number of boreholes in the salt basins of Oman, and has been imaged on seismic reflection profiles, which show it to be confined to large north – south oriented, c. 50-km-wide basins that have been interpreted as rifts (Loosveld et al. 1996). This rift interpretation is supported by sedimentological and stratigraphic evidence derived from field studies (Rabu 1988; Rabu et al. 1993). In the Jabal Akhdar, the sedimentology of the glaciogenic Fiq Formation indicates sediment derivation from two opposing margins (Leather et al. 2002; Allen et al. 2004). The intervening basin has the same scale and orientation as the structurally confined basins imaged on seismic profiles. The phase of sedimentation responsible for the Fiq Formation was preceded by a period of extrusive basaltic volcanism, represented by the Saqlah Member (Rabu 1988; Rabu et al. 1993; Le Guerroue´ et al. 2005). Although the Saqlah volcanic and volcaniclastic rocks and the Fiq glaciogenic and non-glacial sedimentary rocks can be attributed to a continental rift basin, the Ghubrah stratigraphy below the Saqlah Member is far less well constrained. A transgressive, 13C-depleted (down to –5‰) carbonate (Hadash Formation) oversteps basement-cored Abu Mahara basin margins, as seen in the Huqf area, where the carbonate

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 251– 262. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.20

26 N

ARABIAN GULF Makran Fold Belt

Al Noor Athel Fara

Birba

0

Dhahaban

ARA GROUP

IRAN Musandam

CAMBRIAN

Nomenclature

Buah

A

Fig. 2

EDIACARAN

Muscat

Jabal Akhdar

Saih Hatat

Oman Mountains

23O N

F7

Jebel Ja'alan

ka

F6

B

Western Rub Al-Khali Salt Basin

Huqf area

Masirah Island O

OMAN

Kh

a

Hi

gh

a Wa

dB

a

Ea

r ste

n

n Fla

k

ARABIAN SEA

Ba

sin

n-

h sfa

South Oman Salt Basin n si

CRYOGENIAN

20 N

Fiq/Shareef

2000

F1

1600 Saqlah ?

Marsham

ld

1200

ha Ain Sarit Al Hota

Mirbat

C

100 km

Mirbat

W ag

YEMEN

F5 F3

ABU MAHARA GROUP

Ma

Al Jobah

Jabal Akhdar

Ghaba Basin

rem

2800

Hu q

Hi gh

fa xis

Fahud Basin

SAUDI ARABIA

Gh

Khufai

4000

A5 A4 A3 A2 A1 A0

Hadash

Southern Gulf Salt Basin

u ud

Shuram

A6

Masirah Bay 3200

Qalhat

HUQF SUPERGROUP

Qatar Arch

UNITED ARAB EMIRATES

NAFUN GROUP

GULF OF OMAN

Oman composite stratigraphy

Ara cycles

(b)

Subsurface

(a)

Thickness (m)

P. A. ALLEN ET AL.

Jabal Akhdar and Huqf

252

17o N Arkahawl

Evaporites of Ara

Sandstones and conglomerates

Carbonate rocks

Shales and siltstones

Glacial diamictites

800

58o E

Extrusives in basin-fill/ crystalline basement/late plutons

Major transgressive surface

Ayn/Ghubrah

400

Basement

Mirbat and northern Huqf

55oE

52o E

Fig. 20.1. (a) Neoproterozoic outcrops (black) and salt basins (grey) of Oman. Main outcrop areas are Jabal Akhdar of northern Oman (A), Huqf area of east-central Oman (B) and Mirbat area of Dhofar (C), as well as the metamorphic window of Saih Hatat. (After Allen et al. 2004.) (b) Composite stratigraphy of the Huqf Supergroup in Oman. (After Allen 2007 and Rieu et al. 2007a.) Arrows indicate main transgressive surfaces, and F1 to F7 are Fiq units of Leather (2001).

rests via an unconformity lined with silcrete upon felsic ignimbrites and tuffs dated 802 Ma and granitic basement dated 822 and 825 Ma (Allen & Leather 2006; Allen 2007; Bowring et al. 2007). Over basement highs, therefore, the unconformity beneath the Nafun Group stretches from c. 800 Ma to the onset of deposition of the Nafun Group, whereas in the Abu Mahara basins, as in the Jabal Akhdar, the stratigraphic succession is more complete and conformable.

Structural framework The Huqf Supergroup is exposed in the core of a 700-km-long, 30–130-km-wide, east –west-trending antiform in the Jabal Akhdar

of the Oman Mountains, and in the intensely deformed and metamorphosed domal culmination of the Saih Hatat (Glennie et al. 1974; Le Me´tour et al. 1986; Rabu et al. 1993; Gray et al. 2005). The Saih Hatat structural deformation was generated during a top-to-the-NE shear, producing a major, refolded, NE-facing, recumbent anticlinal fold-nappe within pre-Ordovician sedimentary rocks, including lateral equivalents of the Huqf Supergroup. No detailed stratigraphic or sedimentological work has been carried out on account of this deformation and metamorphism, although the stromatolitic Khufai Formation of the Jabal Akhdar has been correlated lithologically with the Hiyam Formation of the Saih Hatat (Rabu et al. 1986). The Huqf Supergroup of the Jabal Akhdar is less intensely deformed than in the Saih Hatat and is exposed in a series of

THE ABU MAHARA GROUP

253

Fig. 20.2. Geological map of the Jabal Akhdar showing log localities, main towns and villages. Log localities: WS1 and WS2, Wadi Sahtan; WH1 and WH2, Wadi Hajir; WBK1, Wadi Bani Kharus; WM1, WM2, WM3 and WM4, Wadi Mistal; MU1 Wadi Mu’aydin; WBJ1, Wadi Bani Jabir. Arrows indicate main entrances to wadis. (Modified from Allen et al. 2004.)

erosional windows or ‘bowls’, chiefly (from west to east) the Sahtan, Kharus and Mistal bowls. Huqf Supergroup sedimentary rocks were tilted (block-faulting and large-amplitude folding most likely associated with the Hercynian orogeny, Rabu et al. 1986) prior to the deep erosion represented by the pre-Permian unconformity, and underwent further deformation during an ‘Alpine’ (c. 90 Ma) event associated with the emplacement of the Samail ophiolite and thrust sheets involving Mesozoic Tethyan carbonate-dominated rocks (Mann & Hanna 1990). Alpine deformation imparted a strong cleavage, particularly in the Ghubrah Formation, where schistosity produces a nearhorizontal pencil cleavage and a roughly north –south stretching lineation measured from the long axes of clasts in diamictite units. The gradient in intensity of deformation and metamorphism towards the east and NE of the Oman Mountains (Le Me´tour et al. 1986), reaching a peak in the Saih Hatat, indicates that the same tectonic events affected both the Jabal Akhdar and Saih Hatat regions. The fact that the penetrative cleavage that affects finegrained units of the Huqf Supergroup extends into the overlying Permian –Mesozoic succession implies that it is Alpine in age, with an increasing shear deformation downwards. The structural geology of the Jabal Akhdar illustrated in the work and map of Rabu et al. (1986), Rabu (1988) and Rabu et al. (1993) and the separation of the pre-Permian from Alpine deformation, awaits a further detailed analysis.

Stratigraphy The Jabal Akhdar region contains excellent exposures of rocks belonging to the Abu Mahara and Nafun Groups of the Huqf Supergroup (Fig. 20.1; see Allen 2007, fig. 4, p. 145, for a summary of stratigraphic nomenclature, and fig. 8, p. 149, for the composite stratigraphic column). Whereas the base of the Huqf Supergroup is not seen, the top is an erosional, angular unconformity beneath the carbonate rocks of the Permian Saiq, which cuts down variably. In some locations (e.g. Wadi Mistal), the Saiq cuts down deeply into the Abu Mahara Group, whereas in others a full succession of Nafun Group is preserved. In Wadi Bani Awf the Nafun Group passes up into the Fara Formation, a surface equivalent of part of the Ara Group, which is truncated by the base– Saiq unconformity. The Abu Mahara Group comprises a lower diamictite-rich Ghubrah Formation, separated from an upper diamictite-rich Fiq Formation by an intervening

laterally discontinuous volcaniclastic and volcanic (pillowed and vesicular basalts) Saqlah Member (Le Guerroue´ et al. 2005). The Fiq Formation is overlain by a well-developed carbonate named the Hadash Formation (McCarron 2000), which is the lowermost unit of the Nafun Group.

Glaciogenic deposits and associated strata Glaciogenic deposits are found in the Abu Mahara Group in the Jabal Akhdar at two different stratigraphic levels. The younger Cryogenian glaciogenic sedimentary rocks are found in the cyclical Fiq Formation (Leather et al. 2002; Allen et al. 2004). The older Cryogenian glaciogenic deposits are found in the much less well-understood Ghubrah Formation (Rabu et al. 1986; Brasier et al. 2000; Leather 2001; Le Guerroue´ et al. 2005).

The Ghubrah Formation Very little is known about the Ghubrah Formation. It was initially recognized by Kapp & Llewellyn (1965) as the lower part of the Mistal Conglomerate, and Rabu et al. (1986) referred to it as the Ghubrah Member of the Mistal Formation. The base is not seen. The Ghubrah Formation crops out in Wadi Mistal (.200 m thick) and Wadi Mu’aydin in the east, and in Wadi Sahtan (.400 m thick) in the west of the Jabal Akhdar. Rabu et al. (1986) contrasted it with the Fiq unit as comprising large thicknesses of poorly stratified diamictic conglomerate, compared with the better-bedded Fiq. Rabu et al. (1986), however, warned that locally, in particular in Wadi Mu’aydin, the sandstones of the Fiq interfinger with conglomerate similar to that of the Ghubrah, rendering the two virtually indistinguishable. Leather (2001) stated that the facies within the Ghubrah and Fiq Formations were ‘notably different’, with the Ghubrah Formation lacking bedding for much of its thickness. McCarron (2000), Brasier et al. (2000) and Leather (2001) all followed Rabu et al. (1986) in believing the Ghubrah to be an older stratigraphic unit separated from the Fiq Formation by the volcanogenic Saqlah Member. Le Guerroue´ et al. (2005) identified an angular relationship between the two formations, proposing an unconformity at the base of the Saqlah. However, the relationship of the Ghubrah to the Fiq, and the sedimentology and structural geology of the Ghubrah, have never been given the attention

254

P. A. ALLEN ET AL.

they deserve. The U –Pb zircon date reported by Brasier et al. (2000), and more accurate U –Pb dates on similar material reported in Bowring et al. (2007), need to be viewed in this context (see below). The facies of the Ghubrah Formation are described briefly by Leather (2001). The Formation comprises a distal glaciomarine facies association made of massive diamictites, massive and graded siltstones, dropstone laminites, very minor carbonate rocks and tuffaceous sandstones. The massive diamictites are up to hundreds of metres thick, with gradational upper and lower boundaries into massive and graded siltstones. The diamictites appear structureless, with some striated and faceted clasts. These massive diamictites may be thick amalgamated debris flows that reworked glacially transported clasts, but the association with dropstone laminites and the large lateral extent (.40 km) suggest a glaciomarine origin by fallout from suspension and rainout of ice-rafted debris, possibly in an outer shelf/slope setting (Anderson et al. 1984; Deynoux 1985; Eyles 1988; Eyles & Lagoe 1990; Moncrieff & Hambrey 1990; Brodzikowski & van Loon 1991). The associated siltstones indicate quiet marine sedimentation when ice rafting was reduced, either because of glacial retreat (Eyles & Lagoe 1990) or movement of iceberg transport lanes (Moncrieff & Hambrey 1990; Miller 1996). The dropstone laminites found in Wadi Mu’aydin indicate a clear ice-rafted component (Powell & Domack 1995). Carbonate

occurs as c. 10-cm-thick boudinaged beds in Wadi Mistal consisting of quartz grains and lithic clasts in a dolomite matrix of ,50 mm rhombs. The tuffaceous and ashy sediments that have provided the U – Pb date reported by Brasier et al. (2000) and Bowring et al. (2007) indicate nearby volcanism during deposition of the Ghubrah Formation.

Sedimentology of the Fiq Formation The Fiq Formation is well exposed in a number of wadis in the Jabal Akhdar of northern Oman, particularly in Wadi Sahtan, Wadi Hajir and Wadi Mistal (Figs 20.2 & 20.3). Although the stratigraphy, first described by PDO geologists in 1965, was mapped and interpreted as partly glaciomarine (Glennie et al. 1974; Beurrier et al. 1986; Rabu et al. 1986; Hughes Clark 1988), the first detailed sedimentological analysis is in the doctoral thesis of Leather (2001). Facies associations. The facies present in the Fiq Formation can be

assigned to four different facies associations: (i) distal glaciomarine, (ii) proximal glaciomarine, (iii) non-glacial sediment gravity flow and (iv) non-glacial shallow marine. The lithofacies associations used have been adapted in part from Eyles et al. (1985), Moncrieff & Hambrey (1990), Brodzikowski & van Loon (1991)

Fig. 20.3. Summary sedimentological logs through the Fiq Formation from 10 localities (see Fig. 20.2), hung from the level of the Hadash Formation. Diamictites are highlighted to illustrate their lateral continuity and vertical superimposition. F1 to F7 are stratigraphic units of Leather (2001), and T1 to T6 are transgressive surfaces. Diamictites in F5 and F7 are basin-wide deposits, whereas older diamictites, such as that in F1, are locally developed. (After Allen et al. 2004.)

THE ABU MAHARA GROUP

and Miller (1996). Lithofacies names have been partly adapted from Eyles et al. (1983), Moncrieff (1989) and Moncrieff & Hambrey (1990). Distal glaciomarine facies association. The distal glaciomarine facies association represents distal glaciomarine environments where ice rafting and non-glacial processes dominate, and is found in Wadi Sahtan, Wadi Bani Awf and Wadi Hajir. It is dominated by massive, sheet-like diamictites (,30 m thick) laterally and vertically transitional to deep-water mudstones and siltstones. Clasts include striated and faceted examples indicating subglacial transport (Wentworth 1936; Boulton 1978; Dowdeswell et al. 1985). An ice-rafted origin is inferred on the basis of the glacially transported clasts, common transitional lower boundaries and lack of internal stratification (Anderson et al. 1984; Eyles 1988; Kellerhals & Matter 2003). The clast-poor nature of many of the diamictites, their sheet-like geometry and their association with deep-water deposits suggests that they formed as a result of a combination of settling of suspended sediment and ice rafting, possibly in an outer shelf/slope setting relatively distant from active ice margins (Deynoux 1985; Eyles & Lagoe 1990; Moncrieff & Hambrey 1990; Brodzikowski & van Loon 1991). Associated dropstone laminites (,50 m thick), found in Wadi Hajir, comprise grey shales and siltstones with sharp-based turbiditic interbeds and rare outsized clasts, indicating a distal shelf to slope setting affected by ice rafting, suspended sediment fallout and dilute turbiditic underflows. Underflows may have been released by subglacial river output (Eyles 1988; Moncrieff & Hambrey 1990) or brine-rich currents, as occurs on the Barents Shelf today (Elverhoi et al. 1989). Proximal glaciomarine facies association. The proximal glaciomar-

ine facies association formed close to the grounded ice margin (Moncrieff & Hambrey 1990; Miller 1996) where rainout of coarse ice-rafted debris and sediment gravity flows interact. Sediment gravity flows are also found in distal glaciomarine and nonglacial environments. In summary, the proximal glaciomarine facies association comprises (i) massive diamictites (from a few metres thick to 100 m thick) containing striated, faceted and ‘flat-iron’-shaped clasts, indicating subglacial transport prior to ice rafting; (ii) subordinate but widespread stratified diamictites, formed by a combination of winnowing of ice-rafted material (Eyles et al. 1985; Moncrieff & Hambrey 1990), or re-deposition of massive diamictites by sediment gravity flows in ice-proximal settings; (iii) clast-supported, laterally discontinuous conglomerates, representing local concentrations of clasts derived from iceberg overturn (Ovenshine 1970), lag deposits indicating intensive current, wave or tide reworking of ice-rafted diamictite deposits in relatively shallow marine conditions (Eyles et al. 1985; Eyles 1988; Moncrieff & Hambrey 1990), and debris flows of remobilized ice-rafted material where the fine sediment has been expelled into an overlying turbulent suspension (Wright & Anderson 1982); (iv) massive or graded, rippled, sharp-based sandstones and siltstones, with lonestones/dropstones at bed boundaries, are turbidite deposits formed in a glaciomarine stetting extending some distance from the grounding line; (v) minor dropstone laminites, due to minor ice rafting in a quiet distal glaciomarine environment; (vi) massive and laminated mudstones and siltstones lacking dropstones, derived from dilute turbidity currents or overflow plumes (Mackiewicz et al. 1984; Powell & Molnia 1989). Taken together, the facies association suggests an ice-proximal environment subject to outflows from subglacial streams, debris flows, sandy turbidites, fallout from suspension from turbid plumes, winnowing and reworking, and ice rafting. Non-glacial sediment gravity flow facies association. The non-glacial sediment gravity flow facies association, which may form units ,250 m thick, comprises (i) erosive-based, commonly lenticular conglomerates and pebbly sandstones, with rounded clasts and

255

large intraformational rafts, representing debris flows or hyperconcentrated flows (Nardin et al. 1979; Costa 1988); (ii) common sharp-based, massive and graded, sheet-like, commonly amalgamated sandstones, locally with rippled tops and fluted bases, which are turbidites; (iii) current-rippled sandstones and siltstones with ubiquitous water-escape structures representing deposition from more dilute turbidity currents, with local wave-generated ripples at the top of shallowing-up cycles; (iv) massive and laminated mudstones and convoluted siltstones, which settled slowly from turbiditic plumes/tails or river plumes/underflows (Stow et al. 1996); (v) thin, brown carbonate on a transgressive surface overlying glacial diamictites (Wadi Sahtan, log WS1, unit F6) suggesting carbonate precipitation at a time of deglacial sealevel rise. Non-glacial shallow marine facies association. The non-glacial shallow marine facies association forms a minor part of the nonglacial sediments of the Fiq Formation. It comprises (i) laterally persistent conglomerates composed of well-rounded pebbles and cobbles, representing transgressive lag (ravinement) deposits winnowed of their finer material during flooding; (ii) extensively wave-rippled sandstones, particularly common in Wadi Sahtan, probably representing shoreface deposition; (iii) rippled wavegenerated heterolithics, passing up into sharp-based, massive sandstones probably representing flows generated by storms.

Depositional history of the Fiq Formation The Fiq Formation is spatially and temporally highly variable (Fig. 20.3). This variability results from the fundamental bathymetric effects produced by the rifted crustal structure. However, distinct genetic units reflecting relative sea-level change can be traced across the entire Jabal Akhdar area. Glacially influenced sedimentary rocks occur at a number of different levels in the Fiq Formation, commonly followed abruptly by a facies change indicating deepening. The facies evolution through time and the relative sea level recorded by these facies can therefore be evaluated (Leather et al. 2002; Allen et al. 2004; Allen & Etienne 2008). Unit F1. Unit F1 is a transitional unit between the underlying vol-

canics of the Saqlah Member and the Fiq Formation. Volcaniclastic detritus was derived from a source to the east, probably in the Saih Hatat area (Le Me´tour et al. 1986; Villey et al. 1986), and was deposited in an already-deep marine basin. The occurrence of diamictites (F1b in Wadi Mistal) containing volcanic clasts indicates glaciation in F1, though its effects appear to be restricted to the east of the outcrop area. Units F2 and F3. In the west of the Jabal Akhdar, quiet, deep-water

conditions became established in F2 times, as shown by increasing interbedding of turbidites derived from the west, followed by shallowing into a wave-rippled shoreface. The overlying F3 proximal diamictite indicates a resumption of glaciation. In the east of Jabal Akhdar, the F1b diamictite was abruptly flooded and succeeded by gravity-flow deposits derived from non-volcanic sources in the east. An overlying diamictite correlates well with F3 in the west of Jabal Akhdar, indicating that the entire outcrop area was influenced by glacial processes at this time. Unit F4. Evidence for an important flooding event following deposition of the F3 diamictite is found across the whole Jabal Akhdar outcrop area, and there is no record of any glacial processes in the basin. Subsequently, sediment gravity flows derived from the east filled the eastern sector of the basin, whereas shallow shoreface conditions existed in the west, indicating asymmetry in subsidence and/or sediment supply. In both the east and west, sediment gravity flows from multiple sources then resumed and dominated sedimentation across the entire basin. A shutdown in sediment

256

P. A. ALLEN ET AL.

supply in the west resulted in the deposition of sediments during transgressive phase T3.

Chemostratigraphy Carbon-isotopic data

Unit F5. Renewed glacial activity across the whole outcrop area

is recorded by proximal diamictites and sediment gravity flows in the east and distal glaciomarine deposition in the west. Unit F6. The end of F5 glaciation is indicated by a major flooding event in the western and central parts of the outcrop area, whereas sediment gravity flows initially dominated the eastern sector of the basin. In the western and central parts of the basin, deposition of unit F6 continued as turbidites derived primarily from western sources, and shallow-water wave-rippled sandstones were deposited in the east. A locally developed diamictite (F6a) indicates a weak glacial advance before significant basin-wide flooding (T5 transgression). The overlying lag conglomerate and thin carbonate rocks and mudstones indicate transgression and establishment of deep-marine conditions before a gradual shallowing took effect towards the top of F6. Unit F7. Coarse-grained gravity-flow deposits are overlain by

proximal glaciomarine diamictites and debris flows and, locally, by subglacial stream outwash, representing the last of the glacial phases in the Fiq Formation. The overlying Hadash Formation is a classic, transgressive postglacial carbonate. This T6 transgression covered the rift margins and flooded other basement highs in the Oman region.

The Hadash Carbonate The ,1.5-km-thick Fiq Formation is overlain by a transgressive, 13 C-depleted dolostone (Hadash Formation) that deepens up into the marine shales and siltstones of the Masirah Bay Formation (Allen & Leather 2006) (Fig. 20.4). The Hadash Formation is dominated by dolostones, and despite being ,15 m thick is laterally extensive over at least 80 km throughout the Jabal Akhdar. Subsurface well penetrations (PDO internal reports) suggest that the Hadash Formation is found throughout the Oman area. Surface outcrop studies also demonstrate that a transgressive carbonate unit overlies basement crystalline or volcanic rocks in the Huqf region (Loosveld et al. 1996; McCarron 2000; Leather 2001; Allen & Leather 2006; Allen 2007). Directly overlying glacial diamictites of the Fiq Formation and with a negative C-isotopic signature, the 8-m-thick carbonate-dominated part of the Hadash Formation shares much in common with cap dolostones worldwide (Williams 1979; Fairchild & Hambrey 1984; Tucker 1986; Narbonne et al. 1994; Kennedy 1996; Hoffman et al. 1998a; Myrow & Kaufman 1999; Hoffman & Schrag 2002; Halverson et al. 2004). The facies of the Hadash Formation are described in Allen et al. (2004). Although the broad depositional environment of the cap carbonate in the Jabal Akhdar is deep water, the high proportion of siliciclastics and palaeocurrent directions in Wadi Bani Jabir, the orientation of roll-up structures (Simonson & Carney 1999) at Hadash and thickness trends between the logged sections collectively suggest that the depositional environment of the carbonate deepened to the centre-west. The cap carbonate therefore partially ‘shales out’ into the deepestwater lithofacies deposited in the centre of the Fiq basin. The most important transition in the basin is between the carbonatedominated eastern Jabal Akhdar area (Hadash 1, Hadash 2 and Wadi Mu’aydin) and the western Jabal Akhdar area, which exhibits a higher concentration of fine-grained siliciclastics (Wadi Hajir 1, Wadi Hajir 2, Wadi Bani Awf, Wadi Sahtan). This was also an important transition during the deposition of the Fiq Member, with thicker, more shale-prone sections evident in the west. It is therefore likely that the basement structure controlled water depth and accommodation across this line.

In total, 271 samples were analysed from seven stratigraphic profiles through the Hadash Formation in the Jabal Akhdar (Fig. 20.4). ‘Least altered’ samples were selected for being well away from areas of intense veining and for having good microscopic fabric preservation revealed from petrographic and cathodoluminescence data. In total, 81 samples from five stratigraphic sections were analysed for major and minor elements (details in Leather 2001). Elemental and isotopic data were used to screen for the effects of diagenesis and metamorphism, following techniques described by Brand & Veizer (1980, 1981), Marshall (1992) and Kaufman & Knoll (1995). Further details can be found in Leather (2001) and Allen et al. (2004). The vertical pattern of C-isotope values (Hadash 1 type section) shows an initial shift to – 4‰ to –8‰ in the basal metre, followed by a gradual drift back to –1‰ at the top of the carbonate, although with considerable scatter. This trend is in common with other Neoproterozoic cap dolostones (Kennedy 1996; Hoffman et al. 1998b; Frimmel et al. 2002; Fo¨lling & Frimmel 2002). However, correlation of chemostratigraphic facies from east to west across the Jabal Akhdar is problematical. The variation in d13C values (c. 2‰) between closely spaced stations is as great as the amplitude of the isotopic excursions used for correlation. Leather (2001) therefore concluded that high-resolution intrabasinal correlation using d13C curves is highly uncertain. Further C-isotopic data from the Nafun Group are found in Burns & Matter (1993), Cozzi & Al Siyabi (2004), Fike et al. (2006), Le Guerroue´ et al. (2006a, b, c) and Le Guerroue´ & Cozzi (2010).

Chemical Index of Alteration The use of the Chemical Index of Alteration (CIA) is described in a separate paper in this volume (Bahlburg & Dobrzinski 2011, Chapter 6), and details of the technique are not repeated here. Throughout the Fiq and Masirah Bay formations there are significant compositional and mineralogical variations. When plotted in Al2O3 – CaO þ N2O – K2O and quartz-plagioclase-K-feldspar space, the data define trends roughly parallel to the Al2O3 – CaO þ N2O and plagioclase-quartz boundaries, suggesting variability in the extent of chemical weathering of the sediment in the source area (Fedo et al. 1995; Nesbitt et al. 1996). The possible influence of changes in grain size, hydrodynamic sorting, provenance and diagenetic alteration on the composition of the sediments is discussed in Rieu et al. (2007b). Both uncorrected CIA values and those corrected for a maximum amount of K-metasomatism are plotted as a function of their stratigraphic height (Fig. 20.5). In both cases, a similar first-order trend is revealed, comprising three intervals during which chemical weathering was reduced as indicated by relatively low CIA and Mineralogical Index of Alteration (MIA) values. Reduced chemical weathering in these intervals is in agreement with the presence of distinctive sedimentary facies (diamictites, dropstone-bearing laminites) that suggest a cold climate. These intervals alternate with units characterized by relatively high CIA and MIA values and that lack evidence for any glacial influence during sedimentation, which are interpreted to represent interglacial periods. Importantly, the end of the entire glacial epoch corresponds to a major increase in CIA and MIA values in the lowermost Masirah Bay Formation. These values are the highest found in the succession (CIA . 80; MIA, 100). A discussion of these trends, and a justification of their interpretation as due to temporal changes in the chemical weathering of contemporary land surfaces, are given by Rieu et al. (2007b, c).

N

10 km

WBJ1

Wadi Hedak Nakhl

S

Wadi Sahtan

Wadi Bani Jabir

Legend

Wadi Mistal Wadi Bani Awf

Fashah Amq

Thin dolomite beds in fine-grained siliciclastic rocks (Facies W3)

WS1

Wadi Wadi Bani Ghubrah Hajir Kharus WM2

Sahtan Bowl WS2

WBA2

WBA1

Al Awabi

WM3 WH2

Sand-grade siliciclastics (Facies E5)

WM1

Ghubrah Bowl

Hajir WBK1

Crystalline limestone and dolospar (Facies E4)

WH1 WM4 Hadash

Permian to Cretaceous cover Huqf Supergroup

Dolomite-dominated deposits of Facies W2, E1, E2 and E3

Wadi Hajir 2

Thinly-interbedded dolomite and siltstone (Facies W1)

Al Ayn

Quaternary outside Jabal Akhdar

Massive glacial diamictites (Fiq unit 7)

MU1

Log locality and section line Entrance to wadi

Wadi Mu'aydin

Village or Town

Wadi Sahtan

Dolomite

Hadash 2 (Wadi Mistal)

Wadi Bani Awf

10

Calcite

Hadash 1 (Wadi Mistal)

W3

THE ABU MAHARA GROUP

Height in m above base of Hadash Formation

11

Carbon isotope data:

Wadi Bani Jabir

9 Wadi Hajir 1 8

W3 W3

7

E4

Wadi Mu'aydin

W3

6

E5

E4 5

W2

E4 E3

4 3 2

E3

W2 W1

E2

W2

W2 W1

0

F7

E2

E1

E1

W1 E1

1

E3 E2

-6

F7

-2 -4 d13C

0

F7

E1

E1

W1 -4

-2

0

d13C

Western Jabal Akhdar

F7

-4

d13C

0

-6

F7

-4

-2

d13C

0

0 -4

-2

F7

d13C

0

-8

-4

F7

d13C

0

-3

F7

-2

-1

0

d13C

Eastern Jabal Akhdar

Fig. 20.4. Summary logs of Hadash Formation from eight localities across the Jabal Akhdar region (see Fig. 20.2), with C-isotopic data shown for seven. Facies labels described in Leather (2001) and Allen et al. (2004). Note the variable horizontal scale for d13C, which demonstrates the significant variation in C-isotopic ratios between sections. Isotopic and geochemical data are in the supplementary material of Allen et al. (2004).

257

258

P. A. ALLEN ET AL.

Fig. 20.5. Variations in chemical and mineralogical indices of alteration (CIA, Al2O3/(Al2O3 þ CaO þ Na2O þ K2O); MIA, Qtz/(Qtz þ Kfs þ Pl)) with stratigraphic height for section in Wadi Sahtan (a) and critical section across the glacial –post-glacial transition at Hadash, Wadi Mistal (b). Numbering of diamictites and flooding surfaces as in Figure 20.3. Errors in CIA due to uncertainties in major element concentrations are ,1.5% (,1 CIA unit). 2s error bars are indicated for MIA values. MB Fm., Masirah Bay Formation. CIA values are corrected to show the maximum possible effect of potassium metasomatism during burial diagenesis, as explained by Rieu et al. (2007c, p. 301). (After Rieu et al. 2007c.)

Boundary relations with overlying and underlying non-glacial units The non-sedimentary basement beneath the Huqf Supergroup is poorly exposed in Oman. No basement rocks are found in the Jabal Akhdar. Crystalline basement is exposed in south Oman in the Mirbat region (Platel et al. 1992a) and on the Al Hallaniyah Islands (Platel et al. 1992b), and in small outcrops in the north of the Huqf area of east-central Oman (at Al Jobah, Allen & Leather 2006; Allen 2007; Bowring et al. 2007) and near Sur in the Jabal Ja’alan (Roger et al. 1992) further to the north. Most knowledge on the pre-Huqf Supergroup geological history of Oman is derived from the Mirbat area, recently summarized by Mercolli et al. (2006) and Allen (2007) (see Allen et al. 2011). The rocks overlying the Abu Mahara Group, termed the Nafun Group, have been correlated confidently between the Jabal Akhdar, the Huqf area of east-central Oman and the subsurface of the Oman salt basins (Cozzi & Al-Siyabi 2004; Allen & Leather 2006; Allen 2007; Le Guerroue´ & Cozzi 2010), in terms of major lithological trends and carbon isotopic profiles. The base of the overlying Nafun Group is marked by a major stratigraphic overstep due to transgression of basin margins, but the age of the Hadash Formation and the bulk of the Nafun Group is not well known. There is no evidence of glaciogenic sedimentary rocks in the Nafun Group, although it contains a very large negative C-isotopic excursion known as the Shuram anomaly (Fike et al. 2006; Le Guerroue´ et al. 2006b; Bristow & Kennedy 2008; Le Guerroue´ & Cozzi 2010).

Palaeolatitude and palaeogeography Kilner et al. (2005) sampled from the Jabal Akhdar, Huqf and Dhofar regions. Six sites yielding 20 samples were taken from the Fiq Formation in the Jabal Akhdar, and the Hadash Formation was sampled at 21 sites giving 60 samples. Kilner et al. (2005) found a component of magnetization (Component B) that passed the fold test and was therefore imparted prior to tectonic deformation in the Cambro-Ordovician (Blendinger et al. 1990; Gass et al. 1990). After correction for tectonic tilt, the declination and inclination for a sample close to the top of the Fiq Formation in Wadi Sahtan (inclination of –218) and for a sample 300 m lower

in the stratigraphy in Wadi Mistal (inclination of þ498) indicate a magnetic reversal during Fiq times. For palaeolatitude determination, a set of samples from the Hadash Formation provide more data (12 sites), with tilt-corrected inclinations between –7.68 and –39.38 for reversed polarities, and between þ218 and þ298 for normal polarities (Kilner et al. 2005, Data Repository Table 1). The average palaeopole and palaeolatitude based on 25 sites throughout Oman quoted by Kilner et al. (2005) is unfortunately a statistical average of samples of ages that are spread from c. 800 Ma to c. 640 Ma (Allen 2007; Allen et al. 2011). Eliminating all but the Fiq and Hadash samples, it is evident that the variation in the B component is large, and the average palaeolatitude of 138 for the Muscat region is therefore open to doubt.

Geochronological constraints Geochronological data (Brasier et al. 2000; Amthor et al. 2003; Allen & Leather 2006; Rieu 2006; Bowring et al. 2007; Rieu et al. 2007a) were integrated to obtain the best available constraints on the ages of stratigraphic units within the Huqf Supergroup by Allen (2007) (Fig. 20.6). Brasier et al. (2000) provided a U –Pb date on zircons recovered from a tuffaceous ash interbedded with thin (centimetre-scale) sandstones and thick, cleaved diamictites from the Ghubrah Formation in Wadi Mistal. The sample yielded an intercept age of 723(þ16/– 10) Ma. Resampling and analysis (Bowring et al. 2007) revealed a zircon population of uniform colour and size, and good crystal forms lacking any obvious evidence of reworking. Furthermore, of the 25 crystals analysed, 22 defined a discordia with an upper intercept age of 713.7 + 0.5 Ma, whereas the final three grains are variably discordant, perhaps due to a small xenocrystic component. The lack of an admixture of older zircons, compared to their abundance in nearby detrital beds (sandstone clast and diamictite matrix, see below), suggests that 714 Ma is a good estimate of the depositional age of the tuffaceous ash. This is a rare and valuable constraint on the age of a Neoproterozoic glaciation because the tuffaceous ash is interbedded with glacial diamictites. A number of detrital zircons have been analysed from the same Ghubrah Formation locality in Wadi Mistal (Bowring et al. 2007). The sandstone clast yielded five grains in the age range

THE ABU MAHARA GROUP

259

ABU MAHARA GROUP, JABAL AKHDAR Neoproterozoic

Mesoproterozoic

Palaeoproterozoic

c.664 (1 grain) 675

810-860 (10 grains c.920 (1 grain) 645 715-718 840-860 790-810 750 700-720

Fiq n=307 (167)

1050 750-826 (detrital grains in ash) 800-830 (4 detrital grains)

c.750 (3 grains)

Saqlah n=95 (28)

860 910

810 735 755 (matrix) 713.7±0.5 (n=22) 710

Ghubrah n=79 (18) 812-826 (5 detrital grains)

810

0.5

1.0 Bowring et al. (2007) data (Lahan core) Detrital grains (diamictites, turbidites) Detrital grains (tuffs, ashes) Syn-magmatic (Ghubrah date)

1.5 Age (Ga)

812– 826 Ma, and a sample of sandy diamictite matrix yielded three grains dated as 755 Ma. Rieu et al. (2007a) sampled a graded siltstone situated 1 m above the thin tuffaceous ash yielding the 713.7 + 0.5 Ma date. Of the 79 detrital zircon ages (derived by LA-ICPMS) compiled (selected as ,25% discordant and 2s errors ,10% for 235 U – 207Pb and 238U – 206Pb), 18 passed a stricter filter (,10% discordant and 2s errors ,20% for 235U – 207Pb and ,10% for 238 U – 206Pb). These most concordant data yield a youngest subpopulation with a peak at 710 Ma, which is in excellent agreement with the age of the underlying tuffaceous ash. An older subpopulation, which makes up 10% of all analysed zircons, ranges from 750 to 910 Ma, but only two grains pass the stricter filter, with ages at 810 Ma. It is clearly hazardous to infer the detrital geochronology of the entire Ghubrah Formation from samples collected from several lithologies at just one locality, and further sampling and analysis is required. With the data currently available, we infer an exclusively Neoproterozoic juvenile source for the Ghubrah zircons, with a syndepositional crystallization age of 714 Ma and a reworked contribution from older sources aged 755– 826 Ma. The c. 755 Ma crystals may represent magmatism associated with the beginning of continental extension responsible for the Ghubrah basins, or derivation from c. 760 Ma intrusives associated with the welding together of arc and gneiss terranes in the Arabian – Nubian Shield to the west. The older zircons (c. 810–825 Ma) have identical ages to basement rocks exposed at Al Jobah (Allen & Leather 2006; Allen 2007) at the northern end of the Huqf region and to the ages of peak metamorphism and magmatism in the gneisses of Mirbat (Mercolli et al. 2006; Allen et al. 2011). A number of attempts have been made to date the volcanic and volcaniclastic Saqlah unit at the base of the Fiq Formation. Bowring et al. (2007) analysed a sample from the Saqlah in Wadi Mistal, recovering seven zircons, three of which yielded an age of c. 750 Ma and four of which yielded ages in the range c. 800 –830 Ma. Further samples of ashy material yielded dates between 750 and 826 Ma, indicating reworked zircons from sources indistinguishable from those of the Ghubrah. The most

2.0

2.5

Fig. 20.6. Zircon geochronology, Abu Mahara Group, Jabal Akhdar (after Allen 2007, supplemented by Bowring et al. 2007). Histograms of zircon ages derived by LA-ICPMS, reported in Rieu (2006) and Rieu et al. (2007a). Grey shaded area shows zircons passing first filter, solid line is zircons passing stricter second filter (see text). Sample number n represents zircons contributing to the histogram; bracketed term is the number passing the first filter but failing the stricter filter. Stars indicate detrital zircon data in Bowring et al. (2007). Solid bars beneath the time scale show the main age ranges of detrital zircons. Note the progressively older average age of the detrital zircon distribution in the younger stratigraphy and the appearance of Mesoand Palaeoproterozoic sources in the Fiq Formation.

concordant of Rieu et al.’s (2007a) dataset (28 dates) give a main peak at 860 Ma, and subpopulation peaks at 730, 815 and 910 Ma. The youngest zircons may have been derived from igneous rocks associated with the accretion of island arc and gneissic terranes in the Arabian –Nubian megasuture (c. 760 Ma), late post-tectonic plutons such as those of the Mirbat area (c. 700– 750 Ma), or the start of continental extension at c. 720 Ma. The older zircons reveal the same sources of plutons formed during a major period of subduction-related crustal growth at .800 Ma. Although disappointing in not providing a syndepositional age for the Saqlah, the results confirm older syn-Ghubrah and juvenile Neoproterozoic basement sources. Bowring et al. (2007) analysed a turbiditic sandstone from the Fiq Formation collected from the northern part of Wadi Mistal. Of 12 single grains analysed, one grain was c. 920 Ma, ten grains were in the range 810– 860 Ma and one grain yielded an age of c. 664 Ma. A further sample of an ashy sediment interbedded with dropstone-bearing laminites and diamictites from core recovered from the Lahan-1 well, thought to be equivalent to a position very near the top of the Fiq, yielded zircons with ages in the range c. 645– 718 Ma, with significant scatter. Only five young grains grouped around 645 Ma were present. Consequently, it is safe to merely conclude that the top of the Fiq Formation must be younger than 645 Ma. Rieu et al.’s (2007a) larger data set of Fiq samples from Wadi Sahtan and Wadi Mistal (307 analyses, 167 of which passed the stricter filter) reveals important Mesoproterozoic (1%), Palaeoproterozoic (5%) and even Archaean (,1%) components. Main peaks of the most concordant data are at c. 790 –810 Ma and c. 840– 860 Ma, with secondary peaks or shoulders at c. 700– 720 Ma and c. 750 Ma. Combining all data, we conclude that the depositional age of the glaciogenic Fiq Formation must be younger than 645 Ma, but that the formation mostly contains detrital zircons derived from the same Neoproterozoic sources as seen in older parts of the Abu Mahara Group, with two additional components: (i) first cycle derivation of very old sources, suggesting progressively deep exhumation/unroofing of accreted terranes in the catchment

260

P. A. ALLEN ET AL.

areas of the Fiq basins, or inheritance of older zircons derived from Neoproterozoic terranes (Hargrove et al. 2006); (ii) minor magmatic sources (.645 Ma) that are most likely contemporaneous with Fiq basin development.

Discussion and uncertainties The Fiq Formation has been used as an example of repeated glacial advance and retreat before post-glacial transgression, signifying dynamic glaciation and vigorously acting sediment routing systems throughout the glacial epoch (Allen & Etienne 2008). Glacially influenced strata, principally proximal and distal rainout diamictites bearing dropstones, lonestones and faceted and striated clasts, are embedded in non-glacial stratigraphy dominated by hemipelagics, sandstone turbidites and debrites, but including wave-rippled sandstones requiring an ice-free sea surface. Further details of the sedimentology of the Fiq Formation and its use as a test for the Snowball Earth hypothesis can be found in Leather (2001), Leather et al. (2002), Allen et al. (2004) and Allen & Etienne (2008). Although the glacial intervals in the Sultanate of Oman have received some prominence in the literature, there remain areas of uncertainty that would benefit from further work. Chief among these are the following: † The age of the Fiq Formation is only constrained to terminate at younger than 645 Ma based on the presence of a small number of detrital zircons in core. A more systematic study of detrital zircon ages in the Fiq Formation is warranted. † The relationship between the Fiq and Ghubrah formations is poorly known and requires structural mapping in the key wadis, especially Wadi Mistal. The absence of a major stratigraphic gap would imply glaciation took place intermittently throughout the period from c. 720 to 640 Ma. The presence of a significant stratigraphic gap and/or angular unconformity would support the idea of discrete glacial epochs in the older and younger Cryogenian. † The Ghubrah Formation is almost entirely unknown sedimentologically. Notwithstanding its commonly highly cleaved nature, there is scope for a sedimentological study of the Formation across the Jabal Akhdar region, including Wadi Mu’aydin.

Conclusions The .400-m-thick Ghubrah Formation is dominated by thick, mostly unstratified marine rainout diamictites, with an interbedded tuffaceous ash that has yielded a U –Pb zircon age of 713.7 + 0.5 Ma. The overlying ,1.5-km-thick Fiq Formation commences locally with the volcanogenic Saqlah unit of submarine pillow basalts, ashes and volcaniclastic rocks, which passes up into the well-bedded cycles of the Fiq Formation, comprising a wide range of glaciogenic and non-glacial sediment gravity flow and shallow marine facies. Detrital zircons from ashy beds in the upper part of the Fiq Formation are as young as 645 Ma. The exact relationship between the Ghubrah Formation and the Fiq Formation is not fully resolved and requires further study. Calculation of CIA from mudstones and the muddy matrix of diamictites reveals large excursions attributable to secular variations in the intensity of chemical weathering on contemporary land surfaces. Such variations were most likely driven by pulses of glaciation separated by interglacial periods. The deposition of the transgressive cap carbonate coincides with a strong increase in CIA, indicative of climatic transit following glaciation. Carbon-isotopic ratios in the Hadash carbonate range from –1 to –8‰, but variability between sample stations is large, making intrabasinal correlation problematic and extrabasinal correlation hazardous.

We are grateful for the financial and logistical support of Petroleum Development Oman over many years and specifically the help of H. al Siyabi, M. Newall, J. Amthor and J. Schreurs. We are also grateful for the additional financial support of Enterprise Ireland and the Research Committee of ETHZu¨rich. A large number of individuals and field parties have provided input on the Abu Mahara Group of the Jabal Akhdar for which we are grateful: B. Levell, A. Heward, P. Hoffman, G. Halverson, A. Matter, and the participants of the field trip associated with the 24th IAS Meeting of Sedimentology in Muscat, January 2005. We are grateful for the reviews of N. Eyles and I. Fairchild, and the editing of E. Arnaud. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Allen, P. A. 2007. The Huqf Supergroup of Oman: Basin development and context for Neoproterozoic glaciation. Earth Science Reviews, 84, 139– 185, doi: 10.1016/j.earscirev.2007.06.005. Allen, P. A. & Leather, J. 2006. Siliciclastic marine sedimentation in the aftermath of a Marinoan glacial epoch: The Masirah Bay Formation, Huqf Supergroup, of Oman. Precambrian Research, 144, 167– 198, doi: 10.1016/j.precamres.2005.10.006. Allen, P. A. & Etienne, J. L. 2008. Sedimentary challenge to snowball Earth. Nature Geoscience, 1, doi: 10.1038/ngeo355. Allen, P. A., Leather, J. & Brasier, M. D. 2004. The Neoproterozoic Fiq glaciation and its aftermath, Huqf Supergroup of Oman. Basin Research, 16, 507– 534, doi: 10.1111/j.1365-2117.2004.00249.x. Allen, P. A., Rieu, R., Etienne, J. L., Matter, A. & Cozzi, A. 2011. The Ayn Formation of the Mirbat Group, Dhofar, Oman. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 239–249. Amthor, J. E., Grotzinger, J. P., Schro¨der, S., Bowring, S. A., Ramezani, J., Martin, M. W. & Matter, A. 2003. Extinction of Cloudina and Namacalathus at the Precambrian – Cambrian boundary in Oman. Geology, 31, 431–434. Anderson, J. B., Brake, C. F. & Myers, N. C. 1984. Sedimentation on the Ross Sea continental shelf, Antarctic. Marine Geology, 57, 295– 334. Bahlburg, H. & Dobrzinski, N. 2011. A review of the Chemical Index of Alteration (CIA) and its application to the study of Neoproterozoic glacial deposits and climate transitions. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 81 – 92. Beurrier, M., Bechennec, F., Rabu, R. & Hutin, G. 1986. Geological Map of Rustaq, Sheet NF 40-3D, scale 1:100,000. Directorate General of Minerals, Oman Ministry of Petroleum and Minerals. Boulton, G. S. 1978. Boulder shapes and grain-size distribution of debris as indicators of transport paths through a glacier and till genesis. Sedimentology, 25, 773– 799. Blendinger, W., van Vliet, A. & Hughes Clarke, M. W. 1990. Updoming, rifting and continental margin development during the Late Palaeozoic in northern Oman. In: Robertson, A. H. F., Searle, M. P. & Ries, A. C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 27 – 37. Bowring, S. A., Grotzinger, J. P., Condon, D. J., Ramezani, J., Newall, M. J. & Allen, P. A. 2007. Geochronological constraints on the chronostratigraphic framework of the Neoproterozoic Huqf Supergroup, Sultanate of Oman. American Journal of Science, 307, 1097– 1145, doi: 10.2475/10.2007.01. Brand, U. & Veizer, J. 1980. Chemical diagenesis of a multicomponent carbonate system 1: Trace elements. Journal of Sedimentary Petrology, 50, 1219– 1236. Brand, U. & Veizer, J. 1981. Chemical diagenesis of a multicomponent carbonate system 2: Stable isotopes. Journal of Sedimentary Petrology, 51, 987– 997. Brasier, M. D., McCarron, G., Tucker, R., Leather, J., Allen, P. A. & Shields, G. 2000. New U– Pb zircon dates for the Neoproterozoic Ghubrah glaciation and for the age of the top of the Huqf Supergroup,

THE ABU MAHARA GROUP

Oman. Geology, 28, 175–178, doi: 10.1130/0091-7613(2000)028 ,0175:NUPZDF.2.2.CO;2. Bristow, T. F. & Kennedy, M. J. 2008. Carbon isotope excursions and the oxidant budget of the Ediacaran atmosphere and ocean. Geology, 36, 863–866, doi: 10.1130/G24968A.1. Brodzikowski, K. & van Loon, A. J. 1991. Glaciogenic Sediments. Developments in Sedimentology, 49, Elsevier, Amsterdam, 674. Burns, S. J. & Matter, A. 1993. Carbon isotopic record of the latest Proterozoic from Oman. Eclogae Geologicae Helvetiae, 86, 595– 607. Costa, J. E. 1988. Rheologic, geomorphic, and sedimentologic differentiation of water floods, hyperconcentrated flows, and debris flows. In: Baker, V. R., Kochel, R. C. & Patton, P. C. (eds) Flood Geomorphology. Wiley, New York, 113–122. Cozzi, A. & Al Siyabi, H. 2004. Sedimentology and play potential of the late Neoproterozoic Buah carbonates of Oman. GeoArabia, 9, 11 – 36. Deynoux, M. 1985. Terrestrial or waterlain diamictites? Three case studies from the Late Precambrian and Late Ordovician glacial drifts in West Africa. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 97 –142. Dowdeswell, J. A., Hambrey, M. J. & Wu, R. 1985. A comparison of clast fabric and shape in Late Precambrian and modern glaciogenic sediments. Journal of Sedimentary Petrology, 55, 691– 704. Elverhoi, A., Pfirman, S. L., Solheim, A. & Larssen, B. B. 1989. Glaciomarine sedimentation in epicontinental seas exemplified by the northern Barents Sea. Marine Geology, 85, 225– 250. Eyles, C. H. 1988. Glacially and tidally influenced shallow marine sedimentation of the Late Precambrian Port Askaig Formation. Palaeogeography, Palaeoclimatology, Palaeoecology, 68, 1– 25. Eyles, C. H. & Lagoe, M. B. 1990. Sedimentation patterns and facies geometries on a temperate glacially-influenced continental shelf: the Yakataga Formation, Middleton Island, Alaska. In: Dowdeswell, J. A. & Scourse, J. D. (eds) Glaciomarine Environments: Processes and Sediments. Geological Society, London, Special Publications, 53, 363– 386. Eyles, N., Eyles, C. H. & Miall, A. D. 1983. Lithofacies types and vertical facies models: an alternative approach to the description and environmental interpretation of glacial diamict and diamict sequences. Sedimentology, 30, 393– 410. Eyles, C. H., Eyles, N. & Miall, A. D. 1985. Models of glaciomarine sedimentation and their application to the interpretation of ancient glacial sequences. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 15 –84. Fairchild, I. J. & Hambrey, M. J. 1984. The Vendian succession of northeastern Spitzbergen: petrogenesis of a dolomite-tillite association. Precambrian Research, 26, 111– 167. Fedo, C. M., Nesbitt, H. W. & Young, G. M. 1995. Unravelling the effects of potassium metasomatism in sedimentary rocks and paleosols, with implications for paleoweathering conditions and provenance. Geology, 23, 921–924, doi: 10.1130/0091-7613(1995)023 ,0921:UTEOPM.2.3.CO;2. Fike, D. A., Grotzinger, J. P., Pratt, L. M. & Summons, R. E. 2006. Oxidation of the Ediacaran Ocean. Nature, 444, 744– 747, doi: 10.1038/nature05345. Fo¨lling, P. G. & Frimmel, H. E. 2002. Chemostratigraphic correlation of carbonate successions in the Gariep and Saldania Belts, Namibia and South Africa. Basin Research, 14, 69– 88. Frimmel, H. E., Fo¨lling, P. G. & Eriksson, P. G. 2002. Neoproterozoic tectonic and climatic evolution recorded in the Gariep Belt, Namibia and South Africa. Basin Research, 14, 55 –68. Gass, I. G., Ries, A. C., Shackleton, R. M. & Smewing, J. D. 1990. Tectonics, geochronology and geochemistry of the Precambrian rocks of Oman. In: Robertson, A. H. F., Searle, M. P. & Ries, A. C. (eds.) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 585– 599. Glennie, K. W. 1977. Outline of the geology of Oman. Me´moires de la Societe´ Ge´ologique de France, hors series, 8, 25 – 31. Glennie, K. W., Boeuf, M. G. A., Hughes Clark, M. W., Moody-Stuart, M., Pilaar, W. F. H. & Reinhardt, B. M. 1974. Geology of the Oman Mountains. KSEPL, Rijswijk, The Netherlands, 423.

261

Gorin, G. E., Raacz, L. G. & Walter, M. R. 1982. Late Precambrian –Cambrian sediments of Huqf Group, Sultanate of Oman. American Association Petroleum Geologists Bulletin, 66, 2609–2627. Gray, D. R., Miller, J. M. & Gregory, R. T. 2005. Strain state and kinematic evolution of a fold-nappe beneath the Samail Opiolite, Oman. Journal Structural Geology, 27, 1986– 2007. Halverson, G. P., Maloof, A. C. & Hoffman, P. F. 2004. The Marinoan glaciation (Neoproterozoic) in northeast Svalbard. Basin Research, 16, 297– 324. Hargrove, U. S., Stern, R. J., Kimura, J.-I., Manton, W. I. & Johnson, P. R. 2006. How juvenile is the Arabian– Nubian Shield? Evidence from Nd isotopes of pre-Neoproterozoic inherited zircon in the Bi’r Umq suture zone, Saudi Arabia. Earth & Planetary Science Letters, 252, 308–326. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129– 155, doi: 10.1046/j.1365-3121.2002.00408.x. Hoffman, P. F., Kaufmann, A. J., Halverson, G. P. & Schrag, D. P. 1998a. A Neoproterozoic snowball Earth. Science, 281, 1342– 1346, doi: 10.1126/science.281.5381.1342. Hoffman, P. F., Kaufman, A. J. & Halverson, G. P. 1998b. Comings and goings of global glaciation on a Neoproterozoic tropical platform in Namibia. GSA Today, 8, 1– 9. Hughes Clarke, M. W. 1988. Stratigraphy and rock nomenclature in the oil-producing area of interior Oman. Journal of Petroleum Geology, 11, 5– 60. Kapp, H. E. & Llewellyn, P. G. 1965. The geology of the Central Oman Mountains. Report S00005-9, Geological Group, Petroleum Development Oman. Kaufman, A. J. & Knoll, A. H. 1995. Neoproterozoic variations in the C-isotopic composition of seawater: stratigraphic and biogeochemical implications. Precambrian Research, 73, 27– 49. Kellerhals, P. & Matter, A. 2003. Facies analysis of a glaciomarine sequence, the Neoproterozoic Mirbat Sandstone Formation, Sultanate of Oman. Eclogae Geologicae Helvetiae, 96, 49– 70. Kennedy, M. J. 1996. Stratigraphy, sedimentology and isotope geochemistry of Australian Neoproterozoic postglacial cap dolostones: deglaciation, d13C excursions, and carbonate precipitation. Journal of Sedimentary Research, 66, 1050–1064. Kilner, B., MacNiocaill, C. & Brasier, M. D. 2005. Low-latitude glaciation in the Neoproterozoic of Oman. Geology, 33, 413– 416, doi: 10.1130/G21227.1. Leather, J. 2001. Sedimentology, chemostratigraphy and geochronology of the lower Huqf Supergroup, Oman. PhD thesis, Trinity College Dublin, vols 1 and 2. Leather, J., Allen, P. A., Brasier, M. D. & Cozzi, A. 2002. Neoproterozoic snowball Earth under scrutiny: evidence from the Fiq glaciation of Oman. Geology, 30, 891– 894. Le Guerroue´, E. 2010. Duration and synchroneity of the largest negative carbon isotope excursion on Earth: The Shuram/Wonoka anomaly. Comptes Rendus Geosciences. doi: 10.1016/j.crte.2009. 12.008. Le Guerroue´, E. & Cozzi, A. 2010. Veracity of Neoproterozoic negative C isotope values: The termination of the Shuram negative excursion. Gondwana Research, 17, 653–661, doi: 10.1016/j.gr.2009.11.002. Le Guerroue´, E., Allen, P. A. & Cozzi, A. 2005. Two distinct glacial successions in the Neoproterozoic of Oman. GeoArabia, 10, 17– 34. Le Guerroue´, E., Allen, P. A. & Cozzi, A. 2006a. Chemostratigraphic and sedimentological framework of the largest negative carbon isotopic excursion in Earth history: the Neoproterozoic Shuram Formation (Nafun Group, Oman). Precambrian Research, 146, 68 – 92, doi: 10.1016/j.precamres.2006.01.007. Le Guerroue´, E., Allen, P. A., Cozzi, A., Etienne, J. L. & Fanning, M. 2006b. 50 million year duration negative carbon isotopic excursion in the Ediacaran ocean. Terra Nova, 18, 147–153, doi: 10.1111/j.1365-3121.2006.00674.x. Le Guerroue´, E., Allen, P. A. & Cozzi, A. 2006c. Parasequence development in the Ediacaran Shuram Formation (Nafun Group, Oman): high resolution stratigraphic test for primary origin of negative carbon

262

P. A. ALLEN ET AL.

isotopic ratios. Basin Research, 18, 205– 220, doi: 10.1111/ j.1365-2117.2006.00292.x. Le Me´tour, J., Villey, M. & de Gramont, X. 1986. Geological map of Quryat, Sheet NF 40-4D, scale 1:100,000. Directorate of Minerals, Oman Ministry of Petroleum and Minerals. Loosveld, R., Bell, A. & Terken, J. 1996. The tectonic evolution of interior Oman. GeoArabia, 1, 28 –50. Mackiewicz, N. E., Powell, R. D., Carlson, P. R. & Molnia, B. F. 1984. Interlaminated ice-proximal glaciomarine sediments in Muir Inlet, Alaska. Marine Geology, 57, 113– 147. Mann, A. & Hanna, S. S. 1990. The tectonic evolution of pre-Permian rocks, Central and Southeastern Oman Mountains. In: Robertson, A. H. F., Searle, M. P. & Ries, A. C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 307– 325. Marshall, J. D. 1992. Climatic and oceanographic isotopic signals from the carbonate rock record and their preservation. Geological Magazine, 129, 143– 160. McCarron, G. M. E. 2000. The Sedimentology and Chemostratigraphy of the Nafun Group, Huqf Supergroup, Oman. PhD thesis, University of Oxford. Mercolli, I., Briner, A. P., Frei, R., Scho¨nberg, R., Nagler, T. F., Kramers, J. & Peters, T. 2006. Lithostratigraphy and geochronology of the Neoproterozoic crystalline basement of Salalah, Dhofar, Sultanate of Oman. Precambrian Research, 145, 182– 206. Miller, J. M. G. 1996. Glacial sediments. In: Reading, H. G. (ed.) Sedimentary Environments: Processes, Facies and Stratigraphy. Blackwell Science Ltd, Oxford, 454– 484. Moncrieff, A. C. M. 1989. Classification of poorly sorted sedimentary rocks. Sedimentary Geology, 65, 191–194. Moncrieff, A. C. M. & Hambrey, M. J. 1990. Marginal marine glacial sedimentation in the late Precambrian succession of east Greenland. In: Dowdeswell, J. A. & Scourse, J. D. (eds) Glaciomarine Environments: Processes and Sediments. Geological Society, London, Special Publications, 53, 387– 410. Myrow, P. M. & Kaufman, A. J. 1999. A newly discovered cap carbonate above Varanger-age glacial deposits in Newfoundland. Canadian Journal of Sedimentary Research, 69, 784– 793. Narbonne, G. M., Kaufman, A. J. & Knoll, A. H. 1994. Integrated chemostratigraphy and biostratigraphy of the Windermere Supergroup, northwestern Canada: implications for Neoproterozoic correlations and the early evolution of animals. Geological Society of America Bulletin, 106, 1281–1292. Nardin, T. R., Hein, F. J., Gorsline, D. S. & Edwards, B. D. 1979. A review of mass movement processes, sediment and acoustic characteristics, and contrasts in slope and base-of-slope systems versus canyon-fan-basin systems. In: Doyle, L. J. & Pilkey, O. H. (eds) Geology of Continental Slopes. Society of Economic Paleontologists and Mineralogists, Special Publication. 27, 61 –73. Nesbitt, H. W., Young, G. M., McLennan, S. M. & Keays, R. R. 1996. Effects of chemical weathering and sorting on the petrogenesis of siliciclastic sediments, with implications for provenance studies. Journal of Geology, 104, 525– 542. Ovenshine, A. T. 1970. Observations of iceberg rafting in Glacier Bay, Alaska, and the identification of ancient ice-rafted deposits. Bulletin Geological Society America, 81, 891– 894. Platel, J. P., Roger, J., Peters, T., Mercolli, I., Kramers, J. D. & Le Me´tour, J. 1992a. Geological Map of Salalah, Sheet NE 40-09, 1: 250000, with explanatory notes. Directorate General of Minerals, Oman Ministry of Petroleum and Minerals. Platel, J. P., Le Me´tour, J., Berthiaux, A., Buerrier, M. & Roger, J. 1992b. Geological map of Juzor Al Halaaniyat, sheet NE 40-10, scale 1:250,000. Directorate General of Minerals, Oman Ministry of Petroleum and Minerals.

Powell, R. D. & Molnia, B. F. 1989. Glaciomarine sedimentary processes, facies and morphology of the south –southeast Alaska Shelf and fjords. Marine Geology, 85, 359– 390. Powell, R. & Domack, E. 1995. Glaciomarine processes and sediments. In: Menzies, J. (ed.) Modern Glacial Environments. ButterworthHeinemann, Oxford. Rabu, D. 1988. Ge´ologie de l’authochthon des montagnes d’Oman, la fenetre du Jabal Akhdar. PhD thesis, Universite´ Pierre et Marie Curie, Paris 6, and documents BRGM 130. Rabu, D., Bechennec, F., Beurrier, M. & Hutin, M. 1986. Geological map of Nakhl. Sheet NF-40-3E, scale 1:100,000. Oman Ministry of Petroleum and Minerals, Directorate General of Minerals. Rabu, D., Nehlig, P. et al. 1993. Stratigraphy and structure of the Oman Mountains. Bureau de Re´cherches Ge´ologiques et Minie`res, 221. Rieu, R. 2006. Sedimentology, Stratigraphy and Geochemistry of the Glacially Influenced Neoproterozoic Mirbat Group, Oman. PhD thesis, ETH-Zu¨rich. Rieu, R., Allen, P. A., Cozzi, A., Kosler, J. & Bussy, F. 2007a. A composite stratigraphy for the Neoproterozoic Huqf Supergroup of Oman: integrating new litho-, chemo- and chronostratigraphic data of the Mirbat area, south Oman. Geological Society London Journal, 164, 997– 1009. Rieu, R., Allen, P. A., Plo¨tze, M. & Pettke, T. 2007b. Compositional and mineralogical variations in a Neoproterozoic glacially influenced succession, Mirbat area, south Oman: Implications for paleoweathering conditions. Precambrian Research, 154, 248– 265, doi: 10.1016/ j.precamres.2007.01.003. Rieu, R., Allen, P. A., Plo¨tze, M. & Pettke, T. 2007c. Climatic cycles during a Neoproterozoic ‘snowball’ glacial epoch. Geology, 35, 299– 302, doi: 10.1130/G23400A.1. Roger, J., Be´chennec, F., Janjou, D., Le Me´tour, J., Wyns, R. & Buerrier, M. 1992. Explanatory Notes to the Geological Map of Ja-alan. 1:100,000 Sheet NF 40-8E. Directorate General of Minerals, Ministry of Petroleum and Minerals. Simonson, B. M. & Carney, K. E. 1999. Roll-up structures: Evidence of in situ microbial mats in Late Archaean deep shelf environments. In: Hagadorn, J. W., Pflueger, F. & Bottjer, D. J. (eds) Unexplored Microbial Worlds. Palaios. 14, 13 –24. Stow, D. A. V., Reading, H. G. & Collinson, J. D. 1996. Deep seas. In: Reading, H. G. (ed.) Sedimentary Environments: Processes, Facies and Stratigraphy. Blackwell Science Ltd, Oxford, 395– 453. Tucker, M. E. 1986. Formerly aragonitic limestones associated with tillites in the Late Proterozoic of Death Valley, California. Journal of Sedimentary Petrology, 56, 818– 830. Tschopp, R. H. 1967. The general geology of Oman. Proceedings of the 7th World Petroleum Congress, Mexico, 2, 231. Villey, M., Le Me´tour, J. & Gramont, X. 1986. Geological map of Fanjah, sheet NF 40-3F, scale 1:100,000. Oman Ministry of Petroleum and Minerals, Directorate General of Minerals. Wentworth, C. K. 1936. An analysis of the shape of glacial cobbles. Journal of Sedimentary Petrology, 6, 85 –96. Williams, G. E. 1979. Sedimentology, stable isotope geochemistry and palaeoenvironment of dolostones capping late Precambrian glacial sequences in Australia. Journal of the Geological Society of Australia, 26, 377–386. Wright, R. & Anderson, J. B. 1982. The importance of sediment gravity flow sediment transport and sorting in a glacial marine environment: eastern Weddell Sea, Antarctica. Bulletin Geological Society America, 93, 951–963. Wright, V. P., Ries, A. C. & Munn, S. G. 1990. Intraplatformal basin-fill from the Infracambrian Huqf Group, east-central Oman. In: Robertson, A. H. F., Searle, M. P. & Ries, A. C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 601–616.

Chapter 21 The Tambien Group, Northern Ethiopia (Tigre) NATHAN R. MILLER1*, DOV AVIGAD2, ROBERT J. STERN3 & MICHAEL BEYTH4 1

Department of Geological Sciences, Jackson School of Geosciences, University of Texas at Austin, Austin, TX 78712-0254, USA 2

Institute of Earth Sciences, Hebrew University of Jerusalem, Jerusalem 91904, Israel

3

Geosciences Department, University of Texas at Dallas, Richardson, TX 75083-0688, USA 4

Geological Survey of Israel, Jerusalem, Jerusalem 95501, Israel *Corresponding author (e-mail: [email protected])

Abstract: The Tambien Group of northern Ethiopia (Tigre), with probable correlatives in Eritrea, is a 2 –3-km-thick siliciclastic– carbonate succession that was deposited in an intra-oceanic arc platform setting within the southern Arabian– Nubian Shield (ANS) area (southern extension of the Nakfa Terrane) of the Mozambique Ocean. Its deposition occurred prior to ocean closure between converging fragments of East and West Gondwana and concomitant structural emergence of the East African Orogen (EAO). The Tambien Group is well exposed and best studied in the Mai Kenetal and Negash synclinoria, where litho- and chemostratigraphy (including d13Ccarb, 87Sr/86Sr) provide the basis for a composite reference section. Two glaciogenic intervals have been suggested from exposures within the Didikama and Matheos Formation in the Negash Synclinorium. No reliable palaeomagnetic data exist to constrain the palaeolatitude of Tambien Group deposition and the southern ANS, but palaeogeographic reconstructions and evaporite pseudomorphs in lower carbonate units (Didikama Formation) imply low to intermediate latitudes (,458). Integration of available geochronological information (regional magmatism and detrital zircon) suggests c. 775– 660 Ma as a plausible window constraining deposition of the prospective glacial intervals. The Tambien Group appears to preserve a coherent chemostratigraphic framework that can be effectively subdivided according to shifts in d13Ccarb polarity [polarity intervals A (þ), B (–), C (þ), D (–)]. Slates underlying and interstratified with polarity interval A carbonate preserve evidence of extreme chemical weathering that lessened prior to deposition of polarity interval B carbonate. Tambien Group carbonate units have sedimentological characteristics consistent with both shallow and deeper marine depositional settings. The lower prospective glacial interval lacks diagnostic sedimentological evidence of synglacial deposition, but is overlain by negative d13C carbonate (polarity interval B) with sedimentological characteristics consistent with well-documented cap-carbonate successions. The upper prospective glacial interval in the Negash Synclinorium (Matheos Diamictite) best exhibits characteristics consistent with glaciogenic deposition (matrix-supported polymictic clasts, possible dropstones, possible bullet-nosed and striated clasts). In contrast to pericratonic rift margin settings that are common for Cryogenian glaciogenic deposits, palaeogeographic reconstructions for the 775 –660 Ma timeframe place northern Ethiopia within an intra-oceanic setting that was likely far removed from cratonic hinterlands. More work on Tambien Group sedimentology, geochronology and palaeogeography is required to better evaluate the extent and timing of glacial conditions associated with the prospective glaciogenic intervals. Supplementary material: Supplementary Table 21.1 of Tambien Group geochronological age constraints is available at http://www. geolsoc.org.uk/SUP18462.

The Tambien Group is exposed (Fig. 21.1) throughout portions of northern Ethiopia (Tigre Province) and Eritrea (NNE extensions from Figure 21.1; Bizen domain and Adobha Abi terrane of Beyth et al. (2003) and De Souza Filho & Drury (1998)), in greenschist-grade terranes comprising the southern portion of the Arabian –Nubian Shield (ANS). It may be equivalent to similar carbonate-rich units in NE Sudan (Bailateb Group; Stern et al. 1994), SW Saudi Arabia (Hali Group; Greenwood et al. 1976), and possibly western Yemen (inferred from its Red Sea conjugate position). Its deposition marks mainly marine siliciclastic and carbonate sedimentation within the Mozambique Ocean, an extinct Neoproterozoic ocean basin destroyed during the late Neoproterozoic consolidation of Greater Gondwana with the emergence of the East African Orogen (EAO) (Stern 1994). The Tambien Group is best studied in northern Ethiopia from the Mai Kenetal (western flank near 138550 N, 388500 E) and Negash (southern extent near 138500 N, 398370 E) synclinoria, where much of the Tambien Group is continuously exposed, and these localities provide the substantial basis for regional litho- and chemostratigraphy. Two possible glaciogenic intervals have been suggested within the Tambien Group (Beyth et al. 2003; Miller et al. 2003, 2009). The lower interval occurs as a greywacke-conglomerate interval within slate of the lower Didikama Formation in the Negash Synclinorium (Fig. 21.2a, column D), and the top of this interval may correlate with the base of the Assem Limestone in the Mai Kenetal Synclinorium (see discussion). The upper prospective glacial interval tops the Tambien Group as diamictite of the

Matheos Formation in the core of the Negash Synclinorium (Fig. 21.2a), and this interval may be equivalent to arkosic sandstone and conglomerate of the Dugub Formation that similarly tops the Tambien Group in the western Shiraro area (Fig. 21.1, Avigad et al. 2007). The upper diamictite unit at Negash was originally defined as the ‘Pebbly slate’ (Beyth 1972). It was subsequently assigned within the Matheos Formation (Garland 1980) and informally described as ‘Pebbly Slate (diamictite)’ in Miller et al. (2003), ‘Slate/Pebbly slate (Diamictite Slate)’ in Alene et al. (2006) and ‘Negash Diamictite’ in Avigad et al. (2007). Miller et al. (2009) subdivide the Matheos Formation into three members, the youngest corresponding to the diamictite facies. In consideration of a possible lower glaciogenic interval in the Negash Synclinorium, the informal term ‘Matheos Diamictite’ is suggested for the upper prospective glacial interval. In addition to these northern Ethiopian localities, Eritrea hosts possible glaciogenic units that have not been systematically studied. The southern ANS is still very much a frontier region in need of systematic sedimentological, geochemical, and geochronological studies of Neoproterozoic units. Much of what is known about the Cryogenian Period for the region has been learned in only the past decade. The earliest suggestion of a Late Proterozoic glaciation was by Bibolini (1920), who described faceted clasts (facce piane) within an unnamed pebbly mudstone-conglomerate unit (conglomerati poligenici) in northern Eritrea. The basic Neoproterozoic stratigraphic framework for northern Ethiopia was established in regional mapping by Beyth (1972), including description of the

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 263– 276. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.21

264

N. R. MILLER ET AL.

Fig. 21.1. Location of key Tambien Group exposures (unshaded units) (A, Shiraro area; B, Mai Kenetal Synclinorium; C, Negash Synclinorium; D, Samre area) within northern Ethiopia (Tigre Province) modified from Miller et al. (2009). Metavolcano–sedimentary block boundaries (Tadesse et al. 1999) occur only within the Tsaliet Group, and are inferred to represent accreted arc terranes or slivers in a supra-subduction zone setting. Stars show geochronological localities for syn-tectonic intrusives (white), post-orogenic intrusives (black) and detrital zircons (white stars with black dots) discussed in the text and shown in Figure 21.5 (also in Supplementary Table 21.1). Prospectively equivalent upper Tambien Group strata of the Gulgula Group occur in western Eritrea (arrow) just west of the Shiraro Area, as well as in the Adobha Abi terrane and Bizen Domain (arrows) to the north in Eritrea. Dashed box east of Mai Kenetal outlines the Werii study area of Sifeta et al. (2005). M. Alem marks the western limb locality (Madahne Alem) within the Negash Synclinorium bearing 774.7 + 4.8 Ma zircons in slate c. 16 m below lowest Tambien Group carbonate beds (Avigad et al. 2007). The Negash Synclinorium is structurally bounded to the east by the Atsbi Horst (AH ). Inset map (below) shows the location of an unnamed pebbly mudstone unit in northeastern Eritrea (Cecioni 1981) that could correlate to the Tambien Group.

units now posited to have glacial associations. Motivated by the Snowball Earth hypothesis, a number of reconnaissance studies have since explored the depositional context of the Tambien Group (Beyth et al. 2003; Miller et al. 2003; Alene et al. 2006; see also Stern et al. 2006). More comprehensive regional investigations, involving U–Pb zircon geochronology, chemical weathering indices, and higher resolution sampling for integrated C- and Sr-isotope stratigraphy, have further refined the age and range of litho- and chemostratigraphic variations in the Tambien Group (Sifeta et al. 2005; Avigad et al. 2007; Miller et al. 2009).

Structural framework (See additional region-specific structural information in the ‘Glaciogenic deposits and associated strata’ section.) The greater

ANS consists of a patchwork of Neoproterozoic tectonostratigraphic terranes, now bisected by the Oligocene and younger Red Sea rift. ANS terranes include significant volumes of juvenile Neoproterozoic crust generated within spreading centers, arc and back-arc settings of the Mozambique Ocean (Stern 1994). Sutures between terranes, many with ophiolites and dated metamorphic assemblages, document collisional deformation and help constrain the timing of terrane amalgamation. Many sutures have appreciable strike – slip offsets, and these may broadly relate to c. 600 Ma escape tectonics (Burke & Sengo¨r 1986), during which ANS terranes were progressively sandwiched by, and offset between, obliquely converging Gondwana cratonic blocks (de Souza Filho & Drury 1998, and references therein). Tambien Group exposures in Tigre occur within the presumed southern extension of the Nakfa terrane of Eritrea. The Nakfa terrane is one of several suture-bounded low-grade volcano-sedimentary

THE TAMBIEN GROUP, NORTHERN ETHIOPIA

Syn-

Enda Zebi Fm

Dugub Fm - - - Mentebtab Fm

Shiraro Group

Didikama Fm

Post-orogenic granitoids

---- Tambien Group ----

Pz & Mz Adigrat Ss Seds

Undiff. Metavolc.

Adigrat Ss Algal* Ls

Shiraro Cgl ? Arkosic ss

B. Mai Kenetal Synclinorium

C. Negash Synclinorium

D. Negash Synclinorium

Beyth 1972 (Tadesse 1999)

Beyth 1972

Miller et al. 2009 (Garland 1980)

Enticho Ss

Enticho Ss

Enticho Ss

Mai Kenetal Ls (200m) Pebbly slate (200m) ? (Tselim Imni Ls) Black detrital ls Tsedia Slate (500m) (300m) (subdivided into Mica dolomite & upper Bilato and lower Logmiti Slates) slate (700m)

Matheos Fm

Relation A. to synShiraro Plains & postRegion tectonic Tadesse Beyth 1972 intrusives 1999

Diamictite Mbr ? Transitional Mbr Black Ls Mbr

Mentebtab Ss Assem LS (300m) (Filafil Ls)

Slate & dolomite (200m)

Purple slate w/grn reduction spots

Werii Slate (1000m) (Segali Slate)

Purple slate w/grn reduction spots; greywacke (300m) ?

Tsaliet Metavolcanics

Tsaliet Metavolcanics

Tsaliet Metavolcanics

Bedded ls Mica dolomite

(b)

Didikama Fm

(a)

265

Dolomite > Slate Slate > Dolomite ? GCI Lower Slate (Atsbi Horst) Tsaliet Group

d13C carb Polarity

2

Mai Kenetal Synform Negash Synform Afro-Arabian Peneplain ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Mai Kenetal LS C (++) Matheos Fm D (-) Diamictite Mbr ? ? ? 3 Transition Mbr C (+) Tsedia Slate C (++) Black LS Mbr ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ unconformity? Assem LS B (-) Didikama Fm B (-) Upper (dol>slate) ~~~~~~~~~~~~~~~ GCI ? ? ? 1 ? ? ? paraconformity? A (+) Lower (slate >dol) Werii Slate A (+?) Lower Slate (Atsbi Horst) Tsaliet Metavolcanics

Tsaliet Metavolcanics

Fig. 21.2. (a) Lithostratigraphic subdivision of the Tambien Group in previous work. Prospective glacial intervals in columns A, C and D are superimposed from Beyth et al. (2003) and Miller et al. (2009). The upper Algal Limestone unit in the Shiraro plains region (column A), originally termed the ‘Algal, stromatoporoidea (?) limestone’ by Beyth (1972) was not mapped in the Axum Sheet (Tadesse 1999) and may occur in the Badme region (north of Shiraro, Fig. 21.1). (b) Chemo- and lithostratigraphic correlations between the Mai Kenetal and Negash synclinoria used to create a composite Tambien Group stratigraphic section (see discussion). Stratigraphic intervals in each section are alphabetized (A –D) according to the succession of associated d13Ccarb polarity (þ or – ) intervals. See Figure 21.3 for associated chemostratigraphic characteristics (circled data sets: d13CTOC, 87Sr/86Sr, [Sr] and chemical weathering index) that support the correlations. Numbers 1–3 correspond to the different segments used in constructing the composite stratigraphy of Figure 21.6. Abbreviations (A&B): Fm, formation; GCI, Greywacke –Conglomerate Interval, the lower prospective glacial interval in the Negash Synclinorium; grn, green; Ls, limestone; Mbr, member; Metavolc, metavolcanics; Mz, Mesozoic; Pz, Palaeozoic; Ss, sandstone; Undiff, undifferentiated. (Modified from Miller et al. 2009.)

terranes comprising the greater Nubian Tokar superterrane (Kro¨ner et al. 1991). The Nakfa terrane supracrustal assemblage records a lower sequence of igneous rocks associated with arc magmatism in and around the southern ANS sector of the Mozambique Ocean, as exemplified by calc-alkaline plutons and associated metavolcanic rocks of the Tsaliet Group. In western Tigre, the Tsaliet Group may be preserved as a series of accreted volcanic arc terranes within a suprasubduction setting (Fig. 21.1; volcano-sedimentary blocks of Tadesse et al. 2000). Cessation of Tsaliet arc volcanism was followed by deposition of weathered arc detritus (Sifeta et al. 2005) and in turn increasing proportions of carbonate sediments, as exemplified by the Tambien Group. The contact between the Tsaliet and Tambien Group is poorly understood, ranging from seemingly conformable and gradational to fault-bounded (Sifeta

et al. 2005). An unconformable basal contact also was postulated from considerable lateral variations observed among basal Tambien metasediments (Arkin et al. 1971; Beyth 1972; Tadesse 1999). Tambien Group deposition, including the two prospective glacial intervals, is thought to have occurred in an intra-oceanic platform setting (above a substratum of consolidated arc terranes) following the main phase of Tsaliet arc magmatism in the region (c. 780 + 30 Ma; Avigad et al. 2007). The extent to which Tsaliet arc subterranes were fully accreted prior to Tambien Group deposition is uncertain, and some syndepositional relief differentiation related to ongoing shortening and/or extension is possible (Miller et al. 2009). The entire Nakfa basement complex was deformed after deposition of the Tambien Group in conjunction with closure of the Mozambique Ocean between fragments of east and west Gondwana,

266

N. R. MILLER ET AL.

and the concomitant structural emergence of the EAO, beginning c. 630 Ma. As a consequence, NNE-trending upright/overturned folds and faults dominate the regional structural grain, with Tambien Group exposures best preserved in synclinoria and grabens. Internal minor or secondary folds and faults complicate measured sections in the Mai Kenetal and Negash synclinoria. Restoring shortening (200%) associated with these collisional structures greatly extends the minimal aerial extent of the Tambien basin or basins. The deformed Tsaliet-through-Tambien Group succession was subsequently intruded by c. 610 Ma (‘Mareb’) granitoids associated with crustal thickening and differentiation of the maturing EAO. Whether or not, and for how long, Tambien Group deposition continued prior to emplacement of the ‘Mareb’ post-orogenic granitoids is unknown. Despite these post-depositional orogenic processes, temperatures and pressures did not exceed lowgreenschist grade metamorphism and well-preserved primary sedimentary structures are common in many Tambien Group carbonate successions (Alene et al. 2006; Miller et al. 2009). Extensive erosion of the EAO resulted in cutting of a widespread regional unconformity prior to Cambro-Ordovician time (Avigad et al. 2005), herein termed the Afro-Arabian Peneplain (AAP). Uppermost Tambien Group exposures in the core of the Negash Synclinorium (Matheos Diamictite) and possibly equivalent siliciclastic deposits of the Dugub Formation in the western Shiraro area (Fig. 21.1) are thought to be northern Ethiopia’s youngest preserved Neoproterozoic metasediments below the AAP.

Stratigraphy Beyth (1972) documented Tambien Group outcrops in the Shiraro area, in several en echelon synclinoria to the east (i.e. Mai Kenetal, Tsedia, Chemit and Negash), and the Escarpment area east of Adigrat (Fig. 21.1, study areas A– C). Apparent regional facies variations within the internal stratigraphy of the Tambien Group were described, and a preliminary correlation was proposed for three areas: Shiraro (west), Mai Kenetal (centre) and Negash (east). Subsequent workers have modified the stratigraphy in various ways as described below and shown in Figure 21.2a. Figure 21.2b shows a recent correlation scheme proposed for the Tambien Group based on integrated litho- and chemostratigraphy of the Mai Kenetal and Negash synclinoria (Miller et al. 2009), which is further elaborated in the discussion section. Above the Tsaliet Metavolcanics, Beyth (1972) assigned four formational designations within the Tambien Group based on the Mai Kenetal type section (in stratigraphic order): Werii Slate, Assem Limestone, Tsedia Slate and Mai Kenetal Limestone. In the Negash Synclinorium, Beyth defined informal facies (in stratigraphic order: greywacke and purple slate with green reduction spots, slate and dolomite, mica dolomite and slate, black detrital limestone, and pebbly slate), which he suggested might correlate with the Mai Kenetal section (Fig 21.2a, columns C v. B). Dolomite-slate successions similar to those of the lower Tambien Group in the Negash Synclinorium were subsequently documented in the Escarpment east of Adigrat, as well as in synclines in the Shiraro area. Garland (1980) formalized Beyth’s Tsaliet Metavolcanics as the Tsaliet Group, and subdivided the Negash Tambien Group succession (Fig. 21.2a, column D) into the lower Didikama Formation (after similar dolomitic rocks in the Shiraro Area) and overlying Matheos Formation (equivalent to Beyth’s informal black detrital limestone and pebbly slate facies). In mapping of the Mai Kenetal synclinorium, Tadesse (1999) used local names to address the possibility that some of Beyth’s original type areas (i.e. Werii Slate, Tsedia Slate) might reside in different volcanosedimentary arc terranes (Fig. 21.1). The main difference between the two lithostratigraphic schemes is differentiation of Beyth’s Tsedia Slate into a lower Logmiti Slate and overlying Bilato Limestone and Slate (Fig. 21.2a, column B). Tadesse (1999) also formalized a new Neoproterozoic stratigraphic scheme for the Shiraro

area, splitting Beyth’s informal Tambien Group facies into the basal Didikama Formation and overlying Shiraro Group (Fig. 21.2a, column A). Designated within the latter were three metasiliciclastic formations (Enda Zebi, Dugub and Mentebtab formations), each with respective lateral facies variations. In the eastern limb of the Negash Synclinorium, Beyth (1972) mapped the base of the Tambien Group as an interval of arkosic sandstone and conglomerate, above a possible tectonic contact with the Astbi Horst (Fig. 21.1, Fig. 21.2a column C). This coarse metasiliciclastic interval corresponds to the lower prospective glacial interval, the Greywacke-Conglomerate interval (GCI) of the Didikama Formation. The basal contact of the GCI has not been studied in detail, but the underlying depositional sequence involves a minimum of several hundred metres of conformably bedded fine-grained marine metasedimentary units (mainly slate with episodic thin carbonate interbeds up to a few centimetres thick). This lower slate succession contrasts from coarse volcanic agglomerate and conglomerate of the Tsaliet Group exposed at the top of the Atsbi Horst (Fig. 21.1), but the contact between these units has not been systematically mapped or studied. Miller et al. (2003, 2009) include this lower slate interval (below the GCI) within the Tambien Group, as the informal Lower Slate member of the Didikama Formation (Fig. 21.2a, column D). Above the GCI, the Didikama Formation transitions into variegated slate that includes the first prominent dolomite beds (informal Slate . Dolomite member) and thicker bedded dolomite at the top of the formation (informal Dolomite . Slate member). The GCI and overlying purple slate (with green reduction spots; the ‘Purple Egg Slate’ of Beyth 1972), with estimated combined thickness of 300 m, forms a distinctive dark (grey-purple) marker unit that can be traced laterally throughout much of the c. 30-km-long exposed eastern synclinorial limb. The Didikama Formation is overlain sharply, and very likely unconformably (Beyth 1972, p. 74; Garland 1980, p. 14), by distinctive black limestone of the Matheos Formation. Miller et al. (2009) subdivide the Matheos Formation into three members. The basal Black Limestone Member is well laminated and partly detrital (intraclasts), and transitions upward into non-calcareous slate (Transitional Member) and eventually pebbly slate (Diamictite Member).

Prospective glaciogenic deposits and associated strata Lower prospective glaciogenic unit Greywacke –Conglomerate Interval, lower Didikama Formation (Lower Tambien Group). Beyth (1972) described the GCI within the

eastern flank of the Negash Synclinorium as poorly sorted, lightgrey to black greywacke, containing mostly medium grained to pebble-sized, subrounded quartz, in a mud (slate) supported matrix (Fig. 21.2a, column C). Thin conglomeratic layers and ripple marks (exposed in bedding planes) also occur within the unit. The unit fines upward into purple slate with distinctive ovate greenish-yellow patches, interpreted as reduction spots. This lower interval (c. 300 m thick, Beyth 1972) is overlain by variegated slate (orange and green), which in turn grades upward (with poor outcrop) into the lowest exposed dolomite beds of the Didikama Formation. Similar purple spotted slate is reported below the Didikama Formation in the Shiraro area (Beyth 1972; Fig. 21.2a, column A). The GCI is underlain by well-layered slate (with occasional sub-decimetre-thick carbonate interbeds and minor foliation) that crops out continuously down-section (eastward) for at least several hundred metres in the hanging wall flank of the Atsbi Horst (Fig. 21.1). Upper Werri Slate (Lower Tambien Group). In the NW flank of the

Mai Kenetal Synclinorium, the Assem Limestone lies sharply and conformably above the Werii Slate and shares

THE TAMBIEN GROUP, NORTHERN ETHIOPIA

sedimentological and isotopic characteristics of well-studied capcarbonate sequences (i.e. abrupt basal lithological transition to high-energy carbonate facies, basal dolomite bed (c. 1 m thick) transitioning upward into limestone, consistently negative d13Ccarb, primary and/or early diagenetic (pre-compaction) prismatic and radial fibrous cements (Miller et al. 2009, figs S2D – G; Hoffman et al. 2007)). The observation that many Cryogenian carbonate successions with negative d13C compositions are affiliated with glaciogenic intervals (e.g. Yoshioka et al. 2003; Halverson et al. 2005; Corsetti et al. 2007 and other arguments, see discussion) raises the possibility that the pre-Assem Limestone depositional sequence could have a glacial affiliation. The underlying Werri Slate is mainly finely laminated, well-foliated, noncalcareous slate that does not exhibit obvious glaciogenic sedimentological characteristics. However, the nature of the upper Werri Slate and its contact with the Assem Limestone has not been studied regionally and poorly sorted metasiliciclastic intervals are documented, at least locally, in the Werri Slate. For example, Beyth (1972) reported occurrences of greywacke within the Werii Slate regionally, and encountered poorly sorted rhyolitic agglomerate with well-rounded metre-scale quartzitic fragments at the southern end of Mai Kenetal syncline (around 138470 N and 388510 E). Greywacke is also observed near its gradational contact with Tsaliet Group (Alene et al. 2006).

Upper prospective glaciogenic unit Diamictite Member, Matheos Formation (Upper Tambien Group). The

Matheos Formation, comprising the interior of the Negash Synclinorium, records a conformable transition from black limestone to diamictite. Ranging from faintly layered (massive) to well-bedded, with near vertical dips, the c. 250-m-thick basal Black Limestone Member (Fig. 21.2a, column D) forms a distinctive ridge that rims the interior of the Negash Synclinorium. Individual beds range in thickness from decimetre- to metre-scale and are typically finely laminated in thin section. Intraclastic grainstone intervals are common. Towards the synclinorial axis, limestone beds thin (becoming sub-decimetre thick) and interstratify with increasing proportions of light to dark grey phyllitic slate. The overlying c. 100 m Transitional Member includes non-calcareous phyllitic slate that passes upward into pebbly slate, without interbedded carbonate. The primary depositional fabric involves mainly fine (,1 to 3 cm scale) horizontal beds. Initial disparate sedimentary clasts are sub-centimetre scale (compositions have not been studied). The overlying c. 200-m-thick Diamictite Member occupies the tightly folded core of the Negash Synclinorium and is distinguished by an overall upward increase in clast abundance and size (typically ,10 cm, but up to 20 cm in diameter). Clasts are matrix-supported and the finer matrix retains horizontal layering, with some lateral pinching and swelling, throughout the member. Syn-orogenic compression of the synclinal core has deformed the diamictite to varying extents. More pelitic parts of the diamictite display foliation and associated clasts are often somewhat elongated with pressure shadows. Despite these superimposed features, clasts that appear to preferentially deform underlying matrix laminae are relatively common. Diamictite clasts have subrounded, elongate and angular shapes (including some bullet-nosed clasts), and some planar clast surfaces may be striated (Miller et al. 2003). Clast lithologies are polymictic, including felsic volcanic rocks, fine-grained black limestone and dolomite (including wellrounded clasts that retain primary sedimentary structures; e.g. oolite), low-grade semipelitic sediments, and rare volcanic conglomerate consistent with the upper Tsaliet Group. Dugub Formation, Shiraro Group (Upper Tambien Group). Conglo-

merate and arkosic metasediments comprise youngest Tambien Group exposures in the lowland Shiraro area of western Tigre (Shiraro Group of Tadesse 1999; Fig. 21.2a, column A). These

267

rocks overlie carbonate units assigned to the Didikama Formation (Tadesse 1999) and crop out within the Shiraro Block (Fig. 21.1), a graben-like depression that is fault-bounded to the east against the Adi Hageray Block and extends westward beyond the studied area into Eritrea. The areally most extensive unit is the Dugub Formation, which we summarize following Tadesse (1999) and our own studies 4.5 km ENE of the town of Shiraro (Fig. 21.1). The Dugub Formation consists of weakly metamorphosed conglomerate, arkosic sandstone and siltstone (Tadesse 1999), which are regionally intercalated at different scales. Well-preserved primary sedimentary structures are common as graded bedding, cross-lamination, ripple marks, and flame and slump structures. Elliptical (rounded to subrounded) and well-sorted metaconglomerate clasts (8 cm) comprise up to 40% of the rock volume. Clast compositions include low-grade (Tsaliet-like) volcanic rocks (chlorite schist, tuffaceous metasediments), phyllite, granite, quartz pegmatite, quartzite and chert. Fine- to medium-grained chlorite, muscovite, feldspar and quartz are notable components of the metaconglomerate groundmass. Carbonate is notably absent, but is reported (as scarce marble layers a few to tens of metres thick) in the underlying Enda Zebi Formation (Tadesse 1999). Dugub metasandstone contains quartz pebbles (up to a few millimetres across) and grades laterally and vertically into non-pebbly sandstone and siltstone. Cross bedding exhibits set heights ranging between 0.1 and 0.5 m; coset heights range up to 1.5 m. Orientations of asymmetrical ripple marks and slump structures suggest west-southwestward transport and palaeoslope directions (Tadesse 1999; Avigad et al. 2007).

Prospective glaciogenic deposits in Eritrea The high-grade Arag terrane in NE Eritrea (c. 400 km NNW of Negash) has a widespread unnamed pebbly mudstone unit, consisting of silicic clasts within an argillaceous-siliceous matrix (Verri 1909; Bibolini 1920, 1921, 1922). The unit is reported as widely exposed between 178N and 178450 N along its north-northwesterly trend (c. 90 km) and penetrated by post-orogenic granitoids (acid aplitic intrusions, Cecioni 1981), which have not been dated. Cecioni (1981) described the pebbly mudstone matrix as a hard, violet, clay-quartz cement, and the clasts as pebbles and boulders of quartzite and silicified schist. Some of the small pebbles (,1 cm) are pitted and knotted; others are angular. Bibolini (1920, 1921) noted the occurrence of faceted clasts and postulated a glaciogenic origin. To our knowledge this is the earliest suggestion of a possible Late Proterozoic (‘algonkiano’) glaciation within the ANS. Cecioni (1981) preferred a gravity flow origin for this unit, but recognized a possible glaciogenic association due to the occurrence of faceted pebbles. This unit trends more or less along strike with Tambien Group exposures in northern Ethiopia and prospective Bizen Domain equivalents in southern Eritrea, but has apparently not been further studied in relation to a glacial association. The Neoproterozoic supracrustal succession of western Eritrea (just west of the Shiraro area in northern Ethiopia, Fig. 21.1) consists of volcano-sedimentary assemblages (Augaro Group) unconformably overlain by basin-fill metasediments of the Gulgula Group (Teklay et al. 2003; Teklay 2006). The Gulgula succession involves greenschist-grade polymictic conglomerate, shale, phyllite with interstratified marble lenses, in addition to carbonaceous sandstone, quartz arenite and allodapic carbonate, which, like the Tambien Group, are deformed as upright, recumbent, and isoclinal folds. Teklay (2006) describes the polymictic conglomerates as clasts, both matrix- and clast-supported, ranging from a few centimetres to a half-metre of granite, slate, phyllite and epidotite. The Gulgula Group merges southwards into the Shiraro block in northern Ethiopia, suggesting a composite Gulgula-Shiraro depositional basin and a lithostratigraphic affiliation with the Shiraro Group

268

N. R. MILLER ET AL.

(upper Tambien Group). The origin of Gulgula Group polymict conglomerates, in addition to those described above in the Shiraro Group (Dugub Formation) require further study to clarify whether or not glacial processes were involved.

Boundary relations with overlying and underlying non-glacial units Lower prospective glaciogenic unit Greywacke-Conglomerate Interval (GCI) of the Didikama Formation, Negash Synclinorium. Where examined in the eastern limb of the

Negash synclinorium, the GCI occurs within the Didikama Formation (Fig. 21.2a, columns C and D) above thick slate (Lower Slate Member) and below variegated (purple/green/orange) slate that, in turn, grades upward into lowest dolomite beds of the Didikama Formation (Slate . Dol Member). The continuity of the transition from the Lower Slate member to the GCI is uncertain due to minor folding and possible faulting associated with the eastward structural transition to the Atsbi Horst (Fig. 21.1), whereas the transition above the GCI appears to be continuous and conformable. Upper Werri Slate, Mai Kenetal Synclinorium. Miller et al. (2009) make chemostratigraphic arguments that the equivalent of the GCI may occur in the Mai Kenetal Synclinorium below the Assem Limestone, either within the upper Werii Slate or between the Werii Slate and Assem Limestone as a paraconformity. The abrupt Werii Slate-Assem Limestone contact can be followed laterally in the field as well as traced (.13 km) in satellite imagery.

Upper prospective glaciogenic unit Matheos Formation Diamictite Member, Negash Synclinorium. The

Matheos Formation Diamictite Member is mapped only within the folded core of the Negash Synclinorium. Within the Matheos Formation, it overlies a distinctive but gradational transition from black limestone (Black Limestone Member) to noncalcareous slate (Transitional Member). Although a direct contact has not been mapped, the oldest overlying sediments are Ordovician Enticho Sandstone, the base of which marks the AAP. Dugub Formation, Shiraro Area. The Dugub Formation in the western Shiraro Area (Fig. 21.1) occupies a grossly similar stratigraphic position within the Tambien Group that could be equivalent to the Matheos Diamictite, but boundary relations for this low-lying unit are poorly known.

Chemostratigraphy Chemostratigraphic data for Tambien Group exposures in Ethiopia (Alene et al. 1999, 2006; Miller et al. 2003; Sifeta et al. 2005) and likely equivalents in Eritrea (Beyth et al. 2003) have become available in only the last decade (Fig. 21.3). For the most part, chemostratigraphic surveys have been limited in terms of types of analyses performed and stratigraphic sampling frequency within a given lithological unit, with fewer than 30 sample intervals considered in any one study (Fig. 21.3b). Miller et al. (2009) substantially expanded the regional chemostratigraphic database, assessing 100 carbonate and 30 slate intervals from the Shiraro, Samre, Mai Kenetal and Negash regions. The integrated Tambien Group record now constitutes a significant chemostratigraphic data set for evaluating Cryogenian marine secular variations (including 71 stratigraphic intervals with 87Sr/86Sr measurements on least-altered units), with the Mai Kenetal and Negash synclinoria

data comprising most (c. 80%) analyses (Fig. 21.3). This compilation reveals significant chemostratigraphic trends, from which regional correlations are proposed (Fig. 21.2b, discussion). Carbonate d13C compositions for the Tambien Group range between –8 and þ8‰ and can be characterized as fluctuating repeatedly upsection between positive and negative polarity (Fig. 21.3). Low-grade Bizen Domain stromatolitic carbonates in SE Eritrea, considered to be Tambien Group equivalents (Beyth et al. 2003), are among the most negative d13Ccarb values reported (range, – 7.2 to –4.8‰; 3 cal, 5 dol) but sample preservation was not critically assessed. Highest d13Ccarb compositions, with maxima near þ7‰, derive from the Mai Kenetal Limestone and Matheos Formation Black Limestone Member. The negative d13Ccarb compositions of the Assem Limestone (c. –1 to – 4‰) are similar to those comprising the lower negative d13Ccarb interval in the Negash Synclinorium (c. – 1 to – 3‰) (despite their contrasting mineralogies), as well as negative d13Ccarb intervals mapped as Assem Limestone in the Chemit ( –4.4 to –3.0‰) and Tsedia (– 4.5 to –0.7‰) synclinoria (Alene et al. 1999, 2006; Miller et al. 2009). Bizen Domain stromatolitic dolomites with negative d13Ccarb (Beyth et al. 2003) may correlate with the negative d13Ccarb interval of the Didikama Formation. The Negash carbonate sequence is so far unique in recording two negative d13Ccarb excursions. Carbonate d18O compositions range between 0 and –16‰, with dolomite typically enriched (by 2– 5‰) relative to stratigraphically proximal (or prospectively equivalent) limestone. Although substantial scatter is apparent, median compositions are mainly between –4 and –10‰, similar to other Cryogenian datasets (Fig. 21.3; Jacobsen & Kaufman 1999; Robb et al. 2004; Halverson et al. 2005, supplementary information). Values more negative than –11‰ may be below the modal range of Cryogenian samples (Kaufman et al. 1993), and could be particularly altered. The Mai Kenetal Limestone and Matheos Formation Black Limestone (units considered best preserved for d13Ccarb and 87Sr/86Sr, see discussion) have median compositions between –7.6 and –3.5‰, with Matheos Black Limestones underlying the transition to diamictite deposition most enriched. Carbon-isotopic compositions of organic matter (d13CTOC) in the Tambien Group, principally from the Mai Kenetal and Negash synclinoria, are mainly in the range of –20 to –30‰, with extremely light values (e.g. , –40‰) suggesting contributions from methanogenic biomass (Fig. 21.3). Both Mai Kenetal and Negash successions show similar stratigraphic enrichment trends in d13CTOC with pre-diamictite values in uppermost Matheos Formation limestones about 3‰ heavier than uppermost Mai Kenetal Limestone exposures. The Tambien Group has a large compilation of 87Sr/86Sr compositions in studies by Miller et al. (2003, n ¼ 9; 2009, n ¼ 71) and Alene et al. (2006, n ¼ 5). Results obtained from the same units are complementary among the studies despite contrasts in sample processing approaches. Least-altered 87Sr/86Sr compositions (normalized to SRM 987 ¼ 0.710240) are mainly between 0.7055 and 0.7068 (Fig. 21.3). The Mai Kenetal succession shows a stratigraphic enrichment trend that compares with a more extensive enrichment trend in the Negash succession. The lower negative d13Ccarb interval at Negash has 87Sr/86Sr compositions (avg: 0.706178 + 40, n ¼ 2) within error of those for Assem Limestone (avg: 0.706175 + 14, n ¼ 14) (Miller et al. 2009). Negative d13Ccarb compositions in the Assem Limestone may initiate with 87Sr/86Sr compositions near 0.70597 (Miller et al. 2009). Tambien Group carbonate units have Sr compositions ranging up to about 4000 ppm, with pronounced stratigraphic enrichment evident in both the Mai Kenetal and Negash synclinoria. Sr compositions less than 1000 ppm typify limestone and dolostone in the Assem Limestone and Didikama Formation (also Didikama Formation in Shiraro and Samre areas), whereas much higher concentrations averaging 2000– 3000 ppm occur in the Mai Kenetal

(a)

THE TAMBIEN GROUP, NORTHERN ETHIOPIA

(b)

269

Fig. 21.3. Summary of chemostratigraphic data sets for the Tambien Group. (a) Statistical summaries of chemostratigraphic (d13Ccarb, d18Ocarb, d13CTOC, 87Sr/86Sr, Sr concentration and chemical weathering index) data for Tambien Group lithostratigraphic units by locality and in stratigraphic order. Letters A –D mark stratigraphic units of alternating d13Ccarb polarity; corresponding circled data sets indicate possible correlations between the Mai Kenetal and Negash synclinoria (black vertical bars) and other Tambien Group localities (see discussion); line types of circled data sets clarify vertical correlations. Black triangles indicate prospective glaciogenic intervals. Data compiled from Miller et al. (2009) and Alene et al. (2006, vein calcite samples excluded) are restricted to samples considered least altered for 87Sr/86Sr. Chemical Index of Weathering (CIA or PIA) data exclude samples containing lithic fragments or carbonate minerals. (b) Distribution of chemostratigraphic analyses for each study used to compile the upper graph statistical summaries. The histogram (top-to-bottom) bar order corresponds to analysis type, as listed from left to right under the ‘Total Analyses/Least Altered Analyses’ compilation.

N. R. MILLER ET AL.

Limestone and Matheos Formation Black Limestone Member. Intermediate Sr compositions (avg: 1198 + 91 ppm, n ¼ 2, Miller et al. 2009) occur in the Tsedia Slate in the Mai Kenetal Synclinorium. Similar transitional compositions are missing in the Negash Synclinorium and instead there is a sudden stratigraphic jump in Sr concentrations between the Didikama Formation dolomite and directly overlying Matheos Formation Black Limestone. A range of Sr concentrations have been measured for matrix carbonate (980 + 712 ppm, n ¼ 11; Miller et al. 2003) and carbonate clasts in the Matheos Formation Diamictite Member (512 ppm, oolite cobble; Miller et al. 2009). The petrological and chemical transition from largely metavolcanic and metasiliciclastic (slate) units in the Tsaliet Group and lower Tambien Group (Werri Slate) to predominant carbonate deposition in the higher Tambien Group (Assem Limestone and Didikama Formation) was investigated by Sifeta et al. (2005) in the Werii area east of the Mai Kenetal Synclinorium (Fig. 21.1). Metavolcanic rocks are sub-alkaline, with chemical fingerprints that are compatible with island arc (primitive or evolved) and/ or MORB settings. Lower Werri Slate compositions indicate

Palaeolatitude and palaeogeography There are no reliable palaeomagnetic constraints for Tambien Group deposition within the southern ANS. Strike –slip displacement and structural shortening associated with closure of the Mozambique Ocean and formation of the EAO constitute significant challenges for early Cryogenian palaeogeographic reconstructions within the ANS. Palaeogeographic reconstructions spanning the 775–660 Ma interval (Fig. 21.4a–c) generally place the ANS (as inferred oceanic arcs) within the greater Mozambique Ocean between subsequently flanking Gondwana fragments (e.g. between India and

(b)

Am Postulated subduction zone

WA

WA

800 Ma (d)

aste r

La

RP o

n ilia

az

Br

WA Schematic continental collision

630 Ma

Collins & Pisarevsky 2005

West Gondwana

Mz-Cz Orogen

Ada m

Sah

750 Ma

After Meert 2003; Meert & Torsvik 2003

ic

Co SF

o an

ili

az

Br

cif

Az

EAO

Ada m

aste r

Adola

RP

Brasiliano Ocean

RP

Pa

e

mbiqu

Moza

Am

SF

SF

Ka

a

Co

Co B

In

ific

-M

L Ka

Pac

Az

EA

Mozambique Ocean

Ka

e

iqu

mb

za Mo

Au

In

Ma

M

S

SC

(c) Au -

Au

In

Continents with c. 750 Ma palaeomagnetic data

La

(a)

derivations from comparable juvenile crustal sources. Associated chemical weathering indices (Fig. 21.3), which proxy the degree of weathering from fresh rock (50) to full clay conversion (100), are generally between 70 and 90 for the Chemical Index of Alteration (CIA, Nesbitt & Young 1982) and .75 for the Plagioclase Index of Alteration (PIA, Fedo et al. 1995).

Am

270

+ +

AAP

ANS

+

+

+

+

Indian Shield

Australia

+

+

+

N

+

+

+

S

+

+

+

+

+

+ +

+

East Antarctic Shield

+

W.

+

+

+

+

+

+

+

+ + +

+

+

+

+

+

Kalahari Craton

SL

+

Mad

+

+

+

RP

East Gondwana

+

+

+

+

+

+

+

+

+

+

+

SF

~ Antarc

+ +

Amazonian Craton

Congo Craton

tic ~ E AO

+

AAP

Saharan Metacraton

+

+

West African Cration

Neoproterozoic Orogen

+

+

+

Pz Orogen

Tambien Group ( ) c. 550 Ma

?

Pz-Mz Orogen

+

Subduction Zone Southern limit of Early Palaeozoic sandstone marking African-Arabian peneplain (AAP) Modified: Meert & Lieberman 2008

1000 km

Fig. 21.4. Global palaeogeographic reconstructions for (a) 800 Ma, (b) 750 Ma, (c) 630 Ma and (d) 550 Ma, showing the inferred palaeogeographic location and structural-tectonic setting of the ANS/Tambien Group (black star) during the Cryogenian Period. Prospective glaciogenic intervals in the Tambien Group were likely deposited during the earlier Cryogenian (.630 Ma) prior to the emergence of the East African Orogen (EAO). (a, b) Rifting and break-up of Rodinia (c. 900–750 Ma) was associated with sea-floor spreading, arc and back-arc basin formation, and terrane accretion in the Mozambique Ocean. (c) Accommodation space in the southern ANS basin likely inverted by 630 Ma in response to closure of the Mozambique Ocean between converging elements of West and East Gondwana and emergence of the EAO. Occurrence of undeformed post-orogenic intrusives of this age or older that puncture the deformed Neoproterozoic supracrustal sequence in Ethiopia and Eritrea, suggests that the Tambien Group at 630 Ma was likely deformed and uplifted within the EAO. (d) Location of the Tambien Group at c. 550 Ma within the context of Gondwana amalgamation and emergence of the Antarctic-EAO. The modern southern limit of Early Palaeozoic sandstone (superimposed from Avigad et al. 2005) documents the minimal extent of the Afro-Arabian Peneplain (AAP) and extent of regional uplift and Cambro-Ordovician erosion associated with the emergent EAO (Avigad et al. 2005). Reconstructions modified from (a) Meert 2003; Meert & Torskvik 2003; (b, c) Collins & Pisarevsky 2005; (d) Meert & Lieberman 2008; Grey et al. 2008. Abbreviations for cratons and continents: Am, Amazonia; AuMa, Australia/Mawson Block; Az, Azania; B, Baltica; Co, Congo/Tanzania/Bangweulu Block; EA, East Antarctica; Ka, Kalahari Block; La/L, Laurentia; In, India; Mad, Madigascar; RP, Rio de la Plata; Sah, Saharan Metacraton; SC, South China; SF, Sao Francisco; S, Siberia; WA, West Africa; Adola, Adamastor, Braziliano, Mozambique, Pacific, oceanic basins; Mz, Mesozoic, Cz, Cenozoic; Pz, Palaeozoic.

500 550 600

PHZ Ediacaran C

THE TAMBIEN GROUP, NORTHERN ETHIOPIA

Enticho Ss (Ord) detrital zircon 2 2 age distribution Eritrea Sibta Shire 1

8 9

750 800 850 900

NEOPROTEROZOIC Cryogenian

700

Ton.

Age (Ma)

650

2

Mai Mareb R. Kenetal 1

Dugub Fm (Shiraro)

7

Matheos Diamictite (Negash)

Zircons below 1st dolomite beds

1

E. Sudan

6

6

Hawzien Negash Youngest detrital zircons in upper prospective glacial intervals

E. Sudan

5 4 Eritrea 3 9 10

Post-orogenic plutons

2

Azeho

2

Deset

1 2

Chila

1

2

Hawzien Rama

1

Madahne Alem (Negash)

Tsaliet Group syn-tectonic granites and metavolcanics Relative W-to-E position (Tigre) STUDY

1. Avigad et al. (2007) 2. Tadesse et al. (2000) 3. Tadesse et al. (1999) 4. Teklay (1997) 5. Teklay et al. (2002)

6. Miller et al. (2003) 7. Asrat et al. (2004) 8. Teklay et al. (2001) 9. Kröner et al. (1991) 10. Teklay et al. (2003)

DATE TYPE/METHOD Pb-Pb zircon Rb-Sr Sm-Nd CHIME zircon

Congo cratons). For example, Collins & Pisarevsky (2005) place ophiolite-bearing strata of the Adola Belt (southern Ethiopia) outboard of the Congo Craton (Fig. 21.4b). As the Adola Belt is generally considered to mark the southern extent of the EAO, it is reasonable to infer a similar outboard location for the Tambien Group. The earliest reliable regional palaeomagnetic data derive from the time of Gondwana amalgamation (Fig. 21.4d); late Cryogenian (593 + 15 Ma) Dokhan volcanics of Egypt (Davies et al. 1980; Wilde & Youssef 2000), interpreted to have a subtropical palaeolatitude (P-lat: 20.6 + 5.08, A95: 10.08, Trindade & Macouin 2007; but see Nairn et al. 1987). The Dokhan palaeopole has been widely used in subsequent palaeogeographic reconstructions (e.g. Meert 2003; Meert & Torsvik 2003; Macouin et al. 2004; Trindade & Macouin 2007). Low subtropical palaeolatitudes (9 –138) were also reported for Huqf Supergroup units in Oman thought to represent deposition during and after an upper Cryogenian (Fiq Formation) glaciation before 544 Ma (Kempf et al. 2000; Kilner et al. 2005), when Oman was likely amalgamating with the ANS (detrital zircon data in Rieu et al. 2007). Subsequent work (Rieu et al. 2006; Allen 2007, and references therein) demonstrates that Mirbat Group localities in both palaeomagnetic studies correspond to an older Cryogenian (Ghubrah-Ayn Formations) glaciation, constrained radiometrically to be ,722 Ma and ongoing at 711.8 + 1.6 Ma. The fact that palaeolatitudes for Oman units associated with both glacial intervals (between c. 722 and 544 Ma) are similarly low, and also close to published c. 550 Ma palaeopoles from Gondwana, has yet to be fully reconciled with the available geologic data (Allen 2007). However, even if low palaeolatitudes are confirmed, Oman had yet to accrete with the ANS at the time of the older Cryogenian (Ghubrah-Ayn Formations) glaciation, and therefore these constraints could not be precisely extended to the Tambien Group. To the extent that the ANS occupied a gross latitudinal range similar to the Congo-Sa`o Francisco and/or East-Sahara cratons during the early Cryogenian (c. 750 Ma), as implied by various palaeogeographic reconstructions (e.g. Fig. 21.4a,b; Trindade & Macouin 2007, fig. 3a), the Tambien Group may have occupied low to intermediate latitudes (,458). Occurrence of evaporite pseudomorphs in lower Didikama Formation dolomite (interpreted as ,774.7 + 4.8 Ma, Miller et al. 2009) may support a subtropical palaeolatitude, as palaeomagnetically constrained evaporite basins as old as 2.25 Ga have statistically significant volume-weighted concentrations in the palaeo-subtropics (Evans 2006).

U-Pb zircon Zircon evaporation Youngest detrital zircon (SHRIMP U-Pb)

Tambien Group Carbonate Deposition BSS

271

Fig. 21.5. Radiometric age constraints bearing on the age of the Tambien Group within the magmatic evolution of the southern ANS and EAO. Tsaliet arc magmatic products and ‘Mareb’ post-orogenic pluton ages arranged according to their relative west-to-east positions in Tigre. Magmatic equivalents for localities in Eritrea and Sudan (open squares) are shown to the left. The Enticho Sandstone detrital zircon age spectrum (Avigad et al. 2007) shown at the left of the figure derives from the Enticho area as shown in Figure 21.1. The shaded horizontal bar shows the age range (827+6 to 770 + 7 Ma) for the Bitter Springs stage (BSS) negative d13Ccarb excursion of Australia (Halverson et al. 2005, 2007; and references therein). Note that this range overlaps with the main phase of Tsaliet syn-tectonic arc magmatism; see text for discussion.

Geochronological constraints The age of Tambien Group deposition is bracketed by magmatism associated with formation of the underlying Nakfa basement complex (Tsaliet Group) and orogenic thickening and differentiation related to maturation of the EAO (‘Mareb’ granitoids). Affiliated magmatism is widely dated throughout the ANS (e.g. Meert 2003; Johnson & Kattan 2007, and references therein). Our geochronological review is restricted to southern ANS studies in eastern Sudan (Kro¨ner et al. 1991), Eritrea (Teklay 1997; Teklay et al. 2001, 2003; Andersson et al. 2006), northern Ethiopia (Tadesse et al. 1997, 2000; Miller et al. 2003; Asrat et al. 2004; Avigad et al. 2007) and western Ethiopia (Ayalew et al. 1990) closest to Tambien Group exposures and its prospective equivalents (Figs 21.1, 21.5, Supplementary Table 21.1). Detrital zircon age spectra from mature Ordovician sand (Enticho Sandstone) capping the AAP in Tigre are interpreted by Avigad et al. (2007) to reflect the magmatic history of the regional Neoproterozoic–Ordovician basement complex, as it was uplifted and eroded within the EAO. The Enticho Sandstone zircon age distribution (Fig. 21.5) reinforces that Neoproterozoic magmatism occurred in two main episodes: (i) arc-related calc-alkaline magmatism c. 850–740 Ma (c. 780 Ma peak), largely related to petrogenesis of the Tsaliet Group, and (ii) post-orogenic magmatism c. 660– 580 Ma (c. 630 Ma peak) after Tambien Group deposition. Additional geochronological constraints elaborated below and shown in Figure 21.5 further suggest that the lower and upper prospective glacial intervals within the Tambien Group were deposited during the lull between these two magmatic episodes, with 775 to .660 Ma as a plausible depositional window (see Discussion).

Pre-Tambien Group magmatism Deformed (syn-tectonic) plutons and metavolcanic rocks of the Tsaliet Group (lower calc-alkaline magmatic components of the Nakfa terrane) have been dated by a number of techniques (e.g. Pb/Pb zircon, Rb –Sr, Sm –Nd, CHIME zircon, U –Pb zircon, zircon evaporation) with varying associated uncertainties (Supplementary Table 21.1 – Geochronological Data). These ages provide an indirect lower age limit for Tambien Group carbonate deposition. The oldest rocks in this region include an 862 + 6 Ma granite clast within the Gulgula Group (Fig. 21.1, Teklay et al. 2003), 854 + 3 Ma deformed volcanic rocks and

272

N. R. MILLER ET AL.

811 + 11 Ma granite from Eritrea (Teklay 1997). In the Axum area (western Tigre), syn-tectonic pluton ages range between 806 + 21 Ma and 756 + 33 Ma (Tadesse et al. 2000). A 784 + 14 Ma syn-tectonic granitoid occurs SSW of Hauzien (Fig. 21.1; Avigad et al. 2007). A conformable felsic interval shortly (c. 17 m) below initial dolomite beds of the Didikama Formation, in the western limb of the Negash Synclinorium (Madahne Alem), produced a zircon U – Pb date of 774.7 + 4.8 Ma (Avigad et al. 2007; Figs 21.4 and 21.5).

Post-Tambien Group magmatism Post-orogenic ‘Mareb’ granitoid plutons (Fig. 21.1) penetrate the older deformed Neoproterozoic complex (including the Tambien Group) throughout northern Ethiopia and Eritrea. In Tigre, ages for these commonly rounded and undeformed intrusives range from 613.4 + 0.9 Ma (Mai Kenetal Granite, Avigad et al. 2007) to 545 + 24 Ma (Mareb Granite, Tadesse 1997). The Negash Pluton, just west of the Negash syncline, is c. 606 Ma (606.0 + 0.9 Ma, Miller et al. 2003; 607 + 7 Ma, Asrat et al. 2004). A small granitoid body puncturing the western Negash limb (lower Didikama Formation) is interpreted to have a comparable age, but did not render suitable zircons for geochronology (Miller et al. 2009). Ages of post-orogenic intrusives are somewhat older in Eritrea (628 + 4 Ma, 622 + 1 Ma; Teklay et al. 2001) and Sudan (SE of Tokar: 652 + 14 Ma, Kro¨ner et al. 1991). The good agreement between individually dated syn-tectonic and postorogenic magmatic products with corresponding modal peaks in the Enticho Sandstone zircon age distribution (Fig. 21.5), and the lack of detrital zircons younger than c. 739.2 + 6.3 Ma in either of the prospective glaciogenic units capping the Tambien Group (Dugub Fm, Matheos Diamictite) are strong evidence that the Tambien Group is older than EAO post-tectonic granitic rocks (Avigad et al. 2007). In the vicinity of Mai Kenetal and Negash, this constraint is c. 610 Ma, but its upper age is likely much older considering that regional deformation and crustal thickening must have preceded the ‘Mareb’ intrusives. The oldest dated post-tectonic pluton in the region (652 + 14 Ma; Kro¨ner et al. 1991) and Enticho sandstone detrital zircon record (initiation of the second magmatic phase at c. 660 Ma; Avigad et al. 2007) suggest that the Tambien Group may have been deposited prior to c. 660 Ma (Fig. 21.5).

Discussion Regional chemostratigraphic context of the prospective glacial intervals The observation that many Tambien Group exposures in Tigre exhibit a similar succession of lithofacies that have comparable chemostratigraphic characteristics suggests that the Tambien Group preserves a regionally coherent chemostratigraphic framework (Beyth 1972; Alene et al. 2006; Miller et al. 2009). Tambien Group depositional history, including the depositional context of the prospective glaciogenic intervals, can be effectively considered relative to stratigraphic changes in the polarity of associated d13Ccarb. These relationships are best demonstrated in the Mai Kenetal and Negash synclinoria (Figs 21.2b, 21.3) but are consistent with the stratigraphic succession in other areas (e.g. Shiraro and Samre). The carbonate successions in the Mai Kenetal and Negash synclinoria both initiate above thick marine slate sequences; however, the nature of this transition to carbonate deposition differs in each locality. In Mai Kenetal the transition from Werii Slate to the Assem Limestone is a sharp conformable or paraconformable contact, the Assem Limestone beginning with and maintaining negative d13Ccarb compositions. In Negash, faulting in the lower portion of each synclinorial limb interrupts the stratigraphic

continuity of the slate-to-carbonate transition, leading to negative d13Ccarb compositions above the lower prospective glacial interval (GCI) in the eastern limb and positive d13Ccarb compositions above variegated slate in the western limb (see Miller et al. 2009 for details). The available regional data suggest that Tambien Group carbonate deposition began with positive d13Ccarb compositions (interval A). For Negash this interval (c. 250 m thick) in the western limb may crop out within an east-verging thrust block, whereas the interval in the eastern limb may be faulted out or yet unrecognized below the GCI. The nature and continuity of the transition from Tsaliet Group agglomerates to lower Didikama Formation slate and dolomite remains poorly understood and a systematic stratigraphic transition from lower positive (interval A) to higher negative (interval B) d13Ccarb compositions has yet to be documented in a conformable sequence. Based on similar lithostratigraphic position and character (slate with negligible carbonate content and advanced chemical weathering indices; Fig. 21.3), the informal lower slate member of the Didikama Formation could be equivalent to the Werii Slate and these units are tentatively included in polarity interval A. The lower prospective glacial interval in both sequences underlies a lower carbonate interval with negative d13Ccarb (interval B) that is in turn overlain by a distinctive black limestone succession with the most enriched d13Ccarb compositions of the entire Tambien Group (interval C; Fig. 21.2b). Polarity interval B and C units in both localities have highly complementary d13CTOC, 87Sr/86Sr and Sr compositions that support their regional correlation (Fig. 21.3). The lithological and chemostratigraphic transition from interval B to C appears to be continuous in Mai Kenetal, whereas this transition in Negash is abrupt and likely associated with an unconformity (Miller et al. 2009). Above the black limestone interval in the Negash Synclinorium is the second ‘upper’ negative d13Ccarb excursion (interval D) associated with a conformable transition to the upper prospective glacial unit (Matheos Diamictite).

Composite chemostratigraphic reference section Figure 21.6 presents a composite chemostratigraphic reference section for the Tambien Group based on the correlations between the Mai Kenetal and Negash synclinoria shown in Figure 21.2b and the reasoning above. The Mai Kenetal section above the Werii Slate (Assem Limestone, Tsedia Slate and Mai Kenetal Limestone) essentially replaces the Negash sequence above the basal positive d13Ccarb interval of the lower Didikama Formation and below (or within) the lower portion of the Matheos Formation Black Limestone Member. This composite effectively replaces dolomite lithologies below a probable unconformity (Negash) with a continuous and conformable succession of limestone lithologies; these seem to show gradational chemostratigraphic trends into highly enriched values that correlate well with the lower Matheos Formation Black Limestone Member (Miller et al. 2009). The abrupt transition from positive (polarity interval A) to negative (polarity interval B) d13Ccarb compositions may be a structural contact, but we suggest that the relative stratigraphic order of d13Ccarb polarity is maintained in the composite. The 87Sr/86Sr evolution of Cryogenian seawater (albeit poorly delineated) involves a general trend of increasing 87Sr/86Sr (Halverson et al. 2007). That least altered samples from polarity interval A dolomite beds have lowest 87 Sr/86Sr compositions is consistent with early Cryogenian deposition within the Tambien Group carbonate platform. The evolution of Tambien Group depositional environments is now considered based on the composite reference section (Fig. 21.6).

Palaeoenvironmental changes The Tambien Group involves mainly a transition from weathered Tsaliet arc detritus (lower slate intervals; Sifeta et al. 2005) to

THE TAMBIEN GROUP, NORTHERN ETHIOPIA

d13Ccarb

800

600

0

0

oolite cobble

2000

TOC 4000

0.0

0.1

1.0

Interval devoid of bedded carbonate

Bilato

(b) MTS CC ?

?

Slate >Dol Lower Slate

200

8

(c)

2 400

Sr (ppm)

4

Logmiti

Tsedia Slate

1000

0

3

Assem LS

1200

Tambien Group

Stratigraphic Height (m)

1400

-4

(d)

MK LS

1600

Tr Diam.

Matheos Fm

1800

Black LS

2000

-8

matrix

Enticho Sandstone

273

(a) 774.7±4.7 Ma

Extreme weathering indices (PIA: 92-99) 1

Tsaliet Gp

0.705

(Metavolcanics)

Sheet crack cements

Evaporite pseudomorphs

0.706 87Sr/86Sr

Stromatolites

0.707

-15

-10

-5

0

d Ocarb 18

Internal slump

MTS Molar tooth structures

Probable dropstones

-45

-35

-25

d CTOC 13

CC Cap carbonate-like textures

Fig. 21.6. Composite chemostratigraphic reference sequence for the Tambien Group based on the Mai Kenetal and Negash synclinoria. Segments 1 –3 denote the basis for constructing this composite sequence (cf. Fig. 21.2b) and the black triangles denote the prospective glacial intervals. The composite sequence is differentiable by four changes (a –d, separated by horizontal dashed lines) in the polarity of associated d13Ccarb and consists of limestone units except for segment 1 (dolomite comprising the lower portion of the Didikama Formation in the western limb of Negash). The lower horizontal shaded interval indicates the stratigraphic range of slates having extreme chemical weathering indices (PIA, plagioclase index of alteration of Fedo et al. 1995). The upper horizontal shaded interval corresponds to the upper portion of the Matheos Formation lacking bedded carbonate; the darker grey d13Ccarb field in the Matheos Diamictite shows the range for diamictite matrix carbonate (Miller et al. 2003), whereas the uppermost data point corresponds to an oolite cobble (Miller et al. 2009). Abbreviations: Ls, Limestone; Dol, dolomite; Tr, transitional, Diam-Matheos diamictite; MK, Mai Kenetal.

predominant carbonate deposition in a marine arc-accretion platform setting. The lack of obvious ash beds, volcaniclastic intervals and syn-tectonic intrusive rocks in the main carbonate succession (polarity interval B and higher) suggests deposition during a magmatically quiet interval (Avigad et al. 2007), and the upward waning of slate deposition in favour of carbonate may represent a phasing out of arc magmatic activity in the region. Chemical weathering indices of slates in polarity interval A suggest an overall stratigraphic trend of increasing chemical weathering of source areas to extreme levels, which lessened somewhat prior to deposition of polarity interval B (Fig. 21.6). The association of extreme weathering indices with increasing 87Sr/86Sr is consistent with an interval of rapid chemical weathering of arc terranes in the southern ANS portion of the Mozambique Ocean. The fine grained nature of slates and absence of shallow water indicators suggests a moderately deep-water depositional setting. Above the lower slate intervals, the Tambien Group carbonate pile is broadly differentiable between lower carbonate units (polarity intervals A –B) suggestive of shallow marine depositional settings and upper carbonate units (polarity interval C) suggestive of deeper marine environments. Based on the occurrence of evaporite pseudomorphs, sheetcrack cements, microbialaminates and stromatolites (Miller et al. 2009, fig. S5), carbonates

associated with polarity interval A are interpreted to have been deposited in high alkalinity tidal flat and intertidal settings. Moderate energy shallow intertidal or subtidal settings are interpreted for the Assem Limestone and upper Didikama Formation (polarity interval B) on the basis of prominent domal and interdigitate stromatolites, coarse rip-ups, and cross-bedded grainstones (Miller et al. 2009, figs S2, S3). The base of the Assem Limestone occurs abruptly above the Werii Slate as a metre-thick dolomite bed, with sedimentary features similar to those described for well-studied transgressive cap-carbonate sequences. Lower carbonate units of the Tambien Group (polarity intervals A –B) have low TOC contents with highly variable d13CTOC, including light compositions, which together with the sedimentary characteristics are consistent with nearshore environments (Miller et al. 2009, figs 7B,C and 9). Deeper subtidal environments with lower (but still variable) energy levels and ubiquitous micrite are interpreted for upper carbonate units of the Tambien Group (polarity interval C). The transition appears to begin at the base of the Tsedia Slate in Mai Kenetal (equivalent to the proposed unconformity between the Didikama and Matheos Formations in Negash), and this boundary could mark the base of a transgressive system following regression. Both the Mai Kenetal Limestone and the Matheos

274

N. R. MILLER ET AL.

Formation Black Limestone are distinctive dark grey to black limestone units characterized by fine horizontal layering and high lateral continuity (Miller et al. 2009, figs S3 and S6). The lack of stromatolites and high-energy sedimentary structures (coarsely graded beds, cross-bedding) but occurrence of dark intraclastic intervals and intraformational slumping may indicate slope deposition below storm wave base. These upper carbonate units are relatively enriched in TOC that has less variable d13CTOC, consistent with more open marine environments (Miller et al. 2009, fig. 7B,C). Upper carbonate units are also most enriched in d13Ccarb and d13CTOC, particularly in the upper part of the Matheos Formation Black Limestone Member. This mutual stratigraphic enrichment pattern, together with distinct stratigraphic increases in 87Sr/86Sr and Sr concentrations, is consistent with high rates of organic matter burial and related fractionation of the upper (photic) marine d13CDIC pool in the lead up to Matheos Diamictite deposition. Although d18Ocarb is the marine proxy most likely to undergo post-depositional alteration, it is notable that the intervals with most enriched compositions (consistent with cryogenic sequestering of d16O) occur in association with the two prospective glacial intervals (Fig. 21.6). The possibility that d18O records fluctuations in ice volume is more likely for the upper prospective glacial interval because the lower prospective glacial interval is underlain by dolomite, which is typically enriched by 2 –4‰ over coeval limestone (Jaffre´s et al. 2007, and references therein).

Timing of the prospective glacial intervals Integration of available geochronological information suggest c. 775–660 Ma as a plausible window for Tambien Group deposition bracketing the two prospective glacial intervals. This depositional window is not definitive, but corresponds to the lull between Tsaliet arc magmatism and later magmatism associated with maturation of the EAO, on the basis of regional stratigraphy and crosscutting relationships. Direct dates on magmatic products from each magmatic episode vary locally within Tigre but concur with the detrital zircon age distribution from the immediately overlying Ordovician Enticho Sandstone (Fig. 21.5). Somewhat older arc magmatism is indicated in Eritrea and eastern Sudan. The 774.7 + 4.8 Ma date obtained from zircons recovered from variegated (tuffaceous?) slate, c. 17 m below the first positive d13Ccarb (polarity interval A) dolomite beds of the Didikama Formation, is a possible maximum age constraint for significant (bedded) carbonate deposition in the Tambien Group (Fig. 21.5). The interval was originally interpreted as either a volcaniclastic interval or sill injected into a volcaniclastic interval. Subsequent petrographic analysis of this interval revealed probable sedimentary textures, including rounded quartz grains and a large (3 cm) rounded quartzose clast. If this date holds, the lower prospective glacial interval occurring at least c. 250 m higher is likely to be appreciably younger. The upper prospective glacial interval is older than deformation and later magmatism associated with the EAO. The oldest indication of EAO magmatic activity in the southern ANS is from a 652 + 14 Ma post-orogenic pluton in eastern Sudan (southern Red Sea Hills, c. 500 km NNW of Negash; Kro¨ner et al. 1991), whereas locally dated ‘Mareb’ granitoids in Tigre are c. 610 Ma (avg: 609.8 + 3.8 Ma, Mai Kenetal, Negash, Hauzien plutons). The dated Red Sea Hills locality in eastern Sudan occurs c. 50 km NW of the area of prospective glacial deposits (pebbly mudstones with flattened clasts) in northern Eritrea (Fig. 21.1) (Cecioni 1981). If the pebbly mudstones are glaciogenic and correlative with the upper prospective glacial interval of the Tambien Group, the Red Sea Hills age offers an important minimum age constraint for the Tambien Group. The most precise age estimates on older (pre-pebbly mudstone) associated arc metavolcanic and plutonic rocks in the Red Sea Hills area, obtained by single zircon 207Pb/206Pb methods (c. 870–

827 Ma), are substantially older than ages determined by Rb –Sr whole rock methods (c. 770– 670 Ma). The younger ages have been interpreted to reflect isotopic disturbance due to regional deformation of the arc complex, possibly as it accreted with the African continent (Kro¨ner et al. 1991; Saharan Metacraton?). Structural deformation of the orogen may thus have begun before 670 Ma. As this deformation was post-depositional and preceded the first post-orogenic intrusives (652 + 14 Ma) by tens of millions of years, the Tambien Group minimum age, and the age of the upper prospective glacial interval, could be appreciably older than 652 + 14 Ma.

Evidence for glacial influence on sedimentation The lower prospective glacial interval of the Tambien Group, as the GCI in Negash and a possibly equivalent interval or paraconformity at the top of the Werri Slate is most controversial. Direct sedimentologic evidence of syn-glacial deposition (i.e. definitive dropstones, striated clasts) is so far lacking, with the possible exception of matrix supported horizons in the GCI (Beyth 1972). On the other hand, both of these intervals are overlain by negative d13Ccarb carbonates with similar chemostratigraphic characteristics. Where the base of this carbonate interval is well exposed (Assem Limestone, Mai Kenetal) it exhibits features consistent with well-documented Cryogenian post-glacial cap carbonates. Alene et al. (2006) discounted a glacial association for the Assem Limestone, but did not document or sample the base of this unit above its abrupt contact with Werri Slate. They suggested the Assem Limestone may correspond to a non-glacial negative d13Ccarb interval possibly equivalent to the Bitter Springs stage (BSS) of Australia (as suggested by Halverson et al. (2005, 2007) for similar negative d13Ccarb intervals within early Cryogenian sections from NW Canada (Little Dal Group) and Svalbard (Akademikerbreen Group)). The age for the BSS in Australia requires interbasinal correlations that are difficult to confirm, and the age range has a corresponding large degree of uncertainty between 827 + 6 Ma (Gairdner Dyke Swarm) and 777 + 7 Ma (Boucat Volcanics) among various studies (Halverson et al. 2005, 2007). Although a correlation with the BSS is possible, the c. 827 –770 Ma age range would correspond to a time of active arc magmatism in the southern ANS (Fig. 21.5). The Assem Limestone and higher Tambien Group succession lacks evidence of active magmatism (ash beds/slate intervals/syn-tectonic intrusives) suggesting it was deposited after the c. 780 Ma acme of arc magmatism in the region. The lower Negash negative d13Ccarb interval occurs several hundred metres above the variegated slate interval dated at 774.7 + 4.8 Ma (zircon). This suggests that the lower negative d13Ccarb interval could be appreciably younger than c. 775 Ma. As there are definitively younger post-glacial cap carbonates (e.g. above the c. 755 Ma Kaigas Formation, southern Namibia, Gariep Belt, Hoffmann et al. 2006) we suggest that a lower prospective glacial association within the Tambien Group remains a viable interpretation. It is conceivable that the glaciogenic association could be indirect. For example, the cap-carbonate features and negative d13Ccarb excursion could be products of enhanced upwelling associated with global cooling and/or vigorous circulation associated with post-glacial climate change in a setting that did not earlier accumulate glacially transported clastic sediments. Both of these intervals in the Mai Kenetal and Negash synclinoria warrant additional study to evaluate glacial associations. Of the proposed upper glaciogenic intervals for the Tambien Group, the Matheos Formation Diamictite Member best exhibits sedimentological characteristics consistent with glacially influenced deposition. In addition to matrix supported polymict clasts, clasts that appear to preferentially deform underlying bedding are interpreted as ice-rafted debris. Some bullet-nosed

THE TAMBIEN GROUP, NORTHERN ETHIOPIA

and possibly striated clasts are consistent with mechanical abrasion during glacial entrainment (Kuhn et al. 1993). This interval is further significant because of its gradational contact with the underlying Matheos Black Limestone Member, which records a decline in d13Ccarb and possibly also 87Sr/86Sr, prior to diamictite deposition. If the latter is not the product of diagenetic alteration, the declining Sr-isotope compositions could signal a strong decrease in continental weathering input relative to oceanic hydrothermal input, as consistent with increasing ice cover. The possibly equivalent Dugub Formation in the Shiraro area occupies a gross stratigraphic position similar to the Matheos Diamictite as well as similar detrital zircon age distributions (Avigad et al. 2007), but the underlying stratigraphy is poorly known. Palaeocurrent proxies suggest a north-northeastern source area for clasts, which considering the prospective glacial intervals in northern and eastern Eritrea could be consistent with a regional phase of glaciation.

Tectonic and palaeogeographic setting The intra-oceanic setting of the Tambien Group contrasts distinctly from pericratonic rift margin settings that are often associated with Cryogenian glaciogenic units. Later Cryogenian intervals are unlikely to be directly recorded throughout much of the ANS because the emergence of the EAO (by 660 Ma) would likely have either destroyed accommodation space or uplifted related strata to levels planed by subsequent erosion. The EAO could have hosted alpine/continental glaciation, however, perhaps even instigating initial downcutting of the vast AAP (Fig. 21.4d; Stern et al. 2006). The US– Israel Binational Science Foundation (grant no. 2002337) supported this study. We thank K. Mehari, D. Ku¨ster and T. Tadesse for field expertise and assistance, A. Abraham (Chief Geologist, Geological Survey of Ethiopia) and Mekele University for logistical support during field seasons, and M. AbdelSalem for helpful feedback on geological aspects of the ANS. B. Schilman and A. Ayalon of the Geological Survey of Israel provided stable isotope expertise. The manuscript was improved by thoughtful reviews from P. Johnson, M. Pope and E. Arnaud. This represents a contribution of the IUGS- and UNESCOfunded IGCP (International Geoscience Programme) Project #512.

References Alene, M., Conti, A., Sacchi, R. & Zuppi, G. 1999. Stable isotope composition (13C and 18O) of Neoproterozoic limestones and dolomites from Tigre, North Ethiopia. Bollettino della Societa` Geologica Italiana, 118, 611– 615. Alene, M., Jenkin, G. R. T., Leng, M. J. & Darbyshire, D. P. 2006. The Tambien Group, Ethiopa: an early Cryogenian (ca. 800– 735 Ma) Neoproterozoic sequence in the Arabian– Nubian Shield. Precambrian Research, 149, 79 –89. Allen, P. A. 2007. The Huqf Supergroup of Oman: basin development and context for Neoproterozoic glaciation. Earth Science Reviews, 84, 139– 185. Andersson, U. B., Ghebreab, W. & Teklay, W. 2006. Crustal evolution and metamorphism in east-central Eritrea, south-east Arabian-Nubian Shield. Journal of African Earth Sciences, 44, 45– 65. Arkin, Y., Beyth, M., Dow, D. B., Levitte, D., Haile, T. & Hailu, T. 1971. Geological map of Mekele Sheet area ND 37-11 Tigre Province. Imperial Ethiopian Government, Ministry of Mines, Geological Survey of Ethiopia, scale 1:250,000. Asrat, A., Barbey, P., Ludden, J. N., Reisberg, L., Gleizes, G. & Ayalew, D. 2004. Petrology and isotope geochemistry of the Pan-African Negash Pluton, northern Ethiopia: mafic-felsic magma interactions during the construction of shallow-level calc-alkaline plutons. Journal of Petrology, 45, 1147–1179 Avigad, D., Sandler, A., Kolodner, K., Stern, R. J., McWilliams, M. O., Miller, N. & Beyth, M. 2005. Mass-production of Cambrian quartz-rich sandstone as a consequence of chemical weathering of

275

Pan-African orogens: implications for global environment. Earth and Planetary Science Letters, 240, 818–826. Avigad, D., Stern, R. J., Beyth, M., Miller, N. & McWilliams, M. 2007. Detrital zircon U– Pb geochronology of Cryogenian diamictites and Lower Palaeozoic sandstone in Ethiopia (Tigrai): age constraints on Neoproterozoic glaciation and crustal evolution of the southern Arabian– Nubian Shield. Precambrian Research, 154, 88– 106. Ayalew, T., Bell, K., Moore, J. M. & Parrish, R. R. 1990. U-Pb and Rb-Sr geochronology of the Western Ethiopian Shields. Geological Society of America Bulletin, 102, 1309– 1316. Beyth, M. 1972. The geology of central western Tigre, Ethiopia. PhD thesis, Bonn, University of Bonn. Beyth, M., Avigad, D., Wetzel, H. U., Matthews, A. & Berhe, S. M. 2003. Crustal exhumation and indications for snowball Earth in the East African Orogen: North Ethiopia and East Eritrea. Precambrian Research, 123, 187–201. Bibolini, A. 1920. Risultali preliminary delle osservazioni faite nel Nord-est della Colona Eritrea. Asmara. Bibolini, A. 1921. Sui conglomerati di Rore Babla e dei Monti Haggar in Colonia Eritrea. Bollettino della Societa` Geologica Italiana, 40, 169– 176. Bibolini, A. 1922. Contributions a l’e´tude de la ge´ologi de l’Afrique orientale Italienne. 13th International Geological Congress (1922, Brussels, Belgium), Title Comptes rendus de la XIIIe session, en Belgique, parts 1 –3, 797– 814. Burke, K. & Sengo¨r, C. 1986. Tectonic escape in the evolution of continental crust. In: Reflection seismology – The continental crust American Geophysical Union, Geodynamic series, 14, 41 –53. Cecioni, G. 1981. Precambrian pebbly mudstones in Eritrea, northeastern Ethiopia. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s PrePleistocene Glacial Record. Cambridge University Press, A24, 150. Collins, A. S. & Pisarevsky, S. A. 2005. Amalgamating eastern Gondwana: the evolution of the Circum-Indian Orogens. Earth-Science Reviews, 71, 229–270. Corsetti, F. A., Stewart, J. H. & Hagadorn, J. W. 2007. Neoproterozoic diamictite-cap carbonate succession and d13C chemostratigraphy from eastern Sonora, Mexico. Chemical Geology, 237, 129–142. Davies, J., Nairn, A. E. M. & Ressetar, R. 1980. The palaeomagnetism of certain late Precambrian and early Palaeozoic rocks from the Red Sea Hills, eastern desert, Egypt. Journal of Geophysical Research, 85, 3699–3710. De Souza Filho, C. R. & Drury, S. A. 1998. A Neoproterozoic suprasubduction terrane in northern Eritrea, NE Africa. Journal of the Geological Society, London, 155, 551– 566. Evans, D. 2006. Proterozoic low orbital obliquity and axial-dipolar geomagnetic field from evaporite palaeolatitudes. Nature, 444, 51 – 55. Fedo, C. M., Nesbitt, H. W. & Young, G. M. 1995. Unraveling the effects of potassium metasomatism in sedimentary rocks and palaeosols, with implication for palaeoweathering conditions and provenance. Geology, 23, 921– 924. Garland, C. R. 1980. Geology of the Adigrat area. Ministry of Mines Memoir No. 1, 51, Addis Ababa, Ethiopia. 1:250,000 map. Gray, D. R., Foster, D. A., Meert, J. G., Goscombe, B. D., Armstrong, R., Truow, R. A. J. & Passchier, C. W. 2008. A Damaran Perspective on the Assembly of Southwestern Gondwana. Geological Society, London, Special Publications, 294, 257–278. Greenwood, W. R., Hadley, D. G., Anderson, R. E., Fleck, R. J. & Schmidt, D. L. 1976. Late Proterozoic cratonization in southwestern Saudi Arabia. Philosophical Transactions of the Royal Society of London, 280, 517– 527. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Toward a Neoproterozoic composite carbonisotope record. Geological Society of America Bulletin, 117, 1181–1207. ¨ ., Maloof, A. C. & Bowring, S. A. 2007. Halverson, G. P., Duda´s, F. O Evolution of the 87Sr/86Sr composition of Neoproterozoic seawater. Palaeogeography, Palaeoclimatology, Palaeoecology, 256, 103– 129. Hoffman, P. F., Halverson, G. P., Domack, E. W., Husson, J. M., Higgins, J. A. & Schrag, D. P. 2007. Are basal Ediacaran (635 Ma) post-glacial ‘cap dolostones’ diachronous? Earth and Planetary Science Letters, 258, 114– 131.

276

N. R. MILLER ET AL.

Hoffmann, K.-H., Condon, D. J., Bowring, S. A., Prave, A. R. & Fallick, A. 2006. Lithostratigraphic, carbon (d13C) isotope and U –Pb zircon age constraints on early Neoproterozoic (ca. 755 Ma) glaciation in the Gariep Belt, southern Namibia, Snowball Earth Conference, 16 – 21 July 2006, Ascona, Switzerland [abstract], 51. Jacobsen, S. B. & Kaufman, A. J. 1999. The Sr, C and O isotopic evolution of Neoproterozoic seawater. Chemical Geology, 161, 37 –57. Jaffre´s, J. B. D., Shields, G. A. & Wallmann, K. 2007. The oxygen isotope evolution of seawater: a critical review of a long-standing controversy and an improved geological water cycle model for the past 3.4 billion years. Earth-Science Reviews, 83, 83– 122. Johnson, P. R. & Kattan, F. H. 2007. Geochronologic dataset for Precambrian rocks in the Arabian peninsula. Saudi Geological Survey Open-File Report SGS-OF-2007-3, 21, tables. Kaufman, A. J., Jacobsen, S. B. & Knoll, A. H. 1993. The Vendian record of C- and Sr-isotopic variations: implications for tectonics and paleoclimate. Earth and Planetary Science Letters, 120, 409– 430. Kaufman, A. J., Knoll, A. H. & Narbonne, G. M. 1997. Isotopes, ice ages, and terminal Proterozoic earth history. Proceedings of the National Academy of Sciences USA, 94, 6600– 6605. Kempf, O., Kellerhals, P., Lowrie, W. & Matter, A. 2000. Palaeomagnetic directions in late Precambrian glaciomarine sediments of the Mirbat Sandstone Formation, Oman. Earth and Planetary Science Letters, 175, 181– 190. Kilner, B., Conall, M. N. & Brasier, M. 2005. Low-latitude glaciation in the Neoproterozoic of Oman. Geology, 33, 413–416. Kro¨ner, A., Linnebacher, P., Stern, R. J., Reischmann, T., Manton, W. & Hussein, I. M. 1991. Evolution of the Pan-African island arc assemblages in the southern Red Sea Hills, Sudan, and in southwestern Arabia as exemplified by geochemistry and geochronology. Precambrian Research, 53, 99 –118. Kuhn, G., Melles, M., Ehrmann, W. U., Hambrey, M. J. & Schmiedl, G. 1993. Character of clasts in glaciomarine sediments as an indicator of transport and depositional processes, Weddell and Lazarev Seas, Antarctica. Journal of Sedimentary Petrology, 63, 477– 487. Macouin, M., Besse, J., Ader, M., Gilder, S., Yang, Z., Sun, Z. & Agrinier, P. 2004. Combined palaeomagnetic and isotopic data from the Doushantuo carbonates, South China: implications for the ‘snowball Earth’ hypothesis. Earth and Planetary Science Letters, 224, 387– 398 Meert, J. G. 2003. A synopsis of events related to the assembly of eastern Gondwana. Tectonophysics, 362, 1 –40. Meert, J. G. & Lieberman, B. S. 2008. The Neoproterozoic Assembly of Gondwana and its relationship to the Ediacaran – Cambrian radiation. Gondwana Research, 14, 5– 21. Meert, J. G. & Torsvik, T. H. 2003. The making and unmaking of a supercontinent: Rodinia revisited. Tectonophysics, 375, 261–288. Miller, N. R., Alene, M., Sacchi, R., Stern, R., Conti, A., Kro¨ner, A. & Zuppi, G. 2003. Significance of the Tambien Group (Tigre, N. Ethiopia) for snowball Earth events in the Arabian –Nubian Shield. Precambrian Research, 121, 263– 283. Miller, N. R., Stern, R. J., Avigad, D., Beyth, M. & Schilman, B. 2009. Neoproterozoic carbonate-slate sequences of the Tambien Group, N. Ethiopia (I): pre-‘Sturtian’ chemostratigraphy and regional correlation. Precambrian Research, 170, 129– 156. Nairn, A. E. M., Perry, T. A., Ressetar, R. & Rogers, S. 1987. A palaeomagnetic study of the Dokhan volcanic formation and younger granites, eastern desert of Egypt. Journal of African Earth Sciences, 6, 353–365. Nesbitt, H. W. & Young, G. M. 1982. Early Proterozoic climates and plate motions inferred from major element geochemistry of lutites. Nature, 299, 715–717. Rieu, R., Allen, P. A., Etienne, J. L., Cozzi, A. & Wiechert, U. 2006. A Neoproterozoic glacially influenced basin margin succession and ‘atypical’ cap carbonate associated with bedrock palaeovalleys, Mirbat area, southern Oman. Basin Research, 18, 471– 496. Rieu, R., Allen, P. A., Cozzi, A., Kosler, J. & Bussy, F. 2007. A composite stratigraphy for the Neoproterozoic Huqf Supergroup of Oman: integrating new litho-, chemo- and chronolstratigraphic data

of the Mirbat area, southern Oman. Journal of the Geological Society, London, 164, 997–1009. Robb, L. J., Knoll, A. H., Plumb, K. A., Shields, G. A., Strauss, H. & Veizer, J. 2004. The Precambrian: the Archean and Proterozoic Eons. In: Gradstein, F., Ogg, J. & Smith, A. G. (eds) A Geologic Time Scale 2004. Cambridge University Press, Cambridge, 129– 140. Sifeta, K., Roser, B. P. & Kimura, J. I. 2005. Geochemistry, provenance, and tectonic setting of Neoproterozoic metavolcanic and metasedimentary units, Werri area, Northern Ethiopia. Journal of African Earth Sciences, 41, 212–234. Stern, R. J. 1994. Arc assembly and continental collision in the Neoproterozoic East African Orogen: implications for the consolidation of Gondwanaland. Annual Reviews of Earth and Planetary Sciences, 22, 319– 351. Stern, R. J., Kro¨ner, A., Bender, R., Reischmann, T. & Dawoud, A. S. 1994. Precambrian basement around Wadi Halfa, Sudan: a new perspective on the evolution of the East Saharan Craton. Geologische Rundschau, 83, 564–577. Stern, R. J., Avigad, D., Miller, N. R. & Beyth, M. 2006. Geological Society of Africa Presidential Review: Evidence for the Snowball Earth Hypothesis in the Arabian –Nubian Shield and the East African Orogen. Journal of African Earth Sciences, 44, 1 –20. Tadesse, T. 1997. The Geology of Axum Area (ND 37-6). Ethiopian Institute of Geological Surveys, Addis Ababa (Memoir No. 9). Tadesse, T. 1999. Axum sheet geological map. Geological Survey of Ethiopia, Addis Ababa, Ethiopia. 1:250,000 map. Tadesse, T., Hoshino, M. & Sawada, Y. 1999. Geochemistry of lowgrade metavolcanic rocks from the Pan African of the Axum area, northern Ethiopia. Precambrian Research, 99, 101– 124. Tadesse, T., Hoshino, M., Suzuki, K. & Izumi, S. 2000. Sm– Nd, Rb– Sr and U– Pb zircon ages of syn and post-tectonic grantoids from the Axum area of northern Ethiopia. Journal of African Earth Sciences, 30, 313– 327. Teklay, M. 1997. Petrology, geochemistry and geochronology of Neoproterozoic magmatic arc rocks from Eritrea: implications for crustal evolution in the southern Nubian Shield. Department of Mines, Asmara, Eritrea, Memoir 1, 125. Teklay, M. 2006. Neoproterozoic arc –back-arc system analog to modern arc –back-arc systems: evidence from tholeiite – boninite association, serpentinite mudflows and across-arc geochemical trends in Eritrea, southern Arabian –Nubian shield. Precambrian Research, 15, 81 – 92. Teklay, M., Kro¨ner, A. & Metzger, K. 2001. Geochemistry, geochronology and isotope geology of Nakfa intrusive rocks, northern Eritrea: products of a tectonically thickened Neoproterozoic arc crust. Journal of African Earth Sciences, 33, 283–301. Teklay, M., Kroner, A. & Mezger, K. 2002. Enrichment from plume interaction in the generation of Neoproterozoic arc rocks in northern Eritrea: implications for crustal accretion in the southern ArabianNubian Shield. Chemical Geology, 184, 167– 184. Teklay, M., Haile, T., Kro¨ner, A., Asmerom, Y. & Watson, J. 2003. A back-arc palaeotectonic setting for the Augaro Neoproterozoic magmatic rocks of western Eritrea. Gondwana Research, 6, 629–640. Trindade, R. I. F. & Macouin, M. 2007. Palaeolatitude of glacial deposits and palaeogeography of Neoproterozoic ice ages. Comptes Rendus Geoscience, 339, 200–211. Verri, P. 1909. Contributo allo studio geografio della Colonia Eritrea. Bollettino della Societa` Geografica Italiana, 10, 251–320. Carta Geologica 1:1,500,000. Wilde, S. A. & Youssef, K. 2000. Significance of SHRIMP dating of the Imperial porphyry and associated Dokhan volcanics, Gebel Dokhan, northeastern desert, Egypt. Journal of African Earth Science, 31, 403– 413. Yoshioka, H., Asahara, Y., Tojo, B. & Kawakami, S. 2003. Systematic variations in C, O, and Sr isotopes and elemental concentrations in Neoproterozoic carbonates in Namibia: implications for a glacial to interglacial transition. Precambrian Research, 124, 69– 85.

Chapter 22 Evidence for Early and Mid-Cryogenian glaciation in the Northern Arabian –Nubian Shield (Egypt, Sudan, and western Arabia) ROBERT J. STERN1*, PETER R. JOHNSON2, KAMAL A. ALI1,3 & SUMIT K. MUKHERJEE1,4 1

Geosciences Department, U Texas at Dallas, Richardson TX 75080-3021, USA

2

Johnson and Vranas Associates, Ltd., Geological Consulting, 6016 SW Haines Street, Portland, Oregon 97219-7046, USA 3

Present address: Faculty of Earth Sciences, King Abdulaziz University, Jeddah 21589, Saudi Arabia 4

Present address: BP America Exploration & Production Company, Houston, TX, USA *Corresponding author (e-mail: [email protected])

Abstract: Evidence of Early- to Mid-Cryogenian (c. 780 Ma and c. 740 Ma) glacial activity is summarized for the northern Arabian–Nubian Shield (ANS), including structural framework, stratigraphy, lithological descriptions and relationships with younger and older units, banded iron formation chemostratigraphy, other characteristics, geochronological constraints, and discussion. The ANS is a broad tract of juvenile continental crust, formed from accreted arc-backarc basin terranes developed around the margins of the Mozambique Ocean. As a result, these successions formed in marine environments at some distance from continental margins. Deposits include banded iron formation (BIF) and possibly glacial diamictite scattered over broad regions of the Central Eastern Desert of Egypt, NW Arabia and possible correlative units in NE Sudan. The older (c. 780 Ma) examples (Meritri group, NE Sudan; basal Mahd group, Arabia) occur in the central ANS, on the southern flank of an important lithospheric boundary, an ophiolite-decorated suture zone. Mahd group diamictite is thin (1– 5 m thick) and rests above the earliest (Cryogenian) ANS unconformity. The Meritri group interval near Port Sudan is much thicker and part of a deformed passive margin. Both Mahd and Meritri group deposits need further study before they are accepted as glaciogenic; confirmation of this interpretation would indicate that Neoproterozoic glacial activity began at least as early as 780 Ma ago. The younger (c. 740 Ma) glacial deposits include diamictite and BIF: the Atud diamictite and BIFs of the Central Eastern Desert of Egypt and the correlative Nuwaybah diamictite and BIF of NW Arabia. Northern ANS-BIF is a well-layered chemical sediment of interlaminated hematite-magnetite and jasper. A glacial origin for the Atud-Nuwaybah diamictites is inferred because large clasts and matrix zircons have ages (Palaeoproterozoic and Neoarchean) and compositions (especially quartzite, arkose, and microdiamictite) that require transport from outside the ANS Cryogenian basin. Northern ANS-BIF may also reveal glacial influence, having been deposited in response to reoxygenation of a suboxic ocean. The 740 Ma diamictite and/or BIF may correlate with Tambien Group diamictites in Ethiopia (Miller et al. 2011). Northern ANS diamictite and BIF were deposited in an oceanic basin of unknown size, as indicated by association with abundant ophiolites; they are strongly deformed, obscuring many primary features. There is no strong evidence for or against Ediacaran glaciation in the ANS, largely because the region was uplifted at this time. The c. 600 Ma ANS peneplain may have been partly cut by Ediacaran glaciation. Some of the post-accretionary basins of Arabia could preserve glaciogenic deposits of Ediacaran age, but assessing this possibility requires further investigation.

The Arabian –Nubian Shield (ANS) consists of mostly Neoproterozoic outcrops around the Red Sea in NE Africa and West Arabia, exposed by Oligocene and younger uplift and erosion. The ANS is one of the largest tracts of juvenile continental crust of Neoproterozoic age on Earth; its evolution accompanied a supercontinent cycle that defined Neoproterozoic tectonics, beginning with the break-up of the end-Mesoproterozoic supercontinent Rodinia in early Neoproterozoic time (Stern 2008). ANS juvenile crust (intra-oceanic arcs and oceanic plateaux) was generated around and within the Mozambique Ocean and coalesced as this ocean closed (Stern 1994). The tectonic cycle culminated in a protracted collision beginning c. 630 Ma, forming the East African Orogen (EAO), an important weld in the end-Neoproterozoic supercontinent of ‘Greater Gondwana’ (Stern 1994) or ‘Pannotia’ (Dalziel 1997). In reconstructed Gondwana, the EAO extends from the Mediterranean (Tethys) southward along the eastern margin of Africa and across East Antarctica. Sedimentary evidence of early Cryogenian glacial episodes is likely to be preserved in ANS sequences, because these episodes occurred when ANS crustal components were mostly below sea level (Stern et al. 2006). In contrast, evidence of late Cryogenian to Ediacaran glacial episodes may be absent because this was a time of collision and uplift in the ANS. Some evidence for Ediacaran climate may be preserved in post-accretionary basins of Ediacaran age in Arabia (Johnson 2003). In this chapter we describe possible glaciogenic units from both flanks of the Red Sea in the northern ANS; the interested reader should see the chapter by Miller et al. (2011) for an overview of

possible glacial deposits from the southern ANS as well as Allen et al. (2011a, b) for chapters covering successions in Oman, to the east of the ANS. The units summarized here include (i) Meritri group metaconglomerate (E. Sudan) and basal Mahd group diamictite (central Arabian Shield); (ii) Atud diamictite (E. Egypt) and Za’am group Nuwaybah diamictite (NW Arabian Shield); and (iii) BIF from E. Egypt, NW Saudi Arabia (Sawawin deposit) and NE Sudan (Fodikwan). Locations of these units are shown in Figure 22.1 and co-ordinates are listed in Table 22.1. We know less about the Meritri group and basal Mahd group and are less confident that they are glaciogenic than we would like to be, but if confirmed, they would be the oldest known Neoproterozoic glaciogenic deposits. These are discussed here to encourage further detailed studies. The deposits of E. Sudan and the central Arabian Shield have not been the subject of focused sedimentological study. The Meritri group conglomerate of Sudan was first identified and briefly described by Abdelsalam & Stern (1993). The basal Mahd group diamictite was first mentioned by Johnson et al. (2003) and then discussed in somewhat greater detail by Stern et al. (2006). The diamictite and banded iron formations to the north in Egypt and NW Arabia (Fig. 22.1) have been studied for some time. The Atud Formation was first described as conglomerate (El-Essawy 1964). Since its first recognition, the Atud Formation has only been recognized in eastern Egypt between 258N and 268N. At the type locality near Gebel Atud, it includes schist, metamudstone, and metagreywacke in addition to diamictite, all of which

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 277– 284. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.22

278

R. J. STERN ET AL.

Cairo

suffered greenschist –facies metamorphism (El-Essawy 1964). The term ‘conglomerate’ is not appropriate for the poorly sorted, matrix-supported deposits of the Atud Formation. These better fit the description of diamictite by Flint et al. (1960) as poorly sorted, heterolithic, and very coarse terrigenous sediments. Accordingly, in this chapter we refer to the Atud diamictite (Stern et al. 2006).

Jordan

29°

27° Nil e

Area of Figure 3

Area of Figure 3

Saudi Arabia

Al Muwaylih

Hurghada Safajah

Structural framework

Quseir Al Wajh

25°

Marsa Alam

Egypt

Al Madinah

23°

Red Sea Biír Umq suture

Sudan

F

Jiddah

Mahd gp. diamictite

21°

Port Sudan

N

19°

200 km 31°

re

utu

bs

si aka

33°

Meritri gp. diamictite 35°

37°

Phanerozoic Exposed Neoproterozoic crust Ophiolitic complexes

41° 39° Suture Shear zone Diamictite Banded Iron Formation

Fig. 22.1. Locality map of the northern ANS showing the approximate locations of geological units described in this chapter. The dark, dashed square designates the area shown in Figure 22.3. Early Cryogenian (c. 780 Ma) diamictite sequences (Meritri and Mahd) are confined to the Nakasib and Bir Umq suture areas in the south, whereas Atud and Nuwaybah diamictite and BIF sequences are confined to Egypt and NW Saudi Arabia.

Table 22.1. Locations of Neoproterozoic deposits of possible glacial origin in the northern ANS Nation

Location and lithology

Age (Ma) Latitude (N) Longitude (E)

Egypt Wadi Abu Marwat BIF Wadi Kareim BIF Wadi Kareim diamictite Wadi Muweilha diamictite Wadi El Dabbah BIF Um Gerifat BIF Um Ghamis BIF Wadi Sitra BIF Wadi Mubarak diamictite El Hadid BIF El Imra BIF Um Nar BIF Gebel Atud diamictite Saudi Arabia Wadi Sawawin BIF Nuwaybah diamictite Mahd diamictite Sudan Meritri Gp. conglomerate Fodikwan BIF

c. 750 c. 750 c. 750 c. 750 c. 750 c. 750 c. 750 c. 750 c. 750 c. 750 c. 750 c. 750 c. 750

268310 258560 500 258560 500 258520 258490 258400 258370 258310 258260 258210 258170 258160 258010

338380 348020 348020 348540 3000 348090 348200 348190 348140 348340 348080 348270 348160 348270

c. 750 c. 750 c. 780

278540 268270 238240 2200

358460 368240 408460 5200

c. 780 –

198380 218440

368420 368420

The Meritri group conglomerate (Sudan) is preserved within the B’ir Umq-Nakasib Suture Zone, whereas the Mahd group diamictite (Saudi Arabia) is preserved just south of this suture zone. This is one of the longest and best-defined ophiolite-decorated suture zones in the ANS (Johnson et al. 2003) and extends (with the Red Sea closed) ENE–WSW over 600 km from the Nile in Sudan into the central Arabian Shield. The suture zone itself consists of rocks that originated in a variety of juvenile oceanic environments and include strongly deformed ophiolite nappes, and metavolcanic, metasedimentary, and intrusive rocks. Dating of the ophiolites, volcanic rocks, and pre- and syntectonic plutons indicates that oceanic magmatism in the region was active c. 870–830 Ma, whereas suturing occurred c. 780–760 Ma (Hargrove III et al. 2006). Structural complexities (folding, faulting, shearing, etc.) are especially severe for the Meritri group, which is part of a southwards-directed nappe stack (Fig. 22.2a), discussed in the following section (Abdelsalam & Stern 1993). The Mahd group diamictite is less folded and faulted because it lies south of the suture zone. The Atud and Nuwaybah diamictites and associated BIF in Egypt and NW Saudi Arabia are also folded and faulted. BIF is strongly deformed although much of this deformation may have occurred as a result of slumping of dense, weak sediments in a tectonically unstable basin. BIF and surrounding sediments are metamorphosed to greenschist facies. Tectonic deformation began as a result of collision between arc terranes prior to c. 680 Ma (Ries et al. 1983). Other deformation resulted from pervasive left-lateral strike-slip shearing along the Najd fault system, which was active during Ediacaran time (Sultan et al. 1988). Najd deformation was a far-field manifestation of collision between fragments of east and west Gondwana (Abdelsalam & Stern 1997). Najd faulting imparted a penetrative shear fabric to most Cryogenian supracrustal units in Egypt and NW Arabia.

Stratigraphy Cryogenian supracrustal sequences in the northern ANS are dominated by variably deformed immature clastic metasediments (greywackes) and metavolcanic rocks; rare sedimentary carbonates also occur. Cryogenian stratigraphic reconstructions are complicated because of strong deformation as well as the presence of several accreted terranes (Johnson & Woldehaimanot 2003). Cryogenian metasedimentary successions are also lithologically monotonous. Distinctive units such as sedimentary carbonates are uncommon, although these become increasingly important in the ANS farther south in Sudan, Eritrea, Ethiopia, and SW Arabia. As a result, ANS Cryogenian supracrustal units often lack a useful formal stratigraphic framework. This is especially true for Egypt, where Stern (1981) informally divided the Cryogenian supracrustal succession of the Central Eastern Desert into a basal ‘older (ophiolitic) metavolcanics’, ‘metasediments’, and ‘younger (arc-like) metavolcanics’. Egyptian Cryogenian successions are unconformably overlain by Ediacaran clastic sediments of the Hammamat Group and the Dokhan Volcanics. A similar situation exists for Sudan. In contrast, there are a plethora of stratigraphic names for Cryogenian supracrustal successions of Saudi Arabia, largely based on the results of quadrangle mapping.

EARLY AND MID-CRYOGENIAN GLACIATION IN THE NORTHERN ARABIAN –NUBIAN SHIELD

279

Fig. 22.2. Stratigraphic summaries of possibly glaciogenic Cryogenian units described in this chapter, locations presented in Table 22.1. (a) Stratigraphic chart for possibly glaciogenic conglomerates of the Meritri group, Nakasib suture, NE Sudan. Column A is a measured section for the lower part of the Arba’at volcanic rocks along Khor Arba’at. Column B is a representative section for the sedimentary part of the Salatib group along Khor Salatib. Column C is a representative section for diamictite of the Meritri group along Khor Meritri. Column D summarizes the stacking order observed in the nappes of the Nakasib suture (from Abdelsalam & Stern 1993). (b) Simplified stratigraphic column for Mahd group rocks showing stratigraphic position of basal diamictites (after Lowther 1994). (c) Simplified stratigraphic column for Atud diamictite and associated BIF in Wadi Kareim, Central Eastern Desert of Egypt, from Ali et al. (2010). (d) Simplified stratigraphic column for Nuwaybah diamictite in NW Saudi Arabia (after Davies 1985).

280

R. J. STERN ET AL.

0

33˚E

50

100 km

Banded Iron Formation

28˚N

Diamictite Ophiolitic melange

Egypt

Ophiolites Major city Local Road Highway

HU 27˚N

36˚E

SA

37˚E

Abu Marawat

26˚N

QU

ad

Qift - Quseir Ro

Muweilih

Sawawin ALM

Kareim

25˚N

Saudi Arabia

Dabbah

Um Gerifat Um Ghamis Gebal EL Hadid

28˚

Mobarak 27˚

El Imra MA Nuwaybah

Idfu - Marsa Alam Road

Um Nar Atud HU: Hurghada SA: Safajah QU: Quseir

MA: Marsa Alam AL: Al Wajh ALM: Al Muwaylih

AL 26˚

Fig. 22.3. Location of the Atud and Nuwaybah diamictite and associated BIF, as well as ophiolitic rocks in Egypt and Saudi Arabia, northern ANS, with the Red Sea closed. Diamictites in this area are thought to be early Cryogenian in age. See Figure 22.1 for regional location (modified after Ali et al. 2010) and Table 22.1 for geographical coordinates of sites.

Meritri group metaconglomerate of possible glacial origin in E. Sudan is associated with five informal groups, separated by thrusts so that original depositional relationships are unresolved (Fig. 22.2a). The c. 2-km-thick Arbaat (metavolcanic) group lies at the base of the succession and appears to be autochthonous. Subsequent units are thrust southwards on top of the Arbaat group and one another and include from the base: the Salatib group, a 1.2-km-thick succession of clastic metasediments, carbonates, conglomerate and felsic tuffs, the Meritri group composed of coarse conglomerate, lithic greywacke, limestone, red sandstone and felsic metavolcanics, the Nakasib ophiolite and at the top, the arc-like Shalhout group (Fig. 22.2a; Abdelsalam & Stern 1993). Lateral variations within the Meritri group show changes from granitic-clast dominated in the east to volcanic-clast dominated in the west. Diamictite at the base of the c. 780 Ma Mahd group (Fig. 22.2b) unconformably overlies diorite and tonalite of the c. 810 Ma Dhukhr batholith, indicating an episode of possibly glacial erosion at c. 780 –810 Ma (Johnson et al. 2003). Continuity of unit to the north is truncated by terrane boundary; possible continuity to the south and east is unknown. The basal Mahd group unconformity is one of the oldest unconformities known in the ANS. Another early Cryogenian unconformity, which might be related to the basal Mahd group unconformity, is found farther south in the Arabian Shield, between plutonic rocks of the c. 800 Ma An Nimas batholith and metamorphosed conglomerate, limestone, and sandstone of the 780 –795 Ma Hali group (Cooper et al. 1979). This encourages speculation that heretofore unrecognized Cryogenian glacial deposits might exist in the southern Arabian Shield. The Atud diamictite is a distinctive lithology found in the Egyptian metasedimentary succession. This has been reported from four areas in the Central Eastern Desert of Egypt: Wadi Kareim, Wadi Mobarak, Wadi Muweilih, and the type locality

east of Gebal Atud (Table 22.1, Fig. 22.3). The Wadi Kareim occurrence is especially significant (Figs 22.2c & 22.3), because this is the only locality where the Atud diamictite is found in clear stratigraphic relationship with BIF. The diamictite is part of a supracrustal succession in which metavolcanic rocks are conformably overlain by immature metasedimentary rocks (c. 100 m thick of wackestone-sandstone, siltstone and diamictite) and BIF. Metavolcanic rocks at the base of the section are c. 100 m thick and are truncated below by a thrust fault. Egyptian diamictite and BIF are found in regional association with Cryogenian ophiolites, a tripartite association that indicates that both sedimentary units were deposited in an oceanic basin. Deformation obscures small-scale lateral variations of the Atud Diamictite; the BIF is restricted to the Central Eastern Desert of Egypt, with nine occurrences between 258150 N and 268350 N (Fig. 22.3, Table 22.1). In NW Saudi Arabia, the diamictite and BIF are geographically separate parts of the Za’am group, the oldest known unit in the Midyan terrane. The diamictite (Nuwaybah Formation) is in the upper part of the Zaam group, which is an assemblage of basaltic, andesitic, and subordinate rhyolitic flows and tuffs, abundant volcaniclastic sandstone and siltstone, and volcaniclastic conglomerate (Fig. 22.2d). The Zaam group was deformed and metamorphosed (greenschist facies) prior to c. 660 Ma, and neither original top nor bottom is exposed. It is overlain unconformably by the younger Thalbah group. The significance of the basal Thalbah group unconformity is unknown but could correspond to Cryogenian glacial erosion. The lateral extent of the Nuwaybah diamictite is not yet known. Further north, in Wadi Sawawin, the BIF-bearing succession consists of Ghawjah Formation metavolcanic rocks overlain by Silasia Formation metasedimentary rocks. Ghawjah metavolcanic rocks are correlated with the ‘younger metavolcanic’ rocks of Egypt. The Silasia Formation consists of c. 1-km-thick immature clastic metasedimentary rocks with BIF-bearing units near the top; the sequence is intruded by diabase sills. The stratigraphic and age relationships shown by the successions at Wadi Kareim and Wadi Sawawin are remarkably similar except that the diamictite is absent from the latter succession.

Possible glaciogenic deposits and associated strata (A1) Meritri group metaconglomerate, E. Sudan Abdelsalam & Stern (1993) inferred that the original thickness for the Meritri group in Khor Meritri was c. 2 km. They subdivided the group into four formations comprising, from oldest to youngest: conglomerate (possibly glaciogenic); lithic greywacke; intercalated limestone, red sandstone and felsic tuff; and felsic volcanic rocks. The coarsest clastic beds are made up of abundant polymict conglomerate intercalated with minor lithic greywacke and limestone. The polymict conglomerate is matrix supported, with the matrix made up of lithic greywacke and minor carbonate. Clasts include granite, granodiorite, diorite, rhyolite, ignimbrite and carbonate as well as subordinate clastic metasediments. Intermediate to felsic volcanic clasts become more abundant to the west. In the NE, along Khor Meritri, plutonic clasts are most abundant, comprising c. 50% of total clasts, whereas volcanic clasts are c. 35% and clasts of metasediments make up c. 15%. Abdelsalam (pers. comm. 2007) measured 76 of these clasts, finding a maximum size of 70  40  35 cm3 (clasts are deformed and stretched); clasts .40 cm are common. The lithic greywacke unit is made up of lithic greywacke intercalated with minor felsic volcanic layers and limestones. Locally the lithic greywacke grades into conglomerate with smaller clasts of mostly felsic volcanic rocks. Sedimentary structures include graded bedding in the sandstone, cross-bedding and channels. Channel structures and cross-bedding indicate that the palaeocurrent direction was from SE to NW.

EARLY AND MID-CRYOGENIAN GLACIATION IN THE NORTHERN ARABIAN –NUBIAN SHIELD

(A2) Basal Mahd group diamictite, central Arabian Shield Although dominated by volcanic rocks, the Mahd group has a 1– 5-m-thick diamictite resting on the unconformity. The diamictite is matrix-supported, with a dark-grey, immature, arkosic matrix containing abundant, angular to sub-angular clasts (up to 30 cm across) of granitic and felsic volcanic rocks.

281

Boundary relations with overlying and underlying non-glacial units (A1) Meritri Group metaconglomerate, E. Sudan Contacts with underlying and overlying units are faulted.

(A2) Basal Mahd Group diamictite, central Arabian Shield (B1) Atud diamictite, E. Egypt The Atud diamictite consists of massive poorly sorted and rounded clasts, from gravel to boulder, in a sheared grey matrix. Atud clasts include quartzite, highly altered granitoid, and a distinctive arkosic breccia (microdiamictite) (Ali et al. 2010). Clasts and matrix zircons in the Atud diamictite are dominated by c. 750 Ma granitic rocks and microdiamictite but with a significant amount of Palaeoproterozoic and Neoarchean granitic rocks and quartzite (Ali et al. 2010). We have also examined Atud diamicite from the eastern half of the Wadi Mobarak belt in Egypt. Basement exposures around Wadi Mobarak are dominated by highly deformed ophiolitic fragments of serpentinites, metagabbro, and greenschist-facies mafic metavolcanic rocks, whereas the sedimentary sequence includes tuff, shale, schist and Atud diamictite (Akaad et al. 1995). Diamictite clasts are similar to those observed in Wadi Kareim.

(B2) Nuwaybah Formation (Za’am group) diamictite, NW Saudi Arabia The Nuwaybah locality is located within the Al Wajh quadrangle in the northwestern part of the Arabian Shield. The diamictite locality we studied is beautifully exposed in a roadcut along the Red Sea highway (Table 22.1). Clasts and matrix zircons in the Nuwaybah diamictite are dominated by c. 750 Ma granitic rocks and distinctive arkose/micro-diamictite but with a significant amount of Palaeoproterozoic and Neoarchaean granitic rocks and quartzite (Ali et al. 2010), similar to lithologies and ages of the Atud diamictite in Egypt.

Contacts with underlying units are unconformable. Contacts with overlying units are conformable.

(B1) Atud diamictite, E. Egypt Contacts with underlying units are only clear at Wadi Kareim, where they appear conformable.

(B2) Nuwaybah Formation (Za’am group) diamictite, NW Saudi Arabia The basal contact of the Nuwaybah Formation is thought to be conformable on the underlying Umm Ashsh Formation (part of the Za’am group). The top of the Formation is obscured by granite intrusions and by an angular unconformity with younger sedimentary rocks (Thalbah group).

(C1) BIF, E. Egypt This BIF is found in conformable stratigraphic relationship with underlying immature, possibly tuffaceous greywackes. It is associated with metavolcanic rocks and intrusive diabase.

(C2) Silasia Formation BIF, NW Saudi Arabia The BIF-bearing sediments are conformable above metavolcanic rocks of the Ghawjah Formation. The top of the formation is not exposed at the Sawawin locality because it is intruded by metadiorite.

(C1) Banded iron formation (BIF), E. Egypt

Chemostratigraphy

BIF of early Cryogenian age is found in the Central Eastern Desert of Egypt, with nine occurrences between latitudes 258150 and 268350 N. The Fodikwan BIF in NE Sudan may be related (Table 22.1, Fig. 22.1), but more work is needed to test this possibility. BIF occurs as fairly regular bands interbedded with metasediments and metavolcanics in a zone that originally had a stratigraphic thickness of 100– 200 m, within which the aggregate BIF thickness is about 10 –20 m (Sims & James 1984). Egyptian BIF is mostly an oxide facies, consisting of interlaminated hematite and jasper, and containing 40– 46% Fe (Sims & James 1984). In several locales, BIF-bearing metasediments are intruded by metadiabase sills.

This section is applicable only for the BIF. There are no significant carbonate sediments in the region, although these become increasingly important to the south (in Sudan, S. Saudi Arabia, Eritrea, and Ethiopia, see Allen et al. 2011a, b; Miller et al. 2011). NW Saudi Arabia (Sawawin) BIF has been sampled systematically and stratigraphically, in order to identify the nature of its sources (Mukherjee 2008). Samples of Egyptian BIF were also studied in order to understand regional variations. Rare earth elements (REE) are particularly useful for studying the BIF because these record the composition of equilibrium seawater that the BIF precipitated from. REE patterns for the Sawawin BIFs are similar to a mixture of modern shallow suboxic seawater (German et al. 1991; Webb & Kamber 2000) with low-T hydrothermal vent fluid solution (Bau & Dulski 1999) suggesting a dominant hydrothermal input of REEs and by analogy Fe into the BIFs (Mukherjee 2008). BIF iron input sources are isotopically dominated by hydrothermal vent fluids but continental runoff is also significant as revealed by Ce/Ce* and Eu/Eu* (Mukherjee 2008).

(C2) Silasia Formation BIF, NW Saudi Arabia BIFs in NW Saudi Arabia occupy a more restricted region than do their Egyptian counterparts. Arabian BIF occurs within the Silasia Formation, which, like the Egyptian section, consists of volcanogenic greywackes that appear to rest conformably on metavolcanic rocks (Ghawjah Formation). The exposed thickness of the Silasia Formation is estimated to be c. 1160 m in the reference area of Wadi Sawawin (Goldring 1990). Arabian BIF is mostly present as an oxide facies, consisting of interbedded hematite and jasper, and containing 40–46% Fe (Goldring 1990). Similar to the Egyptian section, the Silasia Formation is intruded by metadiabase sills up to 100 m thick.

Other characteristics Gold mineralization may be associated with BIF at Abu Marawat (Botros 2002). BIF here occurs as sharply defined horizons within a volcanic-sedimentary succession, which is regionally metamorphosed into greenschist facies. Gold concentrations

282

R. J. STERN ET AL.

of up to 2.15 ppm occur in the BIF, either enclosed in the flaky hematite crystals of hematite-rich layers or as fine inclusions in magnetite-rich bands. El-Habaak & Mahmoud (1995) identified spherical bodies as Eosphaera tyleri in jasper-rich layers from Wadi Kareim BIF, Egypt. Those occur as more or less clear spherules of quartz surrounded by thin veneers of very fine hematite granules. They recognized two distinct varieties of E. tyleri, one of which is c. 15 mm in diameter and the other c. 60 mm; El-Habaak & Mahmoud (1995) suggested that biological activity played a role in spherule formation and BIF deposition. The Mahd group basal diamictite is overlain by hypabyssal intrusive to volcaniclastic rocks (caldera complex) that hosts a major epithermal gold deposit (Mahd adh Dhahab Mine). The age of gold mineralization is not known with certainty, but is likely to have been syngenetic with Mahd Group igneous activity (c. 760– c. 780 Ma; Hargrove III et al. 2006). The deposit consists of quartz veins and stockwork that contains copper, zinc, iron, and lead sulphides and very fine-grained gold and silver (averaging 5–30 mm), mostly as tellurides, associated with the sulphides (Moore 1979).

Palaeolatitude and palaeogeography Extensive deformation and metamorphism makes palaeomagnetic determinations unreliable, and for this reason few have been carried out (Reischmann et al. 1992).

Geochronological constraints (A1) Meritri Group metaconglomerate, E. Sudan The age of the Meritri group is constrained by the age of underlying metavolcanics and intrusive plutonic rocks, reported by Stern & Abdelsalam (1998). Conventional multigrain analyses of two size fractions of zircons separated from a metarhyolite lava of the apparently underlying Arba’at Formation yielded a nearly concordant age of 790 + 2 Ma. The supracrustal sequence, including the Meritri group, was intruded by the Arba’at quartz diorite, which yielded a single fraction, nearly concordant U–Pb zircon age of 779 + 3 Ma (Stern & Abdelsalam 1998). Faulting leads to uncertainties about the exact relationship between the intrusion and Meritri group, but these ages suggest that deposition of the Meritri diamictite may be constrained between 779 + 3 and 790 + 2 Ma.

(A2) Basal Mahd Group diamictite, central Arabian Shield The age of this diamictite is constrained by the age of subjacent and superjacent igneous rocks. The Dhukhr complex, which unconformably underlies the diamictite, has robust conventional multigrain U –Pb zircon crystallization ages of 811 + 4 Ma (Stoeser & Stacey 1988) and 816 + 4 Ma (Calvez & Kemp 1982). The Hufayriyah batholith, which also unconformably underlies the Mahd group c. 60 km north –NE of the diamictite locality, has a U –Pb zircon SHRIMP age of 785 + 6 Ma (Hargrove III et al. 2006). Rhyolite in the Mahd group upsection from the basal diamictite has been dated by U –Pb zircon SHRIMP techniques at 777 + 5 Ma (Hargrove III et al. 2006). In addition to the U –Pb zircon SHRIMP age, the Hufayriyah batholith also yields a conventional multigrain U –Pb zircon age of 760 + 10 Ma (Calvez & Kemp 1982); this result may represent a younger pulse of Hufayriyah intrusion and is not necessarily a robust constraint on the age of rocks at the base of the Mahd group.

(B1) Atud diamictite, E. Egypt The age of this unit is constrained to be younger than both the age of the underlying metavolcanics (Fig. 22.2c) and the age of the youngest clast within the diamictite. Ali et al. (2009) report a

SHRIMP U –Pb zircon ion probe age of c. 750 Ma for the metavolcanics that lie beneath the Atud diamictite at Wadi Kareim. One metavolcanic sample yielded a weighted mean 206Pb– 238U age of 769 + 29 Ma. As noted above, the Atud diamictite at Wadi Kareim is dominated by Cryogenian clasts and matrix material but contains abundant clasts of Palaeoproterozoic and Neoarchaean granitic rocks and pre-Neoproterozoic quartzite. Clasts in the diamictite are as young as 754 + 15 Ma (Ali et al. 2010). Because of stratigraphic relationships shown on Figure 22.2c, the maximum age of c. 750 Ma for the metavolcanics and 754 + 12 Ma for the Atud diamictite at Wadi Kareim also provides a maximum age for BIF deposition at this locality (Ali et al. 2009, 2010). Geochronological data support the inference that clasts in the Atud diamictite sample much older rocks than are exposed in the Eastern Desert of Egypt and so must have been transported some distance. Two granitic cobbles from NW of Marsa Alum (also referred to as the Wadi Mobarak metasedimentary unit) yielded highly discordant conventional multigrain U –Pb zircon upper intercept ages of 1120 and 2060 Ma (Dixon 1981). Dixon (1979) obtained a discordant U –Pb zircon upper intercept of 2.3 Ga for a granitic cobble from Atud conglomerate outcrops west of Quesir. SHRIMP geochronological studies by Ali et al. (2010) confirm and extend Dixon’s conclusions that many Atud clasts are pre-Neoproterozoic.

(B2) Nuwaybah Formation (Za’am Group) diamictite, NW Saudi Arabia Like the Atud diamictite, this unit is dominated by Cryogenian material but contains abundant clasts of Palaeoproterozoic and Neoarchaean granitoids and quartzite. The youngest clasts in the Nuwaybah diamictite yield SHRIMP U –Pb zircon ages of 765 + 22 Ma (Ali et al. 2010). Another clast is an arkose with a SHRIMP U – Pb zircon age of 766 + 5 Ma, which we interpret as the age of the basement supplying the arkose, which was rapidly eroded and deposited, lithified, and re-eroded, all apparently within a very few millions of years. The formation has a minimum age of deposition constrained by the age of intrusions in the Zaam Group near the diamictite. These include the Buwaydah complex, dated by the conventional U –Pb zircon method at 725 + 4 Ma (Hedge 1984) and the Imdan complex, dated by the SHRIMP U –Pb zircon method at 676 + 6 Ma (15 data points (Kennedy et al. 2011) and conventional U – Pb zircon method at 660 + 4 Ma (Hedge 1984). Zaam Group felsic tuff 175 km SE of the diamictite locality yields U– Pb zircon SHRIMP ages of 711 + 10 Ma and 708 + 4 Ma (Kennedy et al. 2004, 2005). However, the stratigraphic relationship between the diamictite and these tuffs is poorly understood, thus the significance of these data for constraining the depositional age of the diamictite is not clear.

(C1) BIF, E. Egypt BIF age is best known from Wadi Kareim, where ages of conformably underlying metavolcanics and diamictite indicate that it is younger than c. 750 Ma (Ali et al. 2009, 2010), but perhaps not much younger.

(C2) Silasia Formation BIF, NW Saudi Arabia The Silasia Formation, which hosts this BIF, is intruded by plutonic rocks of the Muwalylih suite, dated by U –Pb zircon techniques at 710–725 Ma (Hedge 1984). Unfortunately, Hedge (1984) did not provide analytical details, the number of zircon populations or statistics about the results (uncertainty and mean square of the weighted deviates). Ali et al. (2011) report a U –Pb zircon ion

EARLY AND MID-CRYOGENIAN GLACIATION IN THE NORTHERN ARABIAN –NUBIAN SHIELD

probe weighted mean 206Pb/238U age of 763 + 25 Ma for Ghawjah metavolcanics beneath the BIF. Silasia Formation was deformed and then intruded by Sawawin complex diorite, which yields a U – Pb zircon concordia age of 661.5 + 2.3 Ma (Ali et al. 2011). Silasia Formation BIF is thus , 763 + 25 Ma, probably c. 750 Ma, similar in age to Egypt BIF.

Discussion The foregoing summary indicates that there is compelling evidence of a c. 740 Ma age glaciation and plausible evidence of a c. 780 Ma glaciation in the northern ANS. Indistinguishable bracketing ages of the Meritri group (790 + 2 Ma to 779 + 3 Ma) and Mahd group (785 + 6 Ma to 777 + 5 Ma) indicate that deposition of these units may have occurred about the same time. A glaciogenic interpretation for c. 780 Ma Mahd-Meritri groups might help explain the enigmatic Bitter Springs d13C excursion of approximately the same age, thought to be associated with glaciation but for which no definitive evidence of glaciation has been identified to date (Halverson et al. 2005, 2007). The association of the Mahd diamicite with the oldest unconformity known from the Arabian Shield is also consistent with a glacial interpretation and is tentative evidence of a c. 780 Ma local continental glaciation, probably sourced to the south (present coordinates). Certainly, more work is needed to map, study sedimentology and stratigraphy, date clasts and matrix, and examine intervening sedimentary and possible igneous environments of Mahd and Meritri groups deposits. Evidence for a c. 740 Ma glaciation is found in the ophioliteAtud diamictite-BIF basin and includes far travelled exotic clasts (based on age and composition) and association of diamictite with BIF. By virtue of their age, these deposits may correlate with early Cryogenian glaciogenic successions elsewhere (Fairchild & Kennedy 2007). The Atud-Nuwaybah diamictite may be strictly correlative, with similar ages of deposition, clast lithology and distribution of Cryogenian, Palaeoproterozoic and Neoarchaean clast ages (Ali et al. 2010). The association of diamictite with BIF is consistent with the genetic coupling suggested by Snowball Earth concepts (Hoffman 2005), although alternative interpretations exist. Based on similar age constraints (the youngest detrital zircons in the Ethiopian diamictites are also c. 0.75 Ga), Atud diamictite may also be broadly synchronous with Negash and Shiraro diamictites in northern Ethiopia (Avigad et al. 2007; Miller et al. 2003, 2011). Neoproterozoic diamictites from the ANS (this chapter and Miller et al. 2011) do not seem to correlate directly with diamicitites of the Abu Mahara Group of the Huqf Supergroup, Oman, which are younger in age (700– 735 Ma) and are dominated by c. 860 Ma zircons that are not common in the Atud/Nuwaybah diamictites (Rieu et al. 2007). Because pre-Neoproterozoic basement is unknown in Egypt east of the Nile, Dixon (1981) concluded that these clasts were derived from older crust to the west or south, perhaps from the Saharan Metacraton (Abdelsalam et al. 2002). Dixon (1979) suggested that Atud clasts were transported such great distances by ice rafting, but other possible explanations (e.g. meteorite impact ejecta blanket, far-travelled submarine debris flow) are possible. There is controversy regarding how the Egyptian BIFs formed, although ideas published in the geological literature were mostly developed prior to the Snowball Earth hypothesis (Fairchild & Kennedy 2007). Sims & James (1984) suggested that BIF formed as chemical precipitates during lulls in dominantly subaqueous, calc-alkaline volcanism, apparently within an intraoceanic island-arc environment. The close association in time and space between volcanic activity and deposition of the BIF suggests a genetic relation (Sims & James 1984). Finally, it should be noted that there is no strong evidence for or against an Ediacaran glaciation in the ANS. The ANS peneplain may have been partly cut during this time, but regional uplift resulted in generally poor preservation (Stern et al. 2006). Some

283

of the post-accretionary basins of Arabia may preserve glaciogenic deposits of Ediacaran age, but focused investigations are needed to establish or refute this possibility (Johnson 2003). We are grateful for support by many individuals and agencies over the years including the Sudan Geologic Survey, Saudi Geological Survey, Nuclear Materials Agency of Egypt, NASA, NSF and the US–Egypt Joint Technical Program. We also thank M. Abdelsalam, D. Avigad, M. Beyth, N. Miller, F. Kattan, V. Pease, M. Whitehouse and A. Kro¨ner for discussions on this topic, A. Collins, V. Pease and D. Avigad for thoughtful reviews of the manuscript, and E. Arnaud for careful final editing. This represents a contribution of the IUGS- and UNESCOfunded IGCP (International Geoscience Programme) Project #512.

References Abdelsalam, M. G. & Stern, R. J. 1993. Tectonic evolution of the Nakasib suture, Red Sea Hills, Sudan: evidence for a late Precambrian Wilson cycle. Journal of the Geological Society, London, 150, 393– 404. Abdelsalam, M. G. & Stern, R. J. 1997. Sutures and shear zones in the Arabian-Nubian Shield. Journal of African Earth Sciences, 23, 289– 310. Abdelsalam, M. G., Lie´geois, J. P. & Stern, R. J. 2002. The Saharan Metacraton. Journal of African Earth Sciences, 34, 119–136. Akaad, M. K., Noweir, A. M. & Abu-El-Ela, A. M. 1995. The volcanosedimentary association and ophiolites of Wadi Mubarak, eastern Desert, Egypt. Proc. Inter. Conf. 30 Years Cooper., 69. Geological Survey of Egypt Special Publication, 231–248. Ali, K. A., Stern, R. J., Manton, W. I., Kimura, J.-I. & Khamis, H. A. 2009. Geochemistry, Nd isotopes, and U –Pb SHRIMP zircon dating of Neoproterozoic volcanic rocks from the Central Eastern Desert of Egypt: new insights into the c. 750 Ma crust-forming event. Precambrian Research, 171, 1 –22. Ali, K. A., Stern, R. J., Manton, W. I., Johnson, P. R. & Mukherjee, S. K. 2010. Neoproterozoic diamictite in the Eastern Desert of Egypt and Northern Saudi Arabia: evidence of c. 750 Ma glaciation in the Arabian– Nubian Shield. International Journal of Earth Sciences, 99, 705– 726. Ali, K. A., Stern, R. J. et al. 2011. Geochemical, U– Pb zircon and Nd isotopic investigations of the Ghawjah metavolcanics of northwestern Saudi Arabia: the 750 Ma crust-forming event correlated across the Red Sea. Lithos, 120, 379– 392. Allen, P. A., Rieu, R., Etienne, J. L., Matter, A. & Cozzi, A. 2011a. The Ayn Formation of the Mirbat Group, Dhofar, Oman. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 239–249. Allen, P. A., Leather, J. et al. 2011b. The Abu Mahara Group (Ghubrah, and Fiq Formations), Jabal Akhdar, Oman. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 251–262. Avigad, D., Stern, R. J., Beyth, M., Miller, N. & McWilliams, M. O. 2007. Detrital zircon U– Pb geochronology of Cryogenian diamictites and Lower Paleozoic sandstone in Ethiopia (Tigrai): age constraints on Neoproterozoic glaciation and crustal evolution of the southern Arabian– Nubian Shield. Precambrian Research, 154, 88– 106. Bau, M. & Dulski, P. 1996. Distribution of yttrium and rare-earth elements in the Penge and Kuruman iron-formations, Transvaal Supergroup, South Africa. Precambrian Research, 79, 37 –55. Bau, M. & Dulski, P. 1999. Comparing yttrium and rare earths in hydrothermal fluids from the Mid-Alantic Ridge: Implications for Y and REE behaviour during near-vent mixing and for the Y/Ho ratio of Proterozoic seawater. Chemical Geology, 155, 77 – 90. Botros, N. S. 2002. Metallogeny of gold in relation to the evolution of the Nubian Shield in Egypt. Ore Geology Reviews, 19, 137– 162. Calvez, J.-Y. & Kemp, J. 1982. Geochronological investigations in the Madh adh Dhahab quadrangle, Central Arabian Shield. BRGM-TR02-5, Saudi Arabian Deputy Ministry for Mineral Resources. Cooper, J. A., Stacey, J. S., Stoeser, D. G. & Fleck, R. J. 1979. An evaluation of the zircon method of isotopic dating in the Southern

284

R. J. STERN ET AL.

Arabian Craton. Contributions to Mineralogy and Petrology, 68, 429– 439. Dalziel, I. W. D. 1997. Neoproterozoic –Paleozoic geography and tectonics: review, hypothesis, environmental speculation. Geological Society of America Bulletin, 109, 16 –42. Davies, F. B. 1985. Geologic map of the Al Wajh quandrangle, sheet 26B, Kingdom of Saudi Arabia. Saudi Arabian Deputy Ministry for Mineral Resouces, Geoscience Map 83A, Jeddah. Dixon, T. H. 1979. The evolution of continental crust in the Late Precambrian Egyptian Shield. Ph.D. Thesis, UC San Diego. Dixon, T. H. 1981. Age and chemical characteristics of some pre-PanAfrican rocks in the Egyptian Shield. Precambrian Research, 14, 119– 133. El-Essawy, M. A. 1964. Geology of the area east of Gabal Atud, Eastern Desert, Egypt. MSc thesis, Assiut University, Assiut, Egypt, 235. El-Habaak, G. H. & Mahmoud, M. S. 1995. Carbonaceous bodies of debatable organic provenance in the banded iron formation of the Wadi Kareim area, Eastern Desert, Egypt. Journal of African Earth Sciences, 19, 125– 133. Fairchild, I. J. & Kennedy, M. J. 2007. Neoproterozoic glaciation in the Earth System. Journal of the Geological Society, London, 164, 895– 921. Flint, R. F., Sanders, J. E. & Rodgers, J. 1960. Diamictite, a substitute term for symmictite. Geological Society of America Bulletin, 71, 1809– 1810. German, C. R., Holiday, B. P. & Elderfield, H. 1991. Redox cycling of rare earth elements in the suboxic zone of the Black Sea. Geochimica et Cosmochimica Acta, 55, 3553–3558. Goldring, D. C. 1990. Banded iron formation of Wadi Sawawin district, Kingdom of Saudi Arabia. Transactions of the Institution of Mining and Metallurgy (Sect. B: Applications for Earth Science), 99, B1 –B14. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Toward a Neoproterozoic composite carbonisotope record. Geological Society of America Bulletin, 117, 1181– 1207. ¨ ., Maloof, A. C. & Bowring, S. A. 2007. Halverson, G. P., Duda´s, F. O Evolution of the 87Sr/86Sr composition of Neoproterozoic seawater. Palaeogeography, Palaeoclimatology, Palaeoecology, 256, 103–129. Hargrove III, U. S., Stern, R. J., Griffin, W. R., Johnson, P. R. & Abdelsalam, M. G. 2006. From Island Arc to Craton: Timescales of Crustal Formation along the Neoproterozoic Bi’r Umq Suture Zone, Kingdom of Saudi Arabia, Saudi Geological Survey. Hedge, C. E. 1984. Precambrian Geochronology of part of northwestern Saudi Arabia, Kingdom of Saudi Arabia, U.S. Geological Survey. Hoffman, P. F. 2005. On Cryogenian (Neoproterozoic) ice-sheet dynamics and the limitations of the glacial sedimentary record. 28th DeBeers Alex. Du Toit Memorial Lecture. South African Journal of Geology, 108, 557– 576. Johnson, P. R. 2003. Post-amalgamation basins of the NE Arabian shield and implications for Neoproterozoic III tectonism in the northern East African orogen. Precambrian Research, 123, 321– 338. Johnson, P. R. & Woldehaimanot, B. 2003. Development of the Arabian –Nubian Shield: perspectives on accretion and deformation in the East African Orogen and the assembly of Gondwana. In: Yoshida, M., Windley, B. F. & Dasgupta, S. (eds) Proterozoic East Gondwana: Supercontinent Assembly and Breakup. Geological Society, London, Special Publications, 206, 289–325. Johnson, P. R., Abdelsalam, M. G. & Stern, R. J. 2003. The Bi’r UmqNakasib suture zone in the Arabian-Nubian shield: a key to understanding crustal growth in the East African Orogen. Gondwana Research, 6, 523–530. Kennedy, A., Johnson, P. R. & Kattan, F. H. 2004. SHRIMP geochronology in the northern Arabian Shield. Part I: Data acquisition. Saudi Geological Survey Open File Report, SGS-OF-2004-11. Kennedy, A., Johnson, P. R. & Kattan, F. H. 2005. SHRIMP geochronology in the northern Arabian Shield. Part II: Data acquisition 2004. Saudi Geological Survey Open File Report, SGS-OF-2005-10.

Kennedy, A., Kozdroj, W., Johnson, P. R. & Kattan, F. H. 2011. SHRIMP geochronology in the northern Arabian Shield. Part III. Data Acquisition, 2006. Saudi Geological Survey, Open-File Report SGS-OF-2007-9. Lowther, J. M. 1994. Mahd adh Dhahab gold deposit. In: Collenette, P. & Grainger, D. J. (eds) Mineral Resources of Saudi Arabia. Saudi Arabian Directorate General of Mineral Resources, Special Publication SP-2, Jeddah, 105– 111. Ludwig, K. R. 2000. SQUID 1.00. A User’s Manual, Berkeley Geochronology Center, Berkeley. Miller, N. R., Alene, M., Sacchi, R., Stern, R., Conti, A., Kro¨ner, A. & Zuppi, G. 2003. Significance of the Tambien group (Tigrai, N. Ethiopia) for Snowball Earth events in the Arabian– Nubian Shield. Precambrian Research, 121, 263– 276. Miller, N. R., Avigad, D., Stern, R. J. & Beyth, M. 2011. The Tambien Group, Northern Ethiopia (Tigre). In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 263–276. Mukherjee, S. K. 2008. Petrography, age (uranium-lead zircon), geochemical and isotopic studies of the Sawawin Banded Iron Formation (BIF), Northwestern Saudi Arabia: Implications for understanding Neoproterozoic climate change. PhD thesis, The University of Texas at Dallas, 147. Reischmann, T., Bachtadse, V., Kro¨ner, A. & Layer, P. 1992. Geochronology and palaeomagnetism of a Late Proterozoic island arc terrane from the Red Sea Hills, northeast Sudan. Earth and Planetary Science Letters, 114, 1– 15. Ries, A. C., Shackleton, R. M., Graham, R. H. & Fitches, W. R. 1983. Pan-African structures, ophiolites and me´lange in the Eastern Desert of Egypt: a traverse at 268N. Journal of the Geological Society of London, 140, 75 –95. Rieu, R., Allen, P. A., Cozzi, A., Kosler, J. & Bussy, F. 2007. A composite stratigraphy for the Neoproterozoic Huqf Supergroup of Oman: integrating new litho-, chemo-, and chronostratigraphic data of the Mirbat area, southern Oman. Journal of the Geological Society, London, 164, 997–1009. Sims, P. K. & James, H. L. 1984. Banded iron-formations of late Proterozoic age in the central Eastern Desert, Egypt; geology and tectonic setting. Economic Geology, 79, 1777– 1784. Stern, R. J. 1981. Petrogenesis and tectonic setting of late Precambrian ensimatic volcanic rocks, Central Eastern Desert of Egypt. Precambrian Research, 16, 195–230. Stern, R. J. 1994. Arc assembly and continental collision in the Neoproterozoic East African orogen: implications for the consolidation of Gondwanaland. Annual Reviews of Earth and Planetary Sciences, 22, 319– 351. Stern, R. J. 2008. Neoproterozoic crustal growth: the solid Earth system during a critical time of Earth history. Gondwana Research, 14, 33 – 50. Stern, R. J. & Abdelsalam, M. G. 1998. Formation of continental crust in the Arabian –Nubian shield: evidence from granitic rocks of the Nakasib suture, NE Sudan. Geologische Rundschau, 87, 150– 160. Stern, R. J., Avigad, D., Miller, N. R. & Beyth, M. 2006. Evidence for the snowball Earth hypothesis in the Arabian –Nubian Shield and the East African Orogen. Journal of African Earth Sciences, 44, 1 – 20. Stoeser, D. B. & Stacey, J. S. 1988. Evolution, U– Pb geochronology, and isotope geology of the Pan-African Nabitah orogenic belt of the Saudi Arabian Shield. In: El-Gaby, S. & Greiling, R. O. (eds) The Pan-African Belt of NE Africa and Adjacent Areas. Friedr. Vieweg & Sohn, Braunschweig, 227– 289. Sultan, M., Arvidson, R. E., Duncan, I. J., Stern, R. J. & Kaliouby, B. E. 1988. Extension of the Najd shear system from Saudi Arabia to the central Eastern Desert of Egypt based on integrated field and Landsat observations. Tectonics, 7, 1291–1306. Webb, G. E. & Kamber, B. S. 2000. Rare earth elements in Holocene reefal microbialites; a new shallow seawater proxy. Geochimica et Cosmochimica Acta, 64, 1557– 1565.

Chapter 23 Glacial deposits of the Bokson Group, East Sayan Mountains, Buryatian Republic, Russian Federation NICKOLAY M. CHUMAKOV Geological Institute of the Russian Academy of Sciences, Pyzhevsky per. 7, Moscow 119017, Russia (e-mail: [email protected]) Abstract: The Bokson Group (Gr.) forms the platform cover in northern part of the Tuva-Mongolian Massif, south of the East Sayan Mountains. Widespread diamictites occur in the lower part of the Bokson Gr., in the Zabit Formation (Fm.). These contain erratic, faceted and striated stones along with dropstones. They have disconformable or gradual lower boundaries and conformable or transitional upper boundaries with overlying deposits. The isotopic d13Ccrb curve records some maximum and minimum values as a background of insignificantly varying values (from – 3 to 2‰). Maximum values (up to 5.5‰), are recorded in deposits underlying diamictites, whereas the overlying beds of the Zabit Formation (Fm.) yielded three peaks or minima down to 24‰. A fourth negative anomaly derived from the middle part of the Bokson Gr. in the lower part (close to base) of the Tabinzurt Fm. is as low as – 5.7‰. The deposits underlying the diamictites contain Ediacaran (Vendian) microfossils. Cloudina sp. was discovered in the granule of diamictites, whereas beds just above include small shelly fossils (SSF) of the Nemakit-Daldynian type (Cambrotubulus decurvatus Miss., Anabarites trisulcatus Miss.), which is overlain by a member with abundant fossils of the Tommotian assemblage. The diamictites represent glaciomarine deposits formed in the uppermost Ediacaran or lowermost Cambrian stage (Upper Vendian).

The Bokson Gr. occurs in the Oka River basin, in the southern part of the East Sayan Mountains, Buryatian Republic of the Russian Federation. Intensive investigation of the region began in the midtwentieth century in search of bauxite (Il’ina 1958) and phosphorite (Volkov et al. 1972), with detailed stratigraphic work published later (e.g. Khomentovskiy et al. 1985; Terleev & Zadorozhnyi 1996; and others). Reconstruction of sedimentation history (Kheraskova & Samygin 1992) with specific emphasis on the nature of the diamictites (Osokin & Tyzhinov 1998) was made later. Initially the Bokson Gr. was defined as a formation but further studies resulted in dividing it into five, sometimes six, formations. Diamictites are usually described from the lower part of the Zabit Fm. (Kheraskova & Samygin 1992; Pokrovskiy et al. 1999; and others) or from the basal, mainly terrigenous, Khushatay Fm. (Osokin & Tyzhinov 1998; Kuzmichev 2004). The type section of the diamictic member of the Zabit Fm. occurs in the Bokson River basin (Fig. 23.1, I) and that of the Khushatay Fm. in the Sarkhoy River basin (Fig. 23.1, II).

Structural framework The Bokson Gr. forms a platform cover of the Tuva-Mongolian Massif, unconformably lying above the Tonian-Cryogenian? (Upper Riphean) folded basement composed of deposits shed in an island arc and active continental margin settings. During the early Palaeozoic Caledonian folding, the Bokson Gr. experienced significant deformation, which makes correlation of units between different sites difficult: deformation includes folding, disjunctive faulting and overthrusting (Terleev & Zadorozhnyi 1996; Fedotova & Khain 2002). In the north, the deposits were metamorphosed to chlorite-sericite phase, less frequently than quartzbiotite shales (Kheraskova & Samygin 1992; Kuzmichev 2004). The Bokson Gr. fills the Bokson-Sarkhoy Synclinorium in the northern part of the Tuva-Mongolian Massif (Fig. 23.1). The lower part of the group is of flyschoid type and the upper part of the group consists of typical carbonate platform sediments.

Stratigraphy Diamictites form the lower part of the Bokson Gr. and unconformably overlie Tonian-Cryogenian? (Upper Riphean) ophiolites, diabases, turbidites and conglomeratic breccias of the Dzhundzhugur Fm. or the sedimentary-volcanogenic Sarkhoy Gr. whose

volcanic rocks have yielded Rb – Sr ages of 718 + 30 Ma (erochrona, Buyakayte et al. 1989). Both thickness and composition of the lower part of Bokson Gr. vary noticeably laterally. In the type section the lowermost Bokson Gr. is the Zabit Fm. A composite section of the lower part of the Bokson Gr. at the type section consists of the following from the base upwards (Kheraskova & Samygin 1992; Osokin & Tyzhinov 1998; Kuzmichev 2004): Sarkhoy Gr. Dzhundzhugur Fm., Sarkhoy Gr. (1) Deformed basic volcanogenic rocks, dark-coloured clayey and siliceous shales, diabases. (2) Weathering crust (a few metres). Angular unconformity The Boxon. The Zabit Fm. (3) Boulder and large pebbled conglomerates composed mainly of fragments of underlying deposits, less frequently granites (3 –40 m). (4) Diamictites with a mudstone matrix with scattered stones of different size, composition and roundness (0 –100 m). (5) Grey and dark grey dolomites, often clastic and oncolitic with abundant sedimentary structures indicative of subaqueous slumping (200 –225 m). (6) Breccia consisting of both angular and rounded fragments of dolomite and carbonate matrix; frequent structures indicative of underwater slumping (5– 15 m). (7) Massive, locally sandy, grey and dark grey dolomites (15 m). (8) Yellow-grey dolomitic phosphorites and phosphatic dolomites (30 –35 m). (9) Dark grey dolomites, partly brecciated and stromatolitic (140 –200 m). Hiatus and palaeokarst The Tubinzurta Fm. (10) Siltstone member containing a bauxite bed. (11) Variegated and black siltstones, breccia-like dolomites. (12) Grey and black silicified dolomites. Hiatus and palaeokarst The Khuzhirtai Fm. (13) Dolomites passing into dark grey limestones with archaeotheans, trilobites and brachiopods characteristic of the Atdabanian Stage of the Lower Cambrian (.500 m).

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 285– 288. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.23

286

N. M. CHUMAKOV

100

o

Irkutsk 52o

Baikal L. 0

(a)

100 km

Khubsugul L. 100 o

104

o

52 o 1

5

2

6

3

7

4

8

0

10

20 km

(b)

In the western part of the Bokson-Sarkhoy Synclinorium, along the Sarkhoy River (Fig. 23.1, II), the Zabit Fm. is separated from the Cryogenian basement by a thick (up to 500 m) red clastic sequence, the Khushatay Fm. of the Bokson Gr. Some researchers (Osokin & Tyzhinov 1998; Kuzmichev 2004) reported the presence of diamictites at the base and top of Khushatay Fm. In the middle part of the Khushatay Fm., black clayey limestones and shales up to 150 m thick occur (Kheraskova & Samygin 1992). Sections of the Khushatay Fm change significantly over short distances. It is likely, that the basal diamictite member of the Zabit Fm. is coeval to the upper part of the Khushatay Fm.

Glaciogenic deposits and associated strata The diamictites of the Zabit and the Khushatay Fm. have a wide lateral distribution (more than 130  360 km2, taking into account the presence of Bokson Gr. equivalents in the northern Mongolian Khubsugul area). Lithology of the diamictite members of the Bokson Gr. was briefly described by Osokin & Tyzhinov (1998) and Kheraskov & Samygin (1992). These diamictites are massive, but sometimes crudely bedded. Thickness of the massive beds varies from 20 to 55 m. These are composed of greenish and cherry-coloured shale or carbonate shale matrix with scattered pebbles, and small and large boulders derived from rocks underlying the Bokson Gr.: sandstones, quartzites, jaspers, diabases and gabbro, as well as pebbles and boulders of sandstones and dolomites containing stromatolites or fossils of the algae Renalcis. Lenses of dolomitic breccias occur locally. Dolomitic boulders and breccias are confined to the sections where the Zabit Fm. is separated from the Cryogenian basement by the Khushatay Fm., which includes dolomites. The diamictites contain erratic stones similar to the granitoids of pre-Mesoproterozoic crystalline basement that crop out on the Gargan Massif to the east. Semi-angular and poorly rounded stones dominate over rounded ones. Some clast are of flat-iron shape, with flattened and striated faces. Stones are dispersed irregularly and sometimes form ‘stone nests’. The diamictites include rare conglomerate interbeds and sandstone lenses. The massive diamictites are associated with lenses and subordinate beds of thin and rhythmically bedded shales and mudstones containing scattered dropstones, which press through or break underlying beds and are themselves in turn draped by overlying beds. Dolomites, which overlie the diamictites, are grey or black, massive or brecciated, and sometimes stromatolitic.

Fig. 23.1. Geographic position of the Bokson Gr. and its main sections. (a) Map showing general location of study area (box). (b) Geological map of described area. 1, granitoides; 2, Middle Cambrian– Ordovician; 3, the Bokson Group, latest Ediacaran (Upper Vendian) to Middle Cambrian; 4, Tonian-Cryogenian? (Upper Riphean); 5, pre-Mesoproterozoic (preRiphean) crystalline basement; 6, main thrusts; 7, main faults; 8, boundary of the Russian Federation and Mongolia; I, Bokson River Basin; II, Sarkhoy River Basin.

horizon (Kuzmichev 2004). In the Sorok River basin and elsewhere, the boundary between the underlying deposits and the diamictites is gradational (Osokin & Tyzhinov 1998). The upper boundary of the diamictite member with overlying dolomites is sharp but conformable and locally this boundary is gradational (Osokin & Tyzhinov 1998; Kuzmichev 2004).

Chemostratigraphy Isotopic analyses (oxygen and carbon) of the Bokson Gr. carbonate rocks were carried out by B.G. Pokrovskiy and coauthors (Pokrovskiy et al. 1999). Results obtained are illustrated in Figure 23.2. Carbonate samples were dissolved in H2PO4 at 25 8C for 1.5 h in the case of the calcitic fraction; dissolution was made at 100 8C for 1 h for the dolomitic fraction. A correction to isotopic oxygen values was introduced on different fractionation of calcite and dolomite (þ1.1‰ for the latter). The d13C range determined for calcite varied from 24.4 to 24.9‰ in beds that underlie the diamictites; from 22.9 to 1.8‰ for the Zabit Fm. units that overlie the diamictites; from – 4 to 1.5‰ for the Tabinzurt Fm.; from –2.1 to 0.3‰ for the Khuzhirtay Fm.; and from –2.1 to 0.7‰ for the lower part of the Nyurga Fm. The isotopic d13C curve demonstrates distinct positive values corresponding to the beds under the diamictites: three negative peaks for the lower carbonate part of the Zabit Fm.; an interval of mainly positive low values for the upper Zabit – lower Tabinzurt Fm. beds, a sharp drop in value for the Tabinzurt beds that overlie the bauxite member; change with some oscillations to almost zero for the uppermost Tabinzurt Fm.; and low negative values for the Khuzhirtay Fm.

Other characteristics The member of dolomites with interbeds of microcrystalline dolomitic phosphorite occurs in the upper part of the Zabit Fm. in the Bokson area. The thickness of this member varies from 35 to 50 m. Total thickness of phosphorite beds is on average 5.5 m. Average contents of P2O5 in this phosphorite is 8.3% (Yanshin & Zharkov 1986). The basal bauxite member of the Tabinzurt Fm. overlies a karst surface of the Zabit Fm. in the Bokson area (Kuzmichev 2004). This member is 4– 23 m thick and consists of massive, brecciated and laminated reddish bauxite consisting of diaspore, boehmite and a higher proportion of Fe2O3 and SiO2 (Il’ina 1958).

Boundary relations with overlying and underlying units Palaeolatitude and palaeogeography In most sections, the diamictites and basal conglomerates unconformably overlie the Sarkhoy Gr. (Osokin & Tyzhinov 1998). In the Bokson River basin their contact is marked by a weathering

No valid results of palaeomagnetic investigations have been obtained and published for the Bokson Gr. and adjacent

BOKSON GROUP, EAST SAYAN MOUNTAINS

287

Fig. 23.2. Lithological and chemostratigraphic succession of the lower part of Bokson Gr. (from Pokrovskiy et al. 1999, with additions). 1, limestones; 2, massive dolostones; 3, stromatolitic dolostones; 4, dark dolostones, calcarenites and calcirudites; 5, phosphatic calcarenites; 6, limey shales; 7, bauxites; 8, argillites and siltstones; 9, diamictites; 10, eluvial debris; 11, silicification; 12a, calcite; 12b, dolomite; ]1, Lower Cambrian; V2, Ediacaran-Cambrian (Upper Vendian); R3, Cryogenian (Upper Riphean) (Sorkhoy Gr.); SS, Nemakit-Daldyn SSF (small shelly fossils); Cl, Cloudina; Mf, microfossils.

deposits. Following regional palaeotectonic reconstructions for the late Cryogenian and Ediacaran (Early and Late Vendian), the Tuva-Mongolian Massif and its constituents seems to be in low latitudes of around 08 (Fedotova & Khain 2002) and 2308 (Smith 2001) respectively.

Geochronological constraints The Rb –Sr age of volcanic rocks in the Sorkhoy Gr. unconformably underlying the Bokson Gr. has been estimated to be 718 + 30 Ma (erochrona, Buyakayte et al. 1989). The basal Khushatay Fm. of the Bokson Gr. has yielded the acritarch Granomarginatasphaeria judomica Pjat. and multiserial sheaths of organic-walled microfossils Polytrichoides lineatus Hermann, suggestive of an Ediacaran (Vendian) age (Veis & Vorob’eva 1993). Cloudina sp. (attributed by A. Y. Zhuravlev) was recovered from granule of the diamictites (Kheraskova & Samygin 1992), and the SSFs Cambrotubulus decurvatus Miss. and Anabarites trisulcatus Miss., characteristic of the Nemakit-Daldynian (Manykaian) Stage of Siberia, were recovered 30 m above the diamictitic member (Terleev & Zadorozhnyi 1996; Postnikov & Terleev 2004). Higher up the section, the phosphoritic member of the Zabit Fm. contains abundant assemblages of SSF (Cambrotubulus decurvatus Miss., Anabarites trisulcatus Miss., Tisitheca sp., Igorella sp., Fomitchella infundibuliformis Miss.), silicified algae (Obruchevella magna Golov., O. parva Reitl., Heliconema sp., Eosynechococcus meorei Hoffm. and other forms), radiolarians (Palaeocenosphaera parva Nazarov, Entasitinia sp.) and other fossils characteristic of the Lower Cambrian Tommotian stage (Postnikov & Terleev 2004).

Discussion Sedimentary environments A glacial origin of the Zabit diamictites is suggested by the wide distribution of massive and bedded diamictites at the same stratigraphic interval, the presence of erratic and ‘flat iron’ shaped stones with striated faces and dropstones in thin-bedded rocks (Osokin & Tyzhinov 1998). Osokin & Tyzhinov (1998) proposed that the diamictites originated in glacial terrestrial, lacustrine and shelf environments. The opinion of these and other authors (Kuzmichev 2004 and others) about the glacial origin for the diamictites seems correct. However, the presence of continental glacial deposits in the Bokson Gr. is doubtful. There are no signs of glaciodislocations, zones of assimilation or glacial pavements at the base of massive diamictites. The preferred orientation of elongated stones typically found in basal tillites does not occur in the Bokson diamictites. On the contrary, the stones are oriented in a random fashion and often form ‘stone nests’. Gradual transitions between alternating diamictites and thin-bedded sediments and numerous indicators of underwater slumping suggest that the Bokson diamictites are glaciomarine deposits, partly reworked by slumps and debris flows, as noted by Kheraskova & Samygin (1992). There are other interpretations of the Bokson Gr. diamictites. Some researchers suggest that these diamictites represent sedimentary olistostromes formed by reworking of Sarkhoy Gr. deposits (Khomentovskiy et al. 1985). Other geologists consider the diamictites to be tectonic olistostromes formed in front of the nappe where the Bokson Gr. overthrusts Palaeozoic deposits (Terleev & Zadorozhnyi 1996; Fedotova & Khain 2002; and others). This

288

N. M. CHUMAKOV

tectonic interpretation contradicts the fact that the Bokson Group has a gradational contact with the underlying beds in some areas, convincingly demonstrated by regional mapping (Khomentovskiy et al. 1985; Kheraskova & Samygin 1992; Kuzmichev 2004). Another argument against tectonic olistostromes is the fact that the carbonate stones in the diamictites have high positive d13C, similar only to carbonate rocks below the diamictites. Carbonate rocks of the Bokson Gr. that overlie the diamictites have only negative d13C values. Osokin & Tyzhinov (1998) suggested a glacial origin for the carbonate breccia of the Zabit Fm. (see bed 6 in stratigraphy section described above). They base this on scattered carbonate gravel and small pebbles that occur in some lenses of rhythmically thin-bedded limestones and dolomites related with the breccias. The description of the composite type section presented above, notes that the breccias and enclosing deposits bear abundant traces of underwater slumping, suggestive that breccias were most likely products of slides and slumps on slopes of the carbonate platform. The rhythmically and thin bedded limestones and dolomites appear to represent sediments deposited from related turbidity currents. A non-glacial origin is supported by the absence of any foreign clasts in the breccias and thin-bedded rocks.

Age In terms of the framework proposed by the International Stratigraphic Committee, the diamictites of Bokson Fm. can be referred to the uppermost part of the Ediacaran or the lowermost part of the Lower Cambrian (Nemakit-Daldyn Stage). In terms of the Russian stratigraphic scale, diamictites of the Bokson Fm. can be referred to the uppermost part of the Kotlin or the lowermost part of the Nemakit-Daldyn Stages of the Vendian. The diamictites of the Zabit Fm. occur above the Khushatai Fm., which contains microfossils, similar to those of the latest Ediacaran (Late Vendian) (Veis & Vorob’eva 1993). The diamictites contain Cloudina sp. (Kheraskova & Samygin 1992) and lie below beds containing Cambrotubulus decurvatus Miss. and Anabarites trisulcatus Miss. (Terleev & Zadorozhnyi 1996; Postnikov & Terleev 2004), indicative of a Nemakit-Daldynian age. These data together with the Tommotian fossils present higher in the Zabit Fm. allow the diamictites to be referred to the latest Ediacaran (Late Vendian) (Kuzmichev 2004), and more specifically, either a slightly antedated Nemakit-Daldyn age or concurrent to its commencement (Chumakov 2009a, b). The Cloudina in the diamictites was likely redeposited from underlying succession. The lowermost negative d13C anomaly in the Zabit Fm. can be correlated to the negative anomaly of d13C at the base of the Nemakit-Daldyn stage (Pokrovskiy & Missarzhevskiy 1993). On an interregional scale, it is possible to correlate the Bokson diamictites to the Baykonur Glacial Horizon in Kazakhstan and Kyrgyzstan (Chumakov 1981, 2011) and to the Hankalchough and Luoguan formations of northern China (Baode et al. 1986), based on their similar stratigraphic position. Research projects were supported by grants of the Russian Fund of Basic Investigations No. 11-05-00232 and Program No. 25 of the Presidium of the Russian Academy of the Sciences. The author would like to acknowledge stimulating discussions arising from participation in IGCP project 512 ‘Neoproterozoic Ice Ages’ (http://groups.google.com/IGCP-512?nl¼enIGCP-512@?hl¼en). This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Buyakayte, M. I., Kuzmichev, A. B. & Sokolov, D. D. 1989. 718 million year – Rb –Sr erochrona Sarkhoy Group of East Sayana. Doklady Academy of Sciences of USSR, 309, 150–154. Chumakov, N. M. 1981. Upper Proterozoic glaciogenic rocks and their stratigraphic significance. Precambrian Research, 15, 373–396. Chumakov, N. M. 2009a. The Baykonurian Glaciohorizon of the Late Vendian. Stratigraphy and Geological Correlation, 17, 373–381. Chumakov, N. M. 2009b. Neoproterozoic Glacial Events in Eurasia. In: Gaucher, C., Sial, A. N., Halverson, G. P. & Frimmel, H. E. (eds) Neoproterozoic-Cambrian Tectonics, Global Change and Evolution: a focus on southwestern Gondwana. Developments in Precambrian Geology, 16, Elsevier, The Netherlands, 389– 403. Chumakov, N. M. 2011. Glacial deposits of the Baykonur Formation, Kazakhstan and Kyrgyzstan. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 303–307. Fedotova, A. A. & Khain, E. V. 2002. Tectonics of the south East Sayan and its position in the Ural– Mongolian Belt (in Russian). Nauchnyi Mir, Moscow. Guan Baode, Wu Ruitang, Hambrey, M. J. & Geng Wu. 1986. Glacial sediments and erosional pavement near the Cambrian – Precambrian boundary in western Henan Province, China. Journal of Geological Society, London, 143, 311–323. Il’ina, A. V. 1958. Geology and origin of Boxon bauxites in East Sayan. In: Dolgopolov, N. N. (ed.) Bauxites their Mineralogenesis Origin (in Russian). Academy of Sciences of USSR Press, Moscow, 267–281. Kheraskova, T. N. & Samygin, S. G. 1992. Tectonic environments of formation of Vendian-Middle Cambrian clastic-carbonates sequence of the East Sayan. Geotectonics, 6, 18 –36. Khomentovskiy, V. V., Pak, K. L., Postnikov, A. A. & Skopintsev, V. G. 1985. Geology structure of the basin of Ukha-Gol River (in Russian). In: Khomentovskiy, V. V. (ed.) Stratigraphy of the Late Precambrian and Lower Palaeozoic of Siberia. Riphean and Vendian. Institute of the Geology and Geophysics, Novosibirsk, 76–106. Kuzmichev, A. B. 2004. Tectonic History of the Tuvino-Mongolian Massif (in Russian). Probel-2000, Moscow. Osokin, P. V. & Tyzhinov, A. V. 1998. Precambrian tilloids OkaKhubsugul phosphorite basin (East Sayan, North-West Mongolia). Lithology and Mineral Resources, 2, 162–176. Pokrovskiy, B. G. & Missarzhevskiy, V. V. 1993. Isotopic correlations boundary sections of Precambrian and Cambrian of Siberian Platform. Doklady Academii Nayk SSSR, 329, 768– 771. Pokrovskiy, B. G., Letnikova, E. F. & Samygin, S. G. 1999. Isotopic stratigraphy of Boxon Group, Vendian-Cambrian of East Sayan. Stratigraphy and Geological Correlation, 7, 23 –41. Postnikov, A. A. & Terleev, A. A. 2004. Stratigraphy of Neoproterozoic of the Altay-Sayan folded region. Geology and Geophysics, 45, 295–309. Smith, A. G. 2001. Palaeomagnetically and tectonically based global maps for Vendian to Mid-Ordovician time. In: Zuravlev, A. Yu & Riding, R. (eds) The Ecology of the Cambrian Radiation. Columbia University Press, New York, 11– 46. Terleev, A. A. & Zadorozhnyi, V. M. 1996. Discovery of Palaeozoic Foraminifera in ‘Precambrian’ of East Sayan (Sarkhoy River). Doklady Academy of Sciences of Russia, 351, 373–374. Veis, A. F. & Vorob’eva, N. G. 1993. First finds wall-organic microfossils in Upper Precambrian of Boxon-Sarkhoy basin (East Sayan). Stratigraphy and Geological Correlation, 1, 27 –32. Volkov, R. I., Zaytsev, N. S., Il’in, A. V. & Osokin, P. V. 1972. Ukhangol phosphorite field of East Sayan. Soviet Geology, 2, 94 – 107. Yanshin, A. L. & Zharkov, M. A. 1986. Phosphorus and Potassium in Nature (in Russian). Nauka, Novosibirsk.

Chapter 24 The Neoproterozoic glacial formations of the North and Middle Urals NICKOLAY M. CHUMAKOV Geological Institute Russian Academy of Sciences, Pyzhevskiy per. 7, Moscow 109017, Russia (e-mail: [email protected]) Abstract: In the North and Middle Urals, four Neoproterozoic diamictite-dominating units are known: the Churochnaya, Tany and Koyva formations and Lower Starye Pechi Subformation. A glacial origin for the diamictites is indicated by a number of characteristic features such as erratic stones, striated and faceted clasts, shales with dropstones, preferential orientation of elongated stones, the wide distribution and confinement to certain stratigraphic levels of the deposits, and the association with distinct post-glacial dolostones. Most diamictites were deposited in a glaciomarine environment and sourced from an ice sheet on the East European craton. The Churochnaya Formation also contains subordinate terrestrial and probably seasonal sea ice deposits. This formation accumulated on a glaciated continental shelf, while the Tany, Koyva and Starye Pechi formations were deposited on the outer shelf and continental slope of the eastern margin of the East European Craton. The Middle Ural sections contain the White Sea (Ediacaran) Metazoa assemblage, typical late Neoproterozoic microfossils, stromatolite associations, and radiometric dates obtained from volcanic tuffs, granosyenites, trachytes and trachyandesites. The combined data suggest that the Lower Starye Pechi Subformation was deposited in the middle of the Ediacaran Period (Early Vendian). The Churochnaya Formation of the North Urals and the Tany and Koyva formations of the Middle Urals were likely deposited in the late Cryogenian period (Early Vendian).

Neoproterozoic glacial deposits are known in the western parts of the North, Middle and South Urals. They are best exposed and studied along the Pulyudov Ridge and Churochnaya River in the northern Urals and along tributaries of the Chusovaya River in the Middle Urals (Fig. 24.1). This chapter focuses on glacial deposits of the North and Middle Urals; glacial deposits of the South Urals were recently described and are reviewed elsewhere (Chumakov 1981a, b, c, 1992, 1998; Maslov 2000). Until the mid-1960s, the late Precambrian strata of the North and Middle Urals were regarded as Palaeozoic in age and the occurrence of associated ‘tillite-like conglomerates’ (diamictites) was explained by subaqueous slumping. This view was reflected in many publications and geological maps of the Urals and USSR territory. Long regional investigations resulted in the elaboration of the late Precambrian timescale as reviewed by Ablizin et al. (1982). Modern stratigraphic investigations of Neoproterozoic sedimentary sequences in the Urals, including glacial deposits, began in the second half of the twentieth century during lithological investigations (the Polyudov Ridge; Borovko 1967) and geological mapping (the Middle Urals, Ablizin et al. 1982) that was verified by Ehlakov & Morozov (2006). Glacial deposits of the Polyudov Ridge in the North Urals are confined to the Churochnaya Formation. Three glacial levels were established in the Middle Urals: the Tany and Koyva formations of the Serebryanka Group and in the Lower Starye Pechi Subformation of the overlying Sylvitsa Group (Fig. 24.2). The Tany and Koyva formations are replaced eastward by the Wil’va Formation, which contains some diamictite beds. The possible occurrence of diamictites in the Us’va, Fedotovka and Kernos formations has not been confirmed by recent investigations. The Us’va and Fedotovka shales locally contain single lonestones.

Structural framework The Neoproterozoic deposits of the Polyudov Ridge accumulated in the marginal shelf zone of the Russian platform and deformed during the Timanian and, more significantly, during the Hercynian orogeny. Now they are exposed in the cores of marginal brachyanticlines of the Uralian folded zone, disrupted by many faults, and variably foliated. In the Middle Urals, Neoproterozoic sedimentation occurred in the outer shelf and continental slope environments. Sections of this region are

three times thicker than those of the Polyudov Ridge and contain a higher percentage of siliciclastics in the form of turbidites and shales (i.e. flysch). The Neoproterozoic deposits of the middle Urals, along with younger deposits, were locally heavily deformed during the Hercynian, with isoclinal folds and numerous faults and overthrusts. The Tany, Koyva and Starye Pechi formations were metamorphosed to lower greenschists facies at this time. The Wil’va Formation was metamorphosed to greenschist facies.

Stratigraphy Neoproterozoic stratigraphy, lithologies and pertinent radiometric ages of the northern and middle Urals under consideration are schematically shown in Figure 24.2. The Polyudov Ridge section begins with the thick (800 –900 m) siliciclastic Rassolny Formation (not shown in Fig. 24.2), which is overlain by pink limestones and marls of the Deminskaya Formation. Above is the Niz’va Formation, comprising mainly calcareous dolomites with stromatolites similar to those of the late Mesoproterozoic –early Neoproterozoic (Upper Riphean) Min’yar and Uk formations of the South Urals (Raaben 2007). The Niz’va Formation is unconformably overlain by the clastic Ust-Churochnaya Formation, which, in turn, is unconformably overlain by the Churochnaya Formation. The latter consists of red sandstones in the lower half and diamictites with an interval of dolomite and black shales in the upper part. The black shales are overlain by a thick sandy-shaly sequence, which is divided by an unconformity into the Il’ya-Vosh and Kocheshor formations. The late Precambrian succession of the Middle Urals is more complicated. Sequences that are approximately coeval with those of the Polyudov Ridge are considerably thicker, more lithologically diverse, and divisible into four groups, each separated by a depositional hiatus. Each group is further subdivided into formations as shown in Figure 24.2.

Glaciogenic deposits and associated strata The Churochnaya Formation This formation is 350–500 m thick. Its lower part (c. 150 m) consists of red sandstones. The upper part of the formation

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 289– 296. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.24

290

o

58

o

Po

Vi she ra

R.

56

N. M. CHUMAKOV

d ly ov

Moscow

Perm

dg Ri

Krasnovishersk

e

m Ka .

aR

60

o

Phanerozoic Solikamck

Sylvitsa Gr., Kocheshor, Il’ay-Vozh Fms Serebryanka Gr., Churochnzya, Ust-Churochnaya Fms Riphean main disjunctive dislocations Main Neoproterozoic sections Mezhevaya Utka R. Serebryanka R. Sylvitsa R. Koyva R. Vil’va R. Us’va R. Churochnaya R.

.

aR

’v os

K

.

va R

Us’

aya

R.

ov hus

C

58 o

Perm 0

25

50 km

(c. 200–350 m) is composed of massive and bedded green-grey or reddish diamictites containing some beds of sandstones, shales and conglomerates (Fig. 24.3). The diamictites are overlain by the dolomite member (10 m) and above it by black shales (.20 m). The diamictites have a sand-silty matrix with dispersed stones (2 –10% of the diamictite volume) of variable size, roundness and composition. Some diamictites show indistinct bedding, thin lenses and interbeds of sandstones and shales, and nested clasts. In addition to the predominant sandstone and quartzite, there are many erratic stones, including granites, gneisses and schists derived from the crystalline basement of the Russian plate. The proportion of erratic stones increases upsection. Between 3 and 7% of stones are striated, grooved and faceted. Rare diamictites beds contain a large proportion of well-rounded stones of medium size. Some diamictites preserved a preferred orientation among elongated stones. Rhythmically banded shales with dropstones occur at the base of the diamictites. The diamictites are overlain by a dolomite member (up to 10 m). This dolostone is pale, slightly calcareous (1 –7% CaCO3), distinct in structure and texture, and commonly massive or brecciated. In places it is thinly bedded and distorted by subaqueous slumps. Abundant carbonate, chert and quartz veins cut the dolomite

Fig. 24.1. Late Mesoproterozoic to Neoproterozoic (Vendian and Upper Riphean deposits of the North and Middle Urals).

and encrust fragments of breccia. Thin, intercalated black chert (2– 3 cm) overlies the dolostone. The chert is followed by .20 m of black shale with a relatively high phosphorous content (0.1 –1%). The black shales are overlain by multicolour shales with sandstone intercalations. Boundary relations with overlying and underlying non-glacial units. Conglomerates at the base of the Churochnaya Formation

rest on a probable unconformity. Breccias in the upper part of the cap dolomite are apparently related to weathering and karstification. Nevertheless, the hiatus between the dolomite and overlying chert was probably not significant, as a 1.5-m-thick diamictite bed occurs 12 m above the base of the black shale member. Chemostratigraphy. Dolomites and limestones of the Niz’va Formation, which unconformably underlies the Churochnaya Formation, are characterized by slightly variable but on average low, positive d13C values ( –0.8 to þ2.5‰) and stable d18O values ( –10 to – 6‰). Dolomites overlying the diamictites of the Churochnaya Formation have stable negative d13C values (from –3 to – 5‰).

NEOPROTEROZOIC OF THE NORTH AND MIDDLE URALS

Polyudov Ridge O3

Central Ural O3

. .. .. .. .. . . .. .. .. .. . .....

500 m .. .. .. .. ..

.....

545 K-Ar, g 570 K-Ar, g

..... ..... ..... .....

Ust’-Sylvitsa Fm.

Sylvitsa Gr.

Kocheshor Fm.

350 m

Chernyi Kamen’ Fm. 557 ± 13 U-Pb, z 1700 m Perevolok Fm. 300 m

Starye Pechi Fm.

. .. .. .. .. . . .. .. .. .. . ..... . .. .. .. .. . ..... .. .. .. .. ..

..... .....

300 m

Kernos Fm.

Il’ay-Vozh Fm.

300 m

Churochnaya Fm. . .. .. .. .. . .....

Serebrayanka Gr.

480 m .. .. .. .. ..

610; 620 K-Ar, g

350 m

291

..... ..... ..... .....

diamictite . .. .. .. .. . .....

sandstone shales dropstone limestone dolstone volcanic rock

Buton Fm. 400 m

Koyva Fm.

cap dolomite

700 m

569 ± 42 Sm; 559 ± 16 Rb Garevka Fm. 200-700 m

Tany Fm.

stromatolite ..... ..... ..... ..... .....

800 m ..... . .. .. .. .. .

Us’va Fm.

..... .....

Fedotovka Fm.

720 m . . . . .

635 K-Ar, g

..... . .. .. .. .. .

Basegi Gr.

Ust’Churochnaya Fm.

..... .....

1200 m . . . . . .....

1200 m

Shcherogrovitsy Fm. 0-900 m

671+24 Rb; 671+24 U-Pb, z Oslyanka Fm.

Niz’va Fm. 950 m

750 K-Ar, g Deminskaya Fm. 300 m

Kedrovka Gr.

300 m

..... ..... .....

Klyktan Fm. Fig. 24.2. Stratigraphic position of Vendian glacials at the North and Middle Urals (Ablizin et al. 1982; Chumakov 1998; Maslov et al. 2007). Radiometric dating techniques: g, glauconite; z, zircon.

1300 m

Sinii Gory Fm. 1000 m

..... ..... ..... .....

Tany Formation

Boundary relations with overlying and underlying non-glacial units. The lower boundary of the Tany Formation is probably

The Tany Formation (up to 800 m thick) consists of two members of massive dark grey diamictites divided by a sandstone member. The base of the thick Lower Member (c. 350–450 m) consists of alternating mafic volcanics, schists, limestones and diamictites (Ablizin et al. 1982). The lower member of diamictites in some section is capped by a dolomite bed 8 m thick (Fig. 24.4). The lithology of diamictites clasts is diverse and includes quartz arentites, granites, gneisses and carbonates. A considerable portion (up to 45%) of the diamictite clasts are felsic igneous rocks. The abundance and size of granites and gneiss clasts increase to the SW (towards the Russian platform) and in the upper part of the lower member (Ablizin et al. 1982). Some blocks of granite and gneiss are up to 3.5 m in diameter. Nests of medium-sized and large stones are common. Some rare shale interbeds contain dropstones. The Middle Member consists of bedded quartzfeldspathic sandstones (80– 150 m). The Upper Member includes massive and bedded diamictites and thin intercalations of laminated shales with dropstones (Chumakov 1992).

erosional judging from the abundance of clasts from the underlying Kedrovka and Basegi groups within the basal diamictites. The upper boundary of the Tany Formation is sharp and conformable.

Koyva Formation The Koyva Formation contains diamictites only to the north of the Sylvitsa River. In general, diamictites are similar to those of the Tany Formation but differ in their reddish colour and confinement of bedded diamictites and laminated shales with dropstones to the upper part of the formation. Father to the north, at the Vil’va, Us’va and Kos’va rivers, the Koyva Formation contains alkali basalts flows. In the stratotype section at the Koyva River, diamictites are overlain by a transitional member of thinbedded variegated shales with small dropstones and then by the dolomite member (up to 6 m thick; Fig. 24.5). Thin (1– 2 cm)

292

N. M. CHUMAKOV

Lower Starye Pechi Subformation massive diamictites sedimentary and quartzites stones granitoid and gneisses stones diamictites with well rounded stones sandstones with pebbles sandstones shales

In the Lower Starye Pechi Subformation, diamictites alternate with sandstones and are overlain by thin-bedded shales containing scattered sand grains and pebble dropstones (Fig. 24.6). The matrix of the diamictites is dark grey sandy-silt, and the diamictites are bedded in the upper part of the formaton. Clasts in the diamictite vary in size and shape; they are dominated by fragments of the underlying rocks, but some erratic pebbles of quartz, quartzites and plagio-granites occur as well. Striated and grooved clasts have been found. Diamictite wedges are found at the base of some of the diamictite beds.

dolomites chert black rocks brown-red rocks striated stones dropstones

Boundary relations with overlying and underlying non-glacial units. The contact between the Lower Starye Pechi Subformation

and the underlying Kernos Formation is generally assumed to be an unconformity (Ablizin et al. 1982). Nevertheless, locally the lower Starye Pechi Formation contains interbedded sandstones, which are typical of the Kernos Formation, and thus imply a transitional contact (Chumakov 1998). The upper boundary of the Starye Pechi Subformation is transitional.

stone fabric ch

.. .. ==

stone nests

rhythmic lamination

==

Geochronological constraints

vague bedding

subaqueous slump structure higher content of phosphorus

== 30

== ..

0 m

Fig. 24.3. Representative stratigraphic section of the upper Ghuroghnaya Formation, Ghurochnaya River.

interbeds of dolomites occur in rhythmically laminated shales overlying the dolomite. Boundary relations with overlying and underlying non-glacial units. The lower and upper boundaries of the Koyva Formation are

transitional (Ablizin et al. 1982). Chemostratigraphy. In the stratotype section at the Koyva River the

upper dolomite member yielded d13C values from –3 to –5‰. South of the Koyva River, where there are no diamictites or shales with dropstones within the Kyva Formation, the basal bed of the upper dolomite member is 2 m thick and has d13C values from – 4 to –7‰ (A.V. Maslov, pers. comm.). Other characteristics. At the Us’va River, the lower part of the

Koyva Formation contains alkali basalts, tuffs and banded hematitic rocks and ores, the latter of which vary from 10 to 40 m thick.

Polyudov Ridge No precise radiometric data are available for the Neoproterozoic formations of the Polyudov Ridge. Numerous radiometric ages were obtained in the middle of last centenary by K –Ar dating of glauconites (Garris et al. 1964; Borovko 1967; Keller & Chumakov 1983). They show a rather wide range of ages. The average K –Ar ages recalculated with recent 40K decay constants (l40Kb- ¼ 4.962  10210 a21; l40Ke- ¼ 0.581  10210 a21) are 750 + 25 Ma for the Niz’va Fm., 660 + 30 Ma for the Ust’Churochnaya Fm., 610 + 20 Ma for the Il’ay-Vozh Fm., and 570 + 17 Ma for the lower and 545 + 15 Ma for the upper part of the Kocheshor Formation, respectively. The Niz’va Fm. contains abundant stromatolites. Their assemblage consists of Poludia mutabilis Raab., Gymnosolen ramsayi Steinm., Conophyton miloradovichi Raab., Poludia polimorpha Raab., Minjaria uralica Kryl., Linella ukka Kryl., and other forms typical of the Cryogenian (uppermost Upper Riphean) of the Middle and South Urals, the Timans and Svalbard (Raaben 2007). In general, K –Ar ages from the Niv’zva Formation are consistent with the chronology implied by stromatolite biostratigraphy.

Middle Urals Abundant non-skeletal Metazoa (Cyclomedusa davdi Sprigg, Dickinsonia tenuis Glaessner and Wade, Irridinitus multiradiatus Fedonkin, Vaizitsinia sophia Sokolov and Fedonkin and other forms) characteristic of the White Sea and Eriacaran biota occur in the Chernyi Kamen’ Formation in the upper Sylvitsa Group (Grazhdankin et al. 2005, 2007; Maslov et al. 2007). A U –Pb (SHRIMP-II) zircon age of 557 + 13 Ma was obtained on a tuff within the Chernyi Kamen’ Formation (Maslov et al. 2007). This age provides a maximum constraint for all of the Neoproterozoic diamictites of the Middle Urals. More controversial dates were obtained for the individual glacial formations of the succession using less reliable radiometric techniques. The diamictites of the Starey Pechi Subformation seem to be younger than 569 + 42 or 559 + 16 Ma based on Sm– Nd (clinopyroxene) and Rb – Sr (whole rock) ages from the Koyva trachyandesites (Petrov et al. 2005). The Sm –Nd age is imprecise, whereas the Rb – Sr age is similar to an age of 557 + 13 Ma

NEOPROTEROZOIC OF THE NORTH AND MIDDLE URALS

Mezhevaya Utka R. Serebryanka R.

293

Us’va R.

Garevka Fm. 3

dimictites, crystalline clasts dimictites, sedimentary clasts 2

bedded diamictites

Tany Fm.

laminated shales with dropstones sandstones shales 1

limestones dolomites striated stones dropstones rhythmic lamination 0m

bedding

50 100 150

Us’va Fm.

obtained by Maslov et al. from 1.5 km upsection (Sylvitsa Group). Macrofossil associations lower in the Sylvitsa Group are similar to the Miaohe biota of the uppermost Doushantou Formation of South China (Grazhdankin et al. 2007), implying a late Ediacaran age. It was long accepted that grano-syenites of the Troitsk massif (the Kos’va River) cut the Tany and Garevka formations of the Serebryanka Group and that they were emplaced between deposition of the Serebryanka and Sylvitsa groups. This hypothesis implies that the diamictites of the Tany Formation would be older than 630 + 20 (Pb/Pb zircon) or 621 + 12 Ma (Rb –Sr whole rock; Petrov et al. 2005). More recent mapping and dating, however, suggest that the Troitsk grano-syenites cut only the much older Shechegrovittsy Formation (Basegi Group) and have an age of 671 + 24 Ma (SHRIMP U –Pb zircon)(Ronkin et al. 2007). The differences between the ages may be related to two phases of magmatic activity in the Troitsk massif. In any case, the diamictites of the Koyva and Tany formations are constrained to between 557 + 13 and 671 + 24 Ma, which is consistent with biostratigraphic data.

Palaeolatitudes and palaeogeography Palaeomagnetic investigations of the late Proterozoic rocks of the South Urals have failed to reveal primary palaeomagnetic directions. According to palaeogeographic reconstructions by Li et al. (2008), the eastern margin of the Russian Platform was located

Fig. 24.4. Stratigraphic sections of the Tany Formation.

at the latitude of around 608S between 630 and 600 Ma, then displaced to 208S between 600 and 550 Ma.

Discussion Sedimentary environments Massive and bedded diamictites contain shale interbeds with dropstones, clasts with typical glacier striae, grooves and facets, erratic stones (including very large ones), and nests of stones. They are overlain by carbonates resembling typical Neoproterozoic ‘cap dolomites’. In combination, these features suggest that most diamictites of the Tany and Koyva formations and the Lower Starye Pechi Subformation are glaciomarine deposits. The same conclusion can be drawn for most diamictites of the Churochnaya Formation, except for members characterized by till-like fabric indicated by preferred orientations in elongated stones and the occurrence of varved shales, both of which suggest a continental origin. Some other members of the Churochnaya Formation contain mainly well-rounded clasts that may have been deposited by seasonal ice. The Niz’va, Ust-churochnaya and Churochnaya formations were deposited in inner shelf environments on the margin of the Russian Plate. The flysch-like nature of sediments lying under, between and over the diamictites of the Serebryanka Group and the Starye Pechi Formation show that they were deposited on the outer shelf to continental slope.

294

N. M. CHUMAKOV

Possible correlations

m diamictites bedded diamictites sandstones 70

multicolour shales black shales dolomites very large boulders raft of diamictites

. .. stone nests dropstones

26

lamination rhythmic lamination 6

20

6

The glaciogenic deposits of Starye Pechi Formation are older than those of the Chernyi Kamen’ Formation (which contain Edacarian Metazoa and a tuff dated as 557 + 13 Ma; Maslov et al. 2007). The Lower Starye Pechi Subformation may therefore be correlated tentatively to middle Ediacaran glacial horizons of the East European craton (the Glussk and Mortensnes formations) and the glacial Gaskiers Formation in Newfoundland. The Koyva Formation shares similar stratigraphy and facies with the Tany Formation, from which it is separated by a turbidite sequence (the Garevka Formation) that is noticeably reduced in thickness in some places and pinches out near the Koyva River. Therefore, the Koyva, Garevka and Tany formations were likely deposited during a single glacial epoch. This composite unit is slightly older than Buton Formation containing Ediacaran Obruchevella (Golovenok et al. 1989) and much younger than the Shchegrovitsy Formation and Troitsk grano-syenites massif (671 + 24 Ma). These are tentatively correlated to the lower glacial horizon of the Lower Vendian (upper Cryogenian) on the East European craton (the Blon’ and Smalfjord formations), the glacial Nantuo Formation in South China, and the Yerelina Subgroup of South Australia. The age and correlation of the Churochnaya Formation is subject to debate. The overlying Il’ya-Vozh and Kocheshor formations are separated by an erosional unconformity (Fig. 24.2), and it is possible that the Kocheshor Formation is equivalent to the lower Sylvitsa Goup. In this case, the Churochnaya and Il’yaVozh formations together may correlate with the upper Serebryanka Group (Maslov 2004). It follows from this hypothesis that the Churochnaya Formation may be coeval with the glacial Koyva Formation (Fig. 24.2). The traditional correlation of the Niz’va Formation to the middle Neoproterozoic of the Karatavian (Upper Riphean) stratotype of the South Urals and the Klytkan Formation of the Middle Urals is supported by similar stromatolites assemblages (Raaben 2007) and isotopic data (d13C values from –0.5 to 2.0‰ for the Niz’va Formation).

Regional palaeogeography

50

. .. 300

...

Fig. 24.5. Stratigraphic section of the upper Koyva Formation, Koyva River.

The lithology of diamictite clasts suggests that crystalline basement and sedimentary cover of the Russian plate were the main sources for the Churochnaya, Tany and Koyva formations. Local sources for the diamictite clasts have also been established, particularly for the Lower Starye Pechi Subformation (Ablizin et al. 1982) and Vil’va Formation in the eastern Urals (Suslov & Teterin 1997). The regional palaeogeographical reconstructions (Chumakov & Sergeev 2004) suggest that an inland ice sheet existed during Tany time on the eastern part of the Russian Plate, whereas a marginal belt of shelf glaciers lay to the east of it (Churochnaya, Tany and Koyva formations). Further east, sedimentation was dominated by turbidity currents and ice rafting in an outer shelf –slope environment (Tany, Koyva and Wil’va formation). The existence of extensive inland glacier sheets in the northeastern part of East European Craton is also evident in a number of other Early Vendian (late Cryogenian) glacial units on the craton. The most complete late Cryogenian glaciogenic sequences are located in the marginal parts of the East European craton and are further described elsewhere in this volume. Within the cratonic interior, glacial sedimentation occurred in aulocogens that developed in the middle-Late Riphean (late Mesoproterozoic– Neoproterozoic) (Bessonova & Chumakov 1969; Chumakov 1971, 1981b, c, 1992). Our research was supported by grant 11-05-00232 from the Russian Fund of Basic Investigations and Program No 25 of the Presidium of the Russian Academy of

NEOPROTEROZOIC OF THE NORTH AND MIDDLE URALS

295

diamictites

m

bedded diamictites sedimentary and quartzite stones granitoid stones

Sp2 30

sandstones with small pebbles calcareous sandstones massive quartz-feldspathic sandstones 4 1,5

Sp

shales shales with dropstones

5,5

dropstones

. .. Sp1

11

vague bedding

. .. 6

10

rhythmic lamination

till wedges

. .. stone nests

20 m

6 10 5 Kr 0 15

Sciences. The authors would like to acknowledge stimulating discussions arising from participation in IGCP project 512 ‘Neoproterozoic Ice Ages.’

References Ablizin, B. D., Klyuzhina, M. L., Kurbatskaya, F. A. & Kurbatskiy, A. M. 1982. Upper Riphean and Vendian of the West Slope of the Middle Ural (in Russian). Nayka, Moscow. Bessonova, V. Ya. & Chumakov, N. M. 1969. Upper Precambrian glacial deposits of western regions of USSR. Lithology and Mineral Resources, 2, 73– 89. Borovko, N. G. 1967. Vendian and Lower Paleozoic of Polyudov Ridge of the North Urals (in Russian). VSEGEI, Leningrad. Chumakov, N. M. 1971. Vendian glaciation of the Europe and North Atlantic. Doklady Academii Nauk SSSR, 198, 419– 422. Chumakov, N. M. 1981a. Late Precambrian Kurgashglya tilloids, soutern Urals. In: Hambrey, M. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 674–677.

Fig. 24.6. Stratigraphic section of lower Starye Pechi Formation, Sylvitsa River. Sp, Starye Pechi Formation; Kr, Kernos Formation.

Chumakov, N. M. 1981b. Late Precambrian glacial deposits of the Vilchitsy Formation of western regions of the USSR. In: Hambrey, M. & Harland, W. B. (eds) Earth’s PrePleistocene Glacial Record. Cambridge University Press, Cambridge, 655– 659. Chumakov, N. M. 1981c. Late Precambrian glacial deposits of the Blon Formation, Belorussia, USSR. In: Hambrey, M. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 660–662. Chumakov, N. M. 1992. The Problems of Old Glaciations (PrePleistocene Glaciogeology in the USSR). Harwood Academic Publishers, Pennsylvania. Chumakov, N. M. 1998. The key section of Vendian glacial deposits in the South Urals (Kurgashly Formation, Krivoluksky graben) in Russian. In: Knipper, A. L., Kurenkov, C. A. & Semikhatov, M. A. (eds) The Urals: Fundamental Problems of Geodynamics and Stratigraphy. Nauka, Moscow, 138 –153 (in Russian). Chumakov, N. M. & Sergeev, V. N. 2004. Problems of climatic zonality of the Late Precambrian. Climate and biotic events (in Russian) In: Semikhatov, M. A. & Chumakov, N. M. (eds) Climate During

296

N. M. CHUMAKOV

the Epochs of Principal Biosphere Rearrangements. Transactions of Geological Institute of RAS, 550, 271. Ehlakov, Yu. A. & Morozov, G. G. 2006. Stratigraphy (in Russian). In: Kudryashov, A. I. (ed.) Mineral-Stuff Resources of Permian territory. Knizhnaya Ploshchad’, Perm, 49– 63. Garris, M. A., Kasakov, G. A. & Keller, B. M. 1964. Geochronological scale of Upper Proterozoic (Riphean and Vendian). In: Absolute Age of Geological Formations. Nauka, Moscow, 431– 455 (in Russian). Golovenok, V. K., Belova, M. Iu. & Kurbatskaia, F. A. 1989. First find of Obruchevella Reitlinger in Vendian sediments of Middle Urals. Doklady Academii Nauk SSSR, 309, 701–705. Grazhdankin, D. V., Maslov, A. V., Mustill, T. M. R. & Krupenin, M. T. 2005. The Ediacaran White Sea biota in the Central Urals. Doklady Earth Sciences, 401, 382– 385. Grazhdankin, D. V., Nagovitsin, K. E. & Maslov, A. V. 2007. Late Vendian Miaohe-type ecological assemblage of the East European platform. Doklady Earth Sciences, 417, 1183–1187. Keller, B. M. & Chumakov, N. M. (eds) 1983. Stratotype of Riphean. Stratigraphy. Geochronology. Nauka, Moscow (in Russian). Li, Z. X., Bogdanova, S. V. et al. 2008 Assembly, configuration, and break-up history of Rodinia: a synthesis. Precambrian Research, 160, 179– 210. Maslov, A. V. 2000. Some features of sedimentation in Early Vendian at the South and Middle Ural. Lithology and Mineral Resources, 6, 624–639.

Maslov, A. V. 2004. Riphean and Vendian sedimentary sequences of the Timanides and Uralides, the eastern periphery of the East European Craton. In: Gee, D. G. & Pease, V. L. (eds) The Neoproterozoic Timanide Orogen of Eastern Baltica. Geological Society, London, Memoirs, 30, 19– 35. Maslov, A. V., Grazhdankin, D. V. et al. 2007. U –Pb (SRIMP II) age of the zircons from ash tuffs of the Chernyi Kamen’ Formation, Sylvitsa Group of Vendian (Middle Ural). Doklady Russian Academy of Sciences, Geosciences, 411, 354– 359. Petrov, G. A., Maslov, A. B. & Ronkin, Yu. L. 2005. Pre-Paleozoic magmatic complexes of the Kvarkush-Kamennogorsk anticlinorium (the Middle Ural): new data on the geochemistry and the geodynamics (in Russian). Lithosphere, 4, 42 –69. Raaben, M. E. 2007. Stromatolitic formations of the Riphean of EastEuropean plateform. Stratigraphy and Geological Correlation, 1, 35 – 46. Ronkin, Yu. L., Maslov, A. V., Petrov, G. A., Matukov, D. I. & Suslov, S. B. 2007. In situ U– Pb (SRIMP II) date of zircons from granosyenites of Tritsk Massiv (Kvarkush-Kamennogorsk Meganticlinorium, Middle Ural). Doklady Russian Academy of Sciences, Geosciences, 412, 87 –92. Suslov, S. B. & Teterin, I. P. 1997. On the origin ‘exotic’ stones of ‘tillite like’ conglomerates of Upper Wil’va Subformation. In: Iblaminov, R. G. (ed.) Geology and Mineral Resources of the Western Ural. Permian State University, Perm, 30 – 31 (in Russian).

Chapter 25 Glacial deposits of the Nichatka Formation, Chara River basin and review of Upper Precambrian diamictites of Central Siberia NICKOLAY M. CHUMAKOV Geological Institute of the Russian Academy of Sciences, Pyzhevsky per. 7, Moscow 119017, Russia (e-mail: [email protected]) Abstract: Outcrops of the Nichatka Formation are located in the southern centroclinal part of the marginal Berezovskaya basin of the Siberian craton. Deposits of the gently sloping southeastern limb of the asymmetrical basin are unaltered, whereas those of the steep, partly overturned western limb exhibit cleavage and are locally slightly metamorphosed. In the Late Precambrian the Berezovskaya basin was a shallow marine area of the passive margin of the Siberian craton. The Nichatka Formation is the basal unit of the Dal’nyaya Tayga Group of the Patom Supergroup. On the southeastern limb of the Berezovskaya Basin, the formation is composed of massive and bedded diamictites, conglomerates, sandstones and thin-bedded rocks with dropstones. The diamictites contain erratic blocks of different size and roundness. The formation thins out eastward. On the western limb, the formation is represented predominantly by graded-bedded and less frequently massive diamictites, sandstones and rare thin bedded members with dropstones. The diamictites contain clasts with glacial facets and striae. The formation unconformably lies on the Ballaganakh Group of the Patom Supergroup or older rocks and grades into a marker member of thin-bedded red dolomitic marls and dolomites. Overlying carbonate-terrigenous formations of the Patom Supergroup are characterized by two negative (down to –6 and – 10.5‰) d13C anomalies and one positive (up to 3.9‰) anomaly. 87 Sr/86Sr ratios in the carbonates above the Nichatka Formation increase upsection from 0.70725 in the overlying deposits to 0.70837 in the Nemakit-Daldyn (Fortunian) horizon of the early Cambrian. The Nichatka Formation as well as the host deposits are assigned to the Lower Vendian (Late Cryogenian) by reliable geological and isotopic correlations to biostratigraphically better studied Upper Precambrian sequences of the Ura Uplift in the Lena River basin. Stratigraphically similar glacial units were traced along the margins of the Baikal-Patom Highland up to the southern end of Lake Baikal and further to the NW towards East Sayan. The Siberian Craton was the main source area for these glacial units, which points to the existence of the great inland glaciers at the southern part of this craton. Some older Upper Precambrian diamictites of unclear age and genesis are also recorded in Central Siberia.

Diamictites, sandstones and conglomerates, which lie unconformably on Lower Proterozoic–Archaean granites and quartzites in the Lake Nichatka area, are named the Nichatka Formation (Zhuravleva et al. 1959). From this area the formation extends to the NE up to the eastern bank of the Chara River and to the north up to the Dzhelinda River and its inflows (Fig. 25.1b). Later, some researchers (Ivanov et al. 1995 and others) incorrectly referred to the Nichatka Formation along the western limb of the Berezovskaya Basin as the Dzhemkukan Formation, based on long-distance correlations with deposits of different structural and facies zones and composition. The stratotype section of the formation is located on the northwestern bank of Lake Nichatka, 3.5 km to the NW of the mouth of the Shirik River (578470 3600 N, 1178370 0900 E– 578480 0600 N, 1178380 3500 E). Although a glacial origin for the formation was assumed in the mid-twentieth century, geological facts confirming this idea have been discovered and published only recently (Chumakov 1993).

Structural framework In the eastern area of distribution of the Nichatka Fm. gently dipping (5 –158) strata compose the southeastern limb of the asymmetrical marginal Berezovskaya Basin of the Siberian Craton (Fig. 25.1b). The deposits have not been altered by secondary processes. In the steep and locally overturned western limb, the Nichatka Fm. is exposed in the cores of faulted anticlines. There the deposits are intensely cleaved and experienced lower greenschist metamorphism. Deformation of the western limb may have been caused by the clockwise rotation of the Aldan Shield during the Middle Palaeozoic (Shatsillo 2006). The Nichatka Fm. accumulated in the shallow eastern margins of the Late Precambrian Patom basin, which was a passive margin of the Siberian Craton (Chumakov et al. 2007). The terrigenous-carbonate Patom Supergroup was formed in the basin during the Late Precambrian. Towards the east, all units of the supergroup, including the Nichatka Formation, are reduced in thickness, and terrigenous formations are partly substituted by

carbonate platforms. Sedimentation was cyclic. The Nichatka Formation is a basal part of the second Dal’naya Taiga cycle. During sedimentation of the Patom Supergroup, the basin transgressed over the craton. This resulted in overlapping of older deposits by younger ones with unconformities at the margins. Because of thinning and probably erosional processes, the Nichatka formation is missing in sections along the left bank of the Chara River.

Stratigraphy A sequence of the Patom Supergroup in the southern Berezovskaya Basin is shown in Figure 25.2. Analogues of the groups and formations of the Patom Supergroups were established there by means of successive correlation of sections (Zhuravleva et al. 1959), which was supported by geological mapping (Ivanov et al. 1995), similarity of stromatolitic and microphytolitic associations (Dol’nik 2005) and recent data on C and Sr isotopes (Pokrovsky et al. 2006). The basal Ballaganakh Gr. is exposed in the western limb of the basin. The group is mainly composed of sandstones and small-pebbled conglomerates. Its upper part is represented by massive and stromatolitic dolomites and limestones of the Mariinsky Fm. An erosional surface of this formation is overlapped by the Nichatka Formation, which is a basal part of the Dal’nyaya Tayga Gr. The Nichatka Fm. is composed of diamictites, thin-bedded clay-rich siltstones with lonestones (dropstones), conglomerates and sandstones. The formation decreases in thickness eastward and thins out or was eroded on the right bank of the Chara River. The Nichatka Fm. is well correlated with the Bol’shoy Patom Fm. of the transverse Ura Uplift, which complicates the northeastern sector of the Patom folded arc. Both formations are similar in stratigraphic position, lithological composition and genesis. In other sectors of the arc, both formations are replaced by sands and shales of the Dzhemkukan Fm. The Nichatka Fm. is conformably overlain by the KumakhUlakh Fm., which begins with a member of red-brown dolomites marls and dolomites. The rest of the Kumakh-Ulakh Fm. is composed of black and variegated shales with limestone interbeds.

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 297– 302. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.25

298

N. M. CHUMAKOV

Fig. 25.1. (a) Geographical position of Nichatka Fm. and coeval glacial units at Baikal-Patom Highland: Nch, Nichatka Fm.; Dz, Dzhemkukan Fm.; Bp, Bol’shoy Patom Fm.; B.P., Bol’shoy Patom R. (b) Geological scheme for the south centroclinal part of Berezovskaya Basin: 1, Mezsozoic syenite-porphyry; 2, Tinnaya, Zherba and Torgo formations; 3, Sen’ and Kumakh-Ulakh formations; 4, Nichatskaya Fm.; 5, more old Proterozoic deposits; 6, Palaeoproterozoic and Archaean granites; 7, Archaean gnesses and granites; 8, faults and thrusts. Numbers in circles indicate type sections of Nichatka Fm.: 1, upper course of Dzhelinda R.; 2, west side of Nichatka L.; 3, Mokryi Kumakh-Ulakh R.

The formation is correlative to the Barakun Fm. of the Patom Supergroup. The Dal’nyaya Tayga Gr. is crowned by the Sen’ Fm. represented by alternating sandstones and dolomites in the lower part and by limestones, massive stromatolitic and oncolitic dolomites in the upper one. Therefore, some geologists divide the Sen’ Fm. into two formations: Imalyk and Tokko (Petrov 1976). The Sen’ Fm. is correlated to the shaly Valyukhta Fm. of the Patom folded arc and to the terrigenous Ura and carbonate Kalancha formations of the Ura Uplift. The Dal’nyaya Tayga Gr. is overlain by the Zhuya Gr. represented by a single Torgo Fm., which comprises a variegated marl member below and aphanitic algal and oolitic limestones and dolomites in the larger upper part. Within the Patom folded arc, the Torgp Fm. corresponds to the Nikol’skoe Fm. of variegated marls and the Chencha Fm. of limestones. The Precambrian deposits of the southern Berezovskaya depression is terminated by the Zherba and Tinnnaya formations, which are identical in composition to synonymous formations of the Patom arc and Ura Uplift. The Zherba Fm. is mostly composed of quartzite-like, glauconite-bearing sandstones, and the Tinnaya Fm. by bituminous limestones, dolomites and dolomitic breccias. In the Ura Uplift, the uppermost beds of the latter contain NemakitDaldyn fauna (Khomentovskiy et al. 2004). The overlying Yuedey Fm. of variegated marls, dolomites and limestones is correlative to the Lower Cambrian Nokhtuysk Fm. of the Ura Uplift containing the Tommotian fauna at the base and to the Pestrotsvetnaya Fm. of the Aldan Shield.

Glaciogenic deposits and associated strata Sections of the Nichatka Fm. are essentially variable in lithological composition. The stratotype section on the northwestern bank of Lake Nichatka (Fig. 25.3, 2) exhibits repeatedly alternating units of massive, less frequently bedded diamictites (5 –10 m thick) and thin-bedded clay-rich mudstones and sandstones (1.5 –20 m thick). The diamictites consist of lilac-grey clayey-silty-sandy matrix containing 10– 15% of scattered stones of different form and size (2– 75 cm, singular up to 170  100 cm). Roundness varies from class 0 to 5 on a five-class scale, but falls into classes 1 or 2 on average because of the dominance of only slightly rounded or unrounded stones. Most stones are fragments of grey granites and reddish quartzites similar to those underlying the

diamictites, as well as foreign gneisses and pink granites. Some stones (a few percent) have striated and grooved surfaces. Longitudinal, variably sized, subparallel striae and grooves are often confined to newly made abrasion facets. Grooves on granite pebbles sometimes resemble a series of small spear-shaped chattermarks (Chumakov 1993, fig. 3). Stone nests (clast clusters) and variously rounded and then split stones are characteristic. Elongated stones are predominantly oriented to the north and NE. Lower parts of the diamictite are usually bedded, suggesting a gradual transition from the underlying thinly bedded rocks with dropstones to massive diamictites. However, at the base of some massive diamictites there are signs of erosion or disharmonic folding and brecciation of underlying thinly bedded rock. In this case, overlying diamictites contain undeformed rafts of underlying rocks and small lenses and boudins of assimilated sandstones. Units between diamictites are composed of alternations of laminated red siltstones and mudstones with graded fine-grained arkosic sandstones. Upper parts of these units are less distinctly bedded and contain scattered pebbles and boulders (0.5 –150 cm across). These tear underlying beds and are enveloped by overlying ones; such dropstones bear abraded faces with striae and grooves. At lateral contacts of pebbles and boulders, synsedimentary deformation structures were observed. In general, lithological changes throughout the section reflect a gradual transition from thinly bedded members to bedded diamictites. A few kilometres to the NE, diamictites and thinly bedded rocks are replaced by conglomerates with subordinate diamictite interbeds. The conglomerates include large boulders consisting of granites, gneisses, quartzites and boulders and plates of finely crystalline and oolitic dolomites up to 2.3  2  1.8 m in size. Further to the north, the conglomerates are rapidly substituted by a sequence of cross-bedded arkosic sandstones and poorly sorted small-pebbled conglomerates (Chumakov 1993). The Nichatka Fm. demonstrates similar composition and structure throughout the southeastern limb of the Berezovskaya Basin. The character of the formation is different in the western limb (Figs. 25.3, 1). The most complete and representative section of the Nichatka Fm. is located at the Creek Opornyi, a left inflow of the Dzhelinda River (588030 2500 N, 1178350 2600 E – 588030 3200 N, 178360 0300 E). There the formation is mainly represented by yellowish-grey and grey diamictites with graded bedding, which frequently constitute single gradational rhythms with sandstones. Massive and bedded diamictites are subordinate. There are also

NICHATKA FORMATION, CHARA RIVER BASIN

299

Fig. 25.2. Stratigraphic position of the Nichatka Fm.: 1, diamictites; 2, conglomerates; 3, sandstones; 4, shales; 5, limestones; 6, dolostones; 7, marls; 8, dolostone breccia; 9, oolitic carbonates; 10, stromatolitic carbonates; 11, lowermost value of 87Sr/86Sr of formations (chemostratigraphic data by Pokrovsky et al. 2006).

interbeds of cross-bedded sandstones and members of thinly bedded shales with dropstones. Beds of conglomerates and breccia are rare. Massive diamictites contain predominantly pebbles, and small and medium-sized boulders, but large boulders and blocks of granites 0.5  1.2 m in size also occur in some members. The stones have a roundness of 1– 2.5 (according to the five class scale). Clasts in the diamictites are mostly represented by local and erratic granites, quartzites and gneisses. Pebbles and poorly rounded boulders of dolomites similar to those of the Mariinsky Fm. are frequent. Some stones bear subparallel striae. Massive diamictites enclose small sand lenses with erosional or gradational lower boundaries as well as signs of underwater slides in the form of plastic folds and slide rolls.

Boundary relations with overlying and underlying non-glacial units On the southeastern limb of the Berezovka Basin, the Nichatka Fm. overlies with angular disconformity Archaean (Fig. 25.1, 3) and Early Proterozoic gneisses and granites, dark red quartzites and

pyrophyllitic shales of the Purpol Fm. On the western limb of the basin the Nichatka Fm. lies on an erosional surface above the Upper Proterozoic Mariinsky or Bugarikhta formations. The base of the Nichatka Fm. is composed in this region of conglomerates and breccia of massive stromatolitic and oncolitic dolomites and limestones of the Mariinsky Fm. Locally, dolomite surfaces show buried relief (the Bogayukhta River, 578500 0200 N, 1178270 5500 E). The upper boundary of the Nichatka Fm. is more uniform. The formation is everywhere gradually overlain by a basal member of the Kumakh-Ulakh Fm., which consists of alternating laminated red dolomitic marls and pink dolomites. The thickness of this member is more than 3 m in the southeastern limb and up to 10 m in the western limb of the Berezovskaya Basin. Contacts of the Nichatka Fm. and this carbonate member are observed in the middle course of the Sen’ River, 2 km below the Uraga mouth (578500 0200 N, 1178270 5500 E) and at the Creek Opornyi. At the first locality the Nichatka diamictites grade into coarse and then fine-grained clayey sandstones (2 m), which are overlain by platy, sandy and then by calcareous mudstones (8 m). The latter are conformably overlain by red-brown thin-bedded dolomitic

300

N. M. CHUMAKOV

Fig. 25.3. Type sections of Nichatka Fm.: 1, conglomerates; 2, massive diamictites; 3, bedded diamictites; 4, shales with dropstones; 5, sandstones; 6, shales; 7, limestones; 8, dolstones; 9, marls; 10, granites; 11, striated and faceted stones; 12, dropstones; 13, glaciotectonic deformations; 14, till-like stone fabric; 15, stone nests; 16, split, semi-rounded stones; 17, rhythmic bedding and lamination; 18, erosional channels; 19, small sandstone lenses; 20, slide and slump structures; 21, slide rolls; 22, lamination; 23, cross-bedding; 24, conglobreccies. PR1, Palaeoproterozoic; km, Kumakh-Ulakh Fm.; mr, Mariinsky Fm.; ncˇ, Nichatka Fm.

marls of the Kumakh-Ulakh Fm. (0.7 m) passing upward into laminated red clayey dolomites (1.5 m apparent thickness). Similar conformable and gradual transition between the Nichatka and Kumakh-Ulakh formations is recorded in the Creek Opornyi section. There the uppermost Nichatka diamictites grade into redbrown sandstones (.5 m) with scattered pebbles and boulders. Alternating with shales, the sandstones are replaced by finely bedded variegated shales (15 m) including thin sandstone interbeds and higher up by a basal member of the Kumakh-Ulakh Fm. (10 m). This member begins with marls and, higher, exhibits alternating beds (2 –10 cm) of red micritic dolomites and dolomite marls.

Chemostratigraphy C, O and Sr isotopic compositions of carbonates were studied by Pokrovsky et al. (2006) in craton sections of the Patom Supergroup along the Chara River, which runs across the southeastern limb of the Berezovskaya Basin. Results for the calcite component are described below and represented in part in Figure 25.2. Values

for the dolomite component are slightly higher but exhibit parallel trends. The d13C curve shows two negative anomalies. The first negative anomaly in the Kumakh-Ulakh Fm. ranges from –2.9 to – 6.3‰ Vienna Pee Dee Belemnite (VPDB) standard, reaching a minimum value in the lower part of the formation and increasing upsection. d18P values vary considerably from 14.9 to 22‰ Vienna Standard Mean Ocean Water (VSMOW). The second negative anomaly, showing a decrease to –8.0 to –10.5‰, is found in the Torgo Fm. d18P varies insignificantly from 21 to 23‰ with some values as low as 19‰. Between these two negative anomalies there is a moderately positive one (from – 2.7 to þ3.4‰) with a maximal d13C value of þ3.4 to þ3.9‰ in the upper part of the Sen’ Fm. d18P values for this formation vary between 22 and 25‰, reaching 27‰ in its upper part. The overlying carbonate Tinnaya Fm. is characterized by gradually increasing d13C values from –0.8 to þ2.1‰ and d18P values from 20 to 23‰. The lowermost 87Sr/86Sr ratio for samples from the KumakhUlakh Fm. with low values of 87Rb – 86Sr (0.001 – 0.005), Mn –Sr (0.07 –0.12) and Fe/Sr (2.05 –3.30) is 0.70725 (three determinations: Sr 881–2013 ppm; range 0.70725 –0.70734). For the

NICHATKA FORMATION, CHARA RIVER BASIN

Torgo Fm. the lowermost 87Sr/86Sr ratio is 0.70837 (one determination: Sr 158 ppm). For the Tinnaya Fm. the lowermost ratio is 0.70799 (four determinations: Sr 299–812 ppm; range of 0.7079 –0.70832).

Palaeolatitude and palaeogeography No palaeomagnetic investigations of Upper Precambrian deposits have been carried out in the Berezovskaya Basin. According to the majority of recent global reconstructions (Powell et al. 2001; Smith 2001; Meert & Torsvik 2003; Shatsillo 2006; and others), the Siberian craton, including this basin, was located at low latitudes at that time.

Geochronological constraints There are no available radiometric and reliable biostratigraphic data on the Upper Precambrian deposits of the Berezovskaya Basin. They can be dated by means of correlation with biostratigraphically better studied sections of the Patom Supergroup of the Ura Uplift (the Lena River). The corrrelations are based on successive section-by-section tracing of formations, stromatolites and isotopic data (Chumakov et al. 2011). d13C anomalies can also provide confident correlations. The lower Kumakh-Ulakh negative d13C anomaly corresponds to the negative anomaly for the upper Bol’shoy Patom –lower Barakun interval of the Ura Uplift. The positive Sen’ anomaly can be correlated with the analogous Barakun –Valyukht anomaly of that uplift, and the significant negative Torgo anomaly can be easily correlated with the analogous Zhuya anomaly (Pokrovsky et al. 2006). These correlations are supported by similar 87Sr/86Sr ratios for both regions. These data suggest that the Nichatka Fm. is correlative to the glacial Bol’shoy Patom Fm. and that the lower Sen’ Fm. corresponds to the Ura Fm. The last formation contains a rich assemblage of acanthomorphic palynomorphs (Chumakov et al. 2007; Vorob’eva et al. 2008) very similar to the Ediacaran (lower Vendian) Kel’tma microbiota of the Russian plate (Veis et al. 2006) and to ‘late Pertatataka’ [¼ECAP (Ediacaran Complex Acanthomorph Palynoflora), Grey 2005] microbiota of Australia. The upper part of the Tinnaya Formation of the Ura Uplift contains small-shelly fauna of Nemakit-Daldyn type (Khomentovskiy et al. 2004).

301

glaciolacustrine. Along strike, the basal tillites are replaced by boulder-sized conglomerates, and further along strike by crossbedded arkosic sandstones. These deposits, which can be interpreted as glaciofluvial, form extensive fans penetrating into the adjacent basin. Distal sediments of the fans occur on the western limb of the Berezovskaya Basin. They are mostly represented there by graded bedded and massive diamictites. Massive diamictites contain rare striated stones and frequent small sand lenses, the tops of which are marked by thin silt interlayers and the bases of which exhibit a gradual transition into the sandy-clayey-silty matrix of diamictite. There are also small sand lenses filling gentle incisions, showing that local low-energy currents washed out clay and silt from the diamictons. The supposition is that massive and bedded diamictites were deposited by floating shelf glaciers, whereas graded bedded diamictites represent glaciogenic deposits reworked by mud and debris flows. The direction of cross-bedding in the sandy interlayers and the lithology of the clasts suggest that material was transported both from the east and SE and partly from the west to the western end of the basin. Diamictites include rare units of thin and rhythmically bedded siltstones and mudstones with dropstones, which may represent distal turbidite deposits where stones dropped from melting icebergs.

Age Reliable stratigraphic correlation between the sections of the Berezovskaya Basin and the Ura Uplift along with biostratigraphic data existing for Ura Uplift succession (Chumakov et al. 2011) allow us to suppose that the Tinnaya and probably Zherba formations can be tentatively assigned to the Nemakit-Daldyn Stage of the Upper Vendian, the Zhuya Group to the Kotlin and Redkino horizons of the Middle Vendian, and the Dal’nyaya Tayga Group (and the Nichatka Fm. correspondingly) to the lower part of Laplandian Horizon of the Lower Vendian. According to the timescale of the International Commission on Stratigraphy, the Tinnaya Fm. must be referred to the pre-Tommotian (Fortunian) interval of the early Cambrian, the Zhuya Group and apparently the upper part of the Laplandian Horizon to the Ediacaran System, and the lower part of the Laplandian Horizon to the Upper Cryogenian. In the context of distant correlations, the data presented above suggest that the Nichatka Fm., like the Bol’shoy Patom Fm., is correlative to the Nantuo Fm. of South China and probably the Yerelina Subgroup of Australia.

Discussion Regional palaeogeography Sedimentary environments The Nichatka Fm. has a great number of features that are typical of glacial deposits. It contains many massive and bedded diamictites with clasts of variable size, roundness and composition. These stones have, on average, roundness falling into classes 1 or 2 (using the five classes scale). Erratic and abraded stones (with clearly striated and grooved faces and spear-like chattermarks) are common. Rounded and subsequently split stones are also frequent. Dropstones and stone nests (clast clusters) occur commonly in laminated rocks, which alternate with bedded diamictites. The disharmonic deformation at the base of some massive diamictites resembles glaciotectonic disruption, and the orientation of elongated stones is similar to that of tills, suggesting that the Nichatka Fm. on the southeastern limb of the Berezovskaya Basin contains basal tillites. The prevailing northern and northeastern orientation of elongated stones and the occurrence of erratic stones of plagiogneisses and micaceous greisses along with local stones indicate the Aldan Shield as a source for Nichatka deposition in the SE of the basin. The basal tillites alternate with thin and rhythmically bedded siltstones and mudstones with dropstones. Such thin-bedded deposits are likely to be

As described earlier (Chumakov 1993, 2009), glacial deposits coeval with the Nichatka Fm. are traceable (with some interruptions) along almost the entire margin of the Baikal-Patom folded zone (Fig. 25.1a). They are known as the Dzemkukan Fm. to the north of the Lake Nichatka between the Dzhelinda and Chencha rivers. Farther on, they are exposed in the Ura Uplift along the Lena River, where they form the Bol’shoy Patom (or Dzemkukan Fm.). To the SW, glacial deposits are exposed along the western margin of the Baikal-Patom folded zone, where they are also called the Dzemkukan Fm. The Dzemkukan Fm. can then be traced from the Lena –Bol’shoy Patom interfluve up to the Vitim –Chaya interfluve. Although slightly thinner, the Dzemkukan Fm. of this region is similar in composition and structure to the Bol’shoy Patom Fm. (Chumakov 1993). To the SW of the Malaya Chuya River head, the Ballaganakh and lower part of the Dal’nyaya Tayga groups with the glacial deposits thin out. At the Chaya River, analogues of the upper part of the Patom Supergroup form the Baikal Supergroup extending up to Lake Baikal. At the southern end of the lake at the base of the supergroup there appears a sequence of diamictites, conglomerates and sandstones, which are called the Bugul’deka Member and are correlated to the

302

N. M. CHUMAKOV

Dzemkukan, Bol’shoy Patom and Nichatka formations (Chumakov 1993; Sovetov & Komlev 2005). These geological correlations have recently been supported by C- and Sr-isotopic studies, which revealed considerable similarity between the Baikal Supergroup and the upper part of the Patom Supergroup (Kuznetsov & Letnikova 2005), and indirectly by Pb/Pb dating of the Uluntuu limestones (550 + 40 Ma, Kuznetsov et al. 2006). All of the above-mentioned glacial units were united into a single Middle Siberian Glacial Horizon (Chumakov 1993). By means of lithological and sequence stratigraphic data, the Bugul’deyka Member can now be correlated to glacial deposits of the Marninsk Fm. of the Sayany region (Sovetov & Komlev 2005). This makes the area of distribution of the Middle Siberian Glacial Horizon much wider. As stated above, the main source for the Nichatka deposits was the Aldan Shield. Adjacent parts of the Siberian Craton also supplied debris for the Bol’shoy Patom Fm. of the Ura Uplift (Chumakov & Krasil’nikov 1991) and the Dzemkukan Fm. at the western margin of the Baikal-Patom folded zone. Glaciers moved from the craton to the western Baikal area (Sovetov & Komlev 2005). Sometimes, debris was transported to the Sayany region from the folded framework of the Siberian craton (Sovetov 2002; Sovetov & Komlev 2005), and also to the western part of the Berezovskaya Basin. In general, The Siberian Craton was the main sourceland for the Middle Siberian Glacial Horizon. This suggests that the main centres of Middle Siberian glaciation were located on the Siberian Craton, and at least the entire southern part of the craton was covered by glaciers, which, as evinced in the wide distribution of glaciomarine facies, flowed into the Baikal Ocean and surrounding seas. The very extensive area of distribution of the glacial deposits (almost 2000 km long) unambiguously indicates the existence of inland glacial sheets at all southern parts of the Siberian Craton during the Middle Siberian glaciation. Research was supported by grants of the Russian Fund of Basic Investigations (No. 11-05-00232), Program No. 25 of Presidium of Russian Academy of the Sciences. The author would like to acknowledge stimulating discussions arising from participation in IGCP project 512 ‘Neoproterozoic Ice Ages’. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) project #512.

References Chumakov, N. M. 1993. Riphean Middle Siberian Glaciohorizon. Stratigraphy and Geological Correlation, 1, 17 –28. Chumakov, N. M. 2009. Neoproterozoic glacial events in Eurasia. In: Gaucher, C., Sial, A. N., Halverso, G. P. & Frimmel, H. R. (eds) Neoproterozoic– Cambrian Tectonics, Global Change and Evolution: a Focus on Southwestern Gondwana. Developments in Precambrian Geology. Elsevier, The Netherlands, 16, 389– 403. Chumakov, N. M. & Krasil’nikov, S. S. 1991. Lithology of Riphean tilloids; Ura Uplift. Lithology and Mineral Resourses, 3, 58 –78. Chumakov, N. M., Pokrovsky, B. G. & Melezhik, V. A. 2007. Geologic history of Patom Supergroup, Late Precambrian, Middle Siberia. Doclady Academii Nayk. Geologiya, 413, 379– 383. Chumakov, N. M., Pokrovsky, V. G. & Melezhik, V. A. 2011. The glaciogenic Bol’shoy Patom Formation, Lena River, central Siberia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G.

(eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 309–316. Dol’nik, T. A. 2005. Stromatolites and Mikrophytolites in Stratigraphy of Riphean and Vendian of Folded Framework of the South Part of Siberian Craton (in Russian). Geo, Novosibirsk. Grey, K. 2005. Ediacaran Palinology of Australia. Memoir 31. Association of Australian Palaeontologists, Canberra. Ivanov, A. I., Lifshits, V. I., Perevalov, O. V., Strakhova, T. M., Yablonovskiy, B. V, Gryzer, M. I., Il’inskaya, Kh. G. & Golovenok, V. K. 1995. Precambrian of Patom Highland (in Russian). Nedra, Moscow. Khomentovskiy, V. V., Postnikov, A. A., Karlova, G. A., Kochnev, B. B., Yakshin, M. S. & Ponomarchuk, V. A. 2004. Vendian of Baikal-Patom Upland (Siberia). Geology and Geophysics, 45, 465– 484. Kuznetsov, A. B. & Letnikova, E. F. 2005. Opening Baikal Branch of Palaeoasian ocean: Sr and C data (in Russian). In: Koryakin, Y. V. (ed.) Tectonics of the Earth Crust and Mantle. The Tectonic Regularities of the Occurrences Mineral Recourses. GEOS, Moscow, I, 352– 355. Kuznetsov, A. B., Ovchinnikova, G. B., Kaurova, O. K. & Letnikova, E. F. 2006. Pb/Pb age and Sr chemostratigraphy of carbonate rocks of the Baikal Group, the south-west Pribaykal’e (in Russian). Isotopic dating of the ore formation, magmatism, sedimentation and metamorphism. GEOS, Moscow, 1, 362– 365. Meert, J. G. & Torsvik, T. H. 2003. The making and unmaking of a supercontinent: Rodinia revisited. Tectonophysics. 375, 261– 288. Petrov, A. F. 1976. Precambrian orogenic complexes of west part of Aldan shield (in Russian). Nayka, Novosibirsk, 120. Pokrovsky, B. G., Melezhik, B. A. & Buyakayte, M. I. 2006. Geochemistry of isotopes C, O, Sr and S, chemostratigraphy and environments of sedimentation of Late Precambrian deposits of Patom trough. Lithology and Mineral Resourses, 5, 505– 530. Powell, C. McA., Pisarevsky, S. A. & Winwate, M. T. D. 2001. An animated history of Rodinia. Geological Society Australia Abstracts, 65, 85 – 87. Shatsillo, A. V. 2006. Palaeomagnetism of the Vendian of the South Part of the Siberian Craton and Some Aspects of Late Precambrian Palaeogeodynamics (in Russia). Institutes Physics of the Earth Russian Academy of Sciences, Moscow. Smith, A. G. 2001. Palaeomagnetically and tectonically based global maps for Vendian to Mid-Ordovician time. In: Zuravlev, A. Yu & Riding, R. (eds) The Ecology of the Cambrian Radiation. Columbia University Press, New York, 11 – 46. Sovetov, J. K. 2002. Vendian foreland basin of the Siberian cratonic margin: Palaeopangean accretionary phases. Russian Journal of Earth Sciences, 4, 363– 387. Sovetov, Yu. K. & Komlev, D. A. 2005. Tillites at the base of Oselok Group Prisayan’ya and lower boundary of Vendian in south-west part of Siberian Craton. Stratigraphy and Geological Correlation. 13, 3 – 34. Veis, A. F., Vorob’eva, N. G. & Golubkova, E. Y. 2006. First find of Lower Vendian microfossils at Russian plate: taxonomy composition and biostratigraphic significance. Stratigraphy and Geological Correlation, 14, 28– 46. Vorob’eva, N. G., Sergeev, V. N. & Chumakov, N. M. 2008. New occurrences of Lower Vendian microfossils in Ura Formation: problem Patom Supergroup of Middle Siberia. Doclady Academii Nayk. Geologiya, 419, 782– 787. Zhuravleva, Z. A., Komar, V. A. & Chumakov, N. M. 1959. Stratigraphic correlations Patom Complex with sedimentary deposits west and north slopes of Aldan Shield. Doclady Academii Nayk USSR, 128, 1026– 1029.

Chapter 26 Glacial deposits of the Baykonur Formation, Kazakhstan and Kyrgyzstan NICKOLAY M. CHUMAKOV Geological Institute of the Russian Academy of Sciences, Pyzhevsky per. 7, Moscow 119017, Russia (e-mail: [email protected]) Abstract: The Baykonur Formation extends with some interruptions for over 1700 km along the eastern and northern margins of the Upper Precambrian Syrdarya and Tarim microcontinents. The Baykonur Formation constitutes the upper part of the Ulutau Group, which lies on an erosional surface of granosyenite with a U–Pb age of 720 + 20 Ma and is overlain by Lower Cambrian vanadiumbearing carbonaceous-siliceous shales. The Baykonur Formation is mostly composed of diamictite with erratic and striated stones, and includes thin-bedded shale beds with lonestones (dropstones) and a ‘cap dolomite’ unit at its top. The lithological composition of this formation implies a glaciomarine origin, while its stratigraphic position suggests a Late Vendian or an Early Cambrian age (the latest Ediacaran or Early Cambrian according to scale of the International Commission on Stratigraphy).

The Baykonur Formation represents a horizon of Upper Precambrian diamictites, which extend with some interruptions as an extended arc for over 1700 km from the Ulutau Mountains of Central Kazakhstan (Fig. 26.1, 1) through the Bol’shoy Karatau Ridge and Dzhebagly Mountains of Southern Kazakhstan (Fig. 26.1, 3, 4, 5), the Chatkal, Sandalash (Fig. 26.1, 6) and Kokiirim-Too (Fig. 26.1, 7) ridges of Western Kyrgyzstan up to the Naryn-Too and Dzhetym-Too ridges of Central Kyrgyzstan (Fig. 26.1, 8) and the Sarydzhaz Ridge of Eastern Kyrgyzstan (Fig. 26.1, 9) and further up to the China boundary. The Baykonur Formation is the upper formation of the group called Ulutau in Kazakhstan and Dzhetym in Kyrgyzstan. Its name is derived from the Baykonur River flowing from the southwestern slope of the Ulutau Mountains where the stratotype section of the formation is located. The second type section, deposits of which are less metamorphosed and better studied, is an outcrop along the Rang River, which flows from the northeastern slope of the Bol’shoy Karatau Ridge. Probable stratigraphic analogues of the Baykonur Formation are the diamictite of the lower part of the Kyrshibakty Formation in the Malyi Karatau Ridge (Korolev & Maksumova 1984) and the upper diamictitic part of the Kopal Formation exposed in the Mointy River basin (Fig. 26.1, 2). The Baykonur diamictites were first recovered in 1924 by D. V. Nalivkin, who referred them as Upper Palaeozoic glacial deposits. Some time later, the Baykonur Formation was considered to be Cambrian ‘facies of the foot of rocky shores’ (Borovikov 1955), proluvium (Makarychev 1967), subsurface sliding deposits (Volin 1966; Kholodov 1973; and others) or subaqueous sliding deposits with an admixture of mountain glacial debris (Knipper 1963). In the 1960s it became evident that the Baykonur diamictites belong to the Upper Precambrian (Korolev 1963). A majority of geologists agreed with their Precambrian age but continued, after Knipper, to interpret them as polygenic deposits of subaqueous slides, clastic and grain flows or, partly, as proluvium and mountain glacial deposits (Zaytsev & Kheraskova 1979; Kheraskova 1981a, 1986; Korolev & Maksumova 1984; and others). At the same time, evidence was gradually accumulating to indicate that the Baykonur diamictites were to a considerable extent formed by glacial activity (Ankinovich 1961; Zubtsov 1972; Chumakov 1978; Azerbaev 1988).

Structural framework The Baykonur Formation is confined to the synclinorium of the Caledonian basement stretching along the ancient eastern active margins of the Syrdarya and northern margin of the

Tarim microcontinents (Kiselev 2001). The Baykonur Formation experienced folding, thrusting and other dislocations during Palaeozoic ‘Caledonian’ orogeny. The deposits are usually cleaved (locally significantly) and weakly metamorphosed up to lower greenschist facies.

Stratigraphy One of the type sections of the upper part of the Ulutau Group is a section at the Rang River (the Bol’shoi Karatau Ridge), where the following members are exposed (from the base upward, Fig. 26.2). The Aksumbe Formation (1) Pale limestones alternating with green and black clay-rich shales (200 m). The Baykonur Formation (2) Dark thin-bedded clay-rich shales with scattered stones of quartz, dolomite, quartz porphyres and shales. The deposits were disturbed by subaqueous slumps (10 m). (3) Diamictite with a muddy matrix and scattered stones of different size, roundness and composition; bedding is disturbed by sliding, ‘quivering’ undulatory lamination and rare dolomite olistoliths (200 m). (4) Dark grey, brown on surface, bedded dolomite with intraformational breccias and signs of sliding (7.5 m). The Kurumsak Formation (5) Black cherts and vanadium-bearing carbonaceous-siliceous shales (250 m). Sections of the Baykonur Formation in the Ulutau Mountains (Fig. 26.2) and other areas of distribution are similar to the stratotype section. Thicknesses (from 600 to 100 m or less) and names of units above and below the Baykonur Formation are variable.

Glaciogenic deposits and associated strata The Baykonur Formation is mostly composed of diamictite with subordinate interbeds and members of shale, sandstone, conglomerate and breccia. The diamictite retains its main characteristic features throughout the region, but may vary in appearance and degree of metamorphism. The diamictite matrix usually comprises dark grey mudstone with variable portions of sand and dolomite grains. Increased content of the latter imparts a brownish-grey

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 303– 307. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.26

304

N. M. CHUMAKOV

Fig. 26.1. Outcrops of Baykonur Formation and coeval glacial units in Middle Asia. a, Outcrops of Baykonur and Kopal Formations; b, boundary of states. Kazakh Republic: 1, Ulutau Mountains; 2, Mointy River; 3, North Bol’shoy Karatau Ridge; 4, Central Bol’shoy Karatau Ridge; 5, South Bol’shoy Karatau Ridge and Dzhebagly Mountains. Kirgiz Republic: 6, Chatkal Mountain Range; 7, Kokirim Mountain Range; 8, Naryntau and Dzhetymtau Mountain Ranges; 9, Sarydzhaz Mountain Range. I, Uzbek Republic; II, People’s Republic of China.

colour to the matrix. There may also be admixture of chlorite, depending on the metamorphic grade. Chlorite imparts a greenish-grey colour to the matrix. Pyrite inclusions are also frequent at times. The matrix contains chaotically scattered clasts of different sizes, roundness and composition that can amount to between 1 and 20% of the total rock. Their predominant size ranges from 1 to 10–15 cm, although more rare larger fragments up to few metres across are also found. Semi-rounded and angular stones dominate over rounded and well-rounded stones. Striated and faceted clasts occur rarely. The clasts are mainly composed of rocks underlying the diamictite, but there are also erratic rocks that are foreign to the underlying successions (Azerbaev 1988).

In the Rang section, the diamictite contains a great number of small and large (up to a few metres long) dolomite olistoliths, brecciated dolomite and fragments of clay-rich sandstone beds. In some places olistoliths are deformed, disrupted and associated with smaller fragments of the same rocks (Korolev & Maksumova 1984). In addition to the dolomites, there are fragments and blocks of arkosic sandstone. Clast clusters (‘stone nests’) are common. Some parts of the diamictite exhibit a massive texture. Varying concentrations of clasts in the Baykonur diamictite may make impressions of coarse and lenticular bedding. In the Rang section, varying dolomite content defines vague bedding exhibiting gradual boundaries every 0.2– 5 m within the diamictite, which

Fig. 26.2. Type sections of Baykonur Formation (Ulutau Mauntins: Knipper 1963; Kheraskova 1986; Dzhetymtau Range: Korolev & Maksumova 1984).

BAYKONUR FORMATION, KAZAKHSTAN AND KYRGYZSTAN

is often characterized by fine plastic crimping. Large-scale softsedimentary deformation provides evidence of subsurface sliding.

Shales Greenish-grey thin-bedded shales constitute definable units (members) 5–40 m thick in the lower part and less commonly in the middle part of the Baykonur Formation. The shales usually include scattered small and medium-sized, and more rarely large pebbles and boulders. The lonestones cut through into the underlying beds, while the overlying deposits drape the stones (Knipper 1963; Azerbaev 1988). The shales bear frequent signs of subaqueous sliding.

Sandstones The sandstones are composed of feldspar and quartz and form thin interbeds, lenses and small members within the diamictite beds. Locally the sandstones have fine (up to 2 cm) rhythmical graded bedding. The rhythms begin with coarse-grained sandstones or small-pebbled conglomerates or grits, which gradually pass into fine-grained sandstones. Many rhythms have an uppermost thin interbed of siltstones with carbonate cement (Knipper 1963).

Dolomite In many sections, the Baykonur Formation is terminated by a dolomite bed. Its thickness reaches 7.5 –10 m in the Rang section. The dolomite is calcareous (up to 10% of CaCO3), slightly sandy (up to 11%), and has a dark grey colour, which is pale yellow or light brown on weathered surfaces (Chumakov 1992). Its structure is massive or bedded, frequently breccia-like, with signs of subaqueous slides. Some dolomite interlayers represent typical intraclast breccias consisting of numerous flat dolomite clasts.

Conglomerates Conglomerates occur as separate beds and lenses. They greatly increase in amount on the western limb of the Baykonur synclinorium (Kheraskova 1986). The clasts in these conglomerates are the same as those in the diamictites.

Boundary relations with overlying and underlying non-glacial units In the Ulutau Mountains, the Baykonur Formation lies conformably on underlying deposits in the axial part of the Baykonur synclinorium and on an erosional surface on the synclinorium limbs (Kheraskova 1981a, 1986). An unconformity at the base of the Baykonur Formation can also be observed at the Rang section of the Bol’shoy Karatau Ridge and at the Dzetymtau Ridge near the Kalmakashu Pass (Korolev & Maksumova 1984). The upper boundary of the Baykonur Formation is distinct but conformable in the Baykonur synclinorium (Kheraskova 1981a) and at Rang River. Further to the south, in the Tien Shan Mountains, there is a hiatus at the top of the formation locally associated with a weathering crust (Korolev & Maksumova 1984). In the Mointy region, the Kopal Formation contains a diamictite member very similar to the Baykonur Formation (Kheraskova 1981b). It lies conformably on top of dolomites of the lower part of the Kopal Formation. In some sections, the Basagin Formation gradually replaces the diamictites. It contains stromatolites of Cambrian type and is covered without hiatus by the Kyzylzhar

305

Formation, which demonstrates a Middle Cambrian fauna (Zaytsev & Kheraskova 1979; Kheraskova 1981b).

Chemostratigraphy No chemostratigraphic investigations of Neoproterozoic deposits have been carried out in regions of the distribution of Baykonur Fm.

Other characteristics A unit of black, thin and rhythmically bedded vanadium-bearing carbonaceous chert occurs 5–30 m above the base of the Kurumsak Formation of the Bol’shoy Karatau. The shales are 10 –12 m thick and include phosphorite concretions. The chert contains 75– 90% SiO2, 1.5 –15% Corg, 0.05–1.5% V2O5 and up to 2% BaO (Kholodov 1973). Under the microscope, fossil radiolarians and sponge spicules can be observed in the chert (Kholodov 1973).

Palaeolatitude and palaeogeography No valid results for palaeomagnetic investigations have been published for the Baykonur Formation and adjacent deposits. According to general palaeotectonic reconstructions, the Kazakhstan and Kyrgyzstan microcontinents were located at low latitudes during the Cryogenian-Ediacaran (Vendian) (Kheraskova et al. 2003).

Geochronological constraints The Baykonur Formation is the uppermost formation of the Ulutau (Dzhetym) Group. In the Bol’shoy Karatau Ridge, the group lies unconformably on deeply eroded granosyenite with a U – Pb zircon age of 720 + 20 Ma. (Kiselev 2001). The formation has a conformable contact with overlying carbonaceous-siliceous shales of the lower Koktal Formation (the Ulutau Mountains) or its stratigraphic and lithological equivalent, the Kurumsak Formation (the Bol’shoy Karatau Ridge). The Koktal carbonaceous chert (c. 30–50 m higher than the base of the formation) yielded acritarchs Micrhystridium aff. dissimilare Volk., M. tornatum Volk., M. lubomlense Kirjan., Cymatiosphaera? membranacea Kirjan., Leosohaeridia sp. ‘This acritarch assemblage is characteristic of the upper Lower Cambrian and very similar to the association from the Vergol Horizon of the East European platform’ (Krylov et al. 1986). The Koktal Formation lies beneath the Kokbulak Formation containing Middle and Late Cambrian trilobites (Ergaliev 1965).

Discussion Sedimentary environments The wide lateral distribution of Baykonur diamictites in the same stratigraphic position, as well as the occurrence of erratic and striated stones, suggests a significant role for glaciers in their genesis. Interbeds and whole units of thin-bedded shales with dropstones and frequent subsurface sliding indicate that the diamictites were formed in glaciomarine environments and subsequently affected by flows of variable density. A characteristic succession of post-glacial deposits is worthy of attention: diamictites are directly overlain by a marker bed of dolomites resembling cap dolomite and this latter unit is overlapped by carbonaceous chert containing phosphorite concretions and considerable

306

N. M. CHUMAKOV

concentrations of organic carbon (1.5 –15%). This facies succession is similar to many Upper Precambrian post-glacial sequences (Chumakov 1992; Zhu et al. 2007).

Age The sharp unconformity between the Ulutau Group and the underlying granosyenites, which have a U – Pb zircon age of 720 + 20 Ma, provides evidence for a considerably younger age for the group as a whole. The Baykonur Formation has a conformable contact with the overlying Kurumsak Formation or its stratigraphic analogues, for example the Koktal Formation. The erosional disconformity at the top of the Baykonur Formation in the Tien Shan sections is therefore unlikely to reflect a great hiatus. This suggestion is supported by persistent, but thin (few metres) cap dolomite of the Baykonur Formation over a long distance along the Bol’shoy Karatau Ridge and in some other sections. The conformable overlapping of the Baykonur Formation by the Lower Cambrian Koktal Formation indicates a relatively young age for the Baykonur Formation. As mentioned above, the microfossils characteristic of the upper Lower Cambrian (the Vergol Horizon correlatable with the Holmia Zone) were found some tens of metres above the Koktal Formation base. Strata between those with Early Cambrian microfossils and the Baykonur cap dolomite were formed in a basin experiencing condensed sedimentation. This suggests that their accumulation, although occurring over a long time, would be unlikely to have taken more than several million years. From these considerations an inference can be made that the Baykonur Formation was formed during the Late Vendian or initial Early Cambrian, that is, during the latest the Ediacaran or Early Cambrian according to the scale of the International Commission on Stratigraphy. Glaciogenic deposits similar to the Baykonur Formation deposits are also known in regions adjacent to Kyrgyzstan in China, namely in the Kuruktag Range (Wang et al. 1981; Chen et al. 1981). The glacial Hankalchough Formation has a very similar stratigraphic position to the Baykonur Formation, just below deposits with Lower Cambrian fossils (Zhu et al. 2011). The basal member of the Lower Cambrian in the Kuruktag Range consists of black carbonaceous chert containing phosphorite, as in Kyrgyzstan and Kazakhstan. A similar stratigraphic position has also been reported for the glacial Hongtiegou Formation of Chaidam Basin (Shen et al. 2010) and Luoquan Formation of North China (Mu 1981; Guan et al. 1986) and glacial deposits of the Bokson Formation (Chumakov 2009, 2011). I have combined all enumerated formations as the Baykonurian Glacial Horizon (Chumakov 1978, 1981). The widespread extent of the Baykonurian Glacial Horizon points to a significant glacial event occurring close to the Precambrian –Cambrian boundary that affected the Kazakhstan, Kyrgyzstan, Tuva-Mongolian and Tarim microcontinents as well as the North China Craton. It is possible that the large negative d13C anomaly that occurred close to the lower boundary of the Nemakit-Daldyn Horizon (Knoll 2000; Zhu et al. 2007; and others) is related to this Baykonurian Glaciation. Research was supported by grants from the Russian Fund of Basic Investigations (Nos 11-05-00232 and 10-05-00294), and Program No. 25 of the Presidium of the Russian Academy of the Sciences. The author would like to acknowledge stimulating discussions arising from participation in IGCP project 512 ‘Neoproterozoic Ice Ages’. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) project #512.

References Ankinovich, S. G. 1961. Lower Paleozoic vanadium-bearing basin of the North Tien Shan and West Margin of Central Kazakhstan (in Russian). Part I, Alma-Ata, 272.

Azerbaev, N. A. 1988. Lithologic features and origin of Vendian conglomerates of Bol’shoy Karatau (in Russian). Izvestia of Academy of Sciences of Kazakh SSR. Geological Series, 2, 53 – 63. Borovikov, L. I. 1955. Lower Paleozoic of Dzhezkazgan-Ulutau Region of West Part of Kazakhstan (in Russian). Nedra, Moscow. Chen, J., Zhang, H., Xing, Y. & Ma, G. 1981. On the Upper Precambrian (Sinian Suberathem) in China. Precambrian Research, 15, 207– 228. Chumakov, N. M. 1978. Precambrian Tillites and Tilloids (in Russian). Nauka, Moscow. Chumakov, N. M. 1981. Upper Proterozoic glaciogenic rocks and their stratigraphic significance. Precambrian Research, 15, 373–396. Chumakov, N. M. 1992. The Problems of Old Glaciations (PrePleistocene Glaciogeology in the USSR). Harwood Academic Publishers, Pennsylvania. Chumakov, N. M. 2009. The Baykonurian Glaciohorizon of the Late Vendian. Stratigraphy and Geological Correlation, 17, 373–381. Chumakov, N. M. 2011. Glacial deposits of the Bokson Group, East Sayan Mountains, Buryatian Republic, Russian Federation. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 285–288. Ergaliev, G. Kh. 1965. On stratigraphy of Vendian and Cambrian Baykonur Karatau-Dzhebagly zone (in Russian). Izvestia of Academy of Sciences of Kazakh SSR. Geological series, 6, 31 –43. Guan, B., Wu, R., Hambrey, M. J. & Geng, W. 1986. Glacial sediments and erosional pavement near the Cambrian – Precambrian boundary in western Henan Province, China. Journal of Geological Society, London, 143, 311–323. Kheraskova, T. H. 1981a. Late Precambrian tilloid of Baykonur Formation in Ulutau Mountains, Central Kazakhstan, USSR. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 348– 352. Kheraskova, T. H. 1981b. Late Precambrian tilloid of Kopal Formation in the Atasu-Mointy interfluve, Central Kazakhstan, USSR. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 358– 360. Kheraskova, T. N. 1986. Vendian-Cambrian Rock Associations of Caledonides of Asia (in Russian). Nayka, Moscow. Kheraskova, T. N., Didenko, A. N., Bush, V. A. & Volozh, Yu. A. 2003. The Vendian– Early Paleozoic history of continental margin of Eastern Paleogondwana, Paleoasian Ocean, and Central Asian Foldbelt. Russian Journal of Earth Sciences, 5, 165–184. Kholodov, B. N. 1973. Sedimentary Ore Genesis and Metallogeny of Vanadium (in Russian). Nauka, Moscow. Kiselev, V. V. 2001. Analogues of the Sinian complex in the central and northern Tien Shan. Geology and Geophysics, 42, 1453– 1463. Knipper, A. L. 1963. Tectonics of Baikonur Synclinorium (Central Kazakhstan) (in Russian). Publishing Office Academy of Sciences of the USSR, Moscow. Knoll, A. H. 2000. Learning to tell Neoproterozoic time. Precambrian Research, 100, 3– 20. Korolev, V. G. 1963. About Cambrian boundaries in the Middle Asia (in Russian). Trudy Frunzinskogo politekhnicheskogo institute, Geologiya, Gornoe delo, 10, 16 –21. Korolev, V. G. & Maksumova, R. A. 1984. Precambrian Tillites and Tilloids of Tien Shan (in Russian). Ilim, Frunze. Krylov, N. N., Sergeev, V. N. & Kheraskova, T. N. 1986. Discovery of Cambrian microfossils in deposits of Baikonur Synclinorium (in Russian). Izvestiya of Academy of Sciences of the USSR, Seriya geologicheskaya, 1, 51– 56. Makarychev, G. I. 1967. Stratigraphy of Protorozoic and Lower Paleozoic deposits of Bol’shoy Karatau (in Russian). Bulleten’ Moskovskogo obshchestva ispytateley prirody, XXXII, 31 –46. Mu, Y. 1981. Luoquan tillite of the Sinian System in China. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 402–413. Shen, B., Xiao, S., Zhou, G., Kaufman, A. J. & Yuan, X. 2010. Carbon and sulphur isotope chemostratigraphy of the Neoproterozoic Quanji Group of the Chaidam Basin, NW China: Basin stratification in the

BAYKONUR FORMATION, KAZAKHSTAN AND KYRGYZSTAN

aftermath of an Ediacaran glaciation postdating the Shuram event? Precambrian Research, 177, 241– 252. Volin, A. V. 1966. Breccias of sliding and tillites in relation with problem glaciations and moving of poles (in Russian). In: Materials k soveshchaniyu ‘Obshchie sakonomernosti geologicheskikh yavleniy’. Leningrad, 1, 31 – 46. Wang, Y., Lu, S., Gao, Z., Lin, W. & Ma, G. 1981. Sinian tillites of China. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s PrePleistocene Glacial Record. Cambridge University Press, Cambridge, 386– 401. Zaytsev, Y. A. & Kheraskova, T. N. 1979. Vendian of Central Kazakhstan (in Russian). Moscow University Publisher, Moscow.

307

Zhu, M. & Wang, H. 2011. Neoproterozoic glacigenic diamictites in the Tarim Block, NW China. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 367– 378. Zhu, M., Strauss, H. & Shields, G. A. 2007. From snowball earth to the Cambrian bioradiation: calibration of Ediacaran –Cambrian earth history in South China. Palaeogeography, Palaeoclimatology, Palaeoecology, 254, 1 –6. Zubtsov, E. I. 1972. Precambrian tillites of Tien Shan and their stratigraphic significance. Bulleten’ Moskovskogo obshchestva ispytateley prirody, XLVII, 42 – 56.

Chapter 27 The glaciogenic Bol’shoy Patom Formation, Lena River, central Siberia NICKOLAY M. CHUMAKOV1*, BORIS G. POKROVSKY1 & VICTOR A. MELEZHIK2 1

Geological Institute Russian Academy of Sciences, Pyzhevskiy 7, Moscow 119017, Russia 2

Geological Survey of Norway, Leiv Eirkssons Vei 39, N-7491 Trondheim, Norway *Corresponding author (e-mail: [email protected])

Abstract: The Bol’shoy Patom Formation (Fm.) is part of the Upper Precambrian Patom Supergroup, which comprises a siliciclastic and carbonate succession divided into (from base to top) the Ballaganakh, Dal’nyaya Tayga and Zhuya Groups (Gr.). The supergroup was deposited within the bay-like passive margins of the Siberian craton. The Bol’shoy Patom Fm. is the lower unit of the Dal’nyaya Tayga Gr. Massive and stratified diamictites with subordinate sandstones, mudstones, siltstones, conglomerates and conglo-breccias form the Bol’shoy Patom Fm. New biostratigraphic (microfossils of Pertatataka type, Ediacaran fossils) and chemostratigraphic data (87Sr/86Sr, d13C) point to a Vendian (Late Cryogenian and Ediacaran) age for the Dal’nyaya Tayga and Zhuya Groups. Unsorted diamictite matrix, a very wide range of size and roundness of dispersed erratic clasts, the presence of glacial grooves on boulders, dropstones and till pellets in laminated mudstones and siltstones are all evidence for intense ice rafting and ice-shelf sedimentation in the northeastern part of the Patom basin during deposition of the Bol’shoy Patom Fm.

A ‘conglomerate’ (diamictite) unit was first identified in the lower reaches of the Bol’shoy Patom River and on the southern banks of the Lena River (10 km upstream of the mouth of the Bol’shoy Patom River) in 1935 by Z. M. Starostina and formally named in 1941 as the Bol’shoy Patom Fm. by A. A. Predtechensky (Chumakov 1959). This name has been used in many publications up to today (Keller 1963; Bobrov 1979; Chumakov 1993; Vorob’eva et al. 2007), although some authors incorrectly refer to it as the Dzhemkukan Fm., based on correlation with a unit of very different lithology (shales and sandstones) in the Zhuya River area (Ivanov et al. 1995 and others). The best outcrops of the Bol’shoy Patom Fm. are in canyons of the Bol’shoy Patom and Ura River. The upper part of the formation is seen along the Lena River. The type section of the formation is situated on the eastern limb of the Zheday anticline in a canyon of the Bol’shoy Patom River. Lungersgauzen (1963) first suggested a glacial origin for the diamictites of the Bol’shoy Patom Fm. (he called it the Ust-Patom Fm.). Lungersgauzen described them as lithified basal moraines and fluvioglacial deposits. Subsequently, evidence for bedding, graded bedding and subaqueous slumps in the diamictites led Lungersgauzen’s interpretation to be modified (Chumakov 1965; Chumakov & Krasil’nikov 1991).

Structural framework The Bol’shoy Patom Fm. forms the cores of two anticlines that exhibit a northeastern trend and an en-echelon-like arrangement. The anticlines form the apex of the transverse Ura Uplift, which is oriented perpendicular to a huge marginal arch of folds bordering the Patom fold system. The uplift consists of the simply folded, thick (.6 km) Neoproterozoic Patom Supergroup (SGr.). The more southerly Zheday Anticline (Fig. 27.1b) crosses the lowermost course of the Bol’shoy Patom River and the middle course of the Lena River south of the settlement Chapaevo. The Ura anticline (Fig. 27.1a) is located to the east of the Lena River along the Ura River. Both anticlines have flat crests and moderate limb dips (30–608). Rocks of the Bol’shoy Patom Fm have been slightly altered, mostly by early epigenetic processes. Faint cleavage of the shale is developed only at the southwestern end of the Zheday fold. The Patom SGr. was deposited within the bay-like passive margins of the Siberian craton. The supergroup disconformably overlies older rift structures filled by conglomerates, sandstones

and volcanic rocks (metamorphic andesites and andesito-basalts). The supergroup succession points to the cyclic progradation of a sedimentary wedge from craton margins into the basin (Chumakov et al. 2007).

Stratigraphy The Patom SGr. comprises a siliciclastic and carbonate succession divided into three parts (from base to top): the Ballaganakh, Dal’nyaya Tayga and the Zhuya Groups (Gr.). The general succession and lithostratigraphy of the two upper groups are shown in Figure 27.2 and Table 27.1. The Bol’shoy Patom Fm. is the lower unit of the Dal’nyaya Tayga Gr. Lateral step-by-step lithostratigraphic correlations of the groups and their formations, using some robust lithological and isotopic markers, allow most formations to be traced more than 1000 km along the marginal folded arch of the Patom fold system. Beside the Ura Uplift, diamictite formations also occur in the lower part of the Dal’nyaya Tayga Gr. in two other uplifted areas of the folded arch: at its northwestern flank (Bol’shaya Chuya R. – ‘Dzhemkukan’ Fm.) and its eastern flank (Dzhelinda and Chara Rs. – Nichatka Fm.). There were three main cycles of progradation of Patom SGr. sedimentation represented by the Ballaganakh, Dal’nyaya Tayga and Zhuya Groups, respectively (Chumakov et al. 2007). The lower Ballaganakh Gr. is up to 6.5 km thick in the inner part of the Patom arch. It thins to 0.5 km at the craton margin and thins out completely on the craton. The group consists of conglomerates at the base, sandstones in the more substantial middle part and carbonates in the upper part (Mariinskiy Fm.). The Mariinskiy Fm. contains limestones and shales in the inner part of the Patom arch and stromatolitic and oolitic dolomites of carbonate platforms at the margins of the craton. The Dal’nyaya Tayga Gr. was deposited conformably on the Mariinskiy Fm. in the deep anoxic basin of the Patom zone and bordering arch. Margin disconformities occur at the base of the Dal’nyaya Tayga Gr. Near the edge of the craton the Mariinskiy Fm. was preserved as thin erosional relicts only. The Dal’nyaya Tayga Gr. consists of black slates and limestones and some sandstones in the Patom basin and in its margin fold arch. The Dal’nyaya Tayga Gr. is divided in this region into three formations (from the bottom): Dzhemkukan, Barakun and Valyukhta. To the NE, in the Ura uplift, black shales and sandstones of the Dzhemkukan Fm. give way to diamictites and associated rocks of the Bol’shoy Patom Fm. The lower part of the

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 309– 316. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.27

310

N. M. CHUMAKOV ET AL.

Fig. 27.1. Geographic distribution and structural framework of the Bol’shoy Patom Fm. Key: 1, Jurassic System and post Jurassic deposits; 2, Ordovician System; 3, Cambrian System; 4 –8 Neoproterozoic: 4, Tinnaya and Zherba Fms.; 5, Zhuya Gr.; 6, Upper and middle parts of Dal’naya Tayga Gr.; 7, Bol’shoy Patom Fm. (lower formation of Dal’naya Tayga Gr.); 8, Ballaganakh Gr.; 9, main faults; a, Ura Anticline; b, Zheday Anticline.

Valyukhta Fm. is replaced here by siltstones and limestones (Ura Fm.) and its upper part is replaced by stromatolitic and oncolitic dolomites and limestones of the carbonate platform (Kalancha Fm.). Similar lateral facies changes in the lower and upper parts of the Dal’nyaya Tayga Gr. can also be observed on uplifted parts of the eastern and western flanks of the Patom arch (Dzhelinda and Bol’shaya Chuya Rivers, accordingly) and at the western edge of the Aldan Shield beside the Patom arch (Chara River). Carbonate platforms of the upper part of the Zhuya Gr. drape all older deposits from the Patom arch up to the edge of the Aldan shield, thinning in this direction from 600 m to 200 m.

Glaciogenic deposits and associated strata The thickness of the Bol’shoy Patom Fm. is up to 1100 m. Different kinds of diamictites make up c. 60– 80% of the sections as well as subordinate sandstones, mudstones, siltstones, conglomerates and conglo-breccias. It is impossible to describe here the detailed succession of the formation owing to its great thickness and the frequent intercalation of rocks of different types (Fig. 27.3). In general, the formation can be divided into three informal members. Following geological sections of the Bol’shoy Patom Fm and enclosing deposits in the Ura Uplift (from top to base): Lower part of Barakun Fm. Member II: black shales with intercalations of black limestones (.200 m) Member I: dark grey, thinly bedded pelitomorphic dolomite that conformably overlies unit 4 (c. 6 m) Bol’shoy Patom Fm. Member III: at the upper boundary of the member, thin polymictic silty, fine sandstone occurs with rare coarse grains (1 mm); below are dark-grey massive and sometimes faintly bedded diamictites; rare slump structures, small sandstone and conglomerate lenses and interbeds (120 –180 m)

Fig. 27.2. Lithostratigraphic section, palaeontological data and C- and Sr-isotope variations in carbonate rocks of the upper part of the Patom SGr. and lowermost Cambrian system. Key: 1, diamictites (glacial and reworked glacial deposits); 2, carbonate conglo-breccias; 3, conglomerates; 4, sandstones; 5, siltstones; 6, limestones; 7, sandy limestones; 8, oolitic limestones; 9, stromatolitic limestones; 10, dolomites; 11, marls; 12, Pertatatataka microfossil assemblage; 13, Nemakit-Daldyn small shelly fossils; 14, Early Cambrian fossils (sunnaginicus biozone); 15, erosional boundaries; 16–18, d13C variations in sections of the Patom SGr. (Pokrovsky et al. 2006): 16, Ura Uplift; 17, Zhuya R.; 18, Boreholes at the left bank of Lena R.; d13C variations in late Cryogenian –Cambrian (Vendian-Cambrian) sections: 19, Ura Uplift (Sochava et al. 1996; Pelechaty 1998); 87Sr/86Sr: – 0.70725 – 0.70855 (Gorokhov et al. 1995; Vinogradov et al. 1996; Pokrovsky et al. 2006).

Member II: frequent intercalation of graded bedded sandstones, siltstones, carbonate conglo-breccias and dark grey faintly bedded diamictites with slump structures, rare massive diamictites, graded diamictites; there are intercalations of cross-bedded fine sandstones; lenses and scours filled by sandstones and conglomerates; rare dropstones, slump structures (150 –320 m) Member I: intercalation of dark grey massive and faintly bedded diamictites with graded bedded sandstone and mudstones beds and small sandstone lenses. Rare slump and small cross bedding structures (450 –700 m) Upper part of Mariinskiy Fm. Grey sandy limestone (.200 m)

BOL’SHOY PATOM FORMATION

311

Table 27.1. Succession of the Patom SGr. in the Ura Uplift Zhuya Gr. Chencha Fm. Nikol’skoe Fm.

White, pale and pink micritic, stromatolitic and oolitic Limestones, silty and sandy limestone in upper part 450– 600 m Red and green thin-bedded marls with intercalations of limestone; sandstone member at base 500– 600 m

Dal’nyaya Tayga Gr. Kalancha Fm. Dark grey limestone and pale stromatolitic, oncolitic and oolitic dolomites; siltstone intercalations 550– 60 m Ura Fm. Grey bedded mudstones, siltstones with intercalations of dark limestone in upper part and fine sandstones in lower part 350 m Barakun Fm. Black limestone, partly oolitic and brecciated in upper part; black shales with intercalation of black limestone in lower part; thin-bedded dark dolomite at the base 600 m Bol’shoy Patom Fm. Diamictites, shales, sandstones, minor conglomerates 900–1100 m Ballaganakh Gr. Mariinsky Fm. Bugarikhta Fm.

Grey limestones, sandy limestones and black shales 500 m Sandstones, grey shales, minor gravelly sandstones .1000 m

Diamictites Diamictites have a grey or dark grey (almost black) fine matrix with randomly dispersed, usually paler clasts. The matrix ranges between silty sandstone to sandy mudstone. Diamictites with a mudstone matrix are darker. Matrix is arkosic (quartz grains, 30– 70%; plagioclase, 20–58%; microcline, up to 18%). Minor components include quartzites (1 –12%), gneisses and granitoids (2 –10%), carbonates (1 –6%) and mica (1– 5%). Roundness of sand grains varies, but in general is low (less than 1.5 on the five-number scale of Khabakov 1933, and 0.45 on the scale of Krumbein & Sloss 1951). Grains are cemented by sericite, rarely carbonate, chlorite and detrital biotite. Clast content varies from 1 to 15%, commonly 3 –10%. Small and medium-sized pebbles are most typical, cobbles and boulders are also common with a few blocks reaching 1.1–1.7 m across. Clasts are randomly scattered in the matrix, often forming nests of 3–4 stones and occasionally thin stone lags. Clast composition is quite uniform: granitoids (45 –65%), granite-gneisses (20– 35%) and gneisses (5 –30%). Other clast types (5 –15%) include limestones, dolostones, sandstones, amphibolites, quartzites, quartz and schists. The source area for the granitoids and gneisses was

the northwestern Aldan Shield, which at that time would have been situated no less than 100 km to the east of the Ura uplift and crystalline basement of the southern part of Siberian craton. Clast roundness varied between 0 and 4, but mean roundness is generally between 0.7 to 1.2 (using the Khabokov 1933 scale). In general, crystalline stones are more rounded (1 –3), although some angular and subangular slabs of granites and gneisses are found too. Two main types of diamictites can be identified: massive and stratified (Chumakov & Krasil’nikov 1991). Massive diamictites. Massive diamictites are the most typical rocks

of the Bol’shoy Patom Fm., particularly in the lower and upper members. They form 50 –70% of formation sections. The thicknesses of individual units of these diamictites vary from 3 to 70 m. Only indistinct clast concentrations can be recognized sometimes as diffuse horizons or lenticular bodies. Large boulders occur individually or as nests at certain stratigraphic levels. Small lenses (0.1 –0.3  1– 2 m) of sandstones or sandy siltstones of slightly concavo-convex shape occur rarely in some massive diamictite units. Commonly, the lenses have a gradational lower boundary and a sharp upper boundary sometimes with very thin silt

Fig. 27.3. Composite sections of the Bol’shoy Patom Formation (parts of sections not limited by line from right side are based on eluvial debris and small dispersed outcrops, white indicates absence of outcrops). Key: 1, massive diamictites; 2, faint bedded diamictites and diamictites with relicts of bedding; 3, carbonate conglo-breccies; 4, conglomerates; 5, gritstones; 6, gravelly sandstones; 7, sandstones; 8, sandstones with traces of thermal metamorphism; 9, siltstones; 10, sandy limestones, sandstones and siltstones containing many carbonate grains; 11, limestones; 12, dolomites; 13, crystalline and sedimentary clasts; 14, carbonate clasts; 15, cross-bedding; 16, wavy bedding; 17, normal graded bedding; 18, reverse graded bedding; 19, sandstone lenses; 20, erosional channels and scours; 21, ripple marks; 22, striated boulders; 23, slump rolls; 24, slump boudins; 25, slumpfolds; 26, dropstones; 27, lonestons; 28, stone nests; 29, broken stones; 30, sole marks; 31, stone fabrics (hatch lines from Lungersgauzen 1963); 32, dominant direction of cross-bedding (hatch lines from Lungersgauzen 1963).

312

N. M. CHUMAKOV ET AL.

laminae. Some lenses have an erosional lower boundary with basal grits, pebble conglomerates or single boulders (Fig. 27.4a). Some of the massive diamictites contain rare small fragments of sandstone or siltstone beds partly reworked by slumps as folds, rolls, boudins, ragged fragments and other relicts of sedimentary beds. Some slump fragments have faint and gradational boundaries with the matrix, forming clods, clots and ‘noses’, implying slumps of semi- and unlithified sediments. Massive diamictites containing slump structures constitute 4– 30% of sections. Stone nests and broken stones are common. Flat surfaces of single quartzite boulders show one or two systems of longitudinal sub-parallel grooves and striae of different sizes. Elongated stones in some massive diamictites had primary orientations in one predominant direction and one or more crossing it in subdominant direction(s) (Fig. 27.3). At Lena River, the predominant direction has a NW –SE or north –south, or more rarely a NE –SW or WNW – ESE orientation. At Ura River, the predominant direction has a north – south orientation. Stratified diamictites. Stratified diamictites are of two types: bedded and graded-bedded. Bedded diamictites have matrices that are muddy, silty or sandy to varying degrees, and exhibit a range of clast sizes. Some bedded diamictites have a laminated muddy matrix. The thicknesses of individual laminae vary from 0.01 to 1 cm. Rare pebbles in these rocks are angular or subangular. Another form of bedded diamictites has a silty to sandy matrix with bed thicknesses from 0.1 to 0.4 m or more. Most such beds have a lenticular or irregular form (Fig. 27.4b). Graded-bedded diamictites are characterized by clear vertical changes in the granulometry of matrix and clasts. The thicknesses of graded-bedded diamictites vary from 1 to 60 cm. These diamictites are often related to the lower or middle parts of cyclic alternations of diamictites, sandstones and siltstones (Fig. 27.4c, d). The thickness of individual cycles varies from 0.5 to 3 m, while the total thickness of the cyclic units can reach tens of metres. Some cycles have normal grading (Fig. 27.4c), whereas others have reverse grading in the lower parts and normal grading in the upper parts of the cycles (Fig. 27.4d). In this latter case there are often rip-up structures with abundant mudstone chips at the base of cycles. Large stones or fragments of sedimentary rocks, similar to rafts of debris flow, occur sometimes in the upper or middle parts of the cycles.

Associated sedimentary rocks Beside diamictites, the Bol’shoy Patom Fm. contains subordinate rocks: sandstones and grits, conglomerates, siltstones and conglo-breccias.

Sandstones and grits. These make up between 10 and 25% of sec-

tions. The mineral composition of sandstones is similar to that of the diamictite matrix. Limestone grains occur only in the middle member of the formation on the western limb of the Zheday fold, where beds of fine sandstones occur with between 10 and 30% limestone clasts. The thicknesses of sandstone beds generally range from 2 to 10 m. In the Lena River section, fine- and mediumgrained sandstone forms units up to 40 m thick. There are three types of sandstones. The first main type of sandstones exhibits pronounced normally graded bedding: from grits and coarse sandstones at the base to medium and fine-grained sandstones at the top. Thin siltstone laminae occasionally overlie these sandstones. Graded-bedded sandstones are often associated with gradedbedded diamictites. Thin beds of reverse graded-bedded sandstones occur rarely. The second type of sandstones has restricted development in the middle member of the Bol’shoy Patom Fm. They are thinly bedded silty, fine-grained sandstones, often with small climbing ripple cross-bedding. These sandstones form individual units up to 40 m thick. The cross-beds dip generally to the south at Ura River and to the NW at Lena River. The third type of sandstones includes fine, thinly bedded sandstones with abundant carbonate grains. Conglomerates. These occur rarely as thin lenses. It seems that they

are residual, a result of the winnowing of diamictites by bottom currents. Other rare types of coarse sandstones, grits and conglomerates fill erosion channels 0.5–10 m deep and several tens of metres wide. In places, the channels have rather steep slopes. Mudstones and siltstones. These form beds, and more rarely units, of between a few metres and a few tens of metres thick. The rocks exhibit rhythmic lamination and contain dispersed lonestones and small till pellets. Some lonestones can be identified as dropstones, displaying splash features, with rupture of laminae below and draping of laminae above the clasts (Fig. 27.4e). Carbonate conglo-breccies. These form a number of beds from 0.1 to 0.5 m thick in the middle member of the formation. They have a siliciclastic, carbonate-bearing sandstone matrix and contain abundant (from 30 to 70%) angular and subangular fragments (from small pebbles to blocks 1.2 m across) of oolitic and stromatolitic dolomites and limestones. The matrix of the conglo-breccies has normal and reverse graded bedding. Some large stones occur in the middle or upper parts of the beds and are similar to rafts of debris flows.

Fig. 27.4. Some typical structures in diamictites of the Bol’shoy Patom Formation. (a) Sandstone lens with boulders in silty diamictites. (b) Two thin irregular beds of silty-sandy diamictites occur in silty diamictite. (c) Normal grading of cycle. (d) Reverse grading in lower and normal grading in upper part of cycle. (e) Dropstone and till pellets in laminated mudstone.1, stones of granite and sandstone; 2, massive dimictites; 3, bedded diamictites; 4, shale chips; 5, sandstone; 6, shale.

BOL’SHOY PATOM FORMATION

Boundary relations with overlying and underlying non-glacial units The upper boundary of the Bol’shoy Patom Fm. is conformable. It was observed in two places. The best locality is situated 7.5 km from the mouth of the Ura River on its northern bank, and the second one is located at the beach on the southern side of the Lena River 6 km upstream of the Bol’shoy Patom River mouth. At Ura River, diamictite with a sandy mudstone matrix and rare clasts gradually change upsection, at first to silty coarse sandstone with dispersed pebbles and then to silty fine polymictic sandstone with rare coarse grains. Transitional beds are c. 1 m thick. Fine sandstone is conformably overlain by the first dolomite member of the Barakun Fm. This dark grey or black laminated dolomite is peloidal, with small asymmetric anticline-like structures (0.5  0.5 m) with sharp angular aggrading crests resembling ‘tepees’. The lower contact of the Bol’shoy Patom Fm. has not been observed in outcrop, but the lower boundary of the Dal’nyaya Tayga Gr. is believed to be erosional, at least along the basin margins. This opinion is based on the existence in the Bol’shoy Patom Fm. and in the lower part of the coeval ‘Dzhemkukan’ and Nichatka Fms. of big carbonate clasts of rocks of the Mariinskiy Fm. (Chumakov 1993; Ivanov et al. 1995 and others).

Chemostratigraphy The C-isotope profile of carbonates of the Patom SGr. exhibits an alternation of high and low d13C intervals (Fig. 27.2) with two extraordinary features: large-amplitude d13C variation, exceeding 21‰ (– 13.5 to 8.3), and an enormous thickness (up to 1000 m) of 13 C-depleted strata (d13C, , –7.5) Zhuya Gr. (Pokrovsky et al. 2006). The Mariinskiy Fm., underlying the Bol’shoy Patom (Dzhemkukan) Fm., exhibits a 600 m positive ‘plateau’ around d13C  7– 8‰. Negative d13C values have been found in the carbonatebearing sandstone of the upper part of the Bol’shoy Patom Fm. at the Ura River (28.8‰) and in the lower part of Dzhemkukan Fm. at the Zhuya River (29.1‰) where diamictites are absent. Cap dolostones at the Ura River have moderately low d13C values from –3.3 to –4.2‰. Upsection there is a second positive ‘plateau’ within the Barakun (d13C  6–8‰) and Valyukhta (d13C  4–6‰) Fms. with an overall thickness near 1.5 km, which is overlain by the 300-m-thick Nikol’skoe Fm. consisting of variegated micritic and peloidal limestones, marls and oncolitic limestones, which are all depleted in 13C, with d13C values as low as –13.5‰. The strongly 13C-depleted (in the range d13C  –7.5 to 10‰) carbonates continue upwards, where they also characterize the Chencha Fm., a c. 600–700-m-thick unit of micritic limestones, stromatolitic limestones and pink oncolitic limestones. Ultrahigh-d13C carbonates of the Dal’niyaya Tayga Gr., and ultralow-d13C of the Zhuya Gr. have been traced around the Patom Basin to the east of this basin (at Chara River and the CharaTokko watershed) without significant variation in C-isotope composition. The thickness of high-d13C dolomites of the Sen’ Fm. (correlative of Ura and Kalancha Fms.) at the Chara River does not exceed 200–250 m, and low-d13C limestones of the Torgo Fm. (correlative of Chencha Fm.) extend over 300– 350 m. Further to the east, the Torgo Fm. (150 m) partly consists of dolomites with the same d13C ( –7.5 to –10), but higher d18O (up to 30‰; Pokrovskii et al. 2006). High Sr contents are typical for limestones of the Patom SGr. They reach 11 700 ppm in the Chencha Fm., 2500 ppm in the Barakun and Valukhta Fms., 2200 ppm in the Mariinskiy Fm.; only in the Nikol’skaya Fm. do they not exceed 400 ppm. The range of measured 87Sr/86Sr ratios (data from Gorokhov et al. 1995; Vinogradov et al. 1996; Pokrovskii et al. 2006 and their unpublished data) in the limestones of the Chencha Fm. is

313

relatively small (0.7079 – 0.7087), but is greater lower down in the succession: Nikol’skaya Fm. (0.7079 –0.7094); Valukhta Fm. (0.70765 –0.7092); Barakun Fm. (0.7025 –0.7102); Mariinskiy Fm. (0.7075 –0.7092). Taking into account the low measured Rb –Sr (,0.01), Mn –Sr (,0.1) and Fe –Sr (,2) ratios, we consider that 87Sr/86Sr (min.) ¼ 87Sr/86Sr (primary) for all formations excluding, probably, the Mariinskiy Formation. In general, isotope stratigraphic data from the upper part of the Patom SGr. show an increase in lowermost 87Sr/86Sr ratios from high-Sr limestones of from 0.70725 in the Barakun Fm. to 0.7079 –0.7080 in the Nikol’sk and Chencha Fms.

Palaeolatitudes and palaeogeography Systematic palaeomagnetic investigations of the Patom SGr. were made twice at the Paleomagnetic Laboratory of the Geological Institute of the RAS, Moscow (by S. V. Shipunov) and then at the Paleomagnetic Laboratory of the US Geological Survey in Flagstaff (by D. Elston). All rocks of the supergroup appear to have been remagnetized during Palaeozoic times.

Geochronological constraints From the beginning of the 1960s up to 2005, the Patom SGr. was dated on the basis of stromatolites and microphytolites to be Middle and Late Riphean (Mesoproterozoic-Cryogenian) (Zhuravleva et al. 1961; Keller 1963; Chumakov 1993; and others) or Late Riphean (Tonian-Cryogenian) (Khomentovskiy et al. 1998). Presumed correlation between the Russian stratigraphic chart on Late Precambrian and the stratigraphic chart of the International Commission on Stratigaphy is shown in Table 27.2. First suggestions of a Vendian (late Cryogenian-Ediacaran) age for the upper part of the Patom SGr. were made by J.K. Sovetov, who correlated tillites of the Prisayan’ya (South Siberia) with the Bol’shoy Patom Fm. (Sovetov 2002). The Zherba and Tinnaya formations overlying the Patom SGr. were suggested to be Vendian (Ediacaran –lowermost Cambrian) by all investigators of the region. This opinion was recently confirmed by the discovery of small shelly fossils of the Nemakit-Daldyn stage in the uppermost part of the Tinnaya Fm. and fossils of the Cambrian Sunnaginicus Biozone of the Tommotian Stage in the lowermost part of the Nokhtuysk Fm. (Khomentovsky et al. 2004). There are no precise radiometric age data pertaining to the Patom SGr. Numerous data were obtained using the K –Ar and Rb –Sr methods on whole sedimentary rocks of the upper part of Dal’nyaya Tayga Gr., revealing strong epigenetic rejuvenation (452 –531 Ma; Vinogradov et al. 1996). A wide range of ages was obtained on gabbro-dolerite sheets intruded between the Bol’shoy Patom and Barakun Fms. The hypabyssal nature of the sheets implies that they were formed at depths of 2 –4 km. Dating by the K –Ar whole-rock method shows ages between 520 and 633 Ma (Oleynikov et al. 1983). These data can represent only very approximate age constraints. Taking into consideration the thickness of overlying deposits, it is possible to suggest that Table 27.2. Correlation between Vendian and Ediacaran Systems Cambrian Lower

Tommotian Stage Nemakit-Daldynian Horizon

Upper

Vendian Lower

Kotlin Horizon Redkino Horizon Laplandian Horizon

Cambrian Ediacaran

Gaskiers

Glusk Upper Blon' Lower Blon'

Upper Riphean glacial horizons

Lower

Cryogenian

314

N. M. CHUMAKOV ET AL.

the age of the sheets is close to the age of the upper part of the Patom SGr. or deposits immediately overlying. A wide range of ages was also obtained using the K –Ar glauconite method for the Zherba Fm. overlying the Patom SGr., giving 570–604 Ma (Ivanov et al. 1995, corrected with constants from Steiger & Ja¨ger 1977). New biostratigraphic and chemostratigraphic data allow us to reinterpret radically the ages of the Dal’nayay Tayga and Zhuya Groups. Ediacaran fossils Beltanelloides sorichevae were found in the lower part of the Barakun Fm. at Ura River 8.5 km from its mouth (Leonov & Rud’ko in press). New maceration and reinvestigation of very rich microfossil assemblages from shales of the Ura Fm. collected by us on the northern bank of the Ura River, 3.5 km from its mouth, point to similar age (Vorob’eva et al. 2007). This assemblage contains abundant large acanthomorphic acritarchs of genera Ericiasphaera, Tanarium, Appendisphaera, Meghystrichosphaeridium, Sinosphaera and taxa very similar to the genera Alicesphaeridium, Variomargosphaeridium, Dicrospinasphaera and others. Ericiaspaera adspersa dominates the biota, while Tanarium conoideum, Appendisphaera tenuis, ?Sinosphaera rupina and Meghistrichosphaeridium hadianensis also occur. There are also numerous morphologically complicated, new acanthomorphic forms. Most of the listed taxa are known from a number of post-glacial formations: the post-Olympic (?post-Marinoan) Pertatataka Fm. of Central Australia (Gray 2005), the post-Nantuo Doushantuo Fm of South China (Zhou & Xiao 2007), the post-Blaini Infrakrol Fm. of the Lesser Himalayas (Tiwary & Knoll 1994), as well as the pre-Redkino Vychegda Fm. of the Russian Plate (Veis et al. 2006; Vorob’eva et al. 2006) and the upper part of the Nepa Horizon of the Siberian Plate. Associations of such microfossils are usually referred to as ‘Pertatatakatype biota’. If confirmed, the Ura biota would appear to be similar to the second microfossil zone of ECAP of K. Grey of Central Australia (Vorob’eva et al. 2007). The presence of acanthomorphic acritarchs in the Ura microbiota has been reported previously by some other researchers. They considered them to be Palaeozoic forms penetrating into the Ura Formation, or interpreted the microbiota as Late Riphean (Cryogenian) but with many Vendian elements (for a review see Nagovitsin et al. 2004; Vorob’eva et al. 2007). The conclusion on the Late Riphean (Cryogenian) age of the Ura microbiota was made by Nagovitsin et al. (2004), particularly from the finding of the form described incorrectly as Trachyhystrichosphaera aff. aimika. In the opinion of Vorob’eva et al. (2007), the forms described in this paper (Nagovitsin et al. 2004) are really fragments of Early Vendian taxa, such as Tanarium, Alicesphaeridium and others. The assignment of the Dal’naya Tayga Group to the Early Vendian or Late Cryogenian –Early Ediacaran has been confirmed by a minimal age of 600 + 10 Ma from detrital zircons from the upper part of the Khomolkho Fm. (U –Th – Pb, LA-ICPMS, Meffre et al. 2008). The Khomolkho Fm. is a stratigraphic analogue of the Valyukhta Fm. (Zhadnova 1961; Keller et al. 1967; Ivanov et al. 1995, and many others). Khomentovsky & Postnikov (2001) even suggest that the Khomolkho Formation is a stratigraphic analogue of the Ura Fm. This date therefore indicates a maximal age of Valyukhta and Ura Fm. These data allow us to suggest correlation of the Bol’shoy Patom Fm. with the Nantuo, Blaini and Olympic Fms. In this case, the positive d13C anomaly in the upper part of the Dal’nyaya Tayga Gr. may be correlated with a similar anomaly in the lower part of Doushantuo Fm., while the negative anomaly in the Zhuya Gr. may be correlative with a similar anomaly ‘Dounce’ in the uppermost member of the Doushantuo Fm. (Zhou & Xiao 2007; Zhu et al. 2007). 87 Sr/86Sr ratios of the uppermost part of the Dal’nyaya Tayga Gr. and of the Doushantuo Fm. (Yang et al. 1999) are consistent with this interpretation (minimum values of 0.70765 and 0.7077, respectively). Correlations with the Nantuo and the Doushantuo formations allows us to constrain the age of the Dal’nyaya Tayga and Zhuya Groups to between 636/660 and 550 Ma.

Ultra-low-d13C carbonates of the Zhuya Gr. could be correlated, besides anomaly ‘Dounce’, with the negative d13C anomalies of the Shuram and Buah formations (Oman) and probably the similar anomaly in the Wonoka Fm. (S. Australia). Sedimentological and geochemical parameters suggest that the Zhuya Gr. had preserved its primary C-isotope composition and was formed over at least 5 million years (a similar duration is suggested for the Dounce anomaly). The negative C-isotope anomaly of the Zhuya Gr. exceeds all known C-isotope anomalies both in amplitude and thickness of strata, providing evidence of unprecedented ‘contamination’ of the water column by low-13C organic carbon. If the Nantuo Formation is really coeval with the Yerelina Subgroup, which remains to be proven, then the upper part of the Dal’yaya Tayga Gr. and the Zhuya Gr are Ediacaran in age. According to the Russian stratigraphic chart, both the Dal’yaya Tayga and Zhuya groups belong to the Lower and Middle Vendian.

Discussion Three rock associations can be recognized in the Lower Vendian (Late Criogenian) Bol’shoy Patom Fm. The first and main rock associations comprise massive diamictites with subordinate bedded diamictites and laminated mudstones and siltstones. This association is typical of the lower and upper members of the Bol’shoy Patom Fm. Intercalation with laminated mudstones and siltstones, bedded diamictites and the existence of sandstone lenses related to winnowing of diamictites by local bottom currents implies that massive diamictites were deposited in a rather quiet basin. Laminated mudstones and siltstones undisturbed by waves and traces of slumping show that the basin floor was rather deep and sloped. It was a large basin, probably open to the sea, as the Bol’shoy Patom Fm. is gradually replaced towards the south and SW by the thick Dzhemkukan Fm. comprising distal turbidite slates with subordinate sandstones (c. 1000 m). Unsorted matrix, a very wide range of size and roundness of dispersed erratic clasts, the presence of glacial grooves on boulders, dropstones and till pellets in laminated mudstones and siltstones are all evidence for intense ice-rafting in the basin. A wide size range and low mean roundness of stones are not typical for debris rafted by seasonal or perennial sea ice, which commonly transports well-rounded and well-sorted stones picked up from coastal beaches. Therefore, it is possible to conclude that the first rock association was deposited mainly by glacier and iceberg rafting of material derived from shelf glaciers and outlet glaciers. Glacier-like clast fabrics in some massive diamictites point to the presence of grounded shelf glaciers moving in north –south and NW – SE directions. The basement of the Siberian Craton was the main source of clastic deposits in the Bol’shoy Patom Fm. This is clear from the predominance of granitoids, granite-gneisses and gneisses, which make up almost all of the cobbles and boulders in the massive diamictites. The source of carbonate conglo-breccias was from carbonate platforms bordering the Patom basin in Mariinsky time. Their erosion was related to a fall in sea level during the Bol’shoy Patom glaciation. Crossbedding in sandstones allows us to suggest that in the Lena – Bol’shoy rivers area, glacial meltwaters flowed towards the north and NW and in the Ura River area to the SW and south. The presence of wet-based, tidal and outlet glaciers was probably the main cause of the rather restricted areal distribution of iceberg facies in the Patom Basin. In such environments, the main discharge of glacial debris from shelf glaciers and icebergs occurs in a proximal position. In general, the first rock association implies glacial and proximal glacial marine palaeoenvironments. The second and third rock associations are found predominantly in the middle member of the Bol’shoy Patom Fm. The second rock association comprises intercalations of turbidite, debris and grain flow sandstones and massive or bedded diamictites. Rare crossbedded coarse sandstones and conglomerates filled channels cut

BOL’SHOY PATOM FORMATION

into graded-bedded strata. Laminated mudstones with dropstones also occur. The second rock association was probably deposited as submarine fans. Their formation may be related to high-activity debris, grain and turbid flows related to sub-glacial or glacial rivers during glacial retreat. The third rock association consists of thinly bedded, finegrained sandstones with small-scale cross-bedding of climbing ripple-type and finely laminated sandstones with abundant carbonate grains. These deposits are probably related to interchannel interfans and distal parts of submarine fans with bottom currents. The sandstones with abundant carbonate grains could be distal turbidites transported by carbonate conglo-breccia debris flows. The third rock association may be related to a rise in sea level. The above-mentioned data allow us to conclude that the Bol’shoy Patom Fm. was deposited in the marine Patom basin during glaciation by and with the strong influence of ice sheets, iceberg rafting and glacial melt waters. The glaciogenic origin of the Bol’shoy Patom Fm. is supported by the presence of a typical cap dolomite conformably overlying it (first member of Barakun Fm., see section ‘Boundary relations with overlying and underlying non-glacial units’) that is very similar to such dolomites elsewhere in Asia, Svalbard, West Africa, Australia and other regions of the world and by well-founded correlation of this formation with the Nichatka Fm., which crops out on the southeastern flank of the folded arch and western margin of the Aldan shield. In Nichatka Fm., continental glacial facies have been discovered (Chumakov 1993). During deposition of the Bol’shoy Patom Fm., the southern part of the Siberian Craton was covered by ice sheets flowing into the Patom Sea. We thank M. I. Buyakayte for help in field and isotopic researches, and N. G. Vorob’eva and V.N. Sergeev for the investigation of microfossils from our collection. This research was supported by grants of the Russian Fund of Basic Investigations (nos 08-05-00433 and 10-05-00294) and Program No 24 of the Presidium of Russian Academy of Sciences. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) project #512.

References Bobrov, A. K. 1979. Stratigraphy and paleogeography of Upper Precambrian deposits of South Yakutiya. Yakutian Publishers, Yakutsk (in Russian). Chumakov, N. M. 1959. Stratigraphy and tectonics of the south-western part of the Viluy Depression. In: Shatsky, N. S. (ed.) Tektonika SSSR. IV. AN SSSR, Moscow, 345– 460 (in Russian). Chumakov, N. M. 1965. Certain Precambrian tillite-like rocks of the USSR. Izvestiya Academii Nauk SSSR, Seriya geologigeskaya, 2, 83 – 101 (in Russian). Chumakov, N. M. 1993. Riphean Middle Siberian Glaciohorizon. Stratigraphy and Geological Correlation, 1, 17 – 28. Chumakov, N. M. & Krasil’nikov, S. S. 1991. Lithology of Riphean tilloids; Ura Uplift. Lithology and Mineral Resourses, 3, 58 –78. Chumakov, N. M., Pokrovsky, B. G. & Melezhik, V. A. 2007. Geologic history of Patom Supergroup, Late Precambrian, Middle Siberia. Doclady Earth Sciences, 413, 379–383. Gorokhov, I. M., Semikhatov, M. A. et al. 1995. Strontium isotopic composition of Riphean, Vendian and Lower Cambrian carbonate rocks of Siberia. Stratigraphy and Geological Correlation, 3, 3– 33. Gray, K. 2005. Ediacaran palynology of Australia. Association of Australian Palaeontologists. Memoir 31. Association of Australian Palaeontologists, Canberra, 439. Ivanov, A. I., Lifshits, V. I. et al. 1995. Precambrian of the Patom Upland. Nedra, Moscow (in Russian). Keller, B. M. (ed.) 1963. Stratigraphy of SSSR. Upper Precambrian. Nedra, Moscow, 716 (in Russian). Keller, B. M., Semikhatov, M. A. & Chumakov, N. M. 1967. Upper Proterozoic of Siberian Craton and its frame. In: Stratigraphy of Precambrian and Cambrian of Middle Siberia. Krasnoyarsk knizhnoe izdatel’stvo, Krasnoyarsk, 71– 83 (in Russian).

315

Khabakov, A. V. 1933. A Brief Guide to Field Studies of Conglomerates. Nauchno-Tekhnicheskoe Izdatel’stvo, Moscow, 11 (in Russian). Khomentovskiy, V. V., Postnikov, A. A. & Fayzulin, M. Sh. 1998. Baikalian of stratotype locality. Geology and Geophysics, 39, 1505–1517. Khomentovskiy, V. V. & Postnikov, A. A. 2001. Neoproterozoic evolution Baikal-Vilyu branch of Palaeoasian Ocean. Geotectonics, 3, 3– 21. Khomentovskiy, V. V., Postnikov, A. A., Karlova, G. A., Kochnev, B. B., Yakshin, M. S. & Ponomarchuk, V. A. 2004. Vendian of Baikal-Patom Upland (Siberia). Geology and Geophysics, 45, 465– 484. Krumbein, W. C. & Sloss, L. L. 1951. Stratigraphy and Sedimentation. Freeman, San Francisco. Leonov, M. V. & Rud’ko, S. V. 2010. The find of Ediacarian fossils in Dal’nyaya Tayga Group (Ediacaran-Vendian of Patom Highland), in press. Lungersgauzen, G. F. 1963. Tillites and tillit-like deposits (in Russian). In: Keller, B. M. (ed.) Stratigraphy of SSSR. Upper Precambrian. Nedra, Moscow, 566–577 (in Russian). Meffre, S., Large, R. R. et al. 2008. Age and pyrite Pb-isotopic composition of the giant Sukhoi Log sediment-hosted gold deposit, Russia. Geochimica et Cosmochimica Acta, doi: 10.1016/ j.sca.2008.03.005. Nagovitsin, K. E., Faizullin, M. Sh. & Yakshin, M. S. 2004. New forms of Baikalian acanthomorhytes from the Ura Formation of Patom Uplift, East Siberia. Novosti paleontologii i stratigrafii, 6– 7, 7– 19. Oleynikov, B. V., Tomshin, M. D. & Kopylova, A. G. 1983. Basic rocks of Ura anticlinorium (in Russian). In: Oleynikov, B. V. (ed.) Petrology and Geochemistry Late Precambrian Intrusive Basic Rocks of Siberian Platform. Nauka, Novosibirsk, 146– 167 (in Russian). Pelechaty, S. M. 1998. Integrated chronostratigraphy of the Vendian System of Siberia: implications for a global stratigraphy. Journal of Geological Society of London, 155, 957– 973. Pokrovsky, B. G., Melezhik, B. A. & Buyakayte, M. I. 2006. Geochemistry of isotopes C, O, Sr and S, chemostratigraphy and environments of sedimentation of Late Precambrian deposits of Patom trough. Lithology and Mineral Resourses, 5, 505–530. Sochava, A. V., Podkovyrov, V. N. & Vinogradov, D. P. 1996. Variation of carbon and oxygen isotopes in Vendian– Lower Cambrian carbonate rocks of the Urinsky anticlinorium (South Siberian Platform). Lithology and Mineral Resources, 31, 248–257. Sovetov, J. K. 2002. Vendian foreland basin of the Siberian cratonic margin: Paleopangean accretionary phases. Russian Journal of Earth Sciences, 4, 363–387. Steiger, R. H. & Ja¨ger, E. 1977. Subcommission on geochronology: convention on the use ofdecay constants in geo- and cosmochronology. Earth and Planetary Sciences Letters, 36, 359– 362. Tiwary, M. & Knoll, A. H. 1994. Large acanthomorphic acritarchs from the Infrakrol Formation of the Lesser Himalaya and their stratigraphic significance. Journal of Himalayan Geology, 5, 193– 201. Veis, A. F., Vorob’eva, N. G. & Golubkova, E. Y. 2006. First find of Lower Vendian microfossils at Russian plate: taxonomy composition and biostratigraphic significance. Stratigraphy and Geological Correlation, 14, C28 – 46. Vinogradov, V. I., Pichugin, L. P., Bikhover, V. N., Golovin, D. I., Muraviev, V. I. & Bujakayte, M. I. 1996. Isotope features and dating of epigenetic alterations of Upper-Precambrian deposits of the Urinsky Uplift. Lithology and Mineral Resources, 31, 60 –69. Vorob’eva, N. G., Sergeev, V. N. & Semikhatov, M. A. 2006. Unicum Lower Vendian Kel’tma microbiota of Timan: new data on palaeontology of Vendian and its global features Doclady Academii nayk, 410, 366–371. Vorob’eva, N. G., Sergeev, V. N. & Chumakov, N. M. 2007. New occurrences of Lower Vendian microfossils in Ura Formation: problem Patom Supergroup of Middle Siberia. Doclady Earth Science, 419, 782– 787. Yang, J., Sun, W., Wang, Z., Xue, Y. & Tao, X. 1999. Variatios in Sr and C isotopes and Ce anomalies in successions from China: evidence for the oxygenation of Neoproterozoic seawater? Precambrian Research, 93, 215– 233.

316

N. M. CHUMAKOV ET AL.

Zhadnova, T. P. 1961. Stratigraphy of north-east part of Patom Highland. Trudy Tsentral’nogo nauchno-issledovatel’skogo geologorasvedochnogo institute, 38, 103– 123 (in Russian). Zhou, G. & Xiao, S. 2007. Ediacaran d13C chemostratigraphy of South China. Chemical Geology, 237, 89 – 108. Zhu, M., Strauss, H. & Shields, G. A. 2007. From snowball earth to the Cambrian bioradiation: calibration of Ediacaran-Cambrian earth

history in South China. Palaeogeography, Palaeoclimatology, Palaeoecology, 254, 1 –6. Zhuravleva, Z. A., Komar, V. A. & Chumakov, N. M. 1961. Succession and correlation of Upper Precambrian deposits of West Yakutiya. In: Bobrov, A. K. (ed.) Data on Geology and Mineral Resources of Yakutiya ASSR. Issue 13. Yakutiya Geological Survey, Yakutsk, 12 – 28 (in Russian).

Chapter 28 Late Cryogenian (Vendian) glaciogenic deposits in the Marnya Formation, Oselok Group, in the foothills of the East Sayan Range, southwestern Siberian Craton J. K. SOVETOV Trofimuk Institute of Petroleum Geology and Geophysics, Russian Academy of Sciences, Siberian Branch, Koptyug Av. 3, Novosibirsk, 630090, Russian Federation (e-mail: [email protected]) Abstract: The glaciogenic deposits of the Ulyakha Member (Mb.) referred to as Ulyakha diamictite (tillite), and the associated Late Cryogenian (Early Vendian) glaciofluvial deposits, are the lower strata of the Marnya Formation (Fm.) of the Oselok Group (Gr.) in the southwestern Siberian craton, which rest erosionally over the Cryogenian (Late Riphean) Karagassy Group. The Oselok Group deposits fill the Peri-Sayan foredeep and their stratigraphic equivalents immediately overlie the basement of the Siberian Platform. The glaciogenic deposits, including diamictite, breccia, boulder conglomerate and sandstone, had multistage sedimentation, were widespread, and underlie a cap dolomite of the lower Ozerki Mb. The Late Cryogenian (Early Vendian) age of the Ulyakha Mb. is defined by traces of shallow-marine soft-bodied animals in the middle part of the Ozerki Mb., and by findings of Ediacaran-like Metazoa in the overlying Bolshaya Aisa Mb. of the Marnya Fm. Deposition of the Ulyakha tillites and associated rocks has been correlated with the Late Cryogenian (Marinoan) glaciation based on the d13C patterns in carbonate deposits in the Marnya and Uda formations and their stratigraphic position much below the Pre-Manykai and Zhuya d13C negative signatures, which have been equated to the Wonoka anomaly.

Glaciogenic and associated deposits in the Sayan region occur at the base of the Oselok Group in the Biryusa province (Fig. 28.1). Their stratigraphic equivalents are the basal strata of the Olkha Fm. in the Irkutsk province (Baikal Group of the Baikal region), and of the Taseeva and Chapa Groups in the Yenisei Range (Sovetov et al. 2007a). All these sediments fill Late CryogenianEdiacaran (Vendian) foredeeps echeloned along the southwestern margin of the Siberian craton. The glacial facies correlate with poorly sorted sediments in the lowermost sections of the Nepa and Vanavara Groups in the craton interior and of the Dalnyaya Taiga Group in the Patom passive margin (Sovetov 2002a, b; Sovetov & Komlev 2005; Sovetov at al. 2007a). Thus, Late Cryogenian (Early Vendian) glacial events left their signature over about 1.8 million square kilometres. The Oselok Group was originally defined as a formation in the Sayan region (Khomentovsky 1950) and was recognized as a group after detailed stratigraphic studies in the 1960s when it was divided into the Marnya, Uda and Aisa formations (Dubin et al. 1969; Khomentovsky et al. 1972). These three units are distinguished from one another and from the underlying rocks by diamictite and breccia at the base of the Marnya Fm., arkose and orthoquartzite conglomerate and sandstone at the base of the Uda Fm. and lithoclastic massive sandstone near the bottom of the Aisa Fm. The succession of the three formations appears to persist continuously from the Tagul River in the NW to the Iya River in the SE of the Biryusa area (Fig. 28.1). The best outcrops of the Marnya Fm. (400 – 660 m) are found along the sides of the Uda valley between the Gladkiy Mys site and the Marnya inlet (Fig. 28.1). Berzin (1967) discovered unsorted mixitite at the base of the Marnya Fm. along the Uda River but found no traces of glacial origin in the diamictite. Dubin et al. (1969) and Khomentovsky et al. (1972) first distinguished a member of the Oselok basal breccia and conglomerate lying over deeply eroded Late Precambrian (Cryogenian or Upper Riphean) strata of the Karagassy Group. The original subdivision of the Marnya Fm. into lithological members 1 –4 (Dubin et al. 1969) and I –V (Bragin 1985) was later updated (Sovetov 2002a), and the members received their proper names and genetic identification. Diamictite was jointly assigned to the Ulyakha Mb. and faceted-boulder breccia to the lower Plity Mb. Diamictite and faceted-boulder breccia in the two members bear numerous traces of glacial origin, including striated boulders and abrupt facies changes, which has prompted

distinguishing local lithological members (Sovetov & Komlev 2005). Besides sedimentology, the Marnya Fm. (and the Oselok Group as a whole) was sampled for palaeontology in search of Ediacaran biota and traces of soft-bodied organisms, and for C- and O-isotope compositions in cap carbonates. On this basis, along with global Late Neoproterozoic correlations, the glaciogenic deposits of the area have been identified as Late Cryogenian tillite (Sovetov & Komlev 2005; Sovetov 2007). The discussion below concerns the sedimentology of the Marnya Fm. (Oselok Group) in its representative sections, composition of glaciogenic deposits, erosional features, pathways of clastics transported by ice and meltwater, biostratigraphic and chemostratigraphic constraints, and palaeomagnetic data with implications for Late Cryogenian –Early Ediacaran (Early Vendian) palaeogeography of the Siberian craton.

Structural framework and Neoproterozoic sedimentary basins of the Sayan region The study area of the Sayan region occupies the Biryusa catchment along the southwestern margin of the Siberian craton where the basement is exposed or locally buried under Neoproterozoic and Cambrian sediments (Fig. 28.1). The craton in this region borders a metamorphic belt of Palaeozoic rocks lying at 15 –208 to the basement along the Main Sayan Fault. The province of Palaeozoic rocks corresponds to the East Sayan uplift and includes blocks with Precambrian continental crust. The Main Sayan Fault is a zone of left-lateral strike– slip faults and mylonites, which apparently initiated in the Neoproterozoic (Cryogenian) when several terranes rifted off the Siberian craton (Berzin 1967). The zone experienced at least two episodes of compression during the Late Precambrian Baikalian (Cadomian) and Early Palaeozoic orogenies that produced a Late Cryogenian (Early Vendian; Sovetov 1977, 2002b; Sovetov & Blagovidov 2004) and Middle Cambrian –Early Ordovician (Anatol’eva 1972; Sovetov unpublished) molasse foreland basins. The sediments –basement contact is marked by a structural and metamorphic unconformity. The stratigraphic gap between the basement and the Cryogenian Karagassy Group reaches 1– 0.8 Ga in the Biryusa and Urik-Iya fault blocks and 2.9– 2.0 Ga within the Sharyzhalgai terrane where Archaean rocks underlie

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 317– 329. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.28

318

J. K. SOVETOV

Fig. 28.1. Precambrian geology of the Sayan region on the south-western Siberian Craton. (a) Geological map of the region showing structural elements and lithostratigraphy. (b) Stratigraphic scheme of the Neoproterozoic deposits in the Foothills of the East Sayan Range and their correlation with deposits in the adjacent regions. Supersequences: Sq 1, Marnya; Sq 2, Greben; Sq 3, Ust-Tagul.

MARNYA FORMATION, OSELOK GROUP

the basal diamictite of the Late Cryogenian –Early Ediacaran Olkha Fm. The oldest basement rocks (granulite, gneiss and schist) belong to the Archaean Sharyzhalgai Group in the Sharyzhalgai terrane (Sez’ko 1988). The Early Proterozoic Biryusa, Kan and Dzhuglyma terranes are composed of migmatized biotite, amphibole, garnet granulite and gneiss, granite gneiss and amphibolite. The Palaeo-Mesoproterozoic Urik-Iya trough basin is filled with 10 000 m of biotite-quartz phyllite, shale, metasandstone, conglomerate, tuff, and volcanic rocks (Berzin 1967; Sez’ko 1988). The metamorphic complexes of the Biryusa terrane are intruded by Palaeoproterozoic granites of the two-stage Sayan complex dated between 1870 –1734 Ma (Donskaya et al. 2002; Turkina et al. 2003). Unmetamorphosed Neoproterozoic sediments fall into two geodynamically different sequences: the rift-related deposits of the Cryogenian Karagassy Group and the Late Cryogenian –Ediacaran Oselok Group, which filled a foreland basin. The Karagassy Group lies, with a large structural unconformity, over metamorphic and igneous complexes of the Biryusa fault block and over low-grade metasediments and volcanic complexes of the Urik-Iya basin. The Oselok Group oversteps deeply eroded surfaces of the Karagassy formations in the Biryusa catchment; its stratigraphic equivalents (Olkha and Moty Groups) rest upon the Archaean Sharyzhalgai metamorphics only outside the rift basin in the Irkut catchment. According to recent data (Sovetov & Blagovidov 2004, 2006; Sovetov et al. 2007a), the Karagassy sediments fill the Cryogenian Iya-Tumanshet failed rift (aulacogen), whereas the Oselok Group is the fill of the Late Cryogenian –Ediacaran Peri-Sayan foredeep. The faultbounded rifts formed at the same time and in the same geodynamic environment as the Teya-Chapa and Vorogovka failed rifts 500 km to the NW in the Yenisei Ridge (Sovetov 1997; Sovetov et al. 2007a). The Karagassy rocks are intruded throughout the area by small bodies of gabbro dolerite sills, dykes and stocks of the Nersa complex with an Ar/Ar plagioclase age of 741 + 2 Ma (Gladkochub et al. 2006). The Late Cryogenian –Ediacaran Peri-Sayan foredeep initiated upon the Iya-Tumanshet rift and the adjacent craton edge after regression and deep glacial erosion during the Late Cryogenian glaciation (Sovetov 2002b; Sovetov & Blagovidov 2004; Sovetov & Komlev 2005). The Late Cryogenian –Ediacaran sedimentary fill of the Peri-Sayan, Peri-Baikal and Yenisei foredeeps, together with the coeval terrigenous and terrigenous-carbonate platform cover, belong to an enormous foreland basin on the western and southwestern margins of the Siberian craton. The Peri-Sayan foredeep basin evolved in two stages recorded in the Late Cryogenian– Early Ediacaran Marnya (1) and Late Ediacaran Greben (2) regional supersequences (Sovetov et al. 2007a). The Marnya supersequence consists of glaciogenic deposits, shallow-marine terrigenous and carbonate rocks, and less abundant fluvial facies. The Early Ediacaran was associated with transgressions from the SW onto the southern and southwestern parts of the Siberian craton as a result of the closing Palaeoasian ocean and clastic transport from the craton interior. During the Late Ediacaran (Greben) syncollisional stage, the basin experienced transgressions from the NE, and the clastic transport was from flanking orogens in the western, southwestern and southern craton margins. The Cambrian history of the Siberian craton began with the break-up of the Ediacaran (Late Vendian) orogens, formation of a terrigenous hanging-wall basin, followed by the development of evaporite and carbonate basins (Sovetov et al. 2007a). The Karagassy and Oselok sediments were deformed by postCambrian to pre-Devonian west –east compression (Berzin 1967) that split the sedimentary basins into blocks displaced along strike –slip, thrust and reverse faults. This deformation event produced shear zones and related folds, but was not accompanied by metamorphism.

319

Stratigraphy The Karagassy and Oselok Groups (Fig. 28.1) in the Biryusa watershed were interpreted jointly as a single complex (Khomentovsky 1950) before a stratigraphic gap between them and a complex stacking pattern was discovered, providing grounds for their subdivision (Dubin et al. 1969).

Karagassy Group The Karagassy Group, varying in thickness from c. 800 to c. 2000 m, consists of the Shangulezh, Izan (currently referred to as Tagul; Egorova et al. 1971; Bragin 1986; Shenfil 1991; Sovetov & Blagovidov 2006) and Ipsit formations (Khomentovsky et al. 1972). The Shangulezh Fm. (250 –600 m) lies with an angular unconformity over the Urik-Iya folded metamorphosed sediments and volcanic rocks and over high-grade metamorphic rocks and Palaeoproterozoic granitoids of the Sayan complex. The three formations are second-order sequences and are separated from one another by gaps and low-stand erosional surfaces. The sequences record facies changes from alluvial fans to a terrigenous carbonate tidal plain and a carbonate tidal shelf with abundant stromatolite buildups, microphytolite dolarenite and dolomicrite. Of broad occurrence are the structures of supratidal, subtidal and peritidal deposits with bimodal flow directions of terrigenouscarbonate and silt material, minor scouring forms, flat stromatolite mats, mud cracks and halite glyptomorphs.

Oselok Group The Oselok Group overlies the Karagassy Group without structural unconformity but with an erosional contact (locally about 1000 m of the Ipsit and partly Tagul deposits have been removed). The group includes the Marnya, Uda and Aisa formations (Dubin et al. 1969; Khomentovsky et al. 1972; Sovetov & Komlev 2005). The eroded top of the Aisa Fm. underlies, without structural unconformity, the basal conglomerates of the Ust’-Tagul Fm. Marnya Formation. The general succession of members, their

replacement and pinching out, have been inferred from correlation and are reported below following Sovetov & Komlev (2005), with later additions (Fig. 28.2). The new subdivision updates the earlier stratigraphic interpretations (members I to V) by Dubin et al. (1969) and Bragin (1984, 1985). The Karapchatui Member (I), the oldest in the Marnya Fm., was discovered by Bragin (1984), and its stratigraphy was updated through studies by Sovetov in 2006 along the right side of the Uda River (Karapchatui Brook mouth). The section comprises 0.1– 0.2 m of ferruginous silicified cataclastic rocks at the base overlain by (1) 3.2 m of massive quartzite faceted-boulder breccia and white cross-bedded orthoquartzite; (2) 22 m of white and yellow cross-bedded orthoquartzite with debris in small fluvial – lacustrine cycles with thin-bedded orthoquartzite and stromatolite dolomite; (3) 30–35 m of variegated thinly laminated quartz sandstone and stromatolitic dolomite; (4) 20– 30 m of quartz and massive cross-bedded feldspar-quartzose sandstone. Orthoquartzites give way to breccia in the SW, and also in the NE, towards the sides of the underlying channelized unconformity. The Nersa Mb. (18.7 m thick) (Mb I, Bragin 1984), found on the west side of the Uda River in the Nersa and Kremenshet interfluve, consists of massive fluvial boulder conglomerate and coarse pebbly sandstone in the upper section, with clasts of the Karagassy arkose quartzitic sandstone, stromatolite dolomite, chert and dolerite. A c. 22-m-thick sill of the Nersa complex was found 2.1 m below the Nersa-Ipsit contact. The Ulyakha Mb. stratotype occurs along the right side of the Uda River 0.3– 1 km upstream of the Ulyakha mouth, where the

Karapchatui

Plity

Bogatyr

Kedrovyi

Pescherny Muksut Mb. Mb. Kagat Mb.

Glacio-fluvial quarzite sandstone with detritus Dolostone of ephemeral postglacial lake Glacio-fluvial litharenite with single pebbles

Unyl Mb.

Laminated black and grey-green mudstone with sandstone lenses Eolian (?) quartzite sanstone dunes Laminated cap-dolomite

Nizhnyaya Uda Mb.

Current-rippled dolostone Plity Mb.

Grainstone sand waves

N=31

Kedrovy Mb.

Plity Ozerki Mb. Mb.

Bol’shaya Aisa Mb.

Quarzite sand waves and bars

N=19 170 m

20 m

Karapchatui Mb.

20 m

Ipsit Fm. Tagul Fm.

99 00

Kedrovyi

r ive aR

Bogatyr Plity

Bi

54 30

us

60

La

ke

Ba

ik

al

Karapchatui

110

25 km

Ipsit Fm.

Hurricane cross-bedded sandstone Fluvial cross-bedded gravel and sandstone quartzite and arkose Delta plain thin bedded siltstone with gravel and sandstone lenses Ooids and stromatolitic limestone Sandstone coarse cross-bedded bars and channels of deep-water rivers Trace-fossils

Ozerki Kirei Gladky Mys

Palaeocurrent directions Red stromatolitic dolostone Sandy tempestite and sand-waves with slumps

Nizhneudinsk

70

Tide-related channelized sandstone -siltstone and micrite limestone tempestite Gravelly sandstone tempestite

N=10

Uda River

Ulyakha Mb.

Tygnei Mb.

N=52

Siberian Craton

Sandstone-siltstone tempestite

Ulyakha Mb.

Nersa Mb.

Tagul Fm.

100

Ipsit Fm.

30 m

N=27

Quarzite foreshore bars

N=53 Plity Mb.

20 m

Ognit Mb.

Ikei Fm. (after Bragin 1985)*

Ozerki

ry

m0

Kirei

Ozerki Mb.

100

Ulyakha

Dolerite sills Clay-rich diamictite with predominant quarzitic cobbles and boulders Dolomitic diamictite with predominant dolomite cobbles and boulders Faceted-boulder breccia

a b c

Moulds and imprints of soft-body animals: a.Cyclozoa, b. Bilaterian, c. Problematics Glacial erosion surface

Ulyakha

Boulder conglomerate

20 m

Non-cropped parts of sections

Fig. 28.2. Stratigraphy and depositional systems of the Marnya and Uda formations at different locations along the Uda and Biryusa rivers, showing the lateral variability and stratigraphic relationships of the members from south to north. Note that the Ikei Formation includes the Aisa Fm. and Kagat and Muksut members, which are differentiated on the basis of colour. The Ikei Fm. is generally no longer referred to in the literature. Inset maps show localities of stratigraphic logs as well as average palaeo-ice flow direction based on sedimentary structures in glacial and glaciofluvial sediments.

J. K. SOVETOV

200

Uda Fm. (after Bragin 1985)

300

Marnya Fm. (after Dubin et al. 1969)

400

320

500

Gladky Mys

MARNYA FORMATION, OSELOK GROUP

Member erosively overlies the Tagul Fm. The contact is perfectly exposed in the Uda side 5 km upstream of the Marnya mouth. The Ulyakha Mb. (up to 53 m thick) (Mb I, Bragin 1985), comprises lenses and upright veins of greenish-grey, dark grey (to black) or red coarse diamictite and dolomite breccia, and less abundant channelized coarse sandstone (Figs 28.2 & 28.3). The member is recognizable by its boulders and basement erratic boulders, abundant glacial striation on clasts, grooves and other traces of erosion on the surface where diamictite overlies the Tagul Fm. The Tygnei Mb. (80 –90 m thick) (Mb IIa, Bragin 1985), composed of black bituminous pyrite siltstone with thin stromatolite dolomite in the upper part and sandstone interbeds and greenishgrey laminated siltstone in the lower part, crops out along the sides of the Uda River. The Plity Mb. (20 –35 m thick) (Mb IIb, Bragin 1985), consists of faceted-boulder quartzite breccia and yellowish-white feldsparquartzose sandstone under massive fluvial and eolian (?) cross-bedded sandstone. The Kedrovy Mb. (40 m thick), oversteps the Ulyakha diamictite on the west side of the Uda River (near Kedrovy Village) and consists of grey coarse extralitharenite sandstone with cobbles and pebbles at the base of fluvial cycles. It underlies white coarse arkose sandstone of the Ognit Mb. and is a stratigraphic equivalent of the Plity Mb. The Ozerki Mb. (50– 60 m thick) (Mb II, Bragin 1985), lies conformably over various members of the Marnya Fm. at different localities (Fig. 28.2). It pinches out in the Sayan region where the pre-Marnya erosion was moderate. The sections of the member are either (i) yellow laminated, wave rippled and tidalcurrent cap dolomite with numerous cavities after pyrite nodules or (ii) laminated stromatolite and cross-bedded sandy dolomite, and feldspar-quartzose sandstone in the middle part with abundant traces of soft-bodied animals (Sovetov 2007; Sovetov & Solovetskaya 2008). At one locality (Ulyakha, Fig. 28.2), the base of the carbonate Ozerki Mb. overlies the diamictite-bearing Ulyakha Mb. and is thought to correlate with the base of the Ediacaran. The Ozerki Mb. is overlain by the Ognit and Bolshaya Aisa members. The Ognit Mb. (up to 84 m thick) (Mb IV, Bragin 1985), consists of dark-grey and black coarse orthoquartzite and arkoses, whereas the Bolshaya Aisa Member (105 –269 m thick) (Mb V, Bragin 1984) consists of shoreface greenish-grey crossbedded sandstone and siltstone and shelf HCS tempestites and tidalites, with less abundant black microphytolite and silty limestone. The Bolshaya Aisa Mb. contains Ediacaran-like biota (Sovetov & Komlev 2005; Sovetov 2007; Sovetov & Solovetskaya 2008).

Uda Fm. The stratigraphic extent of the Uda Fm. long remained unclear after the lowermost grey rocks of the Oselok Group, originally attributed to it, had been included into the Marnya Fm. According to Bragin (1985), its lower section comprised the Ognit and Bolshaya Aisa members but Khomentovsky et al. (1972) and Sovetov & Komlev (2005) identified its lower boundary by arkose and quartz gravelstone of the Nizhnyaya Uda Mb. Together with the latter, the Uda Fm. is currently divided into three members (Fig. 28.2). The Nizhnyaya Uda Mb. (VI, Bragin 1985), most often 17– 30 m, is composed of white and yellow fine-pebble quartz conglomerate, gravelstone and coarse gravelly cross-bedded fluvial sandstone. The member rests upon the deeply (to 4 m) eroded surface of the Bolshaya Aisa Mb. and to the SW its lower 40 m (out of a 70 m) section comprises coastal bar and sandwave deposits. The Unyl Mb. (VI, Bragin 1985), 25 –100 m, is genetically linked to the Nizhnyaya Uda Mb. and consists of grey and

321

dark-grey siltstone with lenses of fine to coarse cross-bedded sandstone deposited in subsidiary channels and river mouth bars. The Peschernyi Mb. (VIII, Bragin 1985), 7 –30 m, is composed of dark-grey and black stromatolitic and microphytolitic limestone and limestone tempestite, is locally replaced by epigenetic dolomite, and varies strongly in thickness, pinching out cratonward.

Aisa Fm. The Aisa (Ikei, Bragin 1985) Fm. is composed of continental molasse of two lithofacies groups: (i) fluvial channel litharenite sandstone and (ii) bar and overbank sandstone, siltstone and, rarely, mudstone. Estimates of its total thickness vary from 1800 –2000 m (Egorova et al. 1971) to 1300 –1600 m (Khomentovsky et al. 1972), or 1300 –1400 m according to measurements by Sovetov in the most complete section on the sides of the Biryusa and Tagul rivers. The formation has a prominent base along the roof of the Peschernyi limestone and a very sharp top at the base of the overlying Ust’-Tagul Fm. conglomerate attributed to the Late Vendian Nemakit-Daldyn regional stage (EdiacaranEarly Cambrian) (Sovetov et al. 2007a). The Aisa Fm. includes four members of different thicknesses: Kagat, Muksut, Serebrovo and Katalchikov. Overbank deposits contain abundant Arumberia plant remnants (Sovetov 2006).

Glaciogenic deposits of the Marnya Fm. and associated strata Diamictite of the Ulyakha Mb. was studied in outcrops along the Uda River between the Marnya and Ulyakha mouth areas and near Kedrovyi Village on the Biryusa River (Nersa section). The maximum thickness of diamictite measured from its lower erosion surface to the base of the Ozerki Mb. cap dolomite is 53 m. The matrix of boulder diamictite (mixtites) of the Ulyakha Mb. (c. 85% of the rock volume) has a sand-silt-clay composition. The rocks bear evident lithological signatures of tillite (Sovetov 2002a; Sovetov & Komlev 2005; Sovetov & Donskaya 2006), namely: (1) iron- and bullet-shaped stones, which are faceted, striated, polished and range in size from gravel to clasts up to 70 cm (occasionally 1.5 m) across; (2) orientation of elongate stones; (3) lodgment stones pressed into dolomite of the Tagul Fm.; (4) dropstones in sandy dolomite of outwash lakes; (5) steepwalled subglacial channels, boulder beds and clustered stones; (6) up to 1.5-m-deep nearly vertical crevasses with breccia; (7) inner stratification of diamictite in the form of four nested lenses, differing in coloration, composition of stones, and containing meltout intercalations; (8) associated fluvial sandstone with debris laterally replaced by breccia; (9) stones of rocks typical of the underlying Tagul and the Ipsit formations (Karagassy Group) and less frequent erratics of gneiss, granite, biotite schist and metarhyolite derived from the Early Proterozoic basement; (10) erosional forms, such as narrow (a few decimetres wide) and shallow furrows, scours tens of metres wide and 3– 4 m deep, local (tens to hundreds of metres across) and regional (tens of kilometres wide) erosional valleys; (11) blunt-end contacts of diamictite with sidewalls of channelized troughs; (12) folds and rupture features at the top of the underlying rocks of the Ipsit and Tagul formations; (13) association with laminated cap dolomite with sharp lower contact; (14) association with laminated siltstone and shale alternated with gravelly-sandy deposits of gravity flow; (15) association with boulder conglomerate containing striated pebbles; (16) absence of ash fall, impact events, lahars, volcaniclastic debris flows or turbidity flows on the continental slope and rise; (17) absence of indications typical of tectonic thrusting or melange; (18) occurrence of diamictite above a regional unconformity. The architectural elements in Marnya Fm. diamictite were distinguished by mapping outcrops. Figure 28.3 shows cross-sections

322

I

III

VI

V

IV

VII

VIII

IX

XI

X

Ozerki Mb.

II

S n=649

20 m

n=111

n=141

n=226

n=112

(b)

2m

Ozerki Mb.

10 m

N Profile direction 325 n=209

I

n=575

III

II

Plity Member

Marnya Formation

N

dolomite chert arkose and orthoquartzite basement rocks

stones in diamictite

Marnya Formation Ulyakha Member

10 m

S

VI

V

IV

VII (c)

III

2,5 m 5m

VIII VIII

IX

IX

X

XII (d)

XI

20 m

200

Vergence of glacial dislocations

Key to symbols:

Ulyakha Mb.

Profile direction 50 Black argillaceous diamictite and stones 20 cm Green-grey argillaceous diamictite Red-brown sand-clay mixitite, no stones Dolomite diamictite and breccia Intraglacial and supraglacial breccias and gravelstone in channels and crevasses

Black cross-bedded sandstone in intraglacial and supraglacial channels Elongation direction of cobbles and stones Elongation direction of lodgment cobbles and boulders Shear zones and boundaries of diamictite lenses Basal shear zone

Deglacial fine detrital diamictite

Outwash-plain cross-bedded glaciofluvial sandstone with cobbles (zandur)

III

Laminated cap dolomite

Pebbly sandstone from currents (underwater fans and fan-deltas)

Massive and layered stromatolite and grainstone dolomite

Glacial shears

Dollutite

Percentages of different lithologies of stones 4-40 mm

Laminated siltstone and mudstone in ephemeral lake

Closed parts of section

XI

Diamictite lenses separated by a shear zone Reference sections of outcrop logs

Fig. 28.3. Outcrop logs and field sketch showing different facies and lithology of diamictite clasts in the upper and lower Ulyakha Member along the Uda River. (a) Vertical section of an outcrop on the left side of the Uda River 1 km upstream of Ulyakha Brook. (b) Enlarged left side of section (a) showing the geometry of sandstone and conglomerate lenses within black argillaceous diamictite. (c, d) Single vertical section of the lower Ulyakha Mb. at the Ozerki locality on the left side of the Uda River, 5 km upstream of the Marnya River.

J. K. SOVETOV

Tygnei Member

I II

III

VIII

Tagul Fm.

XV

XIV

IV

2m

Tagul Fm.

Glaciogenic deposits in Ulakha river location

XIII

(a)

III

Glaciogenic deposits in Ozerki location

XII

MARNYA FORMATION, OSELOK GROUP

and logs of the upper and lower Ulyakha Mb. along the Uda River. For complete stratigraphy of the member, see Figure 28.2. The upper Ulyakha Mb. (Fig. 28.3a) consists of four nested lenses of diamictite with shear boundaries (I to IV). They are named black and green argillaceous diamictite after their matrix colour and siliciclastic composition. Diamictite consist of outsized clasts from sand- and fine-gravel-sized rocks to elongate iron-shaped boulders with striation oriented at 100 –2158. Lenses III and IV are cut with 3–17-m-wide channels filled with either dolomite diamictite, conglomerate or breccia. The lenses are massive or have inclined bedding with inclusions of black coarse cross-bedded sandstone; the channels have steep or overhanging walls, some resemble crevasses (Fig. 28.3b). Palaeocurrent direction measured on cross-bedding and channel-bed slopes suggests currents moved towards 230 –3308. The Ulyakha Mb. at the Ozerki outcrop consists of three lenses separated by shear zones or shear planes (Fig. 28.3c, d). These deposits host abundant boulders of the Ipsit Fm. quartzite with striation, less abundant boulders of the Tagul Fm., silicified stromatolitic dolomite, and a few erratic boulders of basement granite and gneiss. The boulders and their striation are oriented mainly at 150– 2408. The direction of glacier advance has been inferred from the 150– 2708 orientation of lodgment boulders pressed into the Tagul dolomite. A similar interpretative orientation (2008) has been found in folds of the Tagul dolomite and siltstone beds beneath the diamictite. Small folds only occur at the Ozerky location above a horizontal shear zone 8 –10 m below the Ulyakha Mb., and the underlying part of the Tagul Fm. is rather undeformed. The upper Ulyakha Mb. includes black siltstone and lenses of dolomite diamictite and breccia in U-shaped channels. According to correlation (no visible contact),

323

diamictites are thought to be buried under black bituminous siltstone deposited in ephemeral lakes (Tygnei Mb.), which in turn are overlain by laminated, wave-rippled and stromatolite dolomite of the Ozerki Mb. The Karapchatui Mb. fills a 60–70-m-deep valley, incised into the Ipsit and Tagul Formations (Figs 28.2 & 28.4). The valley is filled with cross-bedded sandstone, sandy stromatolitic dolomite and faceted-boulder breccia. The breccia occur in the valley floor and its east side, and the centre is occupied by an outwashplain fan oriented at 340– 08 in the NW – north direction, according to numerous measurements of oblique foresets in streams, large ripples and dunes.

Boundary relations with overlying and underlying non-glacial units All members of glaciogenic deposits at the base of the Marnya Fm. (diamictite, faceted-boulder breccia, conglomerate and sandstone) have sharp contacts with the underlying deposits of the Ipsit and Tagul formations in the Sayan region, with local scouring features (Figs 28.2–28.4). The deepest erosion exceeds 500 m in the area between the Biryusa and Uda rivers (Fig. 28.4) and is c. 1000 m throughout the Sayan region (Dubin et al. 1969; Bragin 1984; Sovetov & Komlev 2005). The erosive contact of the Ulyakha Mb. black argillaceous diamictite with the underlying dolomite of the Tagul Fm. (Karagassy Group) was mapped on the left side of the Uda River at the Ozerki locality (Fig. 28.3c). The profile runs across a 90–110-m-long and 4.5-m-deep flat-bottomed channel, which is incised into the Tagul

Fig. 28.4. Cryogenian (Early Vendian) glacial erosional forms in the southwestern Siberian Craton. See Figure 28.2, inset, for map of localities.

324

J. K. SOVETOV

Fm. and is filled with dolomite diamictite from the lower Ulyakha Mb. The total erosion depth at the locality approaches 9 m. Grooves and scours with sharp contacts on the surface of the Tagul dolomite, as well as terraces, are oriented at 190– 1958 (Fig. 28.4). A large erosional valley is exposed on the right side of the Uda River at the mouth of the Karapchatui Brook. The Karapchatui Mb. overlies the Tagul Fm., with 10–158 unconformity, in the centre of an erosional valley (Figs 28.2 & 28.4). The eastern valley side dips at 30 –508 and strikes at 3008, and its upper part is composed of the Ipsit Fm. rocks (Fig. 28.4). Another exposed erosional valley occurs along the Uda River at the Plity locality (Fig. 28.2). The 15-m-deep north – south valley is incised into the Ipsit Fm. and filled with faceted-boulder breccia and minor cross-bedded sandstone of the lower Plity Mb. Thick-bedded cross-bedded sandstone of the upper Plity Mb. covers the eastern side of the valley and lies over the Ipsit quartzose arkose siltstone with a sharp contact. The dip of oblique cross-bedding foresets indicates north –south transport of material along the valley. The same relationship between the Plity Mb. and the Ipsit Fm. is found in outcrops along the right side of the Tagul River 6 km upstream of Georgievka Village. The distinct base of the glaciofluvial boulder conglomerate of the Nersa Mb. (Marnya Fm.), with siltstone of the Ipsit Fm., crops out along the Biryusa River (near Kedrovyi Village) (Fig. 28.2). The upper boundary of the Marnya glaciogenic deposits is well exposed along the sides of the Uda River at the Ulyakha and Ozerki localities. The Ulyakha Mb. rocks underlie laminited cap dolomite along a sharp contact (Fig. 28.3). The same prominent boundary between the Plity sandstone and cap dolomite is exposed at the Ozerki locality (Fig. 28.2). Dolomite of the Ozerki Mb. pinches out along the sides of the Uda River (Plity and Gladkiy Mys sites and at Bogatyr’ Hill) and leaves sandstone tempestite of the overlying Bolshaya Aisa Mb. Overlying the Plity Mb. (Fig. 28.2).

Chemostratigraphy Tentative estimates of 87Sr/86Sr ratios were obtained in the Peschernyi Mb. limestone. These Sr ratios are lower than those reported from the Wonoka Fm. (Calver 2000), and the correlated Shuram Fm. of Oman (Burns et al. 1994) (0.70796 – 0.70848 v. 0.7087 –0.7088 and 0.7085 –0.7088, respectively; Fig. 28.5). C- and O-isotope compositions were studied in carbonate rocks of the Marnya and Uda formations (Fig. 28.5; Sovetov & Komlev 2005; Sovetov et al. 2007b). The resulting C-isotope curve includes five characteristic segments A –E (Fig. 28.5; Sovetov & Komlev 2005). The d13C values in segment A, which correspond to samples from the middle and upper parts of the Tagul Fm. (Karagassy Group), vary from – 1.5 to þ3‰. All d13C values in dolomitic diamictite of the Marnya Fm. (segment B) are positive and within þ1.6‰. A negative anomaly occurs in segment C corresponding to the overlying Ozerki dolomite, where d13C gradually decreases to –5.7‰ and then increases smoothly to zero towards the top of the Ozerki Mb. A positive d13C excursion is found in the upper third of the Bolshaya Aisa Mb. (Marnya Fm.; from þ6.3‰ decreasing upward to þ4‰) as well as in the Peschernyi Mb. (Uda Fm.; þ3.1 to þ7.5‰). Based on these data, the Ulyakha diamictite was correlated with the Late Cryogenian Varanger glacial deposits in Norway (Sovetov & Komlev 2005). The chemostratigraphic chart as a whole has the following features (Fig. 28.5): (i) a d13C low in cap carbonates over the Ulyakha Mb., (ii) transition to positive d13C signatures spanning the Bolshaya Aisa and Peschernyi members in the Sayan area (Sovetov & Komlev 2005, figs 6–8), comparable to the anomalies found in the Uluntui Fm. of the Baikal Group, southwestern Baikal area (Khabarov & Ponomarchuk 2005), (iii) a large d13C low in the Kachergat Fm. of the Baikal Group, to – 6‰ (Sovetov et al.

2007b), similar to a low of –10‰ in the Oskoba Fm., Baikit region (Vinogradov et al. 1994; Khomentovsky et al. 1998; Sovetov et al. 2007b), which both appear to correlate with the Zhuya and partly the N2 negative anomalies (Pelechaty 1998; Pokrovsky et al. 2006) in the Patom region and with the Wonoka anomaly identified in various parts of the world (Halverson et al. 2005). The grounds for this correlation are discussed below.

Palaeolatitude and palaeogeography Palaeomagnetic directions were measured in diamictite samples from the Ulyakha Mb. (Marnya Fm.) (54.338N, 98.878E) and the overlying Tygnei black siltstone (54.338N, 98.898E) collected by Sovetov at the Ozerki locality (Uda River). Diamictite and dolomite exhibit nearly horizontal bedding planes at the locality. Measurements were performed using the standard procedure (Butler 1992) of stepwise thermal demagnetization on a magnetometer at the Paleomagnetic Centre of International Petroleum Geology and Geophysics (Novosibirsk, Russia). According to published and preliminary data, the primary palaeomagnetic directions in the rocks of the Marnya Fm. have been overprinted by Early Cambrian geodynamic events (Metelkin et al. 2005; D. V. Metelkin, pers. comm.). The available palaeomagnetic evidence for Neoproterozoic and Early Palaeozoic rocks from the Siberian craton (Smethurst et al. 1998; Pisarevsky & Natapov 2003; Shatsillo et al. 2005; Metelkin et al. 2007) suggest a sub-equatorial location.

Geochronological constraints The minimum age of glaciation comes from Ediacaran (Late Vendian) –Early Cambrian small-shell fauna found within the Ostrovnoi Fm., (Khomentovsky et al. 1998), which is correlated with the upper part of the Ust’Tagul Fm. (Fig. 28.6). The latter erosively overlies the Oselok Group, while its equivalents overlie the Taseeva and Chapa groups in the Yenisei Ridge (Sovetov 2002b, fig. 6; Sovetov & Blagovidov 2004). Furthermore, Metazoan trace fossils were found in quartzite in the middle of the Ozerki Mb. and imprints and moulds of Ediacaran-like organisms have been reported in the Bol’shaya Aisa Mb. (Sovetov & Komlev 2005; Sovetov 2007; Sovetov & Solovetskaya 2008). The maximum age for the Marnya Fm. glaciogenic deposits in the Sayan region has been constrained by two dates: (i) the 741 + 2 Ma 39Ar/40Ar age of plagioclase from the Nersa gabbro dolerite sills (Gladkochub et al. 2005) that intrude the Shangulezh Fm. of the underlying Karagassy Group (gabbro dolerite clasts are also found in the Ulyakha Member diamictite), and (ii) the 39 Ar/40Ar 696 + 8.5 Ma age of biotite from dolerite and tuff (Postnikov et al. 2005) among diamictite of the Chivida Fm. (middle section of Chingasan Group; Yenisei Ridge) that lies stratigraphically lower than the Oselok Gr.

Discussion Main structural and morphological features of the Marnya basal strata were discussed in Sovetov & Komlev (2005) and Sovetov (2008). The formation includes several major glaciogenic sedimentary structures and architectural elements (Miall 1985), namely (i) diamictite with striated boulders and poorly rounded blocks, (ii) faceted-boulder breccia and (iii) intradiamictite- and supra-diamictite glacial channels and crevasses. The subsidiary elements are (iv) outwash –plain fluvial debris sandstone with thin layers and lenses of stromatolite dolomite, (v) sandstone and gravelstone deposits of alluvial plains, fans and fan deltas, (vi) laminated fine silt and mudstone deposited in ephemeral stagnant lakes. Glaciogenic deposition was accompanied by deep erosion of

MARNYA FORMATION, OSELOK GROUP

Sr/ Sr

Pescherny Mb. Unyl Mb. Nizhnyaya Uda Mb. Bol. Aisa Mb. Ognit Mb.

Ozerki Mb.

E

Borakun Fm.

? D

C Dzhemkukan Fm.

Onset glacigenic deposits Ulyakha Mb.

B

Phytolithic bioherms and storm-dominated deposits Contour currents deposits Deep-water delta Siliciclastic deep-water fan

Siliciclastic and carbonate deep-water fans

Siliciclastic deep-water fan and diamictite

Late Cryogenian glaciation

86

Passive continental margin Continental slope and shelf Continental rise

Valukhta Fm. 87

Patom Upland Ura river

Nikolsky Fm. (isotope data by Pokrovsky et al.,2006)

Uda Fm. Marnya Fm.

(isotope data by Sovetov, Komlev, 2005)

Depositional systems

Сhencha Fm.

Kachergat Fm.=Aisa Fm.

2007 b)

Baikal region, Kurtun river (isotope data by Sovetov, et al.,

δ13C

10 8 6 4 2 0 2 4 6 8 10

0.70796-0.70848

Sayan region Uda river

Epicontinental sea, strand plain

Braided alluvial and delta plain

δ13C

10 8 6 4 2 0 2 4 6 8 10

325

Mariinka Fm.

Onset glacigenic deposits Tagul Fm.

A

Fig. 28.5. Chemostratography of the Oselok Gp. and correlation with deep-water deposits of the Patom Upland. Carbon-isotope curve for the Marnya and Uda formations is complemented with some data from the Kachergat Fm., which is a stratigraphic equivalent of the Aisa Fm. The uppermost negative anomaly in the Nikolsky and Chencha formations is thought to correlate with the Wonoka anomaly (Halverson et al. 2005) and the Pre-Manykai anomaly identified by Pokrovsky & Missarzhevsky (1993).

the substrate and multiform scouring, and dislocation of rocks in the underlying bed (Fig. 28.4). Judging by their morphology, the channel-like stone clusters formed in ice during deglaciation and can be identified as kames. Diamictite have all the features of tillite, and are thought to have been produced by rock crushing and transport of clasts by ice and through sub-ice channels by meltwater. Deglaciation is recorded by deposition of black siltstone (to 0.5 m) in ephemeral lakes and, later, by deposition of dark-grey and black laminated cap dolomite of the Ozerki Mb., which grade up the section into stromatolitic dolomite, dolomite tidalites and sand –grainstone sandwaves. The Ulyakha tillite, with the associated faceted boulder breccia, boulder conglomerate and outwash sand, make up a complex succession (Sovetov & Komlev 2005). The palaeocurrent directions suggest the existence of two major glaciers in the Siberian craton: a smaller one marked by the Karapchatui outwash fan in the south and a larger one evident from the Ulyakha Mb. channels and striated lodgment boulders in the NE of the area in the central Siberian Craton. The extent of the latter glacier is recorded in clastic transport, erratic stones from the basement of the Siberian Platform and in sedimentary structures observed in the Sayan and Baikal regions (Sovetov & Donskaya 2006). Lithostratigraphic evidence available for deposits of the Precambrian –Cambrian transition from different areas of the Siberian craton bears some signatures that may serve as ties in regional correlations. Figure 28.6 shows lithostratigraphic correlation of the Oselok Group with the Taseeva Group of the Yenisei Ridge, the Olkha and Moty Groups of the Irkutsk province,

the Baikal Group of the Baikal region, and the Nepa and Tira formations of the Baikit and Nepa-Botuobiya forebulges (Sovetov et al. 2007a). The regional stratigraphy and correlations are controlled by the following marker units and surfaces (Fig. 28.6): (1) transgression and flooding surface at the base of the Katanga and Ostrovnoi formations and its equivalents at the base of the Kurtun Fm. and in the middle of the Shamanka, Ust’-Tagul, and Uglovoi formations, (2) base of the Ust’-Tagul supersequence with river incision, (3) base of the Greben supersequence with river incision, (4) base of the Marnya supersequence defined by a deeply eroded surface under glacial and glaciofluvial deposits. Subsidiary correlation levels are associated with third-order sequences (Sovetov 2002b; Sovetov & Komlev 2005). The Marnya glaciogenic deposits have been correlated with breccia, diamictite at the base of the Olkha Fm. (Mordvin 1972; Chumakov 1993), Baikal Gr. (Sovetov & Komlev 2005) and the base of the Chapa Gr. (Sovetov et al. 2007a), conglomerate and breccia at the base of the Taseeva Gr. (Sovetov et al. 2007a), as well as the Vanavara, and Nepa Formations (Tyschenko 1980) (Fig. 28.6). Carbonate members in the Marnya, Uda, Goloustnaya and Uluntui formations pinch out cratonward from foredeeps, while carbonate members of the Oskoba and Tira formations are replaced in the foredeeps by terrigenous molasse (Fig. 28.6). Sedimentological studies of the upper Oselok Group showed that deposition of the Peschernyi marine limestones was followed by a new evolutionary stage of the foreland basin, with orogens and the bordering alluvial plains originating around the southwestern edge of the Siberian craton (Sovetov & Blagovidov 2004;

326 J. K. SOVETOV

Fig. 28.6. (a) Correlation of Late Neoproterozoic deposits in the southwestern Siberian Craton and stratigraphic position of the tillite units, complemented and modified after Sovetov et al. (2007a). Formation name abbreviations: Ayan, Ayankan Fm.; Bol, Bol’shoi Lug Fm.; Bor, Borakun Fm.; Ch, Chencha Fm.; Kal, Kalancha Fm.; Khuzh, Khuzhir Fm.; Kul, Kulekin Fm.; M, Mariinka Fm.; Nik, Nikol’sky Fm.; Shan, Shanhar Fm.; Sham, Shamanka Fm.; Tb, Tirbes Mb.; Ul, Uluntui Fm.; Ud, Uda Fm.; Vor, Vorogovka Grp.; Mr, Marnya Fm.; Gol, Goloustnaya Fm.; Tb Mbr, Tirbess Mb. Logs: 1, by (Sovetov et al. 2007a), with additions from Karpinskaya, Karpinsky & Ustalov (pers. comm.); 2, modified after Vinogradov et al. (1994); 3, after Tyschenko (1980); 4, by Sovetov et al. (2007a) with addition from Shenfil (1991). (b) Map showing location of stratigraphic logs shown in (a). (c) Map showing distribution of late Cryogenian diamictite and inferred glacial extent for the Siberian Platform. Localities as follows: a, Kutukas river, Enisei Ridge, Chapa Group, Stolbovaya Formation (this paper); b, Chapa and Teya rivers, Enisei Ridge, Chapa Group, Pod’em Formation (this paper); c, Tagul River, Cis’Sayan region, Oselok Group, Marnya Fm. (Sovetov unpublished data); d, Birusa and Uda rivers, Cis’Sayan region, Oselok Group, Marnya Fm. (this paper); e, Irkut River, Irkut Cis’Sayan region, Olkha Fm. (Mordvin 1972); f, Buguldeika River, southwestern Cis’Baikal region, Baikal Group, Goloustnaya Fm. (this paper); g, Right Ulkan River, northwestern Cis’Baikal region, Baikal Group Goloustnaya Fm. (Korobeinikov & Semeikina 1978); h, Patom River, Patom Upland, Dal’nyaya Taiga Group, Dzhemkukan Fm. (this paper); i, Patom River, Patom Upland, Dal’nyaya Taiga Group, Dzhemkukan Fm.; j, Nichatka Lake, Aldan Shield, Nichatka Fm.; k, Siberian Platform, Markha River, Borehole 225–0, Conglomerate suite (Kraevsky & Shishkin 2000).

MARNYA FORMATION, OSELOK GROUP

Sovetov et al. 2007a). Continental molasse deposited during the Ediacaran (Late Vendian) collision belongs to the Greben regional stratigraphic stage. Glaciogenic deposits of the Marnya Fm., interpreted as tillites, and stratigraphic equivalents of the formation belong to the upper diamictite unit in the succession of two regional stratigraphic stages (Sovetov et al. 2007a). The lower unit was deposited during a period that can be correlated with the Early Cryogenian glacial in a setting of rifting, volcanism and subsidence of deep basins (Sovetov 1997). Deposition of the Marnya glacial deposits is thought to be coeval with the early stage of the Late Cryogenian (Early Vendian) foreland basin (Sovetov 2002b; Sovetov & Komlev 2005) and with formation of post-glacial shallow fringing seas in the southwestern part of the craton. The Ulyakha glaciogenic deposits can be equated to a Late Cryogenian glaciation (Halverson et al. 2005), for the following reasons: (i) higher stratigraphic position of the Ulyakha tillite relative to the Chivida diamictite (tillite) of the Yenisei Range (Sovetov 1997; Sovetov et al. 2007a) constrained by the coeval c. 700 Ma volcanism (Postnikov et al. 2005); (ii) occurrences of Vendian-Ediacaran Metazoan in the Ozerki Mb. (Sovetov & Komlev 2005; Sovetov 2007; Sovetov & Solovetskaya 2008); (iii) post-glacial regional transgression (Sovetov 2007) and correlation of related terrigenous deposits (Bolshaya Aisa Mb.) with the Nepa Fm. containing microfossil assemblages as in the Ediacaran Pertatatak Fm. in central Australia (Moczydłowska et al. 1993; Golubkova & Raevskaya 2007). Chemostratigraphy suggests that the glaciation on the Siberian craton recorded in the Marnya Fm. can be equated to the Marinoan epoch, which, according to the modern d13C time scale, is slightly older than the base of Doushantuo cap dolomite 635.2 + 0.6 Ma (Condon et al. 2005). The evidence that the Ulyakha glacial deposits are stratigraphically lower that the Gaskiers tillite is provided by: (i) correlation of the Aisa syncollisional molasse in the Sayan region with the Oskoba Fm. in the Baikit region (Fig. 28.6), with the latter marked by an intense d13C low (referred to as the Pre-Manykai negative anomaly by Pokrovsky 1996 and Sovetov et al. 2007b) that correlates with the Zhuya d13C negative anomaly (Pokrovsky et al. 2006) in the Patom region and with the Wonoka anomaly (Halverson et al. 2005) in South Australia; (ii) position of a d13C positive anomaly in the Peschernyi Member rocks (Uda Fm.), which is thought to be correlated with a similar anomaly below the Gaskiers glaciation; (iii) tectonothermal events in terranes on the craton SW related to their post-glacial accretion at 600– 550 Ma (Nozhkin et al. 2007), which is confirmed by the available 87 Sr/86Sr ratios in the Peschernyi Mb. limestone. Based on the widespread nature of diamictite and breccia at the base of the Ediacaran cover of the Siberian Platform, a thick ice sheet occupied most of the southwestern craton area (Fig. 28.6). Erosion of Neoproterozoic sediments was locally deeper than 1000 m; erosional valleys reached widths up to 50 km and radiated centrifugally to the west, SW and south, following the advance of the glaciers. At the same time, another glaciation centre, inferred from the palaeocurrent directions in an outwash fan of the Karapchatui Mb., was located in the extreme southwestern basement inliers in the Sayan region, and may record the complex structure of a single ice sheet. The continental ice sheet extended as far as remnant deep seas on the Patom passive margin, discharging gravity flows to produce the deep-water fan deposits of the Dzhemkukan Fm. (Chumakov et al. 2006). Funding for the research was provided by the Russian Foundation for Basic Research (grants nos 04-05-65299 and 08-05-00959), which was indispensable for regional sedimentological and stratigraphical studies, and by the Earth Science Department (ESD) of the Russian Academy of Sciences (Project 18 of the ESD Program ‘Biosphere origin and evolution’ and Project 10.1 of the RAS-ESD Program ‘Central Asian orogen: geodynamics and stages of continental crust formation’). I wish to thank D. Metelkin who performed the palaeomagnetic analysis of samples from the Marnya Fm. and my field companions, postgraduate students D. Komlev, M. Medvedev and A. Kulikova. This represents

327

a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) project #512.

References Anatol’eva, A. I. 1972. Pre-Mesozoic Redbed Deposits, Trudy Inst. Geol. Geophiz. Sibir. Divis. Akad. Nauk SSSR, Nauka, Novosibirsk, 190, 323 (in Russian). Berzin, N. A. 1967. Main Fault Zone in the East Sayan. Nauka, Moscow (in Russian). Bragin, S. S. 1984. On relationship of the Precambrian Karagassy and Oselok Groups in the Sayan region. In: Khomentovsky, V. V. (ed.) The Late Precambrian and the Early Palaeozoic Stratigraphy. Central Siberia. IGG, Novosibirsk, 133–147 (in Russian). Bragin, S. S. 1985. Late Precambrian Oselok Group in the Sayan region (stratigraphic division and correlation). In: Khomentovsky, V. V. (ed.) The Late Precambrian and the Early Palaeozoic of Siberia. The Vendian and the Riphean. IGG, Novosibirsk, 44 –57 (in Russian). Bragin, S. S. 1986. Stratigraphy of the Late Riphean Karagassy Group in the Sayan region: Some problems. In: Khomentovsky, V. V. & Shenfil, W. Yu. (eds) The Late Precambrian and the Early Palaeozoic of Siberia. Stratigraphy and Palaeontology. IGG, Novosibirsk, 32– 39 (in Russian). Burns, S. J., Haudenschild, U. & Matter, A. 1994. The strontium isotopic composition of carbonates from the late Precambrian (c. 560– 540 Ma) Hukf Group of Oman. Chemical Geology, 111, 269– 282. Butler, R. F. 1992. Palaeomagnetism: Magnetic Domains to Geologic Terrains. Blackwell Science, Oxford. Calver, C. R. 2000. Isotope stratigraphy of the Ediacaran (Neoproterozoic III) of the Adelaida Rift Complex, Australia, and the overprint of water column stratification. Precambrian Research, 100, 121–150. Chumakov, N. M. 1993. Middle Siberian Riphean glaciohorizon. Stratigraphy. Geological Correlation. Vol. 1. No 1, 21– 34 (in Russian). Chumakov, N. M., Pokrovsky, B. G. & Melezhik, V. A. 2006. Patom Complex geological history, Late Precambrian, Middle Siberia. Doklady Earth Sciences, 413, 379–383 (in Russian). Condon, D., Zhu, M., Bowring, S., Wang, W., Yang, A. & Jin, Y. 2005. U–Pb ages from the Neoproterozoic Doushantuo Fm., China. Science, 308, 95– 98. Donskaya, T. V., Salnikova, E. B. et al. 2002. Early Proterozoic postcollisional magmatism at the southern flank of the Siberian craton: new geochronological data and geodynamic implications. Doklady Earth Sciences, 382, 125–129 (in Russian). Dubin, P. V., Khomentovsky, V. V. & Yakshin, M. S. 1969. New data on Late Precambrian geology of the Sayan region. In: Sokolov, B. S. (ed.) Early Cambrian and Late Precambrian Stratigraphy of the Southern Siberian Craton. Nauka, Moscow, 86 –101 (in Russian). Egorova, O. P., Urumov, Yu. D., Volynets, Yu. N. & Vernoslova, Z. S. 1971. New data on geology and phosphate potential of Upper Riphean strata in the Iya-Tumanshet province of the Sayan region. In: Fainshtain, G. H. (ed.) Geology and Mineral Deposits of the Siberian Craton. Nedra, Moscow, 69– 84 (in Russian). Gladkochub, D. P., Donskaya, T. V., Mazukabzov, A. M., Ponomarchuk, V. A. & Stanevich, A. M. 2005. Neoproterozoic gabbro dolerites of the Biryusa terrane (southern Siberian craton) as possible indicators of the Rodinia dispersal, In: Sklyarov, E. V. (ed.) Geodynamic Evolution of the Lithosphere in the Central Asian Orogen (From Ocean to Continent). Workshop Proceedings. Book 1. IZK, Irkutsk, 59 –62 (in Russian). Gladkochub, D. P., Wingate, M. T. D., Pisarevsky, S. A., Donskaya, T. V., Mazukabzov, A. M., Ponomarchuk, V. A. & Stanevich, A. M. 2006. Mafic intrusions in southwestern Siberia and implications for a Neoproterozoic connection with Laurentia. Precambrian Research, 147, 260–278. Golubkova, E. Yu. & Raevskaya, E. G. 2007. Lower Vendian complex of microfossils from the interior part of the Siberian platform. In: Semikhatov, M. A. (ed.) The Rise and Fall of the Vendian (Ediacaran) Biota. Origin of the Modern Biosphere. Transactions of the

328

J. K. SOVETOV

International Conference of the IGSP 493. GEOS, Moscow, 39– 42 (in Russian). Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Toward a Neoproterozoic composite carbonisotope record. GSA Bulletin, 117, 1181–1207. Khabarov, E. M. & Ponomarchuk, V. A. 2005. Carbon isotopes in the Upper Riphean deposits of the Baikal Group in western Cisbaikalia: stratigraphic implications. Russian Geology and Geophysics, 46, 1019– 1027 (in Russian). Khomentovsky, A. S. 1950. Some data on Geology of Tumanshet saltbearing region (Eastern Siberia). Bulletin of Moscow Society of Nature Researchers, 15, 65 – 79 (in Russian). Khomentovsky, V. V., Shenfil’, V. Yu., Yakshin, M. S. & Butakov, E. P. 1972. Late Precambrian and Early Cambrian Reference Sections in the Siberian Platform. Nauka, Moscow (in Russian). Khomentovsky, V. V., Faizulin, M. Sh. & Karlova, G. A. 1998. The Vendian Nemakit-Daldyn stage in the southwestern Siberian craton. Doklady Earth Sciences, 362, 813– 815 (in Russian). Konstantinova, L. N. 2005. Katanga area. Section II. Vendian. In: Melnikov, N. V. (ed.) Stratigraphy of Oil and Gas Basins of Sibera. Riphean and Vendian of Siberian Platform and its Plaited Border. Academic Publishing House ‘Geo’, Novosibirsk, 187– 193 (in Russian). Korobeinikov, N. K. & Semeikina, L. K. 1978. The nature of interrelation of Baikal Group and Middle Proterozoic volcanic – plutonic complex in the north-western Cis’Baikal region. In: Khomentovsky, V. V. (ed.) News in Stratigraphy and Palaeontology of Late Precambrian in the Eastern and Northern Regions of Siberia. Institute Geology and Geophysics of SB Academy of Sciences of USSR, Novosibirsk, 134– 146 (in Russian). Kraevsky, B. G. & Shishkin, B. B. 2000. Study state and direction of stratigraphy investigation of Riphean deposits in the inner regions of Siberian Platform. In: Krasnov, V. I. (ed.) Stratigraphy and Palaeontology of Siberia. Siberian Scientific Research Institute of Geology, Geophysics and Mineral Deposits, Novosibirsk, 23 –31 (in Russian). Metelkin, D. V., Belonosov, I. V., Gladkochub, D. P., Donskaya, T. V., Mazukabzov, A. M. & Stanevich, A. M. 2005. Palaeomagnetic directions from Nersa intrusions of the Biryusa terrane, Siberian craton, as a reflection of tectonic events in the Neoproterozoic. Russian Geology and Geophysics, 46, 395–411 (in Russian). Metelkin, D. V., Vernikovsky, V. A. & Kazansky, A. Yu. 2007. The Neoproterozoic stage of the Rodinia evolution in the light of new palaeomagnetic data from the western margin of the Siberian craton. Russian Geology and Geophysics 48, 42– 59 (in Russian). Miall, A. D. 1985. Architectural-element analysis: A new method of facies analysis applied to fluvial deposits. Earth-Science Review, 22, 261– 308. Moczydłowska, M., Vidal, G. & Rudavskaya, V. A. 1993. Neoproterozoic (Vendian) phytoplankton from the Siberian platform, Yakutia. Palaeontology, 36, 495–521. Mordvin, A. P. 1972. The Khuzhir ore occurrence of gold-bearing conglomerates, In: Fainshtain, G. H. (ed.) Geology and Gold Potential of the Riphean and Vendian Strata in the Southern Periphery of the Irkutsk Amphitheatre. Vostochno-Sibirskoe Izdatelstvo, Irkutsk, 154– 189 (in Russian). Nozhkin, A. D., Turkina, O. M., Sovetov, Yu. K. & Travin, A. V. 2007. The Vendian accretionary event in the southwestern margin of the Siberian craton. Doklady Earth Sciences, 415A, 866–871 (in Russian). Pelechaty, S. M. 1998. Integrated chronostratigraphy of the Vendian System of Siberia: implications for a global stratigraphy. Journal of the Geological Society, London, 155, 957– 973. Pisarevsky, S. A. & Natapov, L. M. 2003. Siberia in Rodinia. Tectonophysics, 375, 221–245. Pokrovsky, B. G. 1996. Boundary between Proterozoic and Palaeozoic: isotope anomalies in the sedimentary sections of Siberian Platform and global environments change. Lithologia i Poleznye Iskopaemye, 4, 376–392 (in Russian). Pokrovsky, B. G. & Missarzhevsky, V. V. 1993. Isotope correlation of boundary Precambrian/Cambrian beds of the Siberian Platform. Doklady Earth Sciences, 329, 768– 771 (in Russian).

Pokrovsky, B. G., Melezhik, V. A. & Buyakaite, M. I. 2006. Carbon, oxygen, and strontium isotope compositions of Late Precambrian rocks in the Patom complex, Central Siberia. Paper 1. Results, isotopic stratigraphy, and dating problems. Lithologia i Poleznye Iskopaemye, 5, 1 –26 (in Russian). Postnikov, A. A., Nozhkin, A. D. et al. 2005. New data on the age of Neoproterozoic deposits of the Chingasan and Vorogovka Groups in the Yenisei Ridge. In: Sklyarov, E. V. (ed.) Geodynamic Evolution of the Lithosphere in the Central Asian Orogen (From Ocean to Continent). Workshop Proceedings, Book 2. IZK, Irkutsk, 71– 74 (in Russian). Sez’ko, A. I. 1988. Principal stages of continental crust forming in the Cis-Saya. In: Letnikov, F. A. (ed.) Evolution of Crust in Precambrian and Palaeozoic. Sayan-Baikal Mountain Region. Nauka Sib. Division, Novosibirsk, 7 –41 (in Russian). Shenfil, V. Yu. 1991. Late Precambrian Stratigraphy of the Siberian Platform. Nauka, Novosibirsk (in Russian). Shatsillo, A. V., Didenko, A. N. & Pavlov, V. E. 2005. Two competing palaeomagnetic directions in the Late Vendian: new data for the SW region of the Siberian Platform. Russian Journal of Earth Sciences, 7, 3 –24. Smethurst, M. A., Khramov, A. N. & Torsvik, T. H. 1998. The Neoproterozoic and Palaeozoic palaeomagnetic data for the Siberian Platform: From Rodinia to Pangea. Earth Science Research, 43, 1 –24. Sovetov, J. K. 1977. Upper Precambrian Sandstones in the Southwestern Siberian Platform. Nauka, Novosibirsk (in Russian). Sovetov, J. K. 1997. Late Riphean rifting and the Baikal geodynamic cycle of the Siberian Platform. In: Koroteev, V. A. (ed.) The Riphean of Northern Eurasia: Geology and General Problems of Stratigraphy. Institute of Geology and Geochemistry, Ekaterinburg, 223– 230 (in Russian). Sovetov, J. K. 2002a. Vendian glaciation on the Siberian craton. In: Letnikov, F. A. (ed.) Geology, Geochemistry, and Geophysics at the Turn of the 20th-to-21st Century. Proceedings, Russian Conference. IZK, Irkutsk, 122–124 (in Russian). Sovetov, J. K. 2002b. Vendian foreland basin of the Siberian cratonic margin: Palaeopangean accretionary phases. Russian Journal of Earth Sciences, 4, 363– 387. Sovetov, J. K. 2006. Vegetation along Vendian rivers, climate zonation, and palaeogeography of the Siberian craton in the Late Vendian. In: Sklyarov, E. V. (ed.) Geodynamic Evolution of the Lithosphere in the Central Asian Orogen (From Ocean to Continent). Proceedings of Workshop on ESD RAS Integration Research Programs, 2. IZK, Irkutsk, 143– 146 (in Russian). Sovetov, J. K. 2007. New occurrence of soft body Vendian–Ediacaran like Metazoa in south-west of the Siberian Platform, Oselok group, Biruysa Cis Sayan. In: Semikhatov, M. A. (ed.) The Rise and Fall of the Vendian (Ediacaran) Biota. Origin of the Modern Biosphere. Transactions of the International Conference of the IGSP 493. GEOS, Moscow, 33 –37 (in Russian). Sovetov, J. 2008. Marinoan glaciation in the Siberian craton: locality, erosional forms, deposits and constraints to age. CGC-04 Neoproterozoic ice ages: Quo vadis? – Part 2. International Geological Congress. Oslo. 6– 14 August. Abstracts. File://E:/33IGC/ 1343311.html. Sovetov, J. K. & Blagovidov, V. V. 2004. Reconstruction of sedimentary basin: an example of Vendian foredeep – foreland basin in southwest of Siberian Platform. In: Leonov, Yu. G. & Volozh, Yu. A. (eds) Sedimentary Basins: Methods of Research, Structure and Evolution. Transactions, 543. Scientific World, Moscow, 159– 210 (in Russian) Sovetov, J. K. & Komlev, D. A. 2005. Tillite at base of the Oselok Group in the Sayan region and the position of the lower boundary of the Vendian in the southwestern Siberian Platform. Stratigraphy and Geological Correlation, 13, 3– 34. Sovetov, J. K. & Blagovidov, V. V. 2006. Late Riphean sedimentation in Iya-Tumanshet aulacogene (south-west of the Siberian Platform): supersequences and correlation with Riphean stratotype in the Bashkiriyan anticlinorium. In: Maslov, A. V. (ed.) Lithological Aspects of Geology Stratified Environments. Materials of 7th Ural Lithological Meeting. Institute Geology and Geochemistry of Ural

MARNYA FORMATION, OSELOK GROUP

Division of Russian Academy of Sciences, Ekaterinburg, 248–250 (in Russian). Sovetov, J. K. & Donskaya, T. V. 2006. Basement in south-west of Siberian Platform by analysis of erratic stones in the Early Vendian tillite. In: Sklyarov, E. V. (ed.) The Geodynamic Evolution of the Lithosphere in Central Asian Orogen (From Ocean to Continent). Proceedings of Workshop on ESD RAS Integration Research Programs, 2. IZK, Irkutsk, 147– 150 (in Russian). Sovetov, J. & Solovetskay, L. 2008. Oldest Vendian-Ediacaran fossils in the Oselok Group: Contribution to Late Neoproterozoic (Ediacaran) age of sea transgression and origin of the Siberian platform cover. HPF-07 Rise and Fall of the Ediacaran (Vendian) Biota. International Geological Congress, Oslo, 6– 14 August. Abstracts. File://E:/33IGC/1343889.html. Sovetov, J. K., Kulikova, A. E. & Medvedev, M. N. 2007a. Sedimentary basins in the southwestern Siberian craton: Late Neoproterozoic –Early Cambrian rifting and collisional events. In: Linnemann, U., Nance, R. D., Kraft, P. & Zulauf, G. (eds) The Evolution of the Rheic Ocean; From Avalonian-Cadomian Active

329

Margin to Alleghenian– Variscan Collision. Geological Society of America, Special Paper 423, 549– 578. Sovetov, J. K., Blagovidov, V. V. & Talibova, A. G. 2007b. Carbon isotopes in Vendian carbonates in the southwestern Siberian craton: correlation with geodynamic and palaeoclimate events and palaeogeography. In: Halimov, E. M. (ed.) Proceedings of the Vinogradov XVIII Symposium on Isotope Geochemistry. Moscow, 245– 246 (in Russian). Turkina, O. M., Bibikova, E. V. & Nozhkin, A. D. 2003. Stages and geodynamic settings of Early Proterozoic granite formation on the southwestern margin of the Siberian craton. Doklady Earth Sciences, 389, 159–163 (in Russian). Tyschenko, L. F. 1980. Regional correlation of Moty Formation deposits in the Irkutsk amphitheatre. In: Karagodin, Yu. N. (ed.) The Problems of Lithostratigraphy. Nauka, Novosibirsk, 149–158 (in Russian). Vinogradov, V. I., Pokrovsky, B. G. et al. 1994. Isotopic geochemistry and age of Upper Precambrian deposits in the West Siberian Platform. Lithologia i Poleznye Iskopaemye, 4, 49 –76 (in Russian).

Chapter 29 The Tsagaan Oloom Formation, southwestern Mongolia FRANCIS A. MACDONALD Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA 02138, USA (e-mail: [email protected]) Abstract: The Tsagaan Oloom Formation (Fm.) in southwestern Mongolia contains two Neoproterozoic glacial deposits, with diamictite in the Maikhan Ul Member (Mb.) and in the Khongoryn Mb., which are separated by over 500 m of limestone. The Maikhan Ul Mb. ranges in thickness between 5 m and greater than 300 m, expanding in deeper-water sections towards the SW, where it is composed of two massive diamictites separated by over 100 m of sandstone, siltstone and shale. The basal 10 m of the overlying Tayshir Mb. of the Tsagaan Oloom Fm. consists of a fine-laminated, dark grey limestone. The Khongoryn Mb. is composed primarily of limestone clasts in a shale matrix, and is between 0 and 23 m thick. The overlying Ol Mb. contains sedimentary structures characteristic of basal Ediacaran cap carbonates including micropeloids, tubestone stromatolites, giant wave ripples and former aragonite crystal fans. U– Pb evaporation ages from zircons in the underlying Dzabkhan Volcanics constrain the Tsagaan Oloom Fm. to ,773 Ma, and tuffs within the Maikhan Ul and Tayshir members testify to the potential for additional geochronology. The Cryogenian organic-rich limestone of the Tayshir Mb., which lies between the two glacial deposits, is ideally suited for geochemical studies and has been the subject of several carbon, strontium and rare earth element investigations. Limited palaeomagnetic studies suggest a mid- to low-latitude position of the Dzabkhan platform during deposition of the glaciogenic strata, and additional studies are in progress.

Neoproterozoic diamictites are present on the Dzabkhan platform (also referred to as the Zavkhan basin) of southwestern Mongolia, in a .100 km NW – SE trending belt, with additional discontinuous exposures further north. The most complete exposures of the Tsagaan Oloom Fm. are between the Khasagty-Nuru ridge and the Dzabkhan River (Fig. 29.1). The geology of the Dzabkhan platform was first described by Bezzibetsev (1986), who divided the stratigraphy into three formations (the Dzabkhan, Tsagaan Oloom and Bayan Gol). Subsequent work focused on the early Cambrian palaeontology of the Bayan Gol Fm., with an eye for correlation with Siberia; the results of these studies were published entirely in Russian (for a list of these references see Brasier et al. 1996b). The first descriptions in English came in 1996 with the publication of a Geological Magazine issue dedicated to the NeoproterozoicCambrian stratigraphy of southwestern Mongolia (Brasier et al. 1996a). The studies reported therein were the product of two international field excursions, one in 1991 as part of the 21st Joint Soviet – Mongolian Palaeontological Expedition (Zhegallo & Zhuravelev 1991), and a second in 1993 sponsored by IGCP Project 303 and the Mongolian Academy of Sciences (Dorjnamjaa et al. 1993). The results of these excursions included the translation of geological maps and measured sections into English (Khomentovsky & Gibsher 1996), a reconnaissance chemostratigraphic characterization of the Tsagaan Oloom and Bayan Gol formations (Brasier et al. 1996b), and a detailed stratigraphic study of the Maikhan Ul diamictite at Tsagaan Gol (Lindsay et al. 1996). Recently, Macdonald et al. (2009) conducted detailed chemoand litho-stratigraphic studies on previously unstudied sections and discovered an additional diamictite higher in the succession. This work supported the earlier conclusion of Brasier et al. (1996b) that the Maikhan Ul member is early Cryogenian in age and established that the Khongoryn diamictite is an end Cryogenian glacial deposit. With the new C-isotope chemostratigraphic correlations and the documentation of a low-angle unconformity, a .40 million year hiatus was identified within the Tsagaan Oloom Formation, above the Khongoryn diamictite but below the phosphorite horizon (Macdonald et al. 2009). At Tsagaan Gol, the Maikhan Ul Mb. contains two diamictites separated by over 100 m of sandstone, siltstone and shale (Lindsay et al. 1996). This creates a bit of confusion in the literature, particularly with the discovery of an additional diamictite higher in the Tsagaan Oloom Fm., because the two diamictites within the Maikhan Ul Mb. have also been referred to as the upper and lower diamictites of the Tsagaan Oloom Fm.

(Khomentovsky & Gibsher 1996; Lindsay et al. 1996). Macdonald et al. (2009) grouped the lower two diamictites and the intervening clastic units together in the Maikhan Ul Mb., while referring to the diamictite c. 500 m higher in the sequence as the Khongoryn member. Levashova et al. (2010) informally referred to the Maikhan Ul diamictite as the Tayshir Fm. Here, we do not follow this nomenclature because it creates unnecessary confusion, particularly as the overlying carbonates have been previously called the Tayshir Mb. (Macdonald et al. 2009).

Structural framework The Dzabkhan terrane (also referred to as the Baydaric microcontinent when grouped with the Baidrag terrane, Fig. 29.1a) is a composite Precambrian terrane, hosting a heterogeneous Archaean and Proterozoic crystalline basement intruded by c. 805 –770 Ma continental arc volcanism (Badarch et al. 2002; Zhao et al. 2006). Based on similarities in the Neoproterozoic stratigraphy, radiometric ages in the underlying basement (Badarch et al. 1998), and the continuity of aeromagnetic anomalies associated with the fringing Neoproterozoic ophiolites (Buchan et al. 2002), the southwestern margin of the Dzabkhan basin can be traced to the western margin of the Khubsugul basin along the Tuva-Mongolia border (Fig. 29.1a). The tectonic events that transformed the southwestern and western margins of the Dzabkhan and Khubsugul terranes from continental arcs to thermally subsiding passive margins remain unclear. On the southern margin of the Dzabkhan terrane, on the south side of the Khasagty-Nuru ridge, the Dzabkhan and the Tsagaan Oloom Formations are separated by as much as 2 km of canabalizing, rift-related sediments with inter-fingering basalt that are referred to as the Shargyngol complex (Ruzhentsev & Burashnikov 1996). Facies patterns and the orientation of crossbeds in the Tsagaan Oloom Fm. indicate deepening to the SW (Macdonald et al. 2009). In the latest Ediacaran to early Cambrian, the rifted passive margin began to subside again after a depositional hiatus of .40 Ma. It has been proposed that this accommodation space was created by flexure with the arrival of the Khantayshir-Dariv arc (Macdonald et al. 2009). With the early Cambrian arc – continent collision, the Neoproterozoic stratigraphy was shortened and repeated in thrust blocks with a basal detachment beneath the Dzabkhan Fm. The deformation is largely brittle and thin-skinned, without involvement of the basement. Early Palaeozoic granites and narrow NW-trending grabens

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 331– 337. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.29

332

F. A. MACDONALD

Fig. 29.1. (a) Tectonic map of western Mongolia modified from Badarch et al. (2002) and Windley et al. (2007). Teeth on faults indicate the inferred dip of subduction zones. Key: K-D Arc, Khantayshir-Dariv Arc; AZ, accretionary zone including the arc, metamorphic rocks and ophiolitic assemblages. (b) Geological map of the Tayshir region, highlighting the members of the Tsagaan Oloom Fm. and the position of measured sections.

cut the Cambrian NNE-vergent structures. On the outcrop scale, the diamictite and carbonate rocks of the Tsagaan Oloom Fm. are little deformed with no apparent strain. Sedimentary structures are typically preserved in limestone, but are often obfuscated by recrystallization in dolomite.

and ribbonite that include nodular black chert and bed parallel, meandering ichnogenera (Goldring & Jensen 1996). Above the Zunne Arts Mb. of the Tsagaan Oloom Fm., the early Cambrian Bayan Gol Fm. is composed of c. 1000 m of mixed carbonate and siltstone with a rich diversity of ichnogenera, small shelly fossils and calcimicrobial patch reefs (Kruse et al. 1996).

Stratigraphy Glaciogenic deposits and associated strata The stratigraphy of the Dzabkhan basin (Fig. 29.2) begins with .2 km of silicic to intermediate volcanic rocks of the Dzabkhan Fm. On the south side of the Khasagty-Nuru ridge, the Dzabkhan Fm. is succeeded by as much as 2 km of sandstone turbidites and conglomerate that are referred to as the Shargyngol complex (Ruzhentsev & Burashnikov 1996). North of the Khasagty-Nuru ridge, the Shargyngol complex is ,100 m thick and often absent, with the Maikhan Ul Mb. of the Tsagaan Oloom Fm. nonconformably overlying the Dzabkhan Fm. The Maikhan Ul Mb. is composed of diamictite, sandstone and shale, and varies in thickness from 5 m to .275 m. The Maikhan Ul Mb. is overlain with a knifesharp contact by the Tayshir Mb., which consists of c. 570 m of limestone that included three super-sequences (Macdonald et al. 2009). The Tayshir Mb. is succeeded by the Khongoryn diamictite, which is composed of limestone cobble to boulder lonestones in a shale matrix and varies from 0 to 23 m in thickness. The Khongoryn diamictite is in turn overlain by micropeloidal dolostones of the Ol Mb. (Macdonald et al. 2009). In the subsequent transgression, formerly aragonite crystal fans are developed at the dolostone– limestone transition. Above the post-glacial transgression, the Ol Mb. shallows upwards from grey limestone rhythmite into c. 10 m of limestone grainstone. The overlying Ulaan Bulagyn Mb. is up to 500 m thick and is composed primarily of massive weathering dolomite; however, in more distal sections, the Ulaan Bulagyn Mb. thins to less than 100 m and is composed largely of limestone. The Zunne Arts Mb. begins with distinct pink-coloured columnar stromatolites (Boxonia grumulosa) that overly a karstic surface with metre-scale relief (Macdonald et al. 2009). The Boxonia bioherms are overlain by 10 –20 m of violet and green shale that are variably phosphatized and interbedded with lenses of dolomite and microcrystalline to nodular phosphorite. This phosphatic shale is overlain by more than 100 m of blue limestone rhythmite

The Maikhan Ul Member The Maikhan Ul Mb. progressively thickens to the SW (Fig. 29.3), but also displays considerable variability on individual thrust blocks. For example, at the easternmost exposures on the Tayshir Block (F718), the Maikhan Ul Mb. is only 6.7 m thick and is composed predominantly of a massive cobble –boulder clast diamictite, whereas just 1 km to the west (F713) it thickens to 81.6 m with multiple diamictite units separated by 57 m of massive, fine to coarse-grained sandstone. These sandstone bodies are composed of graded centimetre- to metre-thick beds and contain no evidence of tidal influence (Fig. 29.3). Further south, in more distal sections, the Maikhan Ul Mb. continues to thicken. On the Khongoryn Block (F701), the Maikhan Ul Mb. fills palaeo-canyons and varies in thickness between 160 and 283 m. One palaeo-canyon, directly west of F701, is c. 125 m deep and 0.6 km wide. This palaeo-canyon has an erosive base that is mantled with a volcanic-clast, cobble conglomerate and is filled with stratified diamictite units with dropstones and striated clasts (Macdonald et al. 2009), thin-bedded sandstone beds and a c. 0.5-m-thick carbonate bed. Two massive to bedded diamictite units lie above the canyon fill, separated by 62 m of siltstone and sandstone with rare cobble lonestones and two additional c. 0.5-m-thick carbonate beds. At Tsagaan Gol, where the member measures 304 m, again, two diamictite units are separated by a thick sequence of flatbedded shale, siltstone and sandstone (Lindsay et al. 1996). Cobble lonestones are present in both the basal and upper metre of this clastic succession, between the two massive diamictites. Khomentovsky & Gibsher (1996) also reported a measured section from Urtor Tsakhir Mountain, c. 120 km west of Tayshir,

TSAGAAN OLOOM FORMATION

BG

Bayan Gol Fm. c. 1 km of clastic & carbonate rocks

1400

Zunne Arts Member

-8

-4

4

0

8

C P

600

400

Tsagaan Oloom Formation Tayshir Member Ulaan Bulagyn Member K Ol

800

OO OO

The Khongoryn Mb. is thickest one gully east of Tsagaan Gol (F723); however, like the Maikhan Ul Mb., there are significant facies changes both from north to south and from east to west (Fig. 29.4). East of Tsagaan Gol, the diamictite is 23 m thick and composed of pebble- to boulder-sized clasts of blue-grey limestone from the underlying Tayshir member in a dark grey shale matrix that becomes more marly and lighter coloured up-section. Striated clasts and limestone clasts with soft sedimentary deformation are also present. Just 6 km west, near Tsagaan Gol, the diamictite is nearly absent and only 2 m of recessive shale are preserved. The Khongoryn Mb. is also well developed on the Khongoryn block, south of Bayan Gol (F708), where it consists of 14.7 m of sub-rounded limestone pebbles, cobbles and boulders in a grey shale matrix. Both laterally and up-section, clasts are irregularly distributed, varying from clast-poor facies to boulder nests. To the NE of the Khongoryn block, the Khongoryn Mb. is either thin or absent.

OO

C

Maieberg anomaly

OO

Associated carbonate rocks

Tayshir anomaly

C

OO

OO OO

C 200

Rasthof anomaly

DV

MU

-8

0 (m)

where the Maikhan Ul Mb. is even thicker but still preserves this general stratigraphic pattern of two diamictite units separated by sandstone and siltstone. In this area, mudcracks are also well developed near the top of these intervening clastic units. In both the upper and lower diamictite units of the Maikhan Ul Mb., the most common lithology comprises a matrix-dominated diamictite with shale and sandstone encasing sub-rounded cobble derived from the underlying Dzabkhan Formation; granite, metamorphic and carbonate clasts of unknown origin are also present. Also, near the base of the Maikhan Ul Mb. at Tsagaan Gol, clasts of deformed soft sediment have been reported (Lindsay et al. 1996). Clast size varies from grit to blocks .2 m across.

The Khongoryn Member

1200

1000

333

-4

0

4

8

Maikhan Ul diamictite vv vv vv vv

Dzabkhan Formation c. 2 km of voclanic & volcanoclastic rocks

limestone C chert dolomite P phosphorite grainstone & OO giant ooids microbialaminite stromatolite ribbonite trace fossils rhythmite/shale debris flow diamictite flooding siltstone/ exposure sandstone

The basal 10 m of the Tayshir Mb., which overlies the Maikhan Ul Mb., is composed of a dark grey, millimetre-laminated limestone. Overall, the Tayshir Mb. consists of ,650 m of limestone that record three regionally extensive sequences. The base of the first sequence is defined by a c. 10-m-thick, dark grey (weathering to tan), millimetre laminated limestone that is succeeded by c. 100 m of limestone marl and rhythmite, shoaling up-section to c. 20 m of grainstone. The second sequence begins with c. 10 m of limestone marl and rhythmite followed by c. 200 m of massively bedded, blue grainstone and microbialaminite. The third sequence begins with c. 50 m of limestone rhythmite and debris flows with numerous black chert beds and nodules, and then shallows up-section to c. 210 m of dark, fetid limestone microbialaminite and minor grainstone with giant ooids (.0.5 cm diameter). The Ol Mb., which overlies the Khongoryn diamictite, begins with 7–40 m of buff to pink coloured, largely recrystallized, micropeloidal dolostone. Low-angle cross-stratification (Aitken 1991), tubestone stromatolites (Corsetti & Grotzinger 2005), and giant wave ripples (Allen & Hoffman 2005) are also present in the Ol Mb. dolomite (Fig. 29.4). The Ol Mb. transgresses upwards into limestone ribbonite and then rhythmite with

Fig. 29.2. Composite carbon chemo- and lithostratigraphy of the Dzabkhan basin, modified from Macdonald et al. (2009). K, Khongoryn diamictite. The Rasthof anomaly was first well-documented in Namibia (Yoshioka et al. 2003) and, like the Maieberg anomaly (Halverson et al. 2005), has now been reported globally. The Tayshir anomaly was first documented in Mongolia (Macdonald et al. 2009) and a potentially correlative mid-Cryogenian anomaly is also present in the Bonahaven Fm. of the Dalradian Supergroup in Scotland and Ireland (Prave et al. 2009).

334

F. A. MACDONALD

Tayshir Block 1 km

E

W

Khongoryn Block, F701

Tsagaan Block,F724

250

200

150

F718

F713 volcanic massive diamictite

matrix composition

100

sandstone siltstone shale limestone

50

0

metres

clast size boulder cobble gravel clast composition granite clastic carbonate volcanic

Fig. 29.3. Stratigraphy of the Maikhan Ul Mb. Locations of sections are shown in Figure 29.1b. Section F724 is modified from Lindsay et al. (1996). Symbols used that are not in the legend are in Figure 29.2.

c. 5-cm-tall former aragonite crystal fans present at the limestone – dolostone transition. Crystal fans are present both as individual blades growing upwards into the sediment, and as crystal fan shrubs that are over 10 cm across.

Boundary relations with overlying and underlying non-glacial units The Maikhan Ul Mb. rests with an erosive base on the Shargyngol suite and the Dzabkhan Formation (Khomentovsky & Gibsher 1996), and fills palaeo-topography with conglomerates lining palaeo-valleys (Fig. 29.3). According to Lindsay et al. (1996), at Tsagaan Gol, soft-sedimentary deformation is present in sandstone

below the Maikhan Ul Mb., indicating only a limited hiatus. However, it is not clear if this sandstone is part of the Shargyngol suite or should be included within the Maikhan Ul Mb. South of Tsagaan Gol, the clastic units between the Dzabkhan Fm. and the lower Maikhan Ul diamictite unit thicken to over 100 m and lack any evidence of glacial influence on sedimentation. Conversely, to the east and north of Tayshir, both the Maikhan Ul Mb. and the Dzabkhan Fm. thin, with the diamictites of the Maikhan Ul resting on an erosional contact with the Dzabkhan Fm. or the basement rock. Contact between the Maikhan Ul Mb. and the overlying Tayshir Mb. is very sharp. The Tayshir Mb. rests conformably on a laterally persistant, c. 10-cm-thick layer of red clay that marks the top of the Maikhan Ul Mb. The Khongoryn diamictite typically lies above blue-grey, giant ooid grainstones of the lower limestone of the Tsaagan Oloom Fm.; however, on the Khongoryn Block (F708), there is an additional 7.7 m of black shale and rhythmite preserved above the ooids. The erosion of this shale likely provides the detrital matrix for the Khongoryn diamictite. The Khongoryn diamictite is overlain with a sharp yet conformable contact by dolostone and limestone of the Ol Mb.

Chemostratigraphy Strontium isotope values rise from 0.7067 to 0.7073 in the limestones of the Tayshir Mb. In the Ulaan Bulagyn Mb. 87Sr/86Sr values rise from 0.7073 to 0.7077, and then in the Zunne Arts Mb. from 0.7078 to over 0.7080 (Brasier et al. 1996b; Shields et al. 2002). Carbonate d13C values in the dark-grey laminated limestone above the Maikhan Ul Fm. are moderately negative with values increasing upwards through the overlying pink marls to þ8‰ (Fig. 29.2). Values plummet abruptly at the flooding surface in the middle of the Tayshir Mb., reaching a low of – 7.5‰. Macdonald et al. (2009) refer to this sudden drop in d13C values as the Tayshir anomaly. From this nadir, d13C values increase smoothly to þ9‰ for the upper Tayshir Mb., with values reported as high as þ11‰ (Brasier et al. 1996b). Shields et al. (2002) also measured d13C of organic matter in the Tayshir Mb. of the Tsagaan Oloom Fm. and found that trends roughly followed those exhibited by the d13C in carbonate. Overlying the upper diamictite, d13C values in the Ol Mb. begin around –1‰ and follow a sigmoidal profile (Fig. 29.2). Values return to c. –1‰ at the top of the dolostone, and then decrease again at the limestone – dolomite transition, reaching a nadir of –6‰. Above the Ol Mb., d13C values oscillate around þ3‰ for most of the Ulaan Bulagyn Mb., returning to 0‰ below the sub-Zunne Arts Mb. karstic surface. Sheilds et al. (1997, 2002) reported a Ce anomaly in the Tayshir Mb. from samples collected at Tsagaan Gol. In this section, the recessive strata bearing the C-isotope anomaly are not exposed, and thus, they did not document the transgressive sequence or the negative C-isotope values.

Palaeolatitude and palaeogeography Recent palaeomagnetic studies on the 805–770 Ma Dzabkhan Fm. indicate that the Dzabkhan terrane was located at a latitude of 478 þ 168/–128 (Levashova et al. 2010). From palaeomagnetic studies on peri-Siberian terranes, including the early Cambrian Salaany Gol Fm. on the Dzabkhan terrane, Kravchinsky et al. (2001) concluded that the Tuva-Mongolia belt was at low latitude, adjacent to Siberia throughout the Ediacaran and Cambrian. However, this study lacked a robust confidence test (i.e. only a reversal test with few samples and low resolution). Moreover, an earlier study on the Salaany Gol Fm. gave entirely different

TSAGAAN OLOOM FORMATION

Tayshir Block E 50

25

18 km

3 km

10 km

W

Khongoryn Block 18 km E W

335

Tsagaan Block E

15 km

6 km

C C C

C C C

W

C C

C

(metres)

C

0

-25

F710

F716

F715

micropeloidal dolomite

F706

F725

giant wave ripples

F708

F726

tubestone stromatolites

results (Evans et al. 1996), but it was also compromised by uncertainty in the relative ages of the folds used in the fold test and possible magnetic overprints. Further palaeomagnetic studies on the Dzabkhan terrane are necessary, and are in progress (Gregory et al. 2007). Nonetheless, as non-skeletal carbonate is preferentially produced in the warmest parts of the surface ocean (Broecker & Peng 1982), and the Tsaagan Oloom Fm. is dominated by shallow-water carbonates, it is likely that the Dzabkhan terrane was situated at low latitudes (less than 308) throughout the Cryogenian and Ediacaran. Along with other peri-Siberian terranes, it has been suggested that the Dzabkhan terrane occupied a Precambrian position between Siberia and Laurentia (Gladkochub et al. 2006), and rifted away from Siberia in the late Neoproterozoic (Sengor & Natal’in 1996; Kuzmichev et al. 2001; Kuzmichev et al. 2005). Sengor & Natal’in (1996) further posit that throughout the late Neoproterozoic and early Palaeozoic, the Dzabkhan terrane was attached to the Central Mongolian Block, which along with other terranes, stretched to the present day Sea of Okhotsk. Both the Tuva-Mongolia (including the Dzabkhan terrane) and Central Mongolian Blocks host Cambrian trilobites endemic to Siberia (Astashkin et al. 1995) and Silurian brachiopods characteristic of the peri-Siberian realm (Hou & Boucot 1990). Alternatively, citing similarities in SHRIMP ages on zircons, Zhao et al. (2006) and Demoux et al. (2009) have suggested that the Baydrag and Dzabkhan terranes originated from the northern margin of Gondwana. This reconstruction is supported by the palaeomagnetic results of Levashova et al. (2010), which point to Neoproterozoic connections with India, South China, Tarim or Australia.

Geochronological constraints Although the diamictites of the Dzabkhan basin have not been directly dated radiometrically, maximum age constraints on the glacial deposits are provided by zircons from rhyolites within the Dzabkhan Formation of 777 + 6 Ma (Zhao et al. 2006), 803.4 + 8.0 and 773.5 + 3.6 Ma (U– Pb laser evaporation, Levashova et al. 2010).

Discussion A glacial origin of the Maikhan Ul diamictite units is indicated by the presence of faceted and striated clasts, and bullet-shaped

F723 crystal fans

F724

Fig. 29.4. Stratigraphy of the Khongoryn diamictite and overlying Ol Mb. Positions of measured sections are in Figure 29.1b. Symbols used that are not in the legend are in Figures 29.2 and 29.3.

dropstones that penetrate laminated beds. At Tsagaan Gol, cobble dropstones are also present in both the basal and upper metre of the clastic succession, between the two massive diamictite units. Moreover, in more proximal settings, such as on the Khongoryn and Tayshir blocks, rare lonestones are present within the sandstone beds. These observations indicate that the deposition of the clastic units was influenced, at least in part, by glaciation. Macdonald (2009) inferred a pro-glacial environment, including emergent conditions and proglacial lakes, for both the diamictite and the clastic units of the Maikhan Ul Mb. from the presence of mud cracks and 0.5-m-thick carbonate beds. A pro-glacial environment is further supported by high lateral facies variability. Within this context, the intervening clastic units can be interpreted as a step-back of the ice-line, and the upper diamictite as an ice-advance. The rise in 87Sr/86Sr from 0.7067 to 0.7073 in the limestone of the Tayshir Mb. is mirrored in the Rasthof Fm. in Namibia and the Keele Fm. in NW Canada, suggesting that the underlying Maikhan Ul diamictites are early Cryogenian glacial deposits (Halverson et al. 2007). The black laminated cap carbonate above the Maikhan Ul diamictites also contains a modest negative C-isotope anomaly similar to the Rasthof Fm. (Yoshioka et al. 2003); the extremely enriched values of the Tayshir Mb. are also consistent with a Cryogenian age (Hoffman & Schrag 2002; Halverson et al. 2005). The Tayshir anomaly (Macdonald et al. 2009) can be correlated to the moderately negative 13dC values obtained from the exposure-surface riddled Gruis Fm. of northern Namibia (Halverson et al. 2005) and the Cryogenian Bonahaven Dolomite of the British-Irish Caledonides (McCay et al. 2006), or to the Trezona anomaly in Australia (McKirdy et al. 2001) and Namibia (Halverson et al. 2005). A glacial origin of the Khongoryn diamictite is indicated by the presence of striated clasts and dropstones that penetrate laminated beds. The Khongoryn diamictite is thin or absent on the most proximal sections to the NE of the map area (Fig. 29.1). In more distal sections to the SW, the diamictite is composed of cobble to boulder clasts of the underlying limestone within a weakly bedded shale to marl matrix. This shale matrix was likely derived via erosion of the shale unit in the upper portion of the Tayshir Mb., which is only present on the Khongoryn and Tsagaan blocks. The lack of stratigraphic architecture within the deposit, the irregular distribution of ice-rafted debris, such as boulder nests, and the conformable overlying contact with the Ol Mb. indicate that this deposit formed as a single rainout during the terminal deglaciation.

336

F. A. MACDONALD

The overlying basal dolostone of the Ol Mb. is composed of fine-laminated micropeloids and contains tubestone stromatolites, giant wave ripples and pseudomorphosed crystal fans. These peculiar sedimentary structures, their specific order, and the distinct, sigmoidal C-isotope profile are characteristic of basal Ediacaran cap carbonates globally (Hoffman et al. 2007). This suggests that the underlying Khongoryn diamictite is an end-Cryogenian glacial deposit (Macdonald et al. 2009), with the termination bracketed elsewhere by U –Pb ages of 635.51 + 0.54 Ma and 635.23 + 0.57 Ma (Condon et al. 2005). The phosphorites in the Zunne Arts Mb. rest above a low-angle unconformity. Ediacaran chemostratigraphic correlations indicate that this surface represents a .40 Ma depositional hiatus, and as such is unrelated to the glacial deposits in the Khongoryn member (Macdonald et al. 2009). Further sedimentological studies on the Maikhan Ul Mb. are needed to understand the significance of the interbedded sandstone and siltstone, and to determine if the Maikhan Ul diamictites represent a single episode of deglaciation or multiple ice advances and retreats. Stratigraphic studies are also needed to better understand the basin dynamics of the Neoproterozoic– Cambrian margins of the Dzabkhan terrane, and the relationships with other Mongolian terranes. The palaeogeography of the PeriSiberian terranes also remains speculative. It is clear, however, that island arcs surrounded the Dzabkhan terrane for much of the Neoproterozoic and Cambrian, and therefore there is excellent potential for U – Pb zircon geochronology studies in the Dzabkhan basin. Furthermore, the low-grade and high organic content of the limestone in the Tayshir Mb. is ideally suited for multi-proxy studies to better constrain the geochemical evolution of Cryogenian oceans. I thank field assistants B. Boldoo, E. Oyun, T. Adiya, J. Otgonhuu and U. Bold. I also thank Bayassa and A. Amaglaan for help with logistics. I thank D. Jones and P. Hoffman for comments and help in the field. I am grateful to P. Hoffman and NSF grant EAR-0417422 (to PFH) for support of this work. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) project #512.

References Aitken, J. D. 1991. The Ice Brook Formation and Post-Rapitan, Late Proterozoic glaciation, Mackenzie Mountains, Northwest Territories. Geological Survey of Canada Bulletin, 404, 1 –43. Allen, P. A. & Hoffman, P. F. 2005. Extreme winds and waves in the aftermath of a Neoproterozoic glaciation. Nature, 433, 123– 127. Astashkin, V. A., Pegel, T. V. et al. 1995. The Cambrian System of the Foldbelts of Russia and Mongolia. International Union of Geological Sciences, London. Badarch, G., Byamba, J. et al. 1998. Geological Map of Mongolia. Mineral Resources Authority of Mongolia, Ulaan Baator. Badarch, G., Cunningham, W. D. & Windley, B. 2002. A new terrane subdivision for Mongolia: implications for the Phanerozoic crustal growth of Central Asia. Journal of Asian Earth Sciences, 21, 87 –110. Bezzubetsev, V. V. 1986. On the Precambrian –Cambrian stratigraphy of the Dzabkhan River Basin. Materials on the Geology of MPR, Gostopotekhizdat, 1963, 29– 42. Brasier, M. D., Dorjnamjaa, D. & Lindsay, J. F. 1996a. The Neoproterozoic to early Cambrian in southwest Mongolia: an introduction. Geological Magazine, 133, 365–369. Brasier, M. D., Shields, G., Kuleshov, V. N. & Zhegallo, E. A. 1996b. Integrated chemo- and biostratigraphic calibration of early animal evolution: Neoproterozoic – early Cambrian of southwest Mongolia. Geological Magazine, 133, 445– 485. Broecker, W. S. & Peng, T. H. 1982. Tracers in the Sea. LDEO Press, Lamont-Doherty Earth Observatory, Palisades, NY. Buchan, C., Pfander, J. et al. 2002. Timing of accretion and collisional deformation in the Central Asian Orogenic Belt: implications

of granite geochronology in the Bayankhongor Ophiolite Zone. Chemical Geology, 192, 23 –45. Condon, D. J., Zhu, M., Bowring, S. A., Wang, W., Yang, A. & Jin, Y. 2005. U– Pb ages from the Neoproterozoic Doushanto Formation, China. Science, 308, 95– 98. Corsetti, F. A. & Grotzinger, J. P. 2005. Origin and significance of tube structures in Neoproterozoic post-glacial cap carbonates: example from Noonday Dolomite, Death Valley, United States. Palaios, 20, 348– 363. Demoux, A., Kroener, A., Badarch, G., Jian, P., Tomurhuu, D. & Wingate, M. T. D. 2009. Zircon ages from the Baydrag block and the Bayankhongor ophiolite zone: time constraints on late Neoproterozoic to Cambrian subduction- and accretion-related magmatism in central Mongolia. Journal of Geology, 117, 377–397. Dorjnamjaa, D., Bat-Ireedui, Y. A., Dashdavaa, Z. & Soelmaa, D. 1993. Precambrian–Cambrian Geology of the Dzavkhan Zone. Earth Sciences Department, Oxford. Evans, D. A. D., Zhuravlev, A. Y., Budney, C. J. & Kirschvink, J. L. 1996. Palaeomagnetism of the Bayan Gol Formation, western Mongolia. Geological Magazine, 133, 487–496. Gladkochub, D. P., Wingate, M. T. D., Pisarevsky, S. A., Donskaya, T. V., Mazukabzov, A. M., Ponomarchuk, V. A. & Stanevich, A. M. 2006. Mafic intrusions in southwestern Siberia and implications for a Neoproterozoic connection with Laurentia. Precambrian Research, 147, 260– 278. Goldring, R. & Jensen, S. 1996. Trace fossils and biofabrics at the Precambrian-Cambrian boundary interval in western Mongolia. Geological Magazine, 133, 403– 415. Gregory, L. C., Meert, J. G., Levashova, N. & Gibsher, A. S. 2007. Paleomagnetic and geochronologic data from Central Asia: inferences for Early Paleozoic tectonic evolution and timing of worldwide glacial events. American Geophysical Union, Fall Meeting, GP43C –1487. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Toward a Neoproterozoic composite carbon-isotope record. Geological Society of America Bulletin, 117, 1181–1207. Halverson, G. P., Duda´s, F. O., Maloof, A. C. & Bowring, S. A. 2007. Evolution of the 87Sr/86Sr composition of Neoproterozoic Seawater. Palaeogeography, Palaeoclimatology, Palaeoecology, 256, 103–129. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis; testing the limits of global change. Terra Nova, 14, 129–155. Hoffman, P. F., Halverson, G. P., Domack, E. W., Husson, J. M., Higgins, J. A. & Schrag, D. P. 2007. Are basal Ediacaran (635 Ma) post-glacial ‘cap dolostones’ diachronous? Earth and Planetary Science Letters, 258, 114– 131. Hou, H. F. & Boucot, A. J. 1990. The Balkhash-Mongolia-Okhotsk Region of the Old World Realm. Geological Society, London, Memoir, 12, 297– 303. Khomentovsky, V. V. & Gibsher, A. S. 1996. The Neoproterozoic – Lower Cambrian in northern Govi-Altai, western Mongolia: regional setting, lithostratigraphy and biostratigraphy. Geological Magazine, 133, 371– 390. Kravchinsky, V. A., Konstantinov, K. M. & Cogne, J.-P. 2001. Palaeomagnetic study of Vendian and Early Cambrian of South Siberia and Central Mongolia: was the Siberian platform assembled at this time? Precambrian Research, 110, 61 –92. Kruse, P. D., Gandin, A., Debrenne, F. & Wood, R. 1996. Early Cambrian bioconstructions in the Zavkhan Basin of western Mongolia. Geological Magazine, 133, 429– 444. Kuzmichev, A., Bibikova, E. V. & Zhuravlev, D. Z. 2001. Neoproterozoic (800 Ma) orogeny in the Tuva-Mongolia Massif (Siberia): island arc-continent collision at the northeast Rodinia margin. Precambrian Research, 110, 109– 126. Kuzmichev, A., Kroener, A., Hegner, E., Dunyi, L. & Yusheng, W. 2005. The Shishkhid ophiolite, nothern Mongolia: a key to the reconstruction of a Neoproterozoic island-arc system in central Asia. Precambrian Research, 138, 125–150. Levashova, N. M., Kalugin, V. M., Gibsher, A. S., Yff, J., Ryabinin, A. B., Meert, J. & Malone, S. J. 2010. The origin of the Baydaric microcontinent, Mongolia: constraints from paleomagnetism and geochronology. Tectonophysics, 485, 306–320.

TSAGAAN OLOOM FORMATION

Lindsay, J. F., Brasier, M., Shields, G., Khomentovsky, V. V. & Bat-Ireedui, Y. A. 1996. Glacial facies associations in a Neoproterozoic back-arc setting, Zavkhan Basin, western Mongolia. Geological Magazine, 133, 391–402. Macdonald, F. A. 2009. Neoproterozoic stratigraphy of Alaska and Mongolia. PhD, Harvard University. Macdonald, F. A., Jones, D. S. & Schrag, D. P. 2009. Stratigraphic and tectonic implications of a new glacial diamictite–cap carbonate couplet in southwestern Mongolia. Geology, 37, 123–126. McCay, G. A., Prave, A. R., Alsop, G. I. & Fallick, A. E. 2006. Glacial trinity: Neoproterozoic Earth history within the British-Irish Caledonides. Geology, 34, 909– 912. McKirdy, D. M., Burgess, J. M. et al. 2001. A chemostratigraphic overview of the late Cryogenian interglacial sequence in the Adelaide fold –thrust belt, South Australia. Precambrian Research, 106, 149– 186. Prave, A. R., Fallick, A. E., Thomas, C. W. & Graham, C. M. 2009. A composite C-isotope profile for the Neoproterozoic Dalradian Supergroup of Scotland and Ireland. Journal of the Geological Society of London, 166, 1 –13. Ruzhentsev, S. V. & Burashnikov, V. V. 1996. Tectonics of the western Mongolian Salairides. Geotectonics, 29, 379– 394. Sengor, A. C. & Natal’in, B. A. 1996. Paleotectonics of Asia: fragments of synthesis. In: Yin, A. & Harrison, M. (eds) The Tectonic Evolution of Asia. Cambridge University Press, Cambridge, 486– 640.

337

Shields, G., Stille, P., Brasier, M. & Atudorei, N.-V. 1997. Stratified oceans and oxygenation of the late Precambrian environment: a post glacial geochemical record from the Neoproterozoic of W. Mongolia. Terra Nova, 9, 218–222. Shields, G. A., Braiser, M. D., Stille, P. & Dorjnamjaa, D. 2002. Factors contributing to high d13C values in Cryogenian limestones of western Mongolia. Earth and Planetary Science Letters, 196, 99– 111. Windley, B. F., Alexeiev, D., Xiao, W., Kroener, A. & Badarch, G. 2007. Tectonic models for accretion of the Central Asian Orogenic Belt. Journal of the Geological Society of London, 164, 31– 47. Yoshioka, H., Asahara, Y., Tojo, B. & Kawakami, S. 2003. Systematic variations in C, O, and Sr isotopes and elemental concentrations in Neoproterozoic carbonates in Namibia: implications for a glacial to interglacial transition. Precambrian Research, 124, 69– 85. Zhao, Y., Song, B. & Zhang, S. H. 2006. The Central Mongolian microcontinent: its Yangtze affinity and tectonic implications. In: Jahn, B. M. & Chung, L. (eds) Symposium on Continental Growth and Orogeny in Asia. Taipei, Taiwan, 135– 136. Zhegallo, L. & Zhuravelev, A. Y. 1991. Guidebook for the International Excursion to the Vendian– Cambrian Deposits of the Dzabkhan Zone of Mongolia. Palaeontological Institute, Moscow.

Chapter 30 The Khubsugul Group, Northern Mongolia FRANCIS A. MACDONALD* & DAVID S. JONES Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA 02138, USA *Corresponding author (e-mail: [email protected]) Abstract: The Khubsugul Group of northern Mongolia contains diamictites in the Ongoluk and Khesen formations that are succeeded by a stratiform phosphorite deposit and .2 km of early Cambrian dolomite. The stratigraphy of the Khubsugul Group, including the two diamictites, can be correlated with that of the Dzabkhan platform in southern Mongolia. By correlation, the Ongoluk diamictite is an early Cryogenian glacial deposit. A glaciogenic origin is inferred from the presence of striated clasts and bed-penetrating dropstones. The younger Khesen diamictite consists predominantly of a massive carbonate-clast diamictite, but also contains bed-penetrating dropstones in rare stratified facies, and is inferred to be end Cryogenian in age. The two diamictites are separated by as much as 250 m of allodapic carbonate. The phosphorite in the upper Khesen Formation (Fm.) is likely latest Ediacaran to early Cambrian in age and is separated from the glacial deposits by a major hiatus. Consequently, no links can be made between the phosphogenesis and the glacial deposits. Only limited geochemical, geochronological and palaeomagnetic results from the Khubsugul basin have been reported to date, but work is ongoing and there is strong potential for future studies.

Diamictites in the Ongoluk and Khesen formations of the Khubsugul Group of northern Mongolia are exposed discontinuously in a c. 250 km north –south belt (Fig. 30.1). The most complete exposures of the two diamictites and the overlying carbonate and phosphorite occur along the Khesen and Ongoluk Gols (tr. Rivers), on the west side of Lake Khubsugul (Fig. 30.2; 50842.50 N, 1008110 E and 50844.30 N, 100812.20 E). Most of the studies in the Khubsugul basin have focused on the phosphorite deposits and the regional tectonics, with the diamictites mentioned only in passing (e.g. Ilyin 1990, 1998). Although much of this work is in the Russian literature, several key manuscripts from Litologiya i Poleznye Iskopaemye have been translated to English in Lithology and Mineral Resources (e.g. Osokin & Tyzhinov 1998; Ilyin 2004). Additionally, a Russian field guide of the Khubsugul Basin was produced for an IGCP excursion to the phosphorite localities (Ilyin & Byamba 1980). This field trip spawned the hypothesis that ice rings orbited Precambrian Earth, that the shadow of these rings initiated the Neoproterozoic glaciations, and that their collapse led to phosphogenesis and precipitated the Cambrian radiation (Sheldon 1984). Geological work commenced in the Khubsugul basin in the mid-1960s with the discovery of ore-grade phosphorites (Donov et al. 1967). The Khubsugul Group was originally described in detail, including the identification of diamictites in the Ongoluk Fm., by Ilyin (1973). Osokin & Tyzhinov (1998) later differentiated a second diamictite in the basal Khesen Fm. and documented the presence of both diamictites throughout the Khubsugul basin. The diamictites in the Khubsugul basin have not been formally named, and the formations within which they occur have been named differently in Mongolia and Siberia (see Chumakov 2011). The diamictite at the base of the Ongoluk Fm. in the Khubsugul basin is likely equivalent to the diamictite in the Khushatai Fm. of the Sarkhoi Group (Osokin & Tyzhinov 1998). Otherwise, the Khubsugul Group in Mongolia is largely correlative with the Boxon Group in Siberia, with the Khesen Fm. roughly equivalent to the Zabit Fm. Correlations of specific diamictites in the Khesen and Zabit formations are complicated by multiple conglomeratic horizons within both formations (Kheraskova & Samygin 1992; Osokin & Tyzhinov 1998). Ilyin (1973, 2004) referred to the upper c. 50 m of the lower diamictite as the ‘perforated shales’ after the holes left behind from eroded carbonate clasts, and he defined the diamictite as

the basal member of the Khubsugul Group; however, the diamictite is underlain by an additional several hundred metres of clastic rocks that he included with the underlying Arasan Fm. Following Osokin & Tyzhinov (1998), we include these clastic rocks with the Ongoluk Fm. of the Khubsugul Group, and refer to the diamictite in the Ongoluk Fm. as the Ongoluk diamictite. The upper diamictite is in the basal Khesen Fm. of the Khubsugul Group and is herein referred to as the Khesen diamictite. The two diamictites are separated by 100– 250 m of carbonate of the upper Ongoluk Fm.

Structural framework The Khubsugul terrane is a composite Precambrian terrane, hosting a heterogeneous Archaean and Proterozoic crystalline basement intruded by c. 800 Ma continental arc volcanism (Badarch et al. 2002). Based on similarities in Neoproterozoic stratigraphy, radiometric ages in the underlying basement (Badarch et al. 2002) and the continuity of aeromagnetic anomalies associated with fringing Neoproterozoic ophiolites (Buchan et al. 2002), the southwestern margin of the Dzabkhan platform can be traced to the western margin of the Khubsugul basin along the Tuva-Mongolia border (Fig. 30.1, Macdonald 2011). The eastern boundary of the Dzabkhan terrane is obscured by Palaeozoic intrusions (Badarch et al. 1998) and, consequently, pre-Ordovician connections with the Baidrag terrane remain ambiguous. Overlap assemblages indicate that the Dzabkhan, Khubsugul, Baidrag and Tarvagatay terranes had amalgamated into a single continental mass by the Devonian (Badarch et al. 2002). On both the Khubsugul and Dzabkhan terranes, Palaeoproterozoic basement is overlain by thick volcanic –volcaniclastic successions (Badarch et al. 2002). On the Khubsugul terrane, the Sarkhoi volcanic rocks have been interpreted as having a continental arc affinity (Kuzmichev et al. 2001). These are unconformably overlain by the rift-related volcanic rocks and clastic sediments of the Dharkhat Group (Ilyin 1990). The Khubsugul terrane transformed from a continental arc to a thermally subsiding passive margin after the c. 800 Ma Shishkhid arc accreted to its western margin and prior to rifting along its eastern margin (Kuzmichev et al. 2005). Ilyin (2004) documented a deepening to the west in the phosphorite-bearing strata of the Khesen Fm. and the overlying

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 339– 345. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.30

340

F. A. MACDONALD & D. S. JONES

basal Khesen diamictite (Ilyin 2004). Poorly sorted, carbonate clast conglomerates and syn-sedimentary folding are also common in the uppermost Khesen formation. The upper c. 2 km of the Khubsugul Group consists of late Ediacaran and Cambrian platformal carbonates of the Erkhelnur Fm.

Glaciogenic deposits and associated strata The Ongoluk diamictite

Fig. 30.1. Tectonic map of western Mongolia modified from Badarch et al. (2002) and Windley et al. (2007). Teeth on faults indicate the inferred dip of subduction zones. K-D Arc, Khantayshir-Dariv Arc; AZ, accretionary zone including arcs, metamorphic rocks and ophiolites; NP, Neoproterozoic; Ord, Ordovician; Carb, Carboniferous.

early Cambrian shelf carbonates, and suggested these were deposited in a rift graben. Macdonald et al. (2009) alternatively posited that most of the Neoproterozoic sediments on the Dzabkhan and Khubsugul terranes were deposited on a thermally subsiding rifted margin, but that the late Ediacaran to early Cambrian phosphorites and overlying early Cambrian deposits formed in a foredeep basin in response to the Salarian orogeny (Ruzhentsev & Burashnikov 1996). Exposures of Cambrian carbonates on the west side of Lake Khubsugul are folded in tight, south-plunging synclines and are cut by Palaeozoic granites (Fig. 30.3), perhaps also related to an extension of the early Cambrian Salarian orogeny (Ruzhentsev & Burashnikov 1996) accompanying the collision between the Agradag arc and the Khubsugul terrane. Palaeozoic intrusions are particularly common in the south of the Khubsugul basin where they effectively obliterate the host stratigraphy (Badarch et al. 1998). Lake Khubsugul and the Darkhat depression formed as a southern arm of the Neogene Baikal rifting episode. Neoproterozoic –Cambrian stratigraphy on the west side of Lake Khubsugul was uplifted and exposed along a rift shoulder. Rifts commonly followed Precambrian structures and were accompanied by voluminous basaltic volcanism (Logatchev 1984).

Stratigraphy The Khubsugul basin sequence begins with rift-related volcanic and clastic rocks of the Darkhat Group that rest unconformably on Precambrian basement and meta-sediments (Ilyin 1973, 1990, 2004). These volcanic rocks are overlain with hundreds of metres of clastic and carbonate rocks that are variably preserved under the sub-Khubsugul Group unconformity (Osokin & Tyzhinov 1998). The Khubsugul Group begins with as much as 100 m of argillite, unsorted sandstone and minor limestone of the basal Ongoluk Fm. and grades upwards into the Ongoluk diamictite, with clasts becoming larger and more abundant upwards. The Ongoluk diamictite is overlain by 100 –250 m of allodapic dolomite of the upper Ongoluk Fm. The Khesen Fm. begins with a second diamictite that is c. 50 m thick and carbonate clastdominated (Osokin & Tyzhinov 1998). The Khesen diamictite is capped with a 3– 5 m dolostone, the top of which hosts centimetre-scale barite fans. Stratiform and granular phosphorite rests above a major flooding surface, less than 50 m above the

Near Lake Khubsugul, the Ongoluk diamictite is composed of a matrix-supported, stratified diamictite that ranges in thickness from c. 200 m to 415 m, thinning to the south and west. The thickest and most complete section of the Ongoluk Fm. is on the ridge north of Khesen Gol (Fig. 30.4). There, the basal contact of the diamictite is gradational with gravel-sized lonestones becoming more common upwards, both in an unsorted sandstone matrix and in a laminated argillite matrix, and, as such, the base is difficult to define. Sub-angular to sub-rounded gravel clasts of dolomite and quartzite are most common, with occasional cobbles of granite. Clasts become larger and more common in the upper c. 100 m of the diamictite (the ‘perforated shale’ member of Ilyin 1973, 2004), consisting predominantly of dolomite cobbles in an argillite or siltstone matrix. At Ongoluk Gol, the base of the Ongoluk diamictite is not exposed, but c. 130 m of clast poor siltstone and very coarse sandstone are present that are very similar to the lower c. 300 m at Khesen Gol. The ‘perforated shale’ member is c. 50 m thick at Ongoluk Gol, beginning with a 5-m-thick, massive, clastsupported dolomite diamictite, and continuing upwards with a dark, laminated argillite matrix with sub-rounded to angular, boulder-sized clasts of granite, quartzite, volcanic and metamorphic rocks, and gravel to boulder-sized clasts of dolomite. Quartzite and volcanic clasts are commonly faceted and striated (Osokin & Tyzhinov 1998).

The Khesen diamictite The Khesen diamictite tends to be thinner than the Ongoluk diamictite, and is composed of a massive, carbonate clast-dominated diamictite. The thickness of the Khesen diamictite ranges from 10 to 65 m, with no systematic geographical trend. The thickest measured section of the Khesen diamictite in the Lake Khubsugul area is exposed on the ridge north of Ongoluk Gol (Figs 30.3 & 30.5) and consists predominately of a massive, carbonate-clast diamictite. Clasts are composed of angular to sub-rounded dolomite and limestone (including giant ooid and stromatolite clasts) that are commonly imbricated, with sizes ranging from pebble to boulders. The Khesen diamictite has a yellow-weathering dolomite matrix, with the massive, unbedded deposits broken only by thin, lenticular, rhythmically bedded marls. At Khesen Gol, quartzite and volcanic pebbles are also present, and the massive diamictite is interrupted with multiple fine-laminated beds that are penetrated by outsized clasts.

Boundary relations with overlying and underlying non-glacial units Near Lake Khubsugul the Ongoluk diamictite has a gradational basal contact with clast-size and abundance decreasing gradually down-section until completely disappearing from the siltstone and sandstone (Osokin & Tyzhinov 1998). At Khesen Gol, the underlying clastic units consist of nearly 100 m of millimetrelaminated argillite and siltstone, with 0.5 m interbeds of unsorted very coarse sandstone and gravel conglomerate, and at least two beds of texture-less limestone. These units rest unconformably

THE KHUBSUGUL GROUP

341

Fig. 30.2. Geology of the western shores of Lake Khubsugul, showing the positions of measured sections.

on dolostone of the Arasan Group. The Ongoluk diamictite is separated from overlying dolomite by a sharp contact, although regionally, deposition appears to be uninterrupted. Along the Bokson River in Siberia, rocks that are equivalent to the Ongoluk diamictite rest unconformably on the Darkhat volcanics with a 15-m-thick weathering crust at the base (Kuzmichev 2001). A similar, though thinner breccia has been documented in

the Darkhat region of Mongolia and along the Sarkhoi River (Osokin & Tyzhinov 1998). The yellow-weathering Khesen diamictite rests disconformably on a sharp contact with the underlying light blue dolomites. The upper contact, however, is commonly broken with a couple of metres of dolomite breccia between the massive diamictite and the thin dolomite overlying the Khesen diamictite. The dolomite

342

F. A. MACDONALD & D. S. JONES

above a disconformable flooding surface (Ilyin 2004) that cuts down into the underlying units (Fig. 30.5).

Khesen Gol, M602, M613 c. 2km not shown

Chemostratigraphy

Erkhelnur Formation

C

400

P P

Palaeolatitude and palaeogeography

Darkhat

0 (m )

su al iat dh er

re

300

inf

Khesen Fm.

P P

200

Ongoluk Formation

100

100

c. 450 m not shown

C

0 (m)

Arasan

200

Khubsugul Group

r fa

300

P P C

Ongoluk Gol, M607, M608, M609

ce

40 0

Chemostratigraphic studies have not been reported for the carbonate units bounding the two diamictites, although work is in progress. Carbon-isotope values through the phosphatic interval of the Khesen Fm. at Ongoluk Gol range from – 7% to þ5% (Ilyin & Kiperman 2000; Ilyin 2004); however, as these data have not yet been reproduced, their utility for correlation is questionable. Strontium-isotope values of c. 0.7080 have also been reported from carbonate interbedded with the Khubsuglul phosphorite deposit (Shields et al. 2000).

vv vv vv vv

grainstone ribbonite allodapic carbonate

c. 2 km not shown rhythmite & siltstone diamictite flooding surface

c. 200 m not shown

C chert P phosphorite barite fans

Fig. 30.3. Stratigraphy of the Khubsugul Group along the Khesen and Ongoluk Gols. The locations of measured sections are shown in Figure 30.2.

is only c. 1 m thick at Ongoluk Gol with centimetre-scale bladed barite fans at the upper dolomite –limestone transition. Approximately 20 km north, the overlying white dolomite reaches a maximum thickness of 6 m, and is interspersed with bed-parallel cements (Fig. 30.5). A transgression continues above the dolomite into as much as 100 m of dark grey limestone rhythmite. The phosphorite-bearing strata is composed largely of dolomite with common olistostromes and mass flow deposits, and comes in

Cocks & Torsvik (2007) provide a review of the palaeomagnetic and Palaeozoic fauna affinity studies of Siberia and the peri-Siberian terranes. However, there are very few reliable palaeomagnetic data on the Khubsugul terrane, particularly in the Neoproterozoic, so the palaeolatitudes presented are highly speculative. From palaeomagnetic studies on peri-Siberian terranes, Kravchinsky et al. (2001) concluded that the Tuva-Mongolia belt was at low latitude, adjacent to Siberia throughout the Ediacaran and Cambrian. However, this study lacked a robust confidence test (i.e. only a reversal test with few samples and low resolution). Along with other peri-Siberian terranes, it has been suggested that the Khubsugul terrane occupied a Precambrian position between Siberia and Laurentia (Gladkochub et al. 2006), and rifted away from Siberia in the late Neoproterozoic (Sengor & Natal’in 1996; Kuzmichev et al. 2001, 2005). Sengor & Natal’in (1996) further posit that throughout the late Neoproterozoic and early Palaeozoic, these terranes were attached to the Central Mongolian Block, which along with other terranes, stretched to the present day Sea of Okhotsk. However, this reconstruction is inconsistent with the presence of Ordovician accretionary zones on the NW margins of the Baidrag and Dzabkhan terranes and a Late Cambrian Dhizda arc on the west margin of the Khubsugul terrane (Badarch et al. 2002). The Khubsugul and Tavargatay terranes host Cambrian trilobites endemic to Siberia (Astashkin et al. 1995) and Silurian brachiopods characteristic of the peri-Siberian realm (Hou & Boucot 1990). Thus, although there is a paucity of reliable palaeomagnetic constraints on the Khubsugul terranes, several lines of evidence indicate that they were adjacent to Siberia in the Neoproterozoic and early Palaeozoic. Pisarevsky et al. (2000) present a strong c. 615 Ma pole on red beds along the Lena River, pinning Siberia at equatorial latitudes in the Neoproterozoic. Further palaeomagnetic studies on the Khubsugul terrane are necessary, and are in progress (J. Meert pers. comm.).

Geochronological constraints Although the diamictites of the Khubsugul Group have not been directly dated, there are at least two radiometric constraints on the maximum age of the deposits. In Siberia, the Bokson Group overlies the Shishkhid arc, which contains magmatic zircons from rhyolites with a concordant U – Pb SHRIMP age of 800 + 2 Ma (Kuzmichev et al. 2005). The Bokson Group also overlies the Sorkhoi Group, which contains volcanic rocks that have been dated with whole-rock Rb –Sr at 718 + 30 Ma (Buyakaite et al. 1989). In Mongolia, the volcanic rocks of the Sorkhoi Group are stratigraphically equivalent to the Darkhat and Dzabkhan volcanics (Macdonald 2011), which have been dated at 850 + 2 and 750 + 3 Ma (Pb/Pb zircon, Burashnikov 1990), and more recently, at 777 + 6 Ma (U – Pb SHRIMP zircon, Zhao

THE KHUBSUGUL GROUP

Khesen Gol, M613, M614 450

Ongoluk Gol, M605, M610

343

Khirbisteg Gol N. Ongoluk Gol Khesen Gol N. Khesen Ridge Bakha Gol M607, M609 M602 M618 M611 M615, M616

P

P

200

100

P

P P P

C

v

P P P P

P

m

400

P P P P P P

150 v

50

350

v v

100

300 50 0 (m)

250

S

0 (m)

200

150

100

50

dolomite facies grainstone allodapic rhythmite diamictite facies massive stratified matrix sandstone siltstone argillite clast size boulder cobble pebble clast composition granite clastic carbonate Ongoluk Formation Arasan Formation

0 (m) Fig. 30.4. Detailed stratigraphy of the Ongoluk diamictite along the Khesen and Ongoluk Gols. The locations of measured sections are shown in Figure 30.2. m, metamorphic clasts; v, volcanic clasts.

2.5 km

3.3 km

1.0 km

14.0 km

N

Fig. 30.5. Detailed stratigraphy of the Khesen diamictite and overlying basal Khesen Fm. on the western shores of Lake Khubsugul. Positions of measured sections are in Figure 30.2. For legend see Figures 30.3 and 30.4.

et al. 2006), and 803.4 + 8.0 and 773.5 + 3.6 Ma (laser evaporation zircon, Levashova et al. 2010). The uppermost Khesen Formation is thought to be latest Ediacaran in age by correlation with the Zabit Fm. in Siberia, which contains Cloudina and Renalsis (Kheraskova & Samygin 1992). The Khesen Fm. is overlain by the early Cambrian, Archaeocyathid-bearing Erkhelnur Fm. (Ilyin & Zhuraveleva 1968), providing a robust minimum age on the diamictites.

Discussion Although the Dzabkhan platform of southern Mongolia may have been geographically separated from the Khubsugul basin, many units in the Tsagaan Oloom Fm. (Macdonald 2011) can be correlated with the Khubsugul Group. Both successions are underlain by riftogenic volcanic rocks and begin with interbedded clastic rocks and diamictites. The Maikhan Ul diamictite is an early Cryogenian glacial deposit (Brasier et al. 1996), and like the basal Ongoluk diamictite in the Khubsugul Group, it is commonly over 100 m thick and dominated by siltstone and coarse sand (Lindsay et al. 1996). The upper Ongoluk Fm. can be correlated with the Tayshir member of the Tsagaan Oloom Fm., and the carbonate-rich basal Khesen diamictite can be correlated with the Khongoryn diamictite (Macdonald et al. 2009). Chemostratigraphy indicates that the phosphorites on both the Khubsugul and Dzabkhan terranes were deposited in the latest Ediacaran to early Cambrian above a major Ediacaran hiatus (Shields et al. 2000; Macdonald et al. 2009). Unconformities in the Ongoluk Fm. developed near basement highs on rift shoulders that were active at least until the onset of

344

F. A. MACDONALD & D. S. JONES

deposition of the Ongoluk diamictite (Osokin & Tyzhinov 1998). While the Khubsugul basin was undergoing extension a large thickness of diamictite accumulated near the present Lake Khubsugul. Evidence for a glaciogenic origin of the Ongoluk diamictite includes exotic clasts with a mixed lithology, bedtruncating lonestones, and faceted and striated clasts (Osokin & Tyzhinov 1998). The contact with the overlying light-blue allodapic dolostone is sharp, and as they appear regionally conformable, a rift –drift transition is inferred within the Ongoluk diamictite. Although striated clasts have not been observed in the younger carbonate-rich Khesen diamictite, and exotic clasts are rare, a glacial origin of the massive deposit is inferred from the presence of bed-penetrating lonestones exposed along Khesen Gol, the geochemistry signature in associated carbonates (unpublished data) and the presence of overlying barite fans, which are present above basal Ediacaran cap dolostones in Australia (Kennedy 1996), Mauritania (Deynoux & Trompette 1976), NW Canada (Hoffman & Schrag 2002) and south China (Jiang et al. 2003). A basal Ediacaran age of the carbonate immediately overlying the Khesen diamictite is inferred from stratigraphic correlation with the Dzabkhan platform (Macdonald et al. 2009), and from the occurrence of the barite fans. Kheraskova & Samygin (1992) rejected a glaciomarine origin of diamictites in the Zabit Formation exposed on the Siberian side of the border, arguing that these deposits represent rift-related submarine slumps and debris flows. They further suggested that the diamictites in the Zabit Fm. (and correlative Khesen Fm.) are latest Ediacaran to early Cambrian in age, citing the presence of Cloudina and Renalsis in the Zabit Fm. On the Mongolian side of the border, there is no evidence of a glacial origin for the diamictite at the base of the Khesen Fm., or the carbonate and carbonate clasts conglomerate associated with the phosphorite in the uppermost Khesen Fm., which were interpreted as olistostromes and slumps. It is possible that, like the Dzabkhan platform to the south (Macdonald et al. 2009), there is a major hiatus in the Ediacaran and that Kheraskova & Samygin (1992) are miscorrelating mass flows in the latest Ediacaran to early Cambrian upper Khesen Fm. with the end Cryogenian Khesen diamictite described here from the base of the Khesen Fm. Dobretsov (1985) considered the Siberian diamictites to be ‘nappe thrust olistostromes’ related to the Salarian orogeny, which has been stratigraphically constrained to the early Cambrian on the Dzabkhan terrane (Ruzhentsev & Burashnikov 1996). Again, it seems likely that there is a conflation between the Khesen diamictite in the lower Khesen Fm., which is interpreted here as glaciogenic, and the conglomerates interpreted as olistostromes and allodapic deposits in the upper Khesen Fm. Sheldon (1984), Osokin & Tyzhinov (1998) and Ilyin (2004) have suggested a genetic relationship between the diamictite and phosphorite. However, Ilyin (2004) documented a disconformity at the base of the phosphorite (Fig. 30.5). Chemo- and lithostratigraphic studies in the correlative Dzabkhan platform (Macdonald et al. 2009; Macdonald 2011) suggest a major hiatus and flooding surface between the Khesen diamictite and the phosphorite series, casting doubt on any genetic relationship. Further chemo- and lithostratigraphic studies are needed to better constrain the basin dynamics and depositional setting of Neoproterozoic strata in Mongolia. The palaeogeography of the peri-Siberian terranes also remains speculative. It is clear, however, that island arcs surrounded the Khubsugul terrane for much of the Neoproterozoic and early Cambrian, and thus, there is excellent potential for geochronology in the Khubsugul basin. We thank our field assistants U. Bold, J. Otgonhuu and Eerie. We also thank A. Bayasgalan and the Mongolian University of Science and Technology for making our fieldwork possible, D. Schrag for use of the Harvard University Laboratory for Geochemical Oceanography, and G. Eischeid for help in the laboratory. We thank the National Science Foundation for funding, and P. Hoffman for inspiration and support. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) project #512.

References Astashkin, V. A., Pegel, T. V. et al. 1995. The Cambrian System of the Foldbelts of Russia and Mongolia. International Union of Geological Sciences, London. Badarch, G., Byamba, J. et al. 1998. Geological Map of Mongolia. Mineral Resources Authority of Mongolia, Ulaan Baator. Badarch, G., Cunningham, W. D. & Windley, B. 2002. A new terrane subdivision for Mongolia: implications for the Phanerozoic crustal growth of Central Asia. Journal of Asian Earth Sciences, 21, 87 – 110. Brasier, M. D., Shields, G., Kuleshov, V. N. & Zhegallo, E. A. 1996. Integrated chemo- and biostratigraphic calibration of early animal evolution: Neoproterozoic –early Cambrian of southwest Mongolia. Geological Magazine, 133, 445– 485. Buchan, C., Pfander, J. et al. 2002. Timing of accretion and collisional deformation in the Central Asian Orogenic Belt: implications of granite geochronology in the Bayankhongor Ophiolite Zone. Chemical Geology, 192, 23 –45. Burashnikov, V. V. 1990. Tectonics of the Urgamal Zone, Early Calidonides of Western Mongolia. Russian Academy of Sciences, Moscow. Buyakaite, M. I., Kuzmichev, A. B. & Sokolov, D. D. 1989. 718 Ma Rb –Sr errorchron of the Sorkhoi Group in the East Sayan. Doklady Akademii Nauk SSSR, 309, 150–154. Chumakov, N. M. 2011. Glacial deposits of Bokson Group, East Sayan Mountains, Buryatian Republic, Russian Federation. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 285–288. Cocks, L. & Torsvik, T. H. 2007. Siberia, the wandering northern terrane, and its changing geography through the Palaeozoic. Earth Science Reviews, 82, 29 –74. Deynoux, M. & Trompette, R. 1976. Late Precambrian mixtite: glacial and/or non-glacial? Dealing especially with the mixtite of West Africa. American Journal of Science, 276, 117–125. Dobretsov, N. L. 1985. Overthrust tectonics of the East Sayans. Geotectonics, 19, 26– 34. Donov, N. A., Edemsky, H. B. & Ilyin, A. V. 1967. Cambrian phosphorites of Mongolia Popular Republic. Sovetskaya Geologia, 3, 55 – 60. Gladkochub, D. P., Wingate, M. T. D., Pisarevsky, S. A., Donskaya, T. V., Mazukabzov, A. M., Ponomarchuk, V. A. & Stanevich, A. M. 2006. Mafic intrusions in southwestern Siberia and implications for a Neoproterozoic connection with Laurentia. Precambrian Research, 147, 260– 278. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis; testing the limits of global change. Terra Nova, 14, 129–155. Hou, H. F. & Boucot, A. J. 1990. The Balkhash-Mongolia-Okhotsk Region of the Old World Realm. Geological Society, London, Memoir, 12, 297– 303. Ilyin, A. V. 1973. Khubsugul Phosphorite-Bearing Basin. Geologicheskiy Institut, Akademiya Nauk SSSR, Moscow. Ilyin, A. V. 1990. Proterozoic supercontinent, its latest Precambrian rifting, breakup, dispersal into smaller continents, and subsidence of their margins: Evidence from Asia. Geology, 18, 1231– 1234. Ilyin, A. V. 1998. Rare-earth geochemistry of ‘old’ phosphorites and probability of syngenetic precipitation and accumulation of phosphate. Chemical Geology, 144, 243–256. Ilyin, A. V. 2004. The Khubsugul phosphate-bearing basin: new data and concepts. Lithology and Mineral Resources, 39, 454–467. Ilyin, A. V. & Byamba, J. 1980. Handbook for the Excursion ‘Phosphorites of the Khubsugul Basin in the Mongolian Peoples’s Republic’. Geologicheskiy Institut, Akademiya Nauk SSSR, Moscow. Ilyin, A. V. & Kiperman, Y. A. 2000. Mass accumulation of biogenic rocks at the Vendian/Cambrian boundary and carbon isotopic anomalies. Soveremenny voprosy geologii (Modern Problems of Geology). Nauchnyi Mir, Moscow. Ilyin, A. V. & Zhuraveleva, I. T. 1968. On the boundary between the Cambrian and the Precambrian at Prikhusugulie (Mongolian PR). Doklady Akademii Nauk SSSR, 182, 1164–1166. Jiang, G., Kennedy, M. J. & Christie-Blick, N. 2003. Stable isotopic evidence for methane seeps in Neoproterozoic postglacial cap carbonates. Nature, 426, 822– 826.

THE KHUBSUGUL GROUP

Kennedy, M. J. 1996. Stratigraphy, sedimentology, and isotope geochemistry of Australian Neoproterozoic postglacial cap dolostones: deglaciation, d13C excursions, and carbonate precipitation. Journal of Sedimentary Research, 66, 1050– 1064. Kherzaskova, T. N. & Samygin, S. G. 1992. Tectonic conditions in the East Sayan Vendian –Middle Cambrian terrigenous carbonate association. Geotectonics, 26, 445– 458. Kravchinsky, V. A., Konstantinov, K. M. & Cogne, J.-P. 2001. Palaeomagnetic study of Vendian and Early Cambrian of South Siberia and Central Mongolia: was the Siberian platform assembled at this time? Precambrian Research, 110, 61 – 92. Kuzmichev, A. 2001. Early Baikalian tectonic events in the TuvaMongolia Massif: arc-microcontinent collision. Geotectonics, 35, 185– 198. Kuzmichev, A., Bibikova, E. V. & Zhuravlev, D. Z. 2001. Neoproterozoic (800 Ma) orogeny in the Tuva-Mongolia Massif (Siberia): island arc-continent collision at the northeast Rodinia margin. Precambrian Research, 110, 109– 126. Kuzmichev, A., Kroener, A., Hegner, E., Dunyi, L. & &Yusheng, W. 2005. The Shishkhid ophiolite, nothern Mongolia: a key to the reconstruction of a Neoproterozoic island-arc system in central Asia. Precambrian Research, 138, 125–150. Levashova, N. M., Kalugin, V. M., Gibsher, A. S., Yff, J., Ryabinin, A. B., Meert, J. & Malone, S. J. 2010. The origin of the Baydaric microcontinent, Mongolia: constraints from palaeomagnetism and geochronology. Tectonophysics, 485, 306–320. Lindsay, J. F., Brasier, M., Shields, G., Khomentovsky, V. V. & Bat-Ireedui, Y. A. 1996. Glacial facies associations in a Neoproterozoic back-arc setting, Zavkhan Basin, western Mongolia. Geological Magazine, 133, 391–402. Logatchev, N. A. 1984. The Baikal rift system. Episodes, 7, 38 –42. Macdonald, F. A. 2011. The Tsagaan Oloom Formation, southwestern Mongolia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G.

345

(eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 331–337. Macdonald, F. A., Jones, D. S. & Schrag, D. P. 2009. Stratigraphic and tectonic implications of a new glacial diamictite-cap carbonate couplet in southwestern Mongolia. Geology, 37, 123–126. Osokin, P. V. & Tyzhinov, A. V. 1998. Precambrian Tilloids of the OkaKhubsugul phosphorite-bearing basin (Eastern Sayan, Northwestern Mongolia). Lithology and Mineral Resources, 33, 142– 154. Pisarevsky, S. A., Komissarova, R. A. & Khramov, A. N. 2000. New palaeomagnetic result from Vendian red sediments in Cisbaikalia and the problem of the relationsip of Siberia and Laurentia in the Vendian. Geophysics Journal International, 140, 598–610. Ruzhentsev, S. V. & Burashnikov, V. V. 1996. Tectonics of the western Mongolian Salairides. Geotectonics, 29, 379–394. Sengor, A. C. & Natal’in, B. A. 1996. Palaeotectonics of Asia: fragments of synthesis. In: Yin, A. & Harrison, M. (eds) The Tectonic Evolution of Asia. Cambridge University Press, Cambridge, 486– 640. Sheldon, R. P. 1984. Ice-ring origin of the Earth’s atmosphere and hydrosphere and late Proterozoic – Cambrian phosphogenesis, Phosphorite, Geological Survey of India Special Publication, 17, Udaipur, Rajasthan, India. Shields, G., Stille, P. & Brasier, M. 2000. Isotopic records across two phosphorite giant episodes compared: the Precambrian –Cambrian and the Late Cretaceous – recent. SEPM Special Publications, 66, 102–115. Windley, B. F., Alexeiev, D., Xiao, W., Kroener, A. & Badarch, G. 2007. Tectonic models for accretion of the Central Asian Orogenic Belt. Journal of the Geological Society of London, 164, 31 –47. Zhao, Y., Song, B. & Zhang, S. H. 2006. The Central Mongolian microcontinent: its Yangtze affinity and tectonic implications. In: Jahn, B. M. & Chung, L. (eds) Symposium on Continental Growth and Orogeny in Asia. Taipei, Taiwan, 135– 136.

Chapter 31 The Blaini Formation of the Lesser Himalaya, NW India JAMES L. ETIENNE1,2*, PHILIP A. ALLEN3, ERWAN LE GUERROUE´4, LARRY HEAMAN5, SUMIT K. GHOSH6 & RAFIQUE ISLAM6 1

Present address: Neftex Petroleum Consultants Ltd, 97 Milton Park, Abingdon, Oxfordshire OX14 4RY, UK 2

Geologisches Institut, Departement Erdwissenschaften, ETH-Zentrum, Haldenbachstrasse 44, CH-8092, Zu¨rich, Switzerland

3

Department of Earth Sciences and Engineering, Imperial College London, South Kensington Campus, London SW7 2AZ, UK 4

Ge´osciences Rennes (UMR 6118 – CNRS), 263 Avenue du General Leclerc. CS 74205, Universite´ de Rennes 1, 35042 Rennes Cedex, France

5

Department of Earth & Atmospheric Sciences, 1 – 26 Earth Sciences Building, University of Alberta, Edmonton, Alberta, T6G 2E3, Canada 6

Wadia Institute of Himalayan Geology, 33 General Mahadeo Singh Road, Dehra Dun, Uttarakhand, India *Corresponding author (e-mail: [email protected])

Abstract: Neoproterozoic glaciogenic deposits crop out widely across the Lesser Himalaya fold and thrust belt in NW India. Underlain by the siliciclastic Simla and Jaunsar Groups, the Blaini Formation (Fm.) includes at least two thick and regionally extensive diamictite units, separated by siliciclastics and argillites and capped by a pink microcrystalline dolomite. A glaciogenic origin is supported by the presence of relatively abundant striated clasts and the local preservation of polished and striated pavement on underlying Simla Group clastics. The cap dolostone is isotopically light with respect to both 13C and 18O, which show strong covariance. The Blaini is ubiquitously deformed, incorporated in regional-scale folds and thrusts, and exhibits locally intensive intra-formational deformation. Until recently, geochronological constraints have remained poor, but new detrital zircon ages from diamictite samples provide a maximum age limit of 692 + 18 Ma (207Pb/206Pb). Reliable palaeomagnetic data are required to constrain the position of this important passive continental margin in palaeogeographical reconstructions.

The Lesser Himalaya fold and thrust belt in NW India comprises up to 10 km of variably deformed low-grade Neoproterozoic to Lower Palaeozoic metasediments (Schelling 1992). Originally described as the ‘Blaini Conglomerate’ from the type section at Baliana Nala in Himachal Pradesh (Medlicott 1864), the Blaini Fm. is an important stratigraphic marker in the Lesser Himalaya and crops out across .300 km of the fold belt (Fig. 31.1). Some of the best exposures occur in the Mussoorie syncline, between Ma¯ldeota and Dhanaulti, in the Giri River valley near Dadahu and around Simla. A panoply of designations occur in the literature, including Blaini Conglomerate, Blaini Group, Blaini Boulder Bed, Blaini Series, Blaini diamictites, Blaini North, Blaini South and Blaini Fm., but the latter is preferred here given the unit’s extensive mappable nature (Pilgrim & West 1928; Fuchs & Sinha 1978) and use in recent publications (e.g. Jiang et al. 2002, 2003a; Kaufman et al. 2006). A glaciogenic origin for diamictites of the Blaini Fm. has long been recognized (e.g. Oldham 1887; Holland 1908; Pilgrim & West 1928; Auden 1946; Saxena & Pande 1969; Gaur & Dave 1971; Bhargava & Bhattacharyya 1975; Bhatia & Prasad 1975; Jain & Varadaraj 1978; Bhatia & Prasad 1981; Jain 1981); however, in the absence of good biostratigraphic data, an assumed correlation with the Permo-Carboniferous Talchir tillites of peninsular India has hampered recognition of the true age and significance of these deposits, despite a Proterozoic age frequently being posited (e.g. see Bhatia & Kanwar 1975 and references therein; Brookfield 1994). Recently published biostratigraphic data and detailed work on the overlying Krol and Tal Groups have firmly established a Proterozoic age for the Blaini Formation (Fig. 31.2). General overviews, including textural descriptions of diamictite lithofacies, are included in Bhargava & Bhattacharyya

(1975), Bhatia & Prasad (1975), Jain & Varadaraj (1978), Bhatia & Prasad (1981) and Jain (1981).

Structural framework The Blaini Fm. crops out around the margins of a series of tectonically complex synclines and in thrust block slices between the Main Boundary Fault and Main Central Thrust. Resting upon a succession of continental deltaic to shallow marine clastics (Nagthat Fm.) and more distal facies of the Simla Group (Ghosh 1991), the Blaini Fm. acts as a key stratigraphic marker horizon in this part of the Himalayas. The Blaini Fm. is overlain by the Infra Krol Fm. and the Krol and Tal groups, a succession that is thought to represent the inner part of a north-facing passive continental margin (Brookfield 1993). Based upon palaeogeographical reconstructions (e.g. Torsvik 2003) and isotopic and sequence stratigraphic similarities, this succession may have opposed the Chinese south-facing passive margin represented by the sedimentary cover of the South China block (Jiang et al. 2003a; Kaufman et al. 2006; McFadden et al. 2008). Palaeocurrent indicators and facies relationships in the underlying Jaunsar Group (Ghosh 1991) suggest a sediment source to the south/SE, and a basin margin oriented broadly NE –SW. Sequence stratigraphic analysis of the Krol Group illustrates a similar trend for the Ediacaran carbonate platform (Jiang et al. 2002; Kaufman et al. 2006). The Blaini Fm. is ubiquitously deformed, and faulted and folded on a regional scale, exhibiting penetrative cleavage and evidence for pressure solution. For these reasons, gravel clast macrofabrics may locally reflect the tectonic grain rather than primary depositional fabrics. In contrast, clast surface features such as glacial

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 347– 355. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.31

348

J. L. ETIENNE ET AL.

SOLAN

4 5

7

8

Siwalik Zone

3 9

1

Subathu-Dagshai

CHOR

2 6

Tal Group

10 11

12

Carbonate formations (Krol, Shali, Deoban etc)

15

13

14

Black slate formations (Infra Krol, Blaini, Shali, Sor)

MUSSOORIE 17

Fossiliferous Upper Palaeozoics of the Landsdowne syncline

18

16

19

Blaini/Mandhali

DEHRA DUN 20 21

Jaunsar Group (Nagthat/Chandpur undivided)

22 23 24

RISHIKESH

Chail

KAUDIYALA Basic metavolcanic rocks Chail/Nagthat undivided

500 km

SRINAGAR

M.

B.T .

STUDY AREA

Simla slates SHILLONG

30 N

DELHI

Crystalline basement

INDIA CALCUTTA

20 N

MUMBAI MADRAS

0

10

10 N 70 E

80 E

20

30

40

NAINI TAL

km

25

90 E

Fig. 31.1. Geological map of the Lesser Himalaya fold-and-thrust belt in NW India. Redrawn from Fuchs & Sinha (1978). Numbers indicate location of measured sections (Table 31.1; Fig. 31.3).

striae and polish have been preserved and are distinct from tectonic striae such as slickenlines. Brookfield (1987) noted the occurrence of shear structures within diamictite lithofacies, which is evident from pressure shadows and augen adjacent to rotated clasts. Although a tectonic origin for these features needs to be considered, it is possible that the shearing reflects soft-sediment deformation generated during sediment transport either in the subglacial environment, where large shear stresses can be generated beneath overriding ice (Boulton 1996), or during downslope resedimentation of debris, which results in similar microfabrics.

Stratigraphy The Blaini Formation is generally stratigraphically simple, broadly characterized by a tripartite diamictite –shale –diamictite succession that is variable in thickness across 300 km of the fold belt between Simla and Nainital. Additional diamictite intervals locally occur in the NW of the region and have also been reported from the area around Nainital (Jiang et al. 2003a). Diamictite units are generally less than 50 m thick (Fig. 31.3) and contain relatively abundant glacioclastic debris bearing cross-cutting glacial striae,

facets and polish. The Blaini is capped by a pink microcrystalline dolomite and is overlain by the Infra-Krol Fm. In the upper part, the cap contains thin grey and red shale partings.

Glaciogenic deposits and associated strata Massive and laminated diamictite lithofacies Diamictites occur as massive or weakly stratified clast-poor to clast-rich units with poorly sorted silty to sandy matrices. Thicknesses vary across the fold belt, but are typically on the order of 10–40 m (Fig. 31.3). Although exposure is poor, sheet-like geometries best explain the regional distribution of diamictite beneath the cap dolostone. Lateral continuity between stratigraphically lower occurrences of diamictite is more difficult to demonstrate, given the few available sections where the base of the formation may be observed. In the eastern part of the Mussoorie syncline, the coarse sediment fraction of the basal diamictite fines northwards, and on the northwestern limb, the upper surface of the uppermost diamictite is characterized by a winnowed transgressive lag immediately beneath the cap dolostone.

THE BLAINI FORMATION

349

Fig. 31.2. Lithostratigraphic and biostratigraphic subdivision of the Neoproterozoic– Cambrian succession of the Lesser Himalaya, summarized from Jiang et al. (2002), Hughes et al. (2005) and this study. Carbon–isotope data from Kaufman et al. (2006) and this study. For details on biostratigraphy, refer to Prasad et al. (1990), Tiwari (1999), Mazumdar & Banerjee (1998), Brasier & Singh (1987), Bhatt & Mathur (1990), Kumar et al. (1983), Mathur & Srivastava (1994), Kumar et al. (1987), Joshi et al. (1989), Jell & Hughes (1997), Mathur & Joshi (1989a, b), Tripathi et al. (1984, 1986), Singh & Rai (1983), Banerjee & Narain (1976), Bhargava et al. (1998), Mathur et al. (1988), Rai (1987), De et al. (1994), Bhargava (1984), Joshi & Mathur (1987), Tiwari & Knoll (1994), Shanker et al. (1997), Mathur & Shanker (1989, 1990), Bhatt (1991) and Hughes et al. (2005).

In the Solan region, at Rahed (section 3, Table 31.1, Fig. 31.3), the lower diamictite is erosionally based, and overlies a striated and polished surface of Simla Group argillites. Within this lithofacies, clast types include subrounded to subangular quartzite, microscale laminated siltstone, chert, vein quartz, gneiss, limestone, sandstone, dolomite, basalt, shale and slate. Striated and polished clasts are relatively abundant at most localities. Statistical clast shape and roundness data have previously been presented in Bhargava & Bhattacharyya (1975) and Bhatia & Prasad

(1975). Some petrographic and heavy mineral data may also be found in Jain & Varadaraj (1978) and Bhatia & Prasad (1981).

Interpretation Numerous interpretations have been posited for the genesis of the Blaini diamictites, emphasized by discussions in Bhatia & Kanwar (1975, and references therein) which revolve primarily

350

J. L. ETIENNE ET AL.

Table 31.1. Locations of measured sections (Figs 31.1 and 31.3) Section no. (Figs 1, 3) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Locality

Northing

Easting

Baliana Nala Deothal from to Rahed Giri River Valley from to Sargaon (i) Deoria from to Sargaon (ii) Rajgarh-Giripool Road Section Pervi River Valley Mareog Dadahu-Sangar Road Section Sangar Dadahu Dubrah-Mahipur Narendranagar Shivpuri Bridge Bhanswari Dhanaulti Dhanaulti-Raipur Road Section Renuka-Sataun Road Section Maldeota Shilla (Sataun-Shilai Road) Bhatoli (Kempte Falls– Yamuna Bridge) Gular Ghatti Nainital– Bhowali Section near Kaliakhan

30859.6280 30850.4900 30850.5450 30854.8410 308540 01.5000 308550 00.5900 30856.1090 308500 24.9300 308500 23.9200 30856.5550 30853.2230

77802.4660 7789.8540 7789.9200 77812.3260 778130 52.7500 778140 49.8600 77816.2460 778110 03.9700 778110 11.6200 77816.7940 77814.2290

30853.6530 308520 13.0800 30839.3110

77814.3970 778120 53.0000 77826.1350

30841.9150 30837.4760 30820.6720 30809.2070 308080 11.2700 30828.6790 30825.8160 30823.9590

77825.2630 77827.8000 78808.2960 78817.9580 788230 16.3000 78810.8030 78814.2280 78817.9550

30835.3750

77829.2690

30820.4420 30837.3040

78808.1840 77842.0350

30830.5980

78800.0360

30807.4010 29823.0890

78825.4380 79829.6610

around glacial v. non-glacial processes such as turbidity currents or mass failure events (see also discussions in Rupke 1968; Valdiya 1973). However, an enhanced understanding of modern glaciomarine depositional systems now recognizes the important role played by debris flows in resedimenting glaciogenic debris in these environments and the fact that the two are not mutually exclusive (cf. Dowdeswell et al. 1996, 1998; Taylor et al. 2002). Diamictites of the Blaini Fm. are interpreted as glaciogenic in origin on the basis that (i) the lithofacies contains abundant glacioclastic debris (striated, faceted and polished clasts) and (ii) the basal diamictite overlies striated and polished pavement (near Rahed, section 3, Table 31.1, Fig. 31.3). Several publications have suggested that the bulk of macroclasts within the Blaini diamictites are attributable to local sources derived from the underlying passive margin sequence (e.g. Auden 1934; Rupke 1968; Niyogi & Bhattacharya 1971; Valdiya 1973; Jain & Varadaraj 1978), including Brookfield (1987), who noted a direct compositional influence on the basal diamictite depending on which units of the Simla Group it overlies. Petrographic and heavy mineral analyses presented in Jain and Varadaraj (1978) are consistent with these interpretations; however, the observation of orthoclase, microcline, perthite, myrmekite, zircon, tourmaline and spinel by those authors are indicative of provenance from acid igneous rocks, although the degree of reworking needs consideration. Because crystalline basement does not crop out in this part of the fold belt, it is possible that

these data support an extra-basinal provenance for some of the material in the diamictites. Where diamictites directly overlie striated pavement, a primary tillite interpretation is considered most likely; however, across much of the fold belt, this association is not observed. If the intraformational shear structures observed within the diamictites are syn-sedimentary in origin, they may reflect either subglacial processes of sediment transport or sediment redistribution by debris flows, and are therefore not diagnostic palaeoenvironmental indicators. There is little evidence for significant glaciotectonic structuration, stratigraphic complexity, lithofacies heterogeneity or ice-proximal processes typical of terrestrial proglacial environments or glaciomarine grounding-line fan assemblages (cf. Benn & Evans 1998; Powell & Cooper 2002) and a more distal setting from the ice margin may be more likely. Coarse-tail grading locally exhibited in the basal diamictite in the Mussoorie syncline may indicate downslope reworking of glaciogenic debris distal from the ice margin, but the locally reworked tops of younger diamictites beneath the cap carbonate probably indicates deposition within neritic water depths. The diamictite lithofacies thus encompasses a range of different glacially influenced facies types, including primary tillites and glaciogenic debrites, which form part of a continuum of deposits interpreted by Brookfield (1987) as basal tills, flow tills and subaqueous meltout tills.

Sandstones, siltstones and shales Between the lower and upper diamictite units, a series of associated lithofacies occur that include massive amalgamated sandstone beds, flaggy siltstones and laminated silty shales. Sandstones overlie the diamictites at the base of the Blaini Fm. and either pass directly upwards into shales or through stacked flaggy siltstones into shales. Dark grey flaggy siltstones with sharp bases and tops are best developed in the Dhanaulti-Raipur (section 19, Table 31.1, Fig. 31.3) and Sataun-Renuka road sections. These thinly bedded (centimetre-scale) units bear no depositional structures, but in the Dhanaulti sections show evidence for thickening-upward, thinning-upward cycles throughout the section. The shales achieve thicknesses in the order of at least several tens of metres, but given an absence of key marker beds (sandstones rarely re-appear in section until just beneath the upper diamictite unit), and the local extent of deformation, stratigraphic thicknesses may locally exceed this. In the Solan region shales are of limited stratigraphic thickness and sandstone beds have thin (10 mm thick) pebbly lag bases.

Interpretation The lower portion of the Blaini Fm. is characterized by a finingupward motif recorded by a transition from diamictite through sandstones and siltstones into shales. This retrogradational stacking pattern indicates a transition to more distal conditions of sedimentation, with massive sandstones and siltstones probably deposited below the storm-weather wave base. More proximal conditions of sedimentation are indicated up-section as the lithofacies stacking pattern turns around and shales pass up into sandstones and diamictites. Coarser clastics and only thinly developed shales indicate more proximal conditions in the NW around Solan compared with thicker developed shale successions in the SE (e.g. in the Mussoorie syncline). Additional support for more proximal conditions in the NW is provided by the preservation of striated pavement at Rahed indicative of grounded ice conditions.

Dolomite Dolomite locally cements siltstones and parts of the uppermost diamictite unit in the Giri Valley to the east of Solan, and has also been

THE BLAINI FORMATION

Fig. 31.3. Lithostratigraphic sections and d13C chemostratigraphy of the Blaini Formation in the Lesser Himalaya of NW India. Numbered sections are located in Figure 31.1 and Table 31.1.

351

352

J. L. ETIENNE ET AL.

observed as discrete interbeds within equivalent units in the Mussoorie syncline (Kaufman et al. 2006), but is most extensively developed as a regional thinly bedded, laminated pink microcrystalline dolomite (Fig. 31.3). Across the fold belt, the dolostone rarely exceeds 9 m in thickness. Parallel or ‘crinkly’ lamination dominate, with the exception of a small patch of brecciated carbonate observed at Dhanaulti in the Mussoorie syncline. Like other cap carbonates, the Blaini dolostones contain some enigmatic ‘tepeelike’ structures. In the Nigalidhar and Mussoorie synclines, these are similar in form to ripple bedforms, with a wavelength of c. 1 m and amplitudes of the order of 0.2 m. Where observed, these structures occur 4–4.5 m above the base of the cap and are underlain and overlain by planar laminated dolomite. Laminae are locally pinched out at ripple crests, but have not been observed to onlap against the ripple trunks. Cracks run through and penetrate the ripple crests in the axial plane. Some millimetre-scale displacement occurs along these fractures, but no significant brecciation of the carbonates occurs in the axial zones.

Interpretation Most exposures of the Blaini cap carbonate are dominated by planar or crinkly laminated dolomite. The latter may indicate an algal origin, but no distinct structured mats, stromatolitic or other biohermal forms have been observed that could provide more definitive evidence of shallow-water sedimentation. With the exception of very locally developed breccia observed at Dhanaulti, a general lack of intraclasts or sedimentary breccias suggests that deposition occurred largely below the storm-weather wave base. The enigmatic tepee structures, which resemble ripple bedforms, are identical to examples observed elsewhere from cap carbonates (e.g. Allen & Hoffman 2005; Alvarenga et al. 2007); however, their interpretation is far from straightforward. No clear examples of laminae onlapping the ripple forms have been observed that require a sedimentary origin, and a tectonic origin has to be considered. Stromatactis-like structures associated with similar features in the Doushantuo Fm. in South China do not appear to occur (which might imply linkage to methane gas or fluid escape; cf. Jiang et al. 2006). Tepees described elsewhere in the literature have been variably interpreted as resulting from different syndepositional to early diagenetic processes of sedimentation, cementation and deformation, a discussion of which may be found in Jiang et al. (2006).

Boundary relations with overlying and underlying non-glacial units Across the fold belt, the Blaini Fm. is significantly deformed and heavily vegetated at altitudes up to c. 2.5 km above sea level. Rheological contrasts between the different lithofacies of the formation and its bounding strata have often led to deformation along contact surfaces, which presents some challenges in characterizing boundaries. Nevertheless, there are a number of key exposures across the fold belt where stratigraphic relationships may be properly established. For example, sections at Ma¯ldeota show the diamictites to disconformably overlie Nagthat Fm. quartzites by a slickenlined faulted contact. Further north, along the road between Ma¯ldeota and Dhanaulti, the contact between the basal diamictite and underlying quartzites remains sharp, but with no evidence for tectonic dislocation. Striated pavement has previously been reported at this contact, but present exposures do not appear to retain glacial striae. However, in the northwestern part of the fold belt, where the diamictites overlie softer mudrocks of the Simla Group, diamictites directly overlie polished striated pavement (e.g. near Rahed). Multiple sets of cross-cutting glacial striae provide

evidence for subglacial abrasion and the base of the formation in this instance is clearly erosional. Recent reports of biostratigraphically significant protoconodant assemblages from the Gangolihat Dolomite have led Azmi & Paul (2004) to adopt a modified version of Valdiya’s (1995) lithostratigraphic scheme for the region, suggesting a correlation between the Deoban and Krol Group carbonates, and invoking a major regional unconformity at the base of the Blaini. This is controversial given the widely held view that the Deoban carbonates lie stratigraphically beneath the Jaunsar Group, so further data are required to test these competing hypotheses.

Chemostratigraphy Abundant d13C and d18O stable isotope data are available for the Blaini cap dolostone (Kaufman et al. 2006; this study) and some strontium and magnesium values have been published (Kaufman et al. 2006). Kaufman et al (2006) identified a negative C-isotopic signature with up-section trends towards more negative values, which recover to near 0‰ in the Garhwal syncline. We collected c. 350 samples from 10 localities across the fold belt to better characterize the nature of the isotopic character and facilitate correlation between sections (Fig. 31.3). As with most cap carbonates, the Blaini dolostones are depleted in both carbon and oxygen, which show strong co-variation. In this sample set, d13Ccarb values range from –0.44 to –6.12, and d18O values vary between –5.49 and –13.27. In Figure 31.3 we present a first attempt at a composite d13Ccarb curve throughout the cap carbonate. The curve was generated by matching absolute values and trends in values between 10 sections in the Solan region, Nigalidhar and Mussoorie synclines. No correction has been attempted to account for stratigraphic condensation or expansion, which may have resulted from variation in sedimentation rates or differential compaction. Our data confirm the trend observed by Kaufman et al. (2006), with a pronounced negative excursion to –5‰ d13Ccarb followed by a recovery towards less negative values (Fig. 31.3). Kaufman et al (2006) also report negative values for carbonate interbedded with the upper diamictites in the Mussoorie syncline.

Significance of isotopic analyses Neoproterozoic cap carbonates are renowned for being isotopically light with respect to 13C, and the use of their negative isotope excursions as chemostratigraphic markers has become a standard tool in both the correlation of glacially influenced strata and the construction of a composite Precambrian C-isotope curve (Halverson 2005). Whether the values observed here reflect secular changes in the primary isotopic composition of seawater is difficult to establish. A lack of pelagic carbonate-secreting organisms makes establishing a proxy for primary seawater composition challenging, particularly if the original grains have been destroyed by recrystallization. The strong covariation of d13C and d18O raises the probability of diagenetic alteration associated with fluids flushing through the cap carbonate. Some support for fluid migration is evident given the ubiquitous occurrence of sparry calcite-filled veins in the cap. If the up-section trends in values reflect variations in the primary composition of seawater, then absolute values are likely to differ between localities given the transgressive nature of the carbonates deposited during post-glacial eustatic recovery. Given the poor continuity between exposures, an absence of distinct markers or fauna to facilitate correlation, sections may only be compared by trends in values (Fig. 31.3). The composite curve that results is interesting in that it broadly honours our understanding of the basin geometry, with progressive onlap (and therefore younging) from NW to SE, normal to the basin margin (see earlier discussion and Kaufman et al. 2006, fig. 2), and little variability between more

THE BLAINI FORMATION

353

Fig. 31.4. Probability–age plot for detrital zircons extracted from massive diamictite sample JE-1-002 (Ma¯ldeota, Mussoorie syncline).

closely spaced sections that occur along strike from one another. Some perturbations to this trend may indicate palaeobathymetric variation resulting from deglaciation of the shelf. Although numerous measurements were made to characterize the trends across the fold belt, extremely negative values have not been observed that may implicate local methane hydrate destabilization (cf. Jiang et al. 2003b; McFadden et al. 2008), although this does not negate dissociation events as a driving mechanism for negative excursions. The trend towards negative values, and recovery towards more positive values is similar to that recorded from the Keilberg cap on the South Congo Craton and generally exhibited by ‘Marinoan’ (c. 635 Ma) cap carbonates elsewhere (Kennedy et al. 1998). The observation of negative values in interdiamictite carbonates by Kaufman et al. (2006) differ from the general thesis of Kennedy et al. (2001), although some matrix samples from the Ghaub Fm. on Fransfontein Ridge exhibit negative values below the Keilberg cap (Kennedy et al. 2001, fig. 1).

Krol Belt, the P]/] boundary is located either in the uppermost part of the Krol Group (Krol E) or at the base of the Tal Group (Banerjee et al. 1997). Crystalline basement is not exposed in the region, so maximum age constraints are poorly defined. However, new detrital zircon ages for samples of the diamictite at the base of the Blaini Fm. (at Ma¯ldeota in the Mussoorie syncline) provide, to our knowledge, the first direct radiometric data for the unit, with a youngest subpopulation of three grains giving a mean weighted-average age of 692 + 18 Ma (207Pb/206Pb; Fig. 31.4). Four grains in the same sample define a slightly older subpopulation with a mean weighted-average age of 770 + 24 Ma (207Pb/206Pb). A number of older detrital populations are evident, which are also recorded in other samples. The full results of the detrital zircon analysis will be published elsewhere.

Conclusions Palaeolatitude and palaeogeography Palaeomagnetic data for the Malani Igneous Suite place Rajasthan along the western margin of Rodinia in mid to high latitudes at c. 750 Ma (Torsvik et al. 2001). Palaeopoles for the Blaini Fm. also indicate a mid- to high-latitude position for the lower diamictite (46.58N, 1098E), but tropical palaeolatitudes for the overlying cap dolostone (3.08N, 98.58E; Klootwijk 1979). Although these palaeolatitudes sit comfortably with Phanerozoic analogues of northern-hemisphere glaciation, positive fold tests are required to demonstrate primary remnant magnetization. Better geochronological control is also required to constrain the position of this important glaciated passive continental margin in palaeogeographic reconstructions.

Geochronological constraints A Neoproterozoic (as opposed to the formerly accepted view of Permo-Carboniferous) age for the Blaini Fm. is supported by the occurrence of Lower Cambrian trace fossils, trilobites, small shelly fauna and acanthomorphic acritarchs in basal Tal Group chert and phosphorites (e.g. Banerjee & Narain 1976; Brasier & Singh 1987; Bhatt 1989, 1991; Tiwari & Knoll 1994; Tiwari 1999; Hughes et al. 2005; Fig. 31.2). A SHRIMP U –Pb detrital zircon age of 525 + 8 Ma from Tal Group sediments at Gopichand ka Mahal in the Mussoorie syncline supports palaeontological data indicative of a Cambrian age, and thus reflects the youngest isotopic age limit for the Blaini Fm. (Myrow et al. 2003). In the

The Blaini Fm. records widespread glaciomarine conditions of sedimentation in a passive continental margin setting. At least two major periods of ice advance are recorded by the widespread distribution of glaciogenic diamictites, which include primary subglacial tillites deposited where the ice was grounded on the continental shelf, and glaciogenic debris flows that redistributed subglacial debris onto submarine fans in outer neritic to slope environments. The huge volumes of glacioclastic debris and their widespread distribution point towards the existence of an extensive continental ice sheet, probably with polythermal characteristics. The limited geochronological data do not presently allow the duration of these glacial events to be constrained, but both periods of ice sheet growth were probably younger than 692 + 18 Ma. The cap dolomite bears stable isotopic characteristics typical of Marinoan (c. 635 Ma) post-glacial cap carbonates elsewhere; however, further data are required to better constrain the age of these deposits. This research was funded by Schweizerischer National Fonds Grant 103502. M. Papp, M. Faccenda and R. Singh are kindly thanked for their assistance in the field. Stable isotope analyses were undertaken by T. Venneman at the University of Lausanne and D. Mrofka at the University of California Riverside. Laser Ablation MC ICP-MS detrital zircon analyses were undertaken at the University of Alberta, Edmonton. This study has benefited from ongoing discussions with M. Kennedy (UCR), R. Rieu (Repsol YPF), G. Jiang (University of Nevada) and A. Cozzi (ENI), and represents a contribution towards the IGCP 512 project. S. K. Ghosh and R. Islam are grateful to the Director, Wadia Institute of Himalayan Geology, for providing the necessary facilities. Special thanks go to D. Banerjee and A. J. Kaufman for constructive reviews, which improved this manuscript.

354

J. L. ETIENNE ET AL.

References Allen, P. A. & Hoffman, P. F. 2005. Extreme winds and waves in the aftermath of a Neoproterozoic glaciation. Nature, 433, 123– 127. Alvarenga, C. J. S. De, Dardenne, M. A. et al. 2007. Isotope stratigraphy of Neoproterozoic cap carbonates in the Araras Group, Brazil. Gondwana Research, 13, 469–479. Auden, J. B. 1934. The Geology of the Krol Belt. Records of the Geological Survey of India, 69, 123–167. Auden, J. B. 1946. Blaini-Talchir. Current Science, 12, 346–348. Azmi, R. J. & Paul, S. K. 2004. Discovery of Precambrian – Cambrian boundary protoconodants from the Gangolihat Dolomite of Inner Kumaun Lesser Himalaya: implication on age and correlation. Current Science, 86, 1653–1660. Banerjee, D. M. & Narain, M. J. 1976. Trace fossils in the Lower Tal Formation of Mussoorie and their environmental significance. Journal of Sedimentary Petrology, 46, 235–239. Banerjee, D. M., Schidlowski, M., Siebert, F. & Brasier, M. D. 1997. Geochemical changes across the Proterozoic –Cambrian transition in the Durmala phosphorite mine section, Mussoorie Hills, Garhwal Himalaya, India. Palaeogeography, Palaeoclimatology, Palaeoecology, 132, 183– 194. Benn, D. I. & Evans, D. J. A. 1998. Glaciers and Glaciation. Arnold, London. Bhargava, O. N. 1984. Trace fossils from the ?Cambrian Tal Group, Sirmur District H.P. and proposed redefinition of the Tal. Journal of the Palaeontological Society of India, 29, 84 –87. Bhargava, O. N. & Bhattacharyya, B. K. 1975. The Blaini Formation of Himachal Pradesh and Uttar Pradesh. Bulletin of the Indian Geologists’ Association, 8, 71 – 99. Bhargava, O. N., Singh, I., Hans, S. K. & Bassi, U. K. 1998. Early Cambrian trace and trilobite fossils from the Nigali Dhar Syncline (Sirmaur District, Himachal Pradesh), lithostratigraphic correlation and fossil content of the Tal Group. Himalayan Geology, 19, 89 –108. Bhatia, M. R. 1981. Appendix: Late Palaeozoic diamictites of Simla Hills, Lesser Himalaya, India; chronostratigraphic age and geochemistry. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 293– 294. Bhatia, S. B. & Kanwar, R. C. 1975. Blaini and related formations. Special Issue. Bulletin of the Indian Geologists’ Association, 8, 279. Bhatia, M. R. & Prasad, A. K. 1975. Some sedimentological, lithostratigraphic and genetic aspects of the Blaini Formation of parts of Simla Hills, Himachal Pradesh, India. Bulletin of the Indian Geologists’ Association, 8, 162– 185. Bhatia, M. R. & Prasad, A. K. 1981. Evolution of Late Paleozoic glacial marine sedimentation in the Simla Hills, Lesser Himalaya, India. Neues Jahrbuch fur Palaontologie Mh, 267– 288. Bhatt, D. K. 1989. Small shelly fossils, Tommotian and Meishucunian stages and the Precambrian –Cambrian boundary – implications of the recent studies in the Himalayan sequences. Journal of the Palaeontological Society of India, 34, 55 – 68. Bhatt, D. K. 1991. The Precambrian –Cambrian transition interval in Himalaya with special reference to small shelly fossils – a review of current status of work. Journal of the Palaeontological Society of India, 36, 109–120. Bhatt, D. K. & Mathur, A. K. 1990. Small shelly fossils of the Precambrian Cambrian boundary beds from the Krol-Tal succession in the Nainital Syncline, Lesser Himalaya. Current Science, 59, 218– 222. Boulton, G. S. 1996. Theory of glacial erosion, transport and deposition as a consequence of subglacial sediment deformation. Journal of Glaciology, 140, 43 –62. Brasier, M. D. & Singh, P. 1987. Microfossils and Precambrian – Cambrian boundary stratigraphy at Maldeota, Lesser Himalaya. Geological Magazine, 124, 323–345. Brookfield, M. E. 1987. Lithostratigraphic correlation of Blaini Formation (late Proterozoic, Lesser Himalaya, India) with other late Proterozoic tillite sequences. Geologische Rundschau, 76, 477–484. Brookfield, M. E. 1993. The Himalayan passive margin from Precambrian to Cretaceous times. Sedimentary Geology, 84, 1 – 35.

Brookfield, M. E. 1994. Problems in applying preservation, facies and sequence models to Sinian (Neoproterozoic) glacial sequences in Australia and Asia. Precambrian Research, 70, 113–143. De, C., Das, D. P. & Andraha, P. K. 1994. Ichnostratigraphic and palaeoenvironmental significance of trace fossils from Tal Formation of Nigali Dhar Syncline, Sirmur District, Himachal Pradesh, India. Indian Journal of Geology, 66, 77 –90. Dowdeswell, J. A., Kenyon, N. H., Elverhøi, A., Laberg, J. S., Hollender, F.-J., Mienert, J. & Siegert, M. J. 1996. Large-scale sedimentation on the glacier-influenced Polar North Atlantic margins: long-range side-scan sonar evidence. Geophysical Research Letters, 23, 3535– 3538. Dowdeswell, J. A., Elverhøi, A. & Spielhagen, R. 1998. Glacimarine sedimentary processes and facies on the polar North Atlantic Margins. Quaternary Science Reviews, 17, 243–272. Fuchs, G. & Sinha, A. K. 1978. The tectonics of the Garhwal-Kumaun Lesser Himalaya. Jahrbuch der Geologischen Bundesanstalt, 121, 219– 241. Gaur, G. C. S. & Dave, V. K. S. 1971. Blaini tillites near Rishikesh and their origin. Journal of the Geological Society of India, 12, 164– 172. Ghosh, S. K. 1991. Palaeoenvironmental analysis of the Late Proterozoic Nagthat Formation, NW Kumaun Lesser Himalaya, India. Sedimentary Geology, 71, 33 –45. Halverson, G. P. 2005. A Neoproterozoic chronology. In: Xiao, S. (ed.) Neoproterozoic Geobiology. Kluwer Academic Publishers, Delft, Netherlands. Holland, T. H. 1908. On the occurrence of striated boulders in the Blaini Formation of Simla with discussion on the geological age of the beds. Records of the Geological Survey of India, 37, 129–135. Hughes, N. C., Peng, S. et al. 2005. Cambrian biostratigraphy of the Tal Group, Lesser Himalaya, India, and early Tsanglangpuan (late early Cambrian) trilobites from the Nigali Dhar syncline. Geological Magazine, 142, 57 –80. Jain, A. K. 1981. Stratigraphy, petrography and palaeogeography of the Late Paleozoic diamictites of the Lesser Himalaya. Sedimentary Geology, 30, 43 –78. Jain, A. K. & Varadaraj, N. 1978. Stratigraphy and provenance of Late Palaeozoic diamictites in parts of Garhwal Lesser Himalaya, India. Geologische Rundschau, 67, 49 – 72. Jell, P. A. & Hughes, N. C. 1997. Himalayan Cambrian trilobites. Special Papers in Palaeontology, 58, 1– 113. Jiang, G., Christie-Blick, N., Kaufman, A. J., Banerjee, D. M. & Rai, V. 2002. Sequence stratigraphy of the Neoproterozoic Infra Krol Formation and Krol Group, Lesser Himalaya, India. Journal of Sedimentary Research, 72, 524– 542. Jiang, G., Sohl, L. E. & Christie-Blick, N. 2003a. Neoproterozoic stratigraphic comparison of the Lesser Himalaya (India) and Yangtze block (south China): Paleogeographic implications. Geology, 31, 917–920. Jiang, G., Kennedy, M. J. & Christie-Blick, N. 2003b. Stable isotopic evidence for methane seeps in Neoproterozoic postglacial cap carbonates. Nature, 426, 822–826. Jiang, G., Kennedy, M. J., Christie-Blick, N., Wu, H. & Zhang, S. 2006. Stratigraphy, sedimentary structures, and textures of the late Neoproterozoic Doushantuo cap carbonate in South China. Journal of Sedimentary Research, 76, 978– 995. Joshi, A. & Mathur, V. K. 1987. Report of Cruziana type trace fossils from the Arenaceous Member of Tal Formation, Mussoorie Synform. Indian Minerals, 41, 61 –65. Joshi, A., Mathur, V. K. & Bhatt, D. K. 1989. Discovery of redlichid trilobites from the Arenaceous Member of the Tal Formation, Garhwal Syncline, Lesser Himalaya, India. Journal of the Geological Society of India, 33, 538– 546. Kaufman, A. J., Jiang, G., Christie-Blick, N., Banerjee, D. M. & Rai, V. 2006. Stable isotope record of the terminal Neoproterozoic Krol platform in the Lesser Himalayas of northern India. Precambrian Research, 147, 156– 185. Kennedy, M. J., Runegar, B., Prave, A. R., Hoffman, K.-H. & Arthur, M. A. 1998. Two of four Neoproterozoic glaciations? Geology, 26, 1059–1063. Kennedy, M. J., Christie-Blick, N. & Prave, A. R. 2001. Carbon isotopic composition of Neoproterozoic glacial carbonates as a test of

THE BLAINI FORMATION

paleoceanographic models for snowball Earth phenomena. Geology, 29, 1135– 1138. Klootwijk, C. T. 1979. Review of palaeomagnetic data from the IndoPakistani fragment of Gondwanaland. In: Farah, A. & De Long, K. A. (eds) Geodynamics of Pakistan. Geological Survey of Pakistan, Queta, 41– 80. Kumar, G., Raina, B. K., Bhatt, D. K. & Jangpangi, B. S. 1983. Lower Cambrian body- and trace-fossils from the Tal Formation, Garhwal Synform, Uttar Pradesh, India. Journal of the Palaeontological Society of India, 28, 106– 11. Kumar, G., Bhatt, D. K. & Raina, B. K. 1987. Skeletal microfauna of Meishucunian and Qiongzhusian (Precambrian– Cambrian boundary) age from the Ganga Valley, Lesser Himalaya, India. Geological Magazine, 124, 167– 171. Mathur, V. K. & Joshi, A. 1989a. Record of inarticulate brachiopods from the Arenaceous Member of the Tal Formation, Garhwal Syncline, Lesser Himalaya, India. Current Science, 58, 446–448. Mathur, V. K. & Joshi, A. 1989b. Record of redlichiid trilobite from the Lower Cambrian Tal Formation, Mussoorie Syncline, Lesser Himalaya, India. Journal of the Geological Society of India, 33, 268– 270. Mathur, V. K. & Shanker, R. 1989. First record of Ediacaran fossils from the Krol Formation, Nainital syncline. Journal of the Geological Society of India, 34, 245– 254. Mathur, V. K. & Shanker, R. 1990. Ediacaran medusoids from Cambrian Tal Formation, Himachal Lesser Himalaya and the Krol Formation, Naini Tal syncline. Journal of the Geological Society of India, 36, 74 – 78. Mathur, V. K. & Srivastava, M. C. 1994. Record of microgastropod from the Arenaceous Member of the Tal Formation, Garhwal Syncline, Lesser Himalaya, India. Current Science, 66, 228–229. Mathur, V. K., Joshi, A. & Kumar, G. 1988. Trace fossils from Cambrian Tal Formation Himachal Lesser Himalaya, India, and their stratigraphic significance. Journal of the Geological Society of India, 31, 467– 475. Mazumdar, A. & Banerjee, D. M. 1998. Siliceous sponge spicules in the Early Cambrian Chert-phosphate Member of the Lower Tal Formation, Krol belt, Lesser Himalaya. Geology, 26, 899– 902. McFadden, K. A., Huang, J. et al. 2008. Pulsed oxidation and biological evolution in the Ediacaran Doushantuo Formation. Proceedings of the National Academy of Sciences of the United States of America, 15, 3197– 3202. Medlicott, H. B. 1864. On the geological structure and relation of the southern portion of the Himalayan ranges between the rivers Ganges and Ravee. Memoirs of the Geological Survey of India, 3, 1 – 212. Myrow, P. M., Hughes, N. C. et al. 2003. Integrated tectonostratigraphic analysis of the Himalaya and implications for its tectonic reconstruction. Earth and Planetary Science Letters, 212, 433–441. Niyogi, D. & Bhattacharya, S. C. 1971. A note on the Blaini boulder beds of the Lower Himalaya. Himalayan Geology, 1, 111– 122. Oldham, R. D. 1887. Notes on some points in Himalayan Geology. Records of the Geological Survey of India, 20, 155–161. Pilgrim, G. E. & West, W. D. 1928. The structure and correlation of Simla rocks. Memoirs of the Geological Survey of India, 53, 1– 140. Powell, R. D. & Cooper, J. M. 2002. A glacial sequence stratigraphic model for temperate, glaciated continental shelves.

355

In: Dowdeswell, J. A. & O. Cofaigh, C. (eds) Glacier-Influenced Sedimentation on High-Latitude Continental Margins. Geological Society, London, Special Publications, 203, 215– 244. Prasad, B., Maithy, P. K., Kumar, G. & Raina, B. K. 1990. Precambrian– Cambrian acritarchs from the Blaini-Krol-Tal sequence of Mussoorie Syncline, Garhwal Lesser Himalaya, India. Memoirs of the Geological Society of India, 16, 19 – 32. Rai, V. 1987. Additional trace fossils from the Tal Formation (Early Cambrian) Mussoorie hills, Uttar Pradesh, India. Journal of the Palaeontological Society of India, 32, 53 –59. Rupke, J. 1968. Note on the Blaini Boulder Bed of Tehri Garhwal, Kumaon Himalayas. Journal of the Geological Society of India, 9, 131– 133. Saxena, M. N. & Pande, C. 1969. The Blaini Tillite from the type area of Simla Himalayas. Bulletin of the Indian Geological Association, 2, 57 – 64. Schelling, D. 1992. The tectonostratigraphy and structure of the eastern Nepal Himalaya. Tectonics, 11, 925– 943. Shanker, R., Mathur, V. K., Kumar, G. & Srivastava, M. C. 1997. Additional Ediacaran biota from the Krol Group, Lesser Himalaya, India and their significance. Geoscience Journal, 18, 79– 94. Singh, I. B. & Rai, V. 1983. Fauna and biogenic structures in Krol-Tal succession (Vendian– Early Cambrian), Lesser Himalaya: their biostratigraphic and palaeoecological significance. Journal of the Palaeontological Society of India, 28, 67 –90. ´ ’Cofaigh, C. 2002. Taylor, J., Dowdeswell, J. A., Kenyon, N. H. & O Late Quaternary architecture of trough-mouth fans: debris flows and suspended sediments on the Norwegian margin. In: Dowdeswell, J. A. & O’Cofaigh, C. (eds) Glacier-Influenced Sedimentation on High-Latitude Continental Margins. Geological Society, London, Special Publications, 203, 55 –71. Tiwari, M. 1999. Organic-walled microfossils from the Chert– phosphorite Member, Tal Formation, Precambrian –Cambrian Boundary, India. Precambrian Geology, 99, 99– 113. Tiwari, M. & Knoll, A. H. 1994. Large acanthomorphic Acritarchs from the Infrakrol Formation of the Lesser Himalaya and their stratigraphic significance. Journal of Himalayan Geology, 5, 193– 201. Torsvik, T. H. 2003. The Rodinia jigsaw puzzle. Science, 300, 1379–1381. Torsvik, T. H., Carter, L. M., Ashwal, L. D., Bhushan, S. K., Pandit, M. K. & Jamtveit, B. 2001. Rodinia refined or obscured: palaeomagnetism of the Malani igneous suite (NW India). Precambrian Research, 108, 319–333. Tripathi, C., Jangpangi, B. S., Bhatt, D. K., Kumar, G. & Raina, B. K. 1984. Early Cambrian brachiopods from ‘Upper Tal’, Mussoorie syncline, DehraDun District, Uttar Pradesh, India. Geophytology, 14, 221– 227. Tripathi, C., Kumar, G., Mehra, S., Bhatt, D. K., Mathur, V. K. & Joshi, A. 1986. Additional Early Cambrian (Botomian) brachiopod fossil localities in Tal Formation, Lesser Himalaya, India, and their significance. Current Science, 55, 585–588. Valdiya, K. S. 1973. Blaini conglomerates of Himachal Pradesh and Garhwal. Recent Researchs in Geology, 1, Hindustan Publishing Corporation, Delhi. Valdiya, K. S. 1995. Proterozoic sedimentation and Pan-African geodynamic development in the Himalaya. Precambrian Research, 74, 35– 55.

Chapter 32 Neoproterozoic glacial records in the Yangtze Region, China QI-RUI ZHANG*, XUE-LEI CHU & LIAN-JUN FENG State Key Laboratory of Lithospheric Evolution, Institute of Geology and Geophysics, Chinese Academy of Sciences, PO Box 9825, Beijing 100029, China *Corresponding author (e-mail: [email protected]) Abstract: The Nanhua rift occurred at c. 820 Ma. The last rifting stage was likely associated with the Neoproterozoic glaciations. The glacially influenced sequence, in ascending order, comprises the Jiangkou Group (c. 720– 663 Ma), the interglacial Datangpo Formation (Fm.) (c. 663 Ma) and the Nantuo Fm. (,656– 635 Ma). In deep-water environments the glacially influenced units exhibit conformable contacts with a transitional lower boundary and a sharp upper boundary. In shallow marine environments, the lower boundary is always unconformable and erosive and the upper boundary is sharp but conformable. The Jiangkou glaciogenic deposits likely comprise the waxing, the maximum and the waning stages of glaciation. The lower part of the Chang’an Fm. corresponds to the waxing stage. It initially inherited the pre-glacial deep-water environment, and later, the thickening and coarsening upward trend likely indicates a lowering of sea level and ice-sheet advance. The predominant, massive diamictite in the upper Chang’an Fm. probably highlights the maximum glacial stage, and likely resulted from an amalgamation of debris flow deposits, suspension and ice-rafted debris. The Fulu Fm. records the waning stage; its lower member is characterized by arkosic sandstone and greywacke indicative of an interstadial between glacial events, while the upper member represents the final glacial event of the Jiangkou glaciation. The carbonaceous black shale and Mn carbonate at the base of the interglacial Datangpo Fm. were likely indicative of sedimentary starvation associated with Jiangkou deglaciation. The Nantuo glaciation is characterized by an abrupt onset and termination with negligible waxing and waning stages. The sediments likely resulted from subglacial deposition, sedimentary gravity flows and ice-rafted debris with reworking by currents. The onset of Jiangkou glaciation took place around c. 720 Ma and terminated at c. 663 Ma. The Nantuo glaciation started at ,c. 656 Ma and terminated at 635 Ma. The palaeolatitude during the Jiangkou glaciation is unconstrained and was 37 + 78 during the Nantuo glaciation. The palaeogeographic reconstruction for the Yangtze Block is controversial: it was either located between Australia and Laurentia or to the NW of Australia and to the north of India. Values of d34Spyrite from the top of the Fulu and the base of Datangpo formations exhibit extraordinarily high values up to þ60‰. Values of d13Ccarb from the base of the Datangpo Fm. range from –7.4‰ to –13.0‰ and from the cap carbonate of the Doushantuo Fm. are generally lower than –5‰; however, methanogenic values as low as –48‰ have been reported. From the Datangpo Fm., dinosterane has been detected, and phytane, triterpanes and hopane-type pentacyclic triterpanes are also abundant. Biomarker data from glacially influenced rocks indicate that the photosynthetic process never ceased and that the marine environment was probably normal during glaciations. The banded iron formation (BIF) at the base of the Fulu Fm. is either basaltic, tuffaceous, siliciclastic or carbonate-hosted. Rhodochrosite at the base of the Datangpo Fm. was probably associated with submarine volcanic and hydrothermal activity, but a detrital oxide origin cannot be excluded. The difference between the Jiangkou and the Nantuo glaciations is significant. Tectonics and associated topography are speculated to be the main controlling factors on the Neoproterozoic glaciations in the Yangtze region.

The Neoproterozoic glacially influenced (GI) rocks in China were first recognized by Willis et al. (1907). Wang et al. (1981) summarized the Neoproterozoic GI record in China and Lu et al. (1985) edited the first memoir of the Neoproterozoic GI rocks in China. By the 1990s, a series of geological memoirs on the regional geology in each province reported the GI successions in detail. This paper focuses on the Neoproterozoic GI records in the Yangtze region of South China. The stratigraphic classification scheme of Neoproterozoic GI deposits used in this paper is shown in Figure 32.1b. The stratigraphic nomenclature changes considerably, so brief explanations are necessary.

The ‘Sinian’ System and Period The term ‘Sinian’ first appeared in the nineteenth century (Richthofen 1882), and since 1983 it has been restricted to the Neoproterozoic in South China (Liu et al. 1991). More recently, however, it was redefined as the stratigraphic unit between the Nantuo Fm. and the Cambrian (NCS 2001), and coincides with the Ediacaran Period (Knoll et al. 2004).

The Nanhuan System and Period The Nanhuan ‘period’ ranges between 800 Ma and 680 Ma (NCS 2001), and so almost coincides with the chronometrically defined

Cryogenian Period (850 –650 Ma) (Remane 2000). However, this paper redefines its range as from c. 720 Ma to 635 Ma on the basis of the most recent geochronological results.

The Jiangkou Group and the Jiangkou glaciation The Jiangkou Group, formerly the Jiangkou Fm. (Yang 1985), is a term for the lower GI deposits and encompasses the Chang’an and Fulu formations (BGMR-Hunan 1997; Zhang et al. 2003). Before the 1980s, the term ‘Chang’an glaciation’ (Wang et al. 1981) was used to represent the early glaciation, and Lu et al. (1985) proposed the term ‘Gucheng glaciation’ to replace it. However, neither is appropriate, so the term Jiangkou glaciation is preferred (Zhang et al. 2003).

The Fulu Fm. The term ‘Fulu Fm.’ has been redefined at least twice. It was first defined as an interglacial unit between the Nantuo and the Chang’an formations (Fig. 32.2b, A) (BGMR-Guangxi 1985). The upper boundary of the Fulu Fm. was then lowered to the base of the later established Datangpo Fm. (Lu et al. 1985). Recently, the upper boundary of the Fulu Fm. was again lowered to the base of the ‘Gucheng Fm.’ (Fig. 32.2b, A), which was emphasized to be the record of a single glaciation (BGMR-Hunan

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 357– 366. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.32

358

Q. R. ZHANG ET AL.

Fig. 32.1. (a) Outcrops of Neoproterozoic glacial deposits in the Yangtze region (adapted from Liu et al. 1991): (i) outline of the mainland of China and (ii) the inset rectangle in (i). The star in (ii) is the location of Zhaoxing section and the pentagon the location of Shixian section. (b) A comprehensive Neoproterozoic stratigraphic succession in southeastern Guizhou (adapted from BGMR-Guizhou 1987). 1, Condon et al. 2005; 2, Zhou et al. 2004; 3, Zhang et al. 2008a; 4, Zhou et al. 2007a; 5, Wang et al. 2003a; *, Hao & Zhai 2004.

1997; Zhu et al. 2007). This paper adopts the definition given by Lu et al. (1985). By 1985, based on the occurrence of BIF and rhodochrosite, the Fulu Fm. was considered to be an interglacial unit sandwiched between the ‘Chang’an glaciation’ and the Nantuo glaciation. After the establishment of the interglacial Datangpo Fm., the former ‘interglacial’ nature of the Fulu Fm. became invalid and in fact it records the waning stage of the Jiangkou glaciation.

Structural framework At c. 1140 –900 Ma, the Yangtze and the Cathaysian plates collided along the southeastern margin of the Yangtze region (Li et al. 2008). The Nanhua rift started at c. 820 Ma (Fig. 32.1b) at the marginal area of the plate (Wang & Li 2003). At c. 830– 795 Ma and c. 780 –745 Ma, two pulses of bimodal anorogenic magmatic activity occurred (Li et al. 2003). The early pulse began prior to, and peaked at the onset of the rift; the later pulse of volcanism accompanied the main rifting stage. The first three rifting phases were likely pre-glacial, and the fourth phase associated with glaciation (Wang & Li 2003). The surface below the Neoproterozoic GI deposits is important and mostly represents a characteristic stratigraphic break. It probably resulted from the uplift and erosion of the rift shoulder. The pre-glacial units immediately below the erosive surface become older towards the north (Fig. 32.2b), while units above the erosive surface become younger (Fig. 32.2b), probably indicating that rift-induced subsidence persisted throughout the Nanhuan period. The Doushantuo cap dolostone likely marks the end of rifting and the onset of thermal subsidence. Magmatism during the Nanhuan was relatively weak and probably represents the final episode of magmatism and rifting. The rifting may reconcile the great contrast between the

uplift and erosion in the interior of the region and the accommodation created in the rift basin. Syn-sedimentary faulting is evident from abrupt changes in thickness of the GI units. The manganese deposits of the Datangpo Fm. were likely related to hydrothermal and submarine volcanic activity (Liu 1994; Yang et al. 1997). The great thickness (.2000 m, Fig. 32.3) of the Nantuo Fm. in the SW of Hunan (BGMR-Hunan 1988) and SE of Guizhou (BGMR-Guizhou 1987) probably signifies the final filling of the rift basin. The ‘Nd isotopic shift’ (c. 765 Ma to 635 Ma) revealed by Li & McCulloch (1996) also reflects the final filling stage of the Nanhua Rift.

Stratigraphy A comprehensive stratigraphic column of the Neoproterozoic is given in Figure 32.1b. The GI units (c. 720– 635 Ma) bracketed by the Ediacaran and the ‘Yangzian’ (Hao & Zhai 2004) are defined as Nanhuan (Cryogenian) in this paper. There are two GI units: the Jiangkou Group and the Nantuo Fm. (Fig. 32.1b). The Jiangkou Group comprises two formations: the Chang’an and the Fulu. The Chang’an Fm. occurs only in the western parts of the SE margin (Figs 32.2 & 32.3). The Fulu Fm. is composed of two members: the Liangjiehe and the Gucheng (sections A to D in Fig. 32.2); the contact between them is generally conformable (section A & B in Fig. 32.2). The Nantuo Fm. (Fig. 32.1b) is the most widespread Neoproterozoic GI unit (Figs 32.2 and 32.3). Sandwiched between the two GI units is the interglacial Datangpo Fm., which is thinner and finer grained. In the east of the Yangtze region, this tripartite structure is also typical, but the two lower units are much thinner. Generally, the Nanhuan in the Yangtze region is better preserved because of the tectonic simplicity and low degree of metamorphism.

THE YANGTZE REGION

359

Fig. 32.2. Neoproterozoic lithostratigraphy in the Yangtze region. (a) Variation along a north–south direction (locations of sections are given as stars in Fig. 32.3). The distance between the two end sections is c. 607 km. Distances between adjacent sections are also given. Vertical scale is the same except for section A. (b) Schematic stratigraphic break structure. Section A is adapted from BGMR-Guangxi (1985). Sections E, F and G are adapted from Sha et al. (1963). Section G shows no Neoproterozoic glacial deposits, and section A is complete.

The sedimentological description and facies analysis of the GI deposits are updated regularly, with the most recent and detailed version being provided by Dobrzinski & Bahlburg (2007). Figure 32.3 shows the distribution of the Nantuo Fm., in which the dashed line outlines the distribution of the Jiangkou Group.

with cross-stratification and loading structures, and lenses of conglomerate and breccia (Fig. 32.4a). Usually it is grey to greenish in colour with 60 to 80% clay minerals. Clast content is less than 5%, with sizes ranging from 0.2 cm to 5 cm and maximum 80 cm, subangular to sub-rounded. Rock types are quartz, sandy mudstone, sandstone, quartzite-like sandstone, minor volcanic rocks and, rarely, striated clasts.

The Chang’an Fm.

The Fulu Fm.

In deep-water environments, the contact between the Chang’an and the pre-glacial Gongdong formations is generally conformable (BGMR-Guangxi 1985; BGMR-Guizhou 1987). The transitional base of the Chang’an Fm. is distinguished from the pre-glacial grey to greenish mudstone by sandy mudstone with clasts. The Chang’an Fm. can be divided into two parts. The majority of the lower part (619 m thick) is mudstone (Fig. 32.4a), with cycles showing 0.1– 0.8 mm stratification. There are also coarse to fine-grained, massive to graded sandstones, with few current ripples and sole flutes, stratified and massive diamictites and minor granular conglomerate at the base. Clast size varies from 0.2 to 5 cm, clast shape from sub-rounded to angular, and the clast types are mainly quartz, sandstone and slate, with minor fragments of basic to acidic igneous rocks. Locally, there are lonestones (likely dropstones), striated clasts and mudstone breccia. Reverse grading is occasionally seen in coarsegrained intervals. The upper part (1279 m) comprises massive diamictite (92% of whole), with subordinate stratified mudstone, sandstone

The Fulu Fm. comprises two members (Fig. 32.4b): the Liangjiehe and the Gucheng (Zhang et al. 2003). The Liangjiehe Member (583 m) comprises mainly massive, fine-grained arkosic arenite or greywacke with lonestones, and occasionally with beds and lenses of granular conglomerate and diamictite. Cross-stratification and ripple marks, as well as graded bedding, are present. Locally at the base there are layers of purple to reddish banded ironstone with graded bedding and cross-stratification. Beds or lenses of micritic dolostone are also reported. From the sedimentary characteristics of the Liangjiehe Member in the Zhaoxing section (Fig. 32.4b), it is hard to determine its climatic character. However, there are sections where the Liangjiehe Member is typified by lonestone and beds of massive diamictite. Therefore, it is reasonable to suggest that it is interstadial. The Gucheng Member in the type section in Changyang County of Hubei Province is only 5.6 m thick, and comprises beds of diamictite with about 30– 40 vol% clasts, and lenses of sandy

Glaciogenic deposits and associated strata

360

Q. R. ZHANG ET AL.

Fig. 32.3. Distributions of Jiangkou Group (outlined by the black dashed line; adapted from BGMR-Guangxi 1985 and Wang & Li 2003) and Nantuo Fm. (grey areas; adapted from Wang & Li 2003). The contours show thickness of the Nantuo Fm. only.

conglomerate, sandy mudstone with lonestones (interpreted as dropstones), and striated clasts (Ma & Wang 1983). In the deepwater Zhaoxing section the Gucheng Member is about 217 m thick (Fig. 32.4b). The Fulu Fm. is also characterized by its calcareous composition (Dobrzinski & Bahlburg 2007) and dolomite mounds or lenses (Zhou et al. 2007b).

The Datangpo Fm. This is named after a section in northeastern Guizhou Province, with more than 500 m (Fig. 32.5) of fine-grained muddy sandstone, muddy siltstone, mudstone and about 5 m of black carbonaceous manganese shale at the base (Ma & Wang 1983). A range of sedimentary structures (gas holes, diapirs, seeping tubes, soft sediment deformation and mud volcanoes) are reported from layers and lenses of rhodochrosite associated with dolomite lenses (Zhou et al. 2007b). The upward coarsening and thickening trend is obvious. To the SE of the Yangtze region, it thins to a thickness of c. 12 m of dark carbonaceous shale (Fig. 32.4c) (BGMR-Guangxi 1985). However, there are sections without black carbonaceous shale but only grey-green massive mudstone.

The Nantuo Fm. The maximum thickness of the Nantuo Fm. is more than 2000 m (Fig. 32.3); it is composed predominantly of diamictite with scarce mudstone, sandstone and conglomerate. Dropstones, and

striated and bullet-shape clasts (Fig. 32.4c) are frequently observed (BGMR-Guangxi 1985). Because more than 90% of its thickness is massive diamictite, it looks similar to the upper part of the Chang’an Fm.; however, its clasts are larger, more common and lithologically more variable (e.g. granite, diabase, volcanic rocks, carbonate etc.), especially in its upper part. Allochthonous sedimentary blocks, in cases up to a maximum thickness of 10 m, frequently occur on shelf and upper slope environments. Diamictites with sharp base and deformed stratification, and occasionally gravel lag deposits, can be recognized in fresh outcrops.

The cap carbonate of the Doushantuo Fm. The cap carbonate represents the first member of the Doushantuo Fm., typically 3– 6 m thick and extending basin wide with the Nantuo Fm. (Fig. 32.3). It can be classified into three intervals: C1, the strongly disrupted and cemented basal layer; C2, the middle laminated layer with local tepee-like structures; and C3, the thinly laminated silty and shaly limestone and dolomite (Jiang et al. 2006). More than 90% of the rock minerals are micritic and microcrystalline dolomite with little quartz, feldspar and muscovite; calcite occurs mainly as filling materials in fissures (Wang et al. 2005). On shelf environments the cap carbonate is thicker, less deformed, with cemented breccias, tepee-like structures, stromatactis-like cavities, sheet cracks and barite fans. On the outer shelf or upper slopes it is thinly laminated and deformed, and in the basal part (C1 of Jiang et al. 2006). The centimetre-scale dolostone layers frequently intercalate with lenses or layers of silicate. The cap carbonate is missing in the lower slope and basin environments.

THE YANGTZE REGION

361

Fig. 32.4. Stratigraphic columns of the (a) Chang’an and (b) Fulu formations along the Zhaoxing section (the star in Fig. 32.1a,ii; adapted from BGMR-Guizhou 1987) and (c) the Datangpo and Nantuo formations along the Shixian section (the pentagon in Fig. 32.1a,ii; adapted from BGMR-Guangxi 1985).

Boundary relations with overlying and underlying non-glacial units The boundaries of the GI units are conformable in slope to basin environments; in shelf environments the upper boundary remains conformable but the lower boundaries are always unconformable.

The upper boundary of the Jiangkou Group The boundary is conformable but lithologically sharp. Below the surface is diamictite (gravel bearing sandy mudstone) of the Gucheng Member and above it is usually the carbonaceous black shale and locally grey-greenish mudstone of the Datangpo Fm.

The lower boundary of the Nantuo Fm. The lower boundary of the Jiangkou Group This is mostly unconformable and in contact with different units; however, it is conformable in the SW deep-water environment (Fig. 32.2b), for example, in the Zhaoxing section (the star in Fig. 32.1a(ii)) it is transitional. The top part of the Gongdong Fm. (Fig. 32.1b) comprises thinly bedded, grey sandy mudstone in which a single layer (c. 4 cm thick) is usually composed of a parallel laminated lower part and a structureless upper part; this is likely to correspond to the Td and Te of the Bouma divisions, respectively. The base of the Chang’an Fm. is, however, characterized by a thickly bedded greywacke, c. 20 m thick, with sporadic pebbles (,5 mm), on top of which is laminated sandy mudstone, where dropstones and clast clusters are occasionally found.

On shelf and upper slope environments, the contact between Nantuo Fm. and the Datangpo Fm. is lithologically sharp and erosive. In deep-water environments, however, it is locally transitional from mudstone to inversely graded diamictite beds as seen in the Pancao section of the Guzhang County in Hunan Province and in the Dacao section of the Ningguo County in Southern Anhui Province.

The upper boundary of the Nantuo Fm. The contact between the Nantuo and Doushantuo formations is sharp, conformable and occasionally pyrite-bearing (Jiang et al. 2006). The top of the Nantuo Fm. usually comprises diamictite

362

Q. R. ZHANG ET AL.

Fig. 32.5. The type section of Datangpo Fm. in Songtao County in the NE of Guizhou Province (adapted from Yang et al. 2002).

in shelf environments; however, in slope and basin environments there could exist a thin interval of sandy mudstone or muddy siltstone.

Isotope stratigraphy

Datangpo Fm. At the base of the Datangpo Fm., values of d13Ccarb

of rhodochrosite range from –7.4‰ to –13.0‰ (Tang & Liu 1999; Yang et al. 2002), so values of d13Ccarb from the base of Datangpo Fm. are probably lower than –5‰ (Chu et al. 2003). Doushantuo Fm. Carbon-isotope data from the cap carbonate of

Fulu Fm. This is about 30 m thick in the Minle section of Hunan

Doushantuo Fm. are abundant. Jiang et al. (2003a) and Wang et al. (2008a) reported very negative C-isotope values as low as –48‰ in the Yangtze Gorges area and in Changyang County, interpreted as evidence for methanogenesis. Generally there is a negative excursion in the cap carbonate of the Doushantuo Fm. (Jiang et al. 2007; Zhou & Xiao 2007; Zhu et al. 2007).

Datangpo Fm. Super heavy values of d34Spyrite (40 –60‰) are

Nd depleted mantle model ages (tDM)

34

d Spyrite Province. d34Spyrite values of þ24.1‰ and þ23.5‰ at c. 1 m and 10 m from the base and þ41.8‰ to þ58.4‰ at c. 23 m and 29 m are reported by Li et al. (1999). Zhou et al. (2007b) speculate that these high positive d34Spyrite values were likely associated with cold methane seepage.

reported from the vicinity of rhodochrosite mineralization (Li et al. 1999, 2006; Tang & Liu 1999) and are attributed alternatively to the anoxic conditions, low level of dissolved sulphate S (Liu et al. 2006) or restricted basin (Li et al. 1999). However, Wang et al. (2005) suggest that these extraordinarily high d34S values were probably associated with cold methane seepage, as they are similar to the values of c. 47.8– 67.1‰ from ODP Site 1146 in the South China Sea. Cap carbonate of the Doushantuo Fm. Values of d34Spyrite ranging

from 36.35‰ to 50.62‰ are reported by Wang et al. (2005), with the high values interpreted as recording sulphate reduction at cold methane seeps.

d13Ccarb Fulu Fm. In the upper part of the Fulu Fm., values of d13Ccarb from

dolomite and pebbly dolomitic sandstone are higher than –2.86‰ and were proposed to be associated with cold spring carbonate (Zhou et al. 2007b).

The tDM ages of sediments from the Neoproterozoic to lower Danzhou Group (c. 820– 765 Ma, Fig. 32.1b) are stable at c. 1.8 Ga (Li & McCulloch 1996). In the upper Danzhou Group (c. 765–720 Ma, Fig. 32.1b) the tDM ages drop to 1.3–1.4 Ga, and during the Nanhuan (c. 720–635 Ma, Fig. 32.1b) return to c. 1.8 Ga. From the Ediacaran onwards (,635 Ma), the tDM ages stabilize at c. 1.8 Ga again. This ‘Nd isotopic shift’ (Li & McCulloch 1996, fig. 3) probably reveals the final filling of the Nanhua rift by predominant juvenile crustal materials.

Other characteristics Banded iron formation (BIF) BIF appears at the base of the Fulu Fm. (Fig. 32.4), and can be traced from the northern Guangxi Region in the west to the middle of the Jiangxi Province in the east of the Yangtze region. Tang et al. (1987) classify the deposits into four types: basaltic, tuffaceous, siliciclastic and carbonate interbedded.

THE YANGTZE REGION

Manganese deposits

Geochronological constraints

The black shale-hosted manganese deposits of the Datangpo Fm. are important in China, although the deposits are relatively small (Roy 2006). They are mostly associated with faults and formed in local rift or graben basins. The content of volcanic materials may reach 25–50% (Yang & Pang 2006). Submarine volcanic and hydrothermal activities are considered to be the major factor in the formation of the deposits (Yang et al. 1997; Qin et al. 2005; Yang & Pang 2006). They are also closely associated with deglacial transgression after the Jiangkou glaciation, so Liu et al. (2006) speculate that they resulted from diagenetic remobilization of lateritic soil residues deposited in restricted basins by lowlatitude glaciation. These two factors are not mutually exclusive, and a combined model is proposed (Tang & Liu 1999).

Onset of the Jiangkou glaciation

363

A sample from the top unit of the Banxi Group (Fig. 32.2b, B) provides a SHRIMP U –Pb zircon age of 725 + 10 Ma (Zhang et al. 2008a). Because the Danzhou Group (Fig. 32.2b, A) is transitional upward into the Jiangkou Group, this age becomes a new maximum age constraint on the onset of the Jiangkou glaciation, and the age of the onset is reasonably estimated to be c. 720 Ma.

Termination of the Jiangkou glaciation The U –Pb zircon age of 663 + 4 Ma from a sample near the base of the Datangpo Fm. (Zhou et al. 2004) approximates to the termination of the Jiangkou glaciation (Fig. 32.1b).

Biomarkers Wang et al. (2003b, 2008b) studied biomarkers from samples of the Nanhuan System. The total organic content (TOC) of diamictites and cap carbonate range from 0.01 to 0.13%, 1–2 orders of magnitude lower than those of the non-GI sedimentary rocks. The positive correlation trend between TOC and the concentrations of phytane and/or pristane implies that photosynthetic autotrophs were probably the primary organisms contributing to the sedimentary organic matter. The unusually low concentrations of phytane plus pristane (0.005 –0.076 ng g21) in diamictites reveal a very weak photosynthetic process, and likely indicate that the photosynthetic process never ceased, allowing the survival of photosynthetic eukaryotes and other organisms. The GI sedimentary environment was probably normal marine (Wang et al. 2003b, 2008b). Dinosterane has been detected from black shale of the Datangpo Fm. (Meng et al. 2003), indicating that dinoflagellates and other eukaryotic algae had evolved by the mid-Neoproterozoic. Xie et al. (1999) report n-alkane distribution curves for manganese deposits of the Datangpo Fm. Phytane is predominant, while the Pr– Ph ratios (0.71 –0.83) indicate a reducing environment. Triterpanes and hopane-type pentacyclic triterpanes are abundant, and the organic matter originated from algae and microbes was likely associated with methane seeps (Zhou et al. 2007b).

Palaeolatitude and palaeogeography Palaeopoles Li et al. (1996) report a palaeolatitude of 38 + 88 for the Liantuo Fm. (766 + 18 Ma) in Hubei Province. Zhang & Piper (1997) achieve a similar palaeolatitude of 37 + 78 for the Nantuo Fm. (635 Ma) in Yunnan Province. The proximity of the palaeopoles of Liantuo and Nantuo formations indicates that either the palaeopoles remained almost unchanged from the Liantuo stage to the Nantuo stage, or that there was a loop in the apparent polar wander path from c. 766 –635 Ma.

Palaeogeographical position The relative position of the Yangtze region is controversial. Li et al. (2008) places the Yangtze region (South China Block) between Australia and Laurentia during the assembly of the supercontinent Rodinia at c. 900–780 Ma and proposes that it moved to the NE of Australia during the break-up of Rodinia at c. 750–630 Ma. Kirschvink (1992), Zhang & Piper (1997), Jiang et al. (2003b) place the South China Block to the NW (present coordinate) of Australia and to the north of India.

Onset of the Nantuo glaciation A recent U– Pb SHRIMP zircon age of 656 + 3 Ma from the top of the Dantangpo Fm. in west Hunan indicates that the onset of the Nantuo glaciation should be younger than c. 656 Ma (Zhang et al. 2008b). Considering the shelf setting in the sampled section and significant erosion at the Nantuo/Datangpo boundary, the Nantuo glaciation could be significantly younger than 656 Ma.

Termination of the Nantuo glaciation After the dating of 635.2 + 0.6 Ma from samples near the base of the cap carbonate on top of the Nantuo Fm. (Condon et al. 2005), the age of the termination of Nantuo glaciation is constrained to c. 635 Ma (Ogg & Gradstein 2008).

Discussion There has been a prolonged debate on the correlation of the Liantuo Fm. with various GI units (Yin et al. 2003 among others), which has brought great difficulty to the investigation of the evolution of Neoproterozoic glaciations in South China. However, a newly achieved age of 725 + 10 Ma (Zhang et al. 2008a) from the top of the Banxi Group, the recalculated age of 766 + 18 Ma of the former 748 + 12 Ma of the Liantuo Fm. (Zheng 2003) and 765 + 14 Ma from the Danzhou Group (Zhou et al. 2007a) seem to have solved the problem and confirm that the Liantuo Fm. is not a GI unit and can only be correlated with the middle of the Banxi (Danzhou) Group. Based on the above evidence, it is almost certain that in the Yangtze region there were two Neoproterozoic glaciations: the Jiangkou (c. 720 to 663 + 4 Ma) and the Nantuo (,656+3 to 635.2 + 0.6 Ma). The salient decrease in tDM ages from c. 765 Ma to c. 720 Ma probably indicates that the uplifted rift shoulder was unroofed, weathered and eroded, and supports the proposition that the atmospheric content of CO2 was significantly drawn down, triggering glaciations (Godde´ris et al. 2007). The ‘Nd isotopic shift’ reveals that a great amount of juvenile mantlederived material was incorporated into the sedimentary provenance (Li & McCulloch 1996). The peculiar hiatus architecture of the GI stratigraphy in the Yangtze region (Fig. 32.2b) testifies independently to the uplifting, erosion and filling of the rift basin. It is therefore the tectonics and greatly evolved topography that finally controlled the geological records of the glaciations.

The depositional environment It was the Nanhua rift that provided the accommodation necessary for GI deposition. The northwardly expanding distribution of

364

Q. R. ZHANG ET AL.

younger GI units (Fig. 32.2b) likely signifies the protracted subsidence of the basin. Except for a few outcrops in the west of the region, the Neoproterozoic GI deposits are believed to be glaciomarine. The Jiangkou glaciations. The maximum thickness is 2698 m (BGMR-Guizhou 1987) and it can be divided into three stages. The first or waxing stage corresponds to the lower part of the Chang’an Fm. (Fig. 32.4a). The laminated sandy mudstones with dropstones and clast-clusters at the base indicate an inherited deep-water environment with the addition of glacially derived sediments. The thickening and coarsening upward trend probably signifies the lowering of the sea level and advance of ice sheets. The predominant massive diamictite with graded sandstones and mudstone in the upper part of the Chang’an Fm. (Fig. 32.4a) probably records the amalgamation of debris-flow deposits and rainout in an ice-distal environment during the second or main glacial stage. The shallow ice-proximal records of these two stages were likely deposited originally, but were not preserved because of erosion by the later glacial events. In the waning or third stage, BIF and massive and cross-stratified greywacke are probably indicative of an interstadial, where the frequent occurrence of lonestones indicates the existence of icebergs and ice sheets. The diamictite in the Gucheng Member was pro-glacial, as described by Dobrzinski & Bahlburg (2007), and likely subglacial or iceproximal too in certain sections. The Datangpo interglacial. The thin starved sediments of the Datangpo Fm. in the Zhaoxing-Shixian area (the star and pentagon in Fig. 32.1a(ii)) are probably indicative of a deep-water environment. With few exceptions, nothing coarser than muddy finegrained sandstone is reported from the unit in this region, so it probably represents a distal marine environment. The basal black shale and Mn dolomite were probably deposited in a restricted shelf environment at the early transgression induced by the deglaciation of the Jiangkou glaciation. The Nantuo glaciation. The Nantuo Fm. in the Yangtze Gorges area

was originally believed to be terrestrial (Wang et al. 1981, among others) and was known as the Nantuo Tillite for more than half a century. In fact, the Nantuo Fm. in the Yangtze Gorges area (Zhang 1995), as well as in the Nanhua rift, was glaciomarine. The predominant diamictite was likely subglacial and ice-proximal in an inner shelf environment, including sediment gravity flow and rainout. In ice-distal environments, amalgamation of debris flow deposits, rainout diamictite and disturbed deposition are dominant. The thin layer of lag deposits (Liu et al. 1999) likely indicates the existence of a winnowing current. The inversely graded pebbly mudstone at the transitional lower boundary was likely hemipelagic, and the sandstone or sandy mudstone at the top was turbiditic in origin; however, the waxing and waning stages are not significant. Jiang et al. (2006) consider that the cap carbonate of the Doushantuo Fm. shares morphological and petrographic attributes with modern and ancient methane seeps, as a result of post-glacial warming of the ocean, and is indicative of relatively deep-water deposition, most likely below storm wave base.

Comparison between the Jiangkou and the Nantuo glaciations The differences between the two glaciations are obvious. The most conspicuous difference is in stratigraphic complexity. The Jiangkou glaciation can be divided into three stages, the waxing, the maximum and the waning. The preservation of the Jiangkou glacial units is highly limited in the southeastern marginal region of the Yangtze block, because the complete succession is restricted to the deep-water environment. The Nantuo glaciation was relatively simple, usually with negligible or no waxing and waning records.

The reasons for these differences are controversial. Dobrzinski & Bahlburg (2007) emphasize the different reaches of the ice sheets and ascribe the differences to the palaeolatitude, intensity and extension of the two glaciations. In fact, the role of the tectonic control implied in Figure 32.2b was likely crucial, and the difference of the evolving topography in different rift stages was probably a secondary controlling factor. The Jiangkou glaciation happened within a background of contrasting relief, while the Nantuo glaciation occurred within a glacially levelled topography. Sedimentary structures (Liu et al. 1999), redox proxies (Dobrzinski et al. 2004), and biomarkers (Wang et al. 2003b, 2008b) probably suggest that hydrological cycle was not shut down during glaciations in the Yangtze region. This paper was supported by project nos 40532012 and 40373011 of the Natural Science Foundation of China (NSFC) and project KZCX3-SW-141 of the Chinese Academy of Sciences (to X.-L. Chu). We thank G. A. Shields and E. Arnaud for their encouragement and help in writing this paper. Detailed and insightful reviews by G.-Q. Jiang, C.-M. Zhou, H. Bahlburg and G. A. Shields-Zhou are gratefully acknowledged. This is a contribution to International Geological Correlation Program Project 512.

References BGMR-GUANGXI (BUREAU OF GEOLOGY AND MINERAL RESOURCES OF GUANGXI ZHUANG AUTONOMOUS REGION) 1985. Regional Geology of Guangxi Zhuang Autonomous Region. Geological Publishing House, Beijing, 853 (in Chinese with English abstract). BGMR-GUIZHOU (BUREAU OF GEOLOGY AND MINERAL RESOURCES OF GUIZHOU PROVINCE) 1987. Regional Geology of Guizhou Province. Geological Publishing House, Beijing, 698 (in Chinese with English abstract). BGMR-HUNAN (BUREAU OF GEOLOGY AND MINERAL RESOURCES OF HUNAN PROVINCE) 1988. Regional Geology of Hunan Province. Geological Publishing House, Beijing, 719 (in Chinese with English abstract). BGMR-HUNAN (BUREAU OF GEOLOGY AND MINERAL RESOURCES OF HUNAN PROVINCE) 1997. Multiple Classification of the Stratigraphy of China (43): Stratigraphy (Lithostratic) of Hunan Province. China University of Geosciences Press, Wuhan, 292 (in Chinese). Chu, X.-L., Zhang, Q.-R., Zhang, T.-G. & Feng, L.-J. 2003. Sulfur and carbon isotopic variations in Neoproterozoic sedimentary rocks from southern China. Progress in Natural Science, 13, 875–880. Condon, D., Zhu, M.-Y., Bowring, S., Wang, W., Yang, A. & Jin, Y. 2005. U– Pb ages from the Neoproterozoic Doushantuo Formation, China. Science, 308, 95– 98. Dobrzinski, N. & Bahlburg, H. 2007. Sedimentology and environmental significance of the Cryogenian successions of the Yangtze Platform, South China block. Palaeogeography, Palaeoclimatology, Palaeoecology, 254, 100–122. Dobrzinski, N., Bahlburg, H., Strauss, H. & Zhang, Q.-R. 2004. Geochemical climate proxies applied to the Neoproterozoic glacial succession on the Yangtze platform, south China. In: Jinkin, G. S., McMenamin, M. A. S., McKay, C. P. & Sohl, L. (eds) The Extreme Proterozoic: Geology, Geochemistry and Climate. American Geophysical Union, Washington DC, Geophysical Monograph, 146, 13– 32. Godde´ris, Y., Donnadieu, Y. et al. 2007. Coupled modeling of global carbon cycle and climate in the Neoproterozoic: links between Rodinia breakup and major glaciations. Comptes Rendus Geoscience, 339, 212– 222. Hao, J. & Zhai, M. 2004. Jinning movement and Sinian System in China: their relationship with Rodinia Supercontinent. Chinese Journal of Geology, 39, 139–152 (in Chinese with English abstract). Jiang, G.-Q., Kennedy, M. J. & Christie-Blick, N. 2003a. Stable isotopic evidence for methane seeps in Neoproterozoic postglacial cap carbonates. Nature, 426, 18– 25. Jiang, G. Q., Sohl, L. E. & Christie-Blick, N. 2003b. Neoproterozoic stratigraphic comparison of the Lesser Himalaya (India) and

THE YANGTZE REGION

Yangtze block (south China): Paleogeographic implications. Geology, 31, 917–920. Jiang, G. Q., Kennedy, M. J., Christie-Blick, N., Wu, H.-C. & Zhang, S.-H. 2006. Stratigraphy, sedimentary structures, and textures of the late Neoproterozoic Doushantuo cap carbonate in South China. Journal of Sedimentary Research, 76, 978–995. Jiang, G. Q., Kaufman, A. J., Christie-Blick, N., Zhang, S.-H. & Wu, H.-C. 2007. Carbon isotope variability across the Ediacaran Yangtze platform in South China: implications for a large surface-to-deep ocean d13C gradient. Earth and Planetary Science Letters, 261, 303–320. Kirschvink, J. L. 1992. A paleogeographic model for Vendian and Cambrian time. In: Schopf, J. W. & Klein, C. (eds) The Proterozoic Biosphere: A Multidisciplinary Study. Cambridge University Press, Cambridge, 569–581. Knoll, A. H., Walter, M. R., Narbonne, G. M. & Christie-Blick, N. 2004. A new period for the geologic time scale. Science, 305, 621–622. Li, X.-H. & McCulloch, M. T. 1996. Secular variation in the Nd isotopic composition of Neoproterozoic sediments from the southern margin of the Yangtze Block: evidence for a Proterozoic continental collision in southeast China. Precambrian Research, 76, 67– 76. Li, Z.-X., Zhang, L. & Powell, C. Mc. A. 1996. Positions of the east Asian cratonic blocks in the Neoproterozoic supercontinent Rodinia. Australian Journal of Earth Science, 43, 593–604. Li, R.-W., Chen, J., Zhang, S., Lei, J., Shen, Y. & Chen, X. 1999. Spatial and temporal variations in carbon and sulfur isotopic compositions of Sinian sedimentary rocks in the Yangtze platform, South China. Precambrian Research, 97, 59– 75. Li, Z.-X., Li, X.-H., Kinny, P. D., Wang, J., Zhang, S. & Zhou, H. 2003. Geochronology of Neoproterozoic syn-rift magmatism in the Yangtze Craton, South China and correlations with other continents: evidence for a mantle superplume that broke up Rodinia. Precambrian Research, 122, 85– 109. Li, Z.-X., Bogdanova, S. V. et al. 2008. Assembly, configuration, and break-up history of Rodinia: a synthesis. Precambrian Research, 160, 179– 210. Liu, J.-S. 1994. Early Sinian sedimentary formations, turbidite and manganese ore genesis in Hunan Province. Geotectonica et Metallogenia, 18, 173–182 (in Chinese). Liu, H.-Y., Dong, R.-S. et al. 1991. The Sinian System of China. Scientific Press, Beijing (in Chinese). Liu, H.-Y., Hao, J. & Li, Y.-J. 1999. The Late Precambrian Stratigraphy and Geological Evolution of Middle-Eastern China. Science Press, Beijing. Liu, T.-B., Maynard, J. B. & Alten, J. 2006. Superheavy S isotopes from glacier-associated sediments of the Neoproterozoic of south China: oceanic anoxia or sulfate limitation? In: Kesler, S. E. & Ohmoto, H. (eds) Evolution of Early Earth’s Atmosphere, Hydrosphere – Constraints from Ore Deposits. Geological Society of America Memoir, 198, 205–222. Lu, S.-N., Ma, G.-G., Gao, Z. J. & Lin, W.-X. 1985. Primary research on glacigenous rocks of Late Precambrian in China. In: PRECAMBRIAN GEOLOGY EDITORIAL COMMITTEE (ed.) Precambrian Geology No. 1, The Collected Works of Late Precambrian Glacigenous Rocks in China. Geological Publishing House, Beijing, 1 –86 (in Chinese with English abstract). Ma, G.-G. & Wang, Y.-G. 1983. Discussion on the glacial geology of the early Sinian in the area adjacent to Hubei, Hunan, Sichuan and Guizhou Provinces of China. Bulletin of Yichang Institute of Geology and Mineral Resources, Chinese Academy of Geological Sciences, 7, 43– 51 (in Chinese with English abstract). Meng, F.-W., Yuan, X. L., Zhou, C.-M. & Chen, Z.-L. 2003. Dinosterane from the Neoproterozoic Datangpo black shales and its biological implications. Acta Micropalaeontologica Sinica, 20, 97 –102. NCS (NATIONAL COMMITTEE ON STRATIGRAPHY) 2001. Stratigraphic Guide in China and Explanation (revised). Geological Publishing House, Beijing, 59 (in Chinese). Ogg, J. G. & Gradstein, F. M. 2008. The Concise Geologic Time Scale. Cambridge University Press, Cambridge. Qin, Y., Zhou, Q. & Zhang, S. 2005. Elementary properties of manganese of Nanhuan in the Northeastern Guizhou. Guizhou Geology, 122, 246– 251.

365

Richthofen, F. von. 1882. China, 2. Ergebnisse Eigener Reisen und Darauf Gegrundeter Studien, Berlin. Remane, J. 2000. International Stratigraphic Chart, with Explanatory Note. Sponsored by ICS, IUGS and UNESCO (distributed at the 31st International Geological Congress, Rio de Janeiro 2000), 16. Roy, S. 2006. Sedimentary manganese metallogenesis in response to the evolution of the Earth system. Earth-Science Reviews, 77, 273– 305. Sha, Q.-A., Liu, H.-Y., Zhang, S.-S. & Chen, M.-E. 1963. Nantuo tillite in the eastern Yangtze Gorges area. Scientia Geologica Sinica, 3, 139– 148 (in Chinese). Tang, J.-F., Fu, H.-Q. & Yu, Z.-Q. 1987. The stratigraphic horizon, type, and formation condition of Precambrian siliceous iron formation in south China. Ore Geology, 6, 1 – 10 (in Chinese). Tang, S.-Y. & Liu, T.-B. 1999. Origin of the early Sinian Minle manganese deposit, Hunan Province, China. Ore Geology Reviews, 15, 71– 78 (in Chinese). Wang, J. & Li, Z.-X. 2003. History of Neoproterozoic rift basins in South China: implications for Rodinia break-up. Precambrian Research, 122, 141–158. Wang, Y.-L., Lu, S.-N., Gao, Z.-J., Lin, W.-X. & Ma, G.-G. 1981. Sinian tillites of China. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, 386–401. Wang, J., Li, X.-H., Duan, T., Liu, D., Song, B., Li, Z. & Gao, Y. 2003a. Zircon SHRIMP U– Pb dating for the Cangshuipu volcanic rocks and its implications for the lower boundary age of the Nanhua strata in South China. Chinese Science Bulletin, 48, 1663–1669. Wang, T.-G., Wang, C.-J., Zhang, W.-B., Shi, Q., Zhu, L. & Chen, J.-Y. 2003b. Initial organic geochemical investigation on Late Neoproterozoic– Early Cambrian sediments in the Yangtze region, China. Progress in Natural Science, 13, 936– 941. Wang, J.-S., Gan, H.-Y., Wei, Q., Hu, G.-W. & Ge, Q. 2005. Stable isotopes of carbon and sulfur of cap dolomite in the Three Gorges and its mechanism discussion. Geoscience, 19, 14 –20 (in Chinese with English abstract). Wang, J.-S., Jiang, G.-Q., Xiao, S.-H., Li, Q. & Wei, Q. 2008a. Carbon isotope evidence for widespread methane seeps in the c. 635 Ma Doushantuo cap carbonate in south China. Geology, 36, 347– 350. Wang, T.-G., Li, M.-J., Wang, C.-J., Wang, G.-L., Zhang, W.-B., Shi, Q. & Zhu, L. 2008b. Organic molecular evidence in the Late Neoproterozoic Tillites for a palaeo-oceanic environment during the snowball Earth era in the Yangtze region, southern China. Precambrian Research, 162, 317– 326. Willis, B., Blackwelder, E. & Sargent, R. H. 1907. Research in China. Vol. 1 part 1. The Carnegie Institution of Washington, Washington, DC. Xie, Q.-L., Chen, D.-F. & Chen, X.-P. 1999. Characteristics of sedimentary organic matter in Songtao manganese deposits, Guizhou. Acta Sedimentologica Sinica, 17, 280– 284. Yang, Y.-J. 1985. The glacigenous strata of lower Sinian in Hunan Province. In: PRECAMBRIAN GEOLOGY EDITORIAL COMMITTEE (ed.) Precambrian Geology No. 1: The Collected Works of Late Precambrian Glacigenous Rocks in China. Geological Publishing House, Beijing, 224–241 (in Chinese with English abstract). Yang, S.-X. & Pang, K.-T. 2006. Mineralization model for the manganese deposits in northwestern Hunan – an example from Minle manganese deposit in Huayuan, Hunan. Sedimentary Geology and Tethyan Geology, 26, 72 –80. Yang, J.-D., Xue, Y.-S., Sun, W.-G., Tao, X.-C., Wang, Z.-Z. & Zhou, C.-M. 1997. The origin and age of the manganese ore of Nantuo stage. Chinese Science Bulletin, 42, 1538– 1541. Yang, R.-D., Ouyang, Z.-Y., Zhu, L.-J., Wang, S.-J., Jiang, L.-J., Zhang, W.-H. & Gao, H. 2002. A new understanding of manganese carbonate deposits in early Sinian Datangpo stage. Acta Mineralogica Sinica, 22, 329– 336. Yin, C.-Y., Liu, D.-Y., Gao, L.-Z., Wang, Z.-Q., Xing, Y.-S., Jian, P. & Shi, Y.-R. 2003. Lower boundary age of the Nanhua System and the Gucheng glacial stage: evidence from SHRIMP dating. Chinese Science Bulletin, 48, 1657– 1662. Zhang, Q.-R. 1995. The origin of the Sinian Nantuo Formation in Yichang County, Hubei Province. Scientia Geologica Sinica, 30, 147– 152 (in Chinese with English abstract).

366

Q. R. ZHANG ET AL.

Zhang, Q. R. & Piper, J. D. A. 1997. Palaeomagnetic study of Neoproterozoic glacial rocks of the Yangzi Block: palaeolatitude and configuration of South China in the late Proterozoic Supercontinent. Precambrian Research, 85, 173– 199. Zhang, Q.-R., Chu, X.-L., Bahlburg, H., Feng, L.-J., Dobrzinski, N. & Zhang, T.-G. 2003. Stratigraphic architecture of the Neoproterozoic glacial rocks in the ‘Xiang-Qian-Gui’ region of the central Yangtze Block, South China. Progress in Natural Science, 13, 783– 787. Zhang, Q.-R., Li, X.-H., Feng, L.-J., Huang, J. & Song, B. 2008a. A new age constraint on the onset of the Neoproterozoic glaciations in the Yangtze Platform, South China. Journal of Geology, 116, 423– 429. Zhang, S.-H., Jiang, G.-Q. & Han, Y.-G. 2008b. The age of the Nantuo Formation and Nantuo glaciation. Terra Nova, 20, 289– 294. Zheng, Y.-F. 2003. Neoproterozoic magmatic activities and global evolution. Kexue Tongbao, 48, 1705–1720 (in Chinese).

Zhou, C.-M. & Xiao, S.-H. 2007. Ediacaran d13C chemostratigraphy of South China. Chemical Geology, 237, 89– 108. Zhou, C.-M., Tucker, R., Xiao, S.-H., Peng, Z.-X., Yuan, X.-L. & Chen, Z. 2004. New constraints on the ages of Neoproterozoic glaciations in south China. Geology, 32, 437– 440. Zhou, J.-B., Li, X.-H., Ge, W.-C. & Li, Z.-X. 2007a. Age and origin of middle Neoproterozoic mafic magmatism in southern Yangtze Block and relevance to the break-up of Rodinia. Gondwana Research, 12, 184– 197. Zhou, Q., Du, Y.-S., Yan, J.-X., Zhang, M.-Q. & Yin, S.-L. 2007b. Geological and geochemical characteristics of the cold seep carbonates in the Early Nanhua System in Datangpo, Songtao, Guizhou Province. Earth Science – Journal of China University of Geosciences, 32, 845–852 (in Chinese with English abstract). Zhu, M.-Y., Zhang, J.-M. & Yang, A.-H. 2007. Integrated Ediacaran (Sinian) chronostratigraphy of South China. Palaeogeography, Palaeoclimatology, Palaeoecology, 254, 7 – 61.

Chapter 33 Neoproterozoic glaciogenic diamictites of the Tarim Block, NW China MAOYAN ZHU* & HAIFENG WANG State Key Laboratory of Palaeobiology and Stratigraphy, Nanjing Institute of Geology and Palaeontology, Chinese Academy of Sciences, Nanjing, 210008, China *Corresponding author (e-mail: [email protected]) Abstract: Neoproterozoic glaciogenic outcrops are mostly limited to the southwestern and northwestern margins of the Tarim basin and the Tianshan Mountains. Only two glaciogenic diamictites are recorded in the Neoproterozoic successions from the Tielikeli and AksuWusi areas of the Tarim basin. The lower diamictite from the Polong Formation in the Tielikeli area is c. 800 m thick. The upper diamictite in the area is very thin (Yutang diamictite, 10 m; Yulmeinak diamictite, 51 m). In contrast to the Tarim basin, the Neoproterozoic successions along the northern margin of the Tarim Block in the Tianshan Mountains record three or four glaciogenic diamictites, which are well developed and crop out in the Qurugtagh area. Although the glaciogenic nature of the oldest Beiyixi diamictite remains questionable, all other diamictites (Altungol, Tereeken and Hankalchough) exhibit distinct glaciogenic sedimentary features. SHRIMP zircon U –Pb dating demonstrate that the Bayisi diamictite was deposited between 740+7 Ma and 725+10 Ma, the Altungol and Tereeken diamictites between 725+10 Ma and 615+6 Ma, respectively, and the Hankalchough diamictite between 615+6 Ma and c. 542 Ma. Carbonates atop the Altungol, Tereeken and Hankalchough diamictites show distinct C-isotopic values that are typical for those recorded in ‘cap carbonates’ on other continents. The striking similarity between the cap carbonate of the Tereeken and Marinoan-age Nantuo diamictites further support a 635 Ma ‘Marinoan’ age for the Tereeken glaciation. Unlike the 582 Ma Gaskiers glaciations of Newfoundland, the Hankalchough diamictite is possibly ,551 Ma, as suggested by C-isotope chemostratigraphy and biostratigraphy, indicating that post-Marinoan glaciations on different continents may be diachronous. Supplementary material: Data are available at: http://www.geolsoc.org.uk/SUP18469.

Neoproterozoic glaciogenic diamictites on the Tarim Block were first recognized by a Swedish geologist, E. Norin, following a joint Chinese –Swedish expedition 1927–1935 in the Qurugtagh area (Norin 1937). Although several detailed studies have been carried out since and three Neoproterozoic glaciogenic diamictites had already been reported in the early 1980s (see Gao et al. 1985a and references therein), these diamictites only recently received wide attention (Xiao et al. 2004). Various research groups have now investigated the Neoproterozoic successions of the Tarim basin margin, particularly in the Qurugtagh region, focusing on the study of Neoproterozoic tectonics, magmatism, lithologies, geochronology, chemostratigraphy and palaeontology. The present chapter aims to review both old and recent studies, focusing on the stratigraphy of the Neoproterozoic diamictites in four palaeogeographical areas located on or close to the Tarim Block (Fig. 33.1). Using detailed descriptions and recent geochronological and chemostratigraphic data, this chapter discusses correlations and ages of the Neoproterozoic diamictites of the Tarim Block. Published palaeomagnetic data and their palaeogeographic implications are also reviewed.

Glaciogenic diamictites in the Tielikeli area, SW Tarim Basin The Tielikeli area is located at the southwestern margin of the Tarim basin, where Neoproterozoic outcrops are well exposed along the northern fringes of the western Kunlun Mountains (Fig. 33.1(1) & Fig. 33.2) (Ma et al. 1991; Wang et al. 2004). The Neoproterozoic strata and tillites of the area, which had originally been mapped as part of the Devonian Qizlafu Group, were first recognized by the local geologist Ma and his colleagues in 1978 and described in detail a decade later (Ma et al. 1989).

Stratigraphy The Cryogenian and Ediacaran successions in the area (Fig. 33.3) start with c. 230 m of conglomerate (Yalaguz Fm.), which

unconformably overly the Sukulok Group. Interbedded within the massive conglomerate unit are lenticular or bedded feldspathic sandstones and thinly bedded silty mudstone. The conglomerate is dominated by sub-rounded or rounded, well-sorted clasts of chert and porphyry. The Yalaguz Fm. is overlain by the Qakmaklik Group which consists in ascending order of the Polong, Kilix and Yutang formations. The Polong Fm. is dominated by purplish-brown massive diamictite of c. 880 m thick. The diamictite is underlain by c. 200 m of grey laminated cherts and silicified shale. The Kilix Fm. consists of interbeds of purplish-green laminated mudstone and siltstone with thin beds of silty micritic limestone (c. 312 m) in the lower part, purple quartz sandstone (42.9 m) in the middle, and purple interbeds of feldspathic sandstone and conglomerate (86.8 m) at the top. The feldspathic sandstone exhibits large-scale cross-stratification and horizontal stratification. The Yutang Fm. is subdivided into three parts. The lower part (c. 50 m) consists of grey-greenish laminated muddy cherts and mudstone with a conglomerate 0.1–0.5 m thick at the base. The middle part (c. 10 m) consists of purplish brown diamictite. The upper part (121.9 m) of the Yutang Fm. consists of reddish and brownish feldspathic sandstone and conglomerate showing large-scale cross-stratification and wave ripples. Overlying the Qakmaklik Group is the Kurkak Fm., which is composed of four parts. A 10-m-thick laminated dolomite marks the contact between the Qakmaklik Group and the Kurkak Fm. The lower part of the Kurkak Fm. is c. 110 m in thickness and consists of black shale. This is succeeded by 100 m of grey quartz sandstone and conglomerate in the middle, and dark grey shale with interbeds of siltstone and rare phosphorite nodules/ concretions or lamina in the upper part of this unit. The Kzisuhum Fm. is composed of four parts in ascending order. The lower interval is 55.6 m thick and consists of dark green quartz sandstone with thin interbeds of feldspathic sandstone and pyrite-bearing siltstone in the lower part, and dark grey siltstone, greyish-green sandstone and grey coarse-grain quartz sandstone with three phosphorite beds in the upper part. The overlying interval comprises 86.1 m of yellowish brown sandy dolostone with interbeds of purple siltstone and minor sandy siderite

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 367– 378. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.33

368

M. ZHU & H. WANG

Fig. 33.1. Geological map of the Tarim Block and adjacent areas (modified from Lu et al. 2008). Key: PH, Phanerozoic rocks; NP, Neoproterozoic rocks; MP, Mesoproterozoic rocks; PP, Palaeoproterozoic rocks; AR, Archaean rocks; RNP, early Neoproterozoic granitoids; F, Faults; IF, Inferred faults; Diamictites, Cryogenian and Ediacaran diamictites; Q, Quaternary desert and sedimentary deposits; Square: 1, Tielikeli area, SW Tarim Basin; 2, Aksu-Wusi area, NW Tarim Basin; 3, Quruqtagh area, southern margin of Eastern Tianshan Mts.; 4, Guozigou-Keguqing Mt. area, western range of Northern Tianshan Mountains.

layers. Approximately 52.2 m quartz sandstone with large-scale planar cross-stratifications overlies the dolostone interval. The upper part ofthe Kzisuhum Fm. is dominated by dolostone (177.4 m). The contact to the overlying Devonian strata is sharp and unconformable.

Description of the diamictites There are two diamictite intervals in the Tielikeli area. Both are interpreted as glaciogenic diamictites or tillites. The lower diamictite interval of the Polong Fm. is dominated by massive diamictite with 5 –30% randomly organized, angular, sub-angular or sub-rounded clasts of porphyry, quartz porphyry, chert, quartzite,

dolostone, granite, gneisses, diabase and so on. Clasts are dominated by pebbles of 2 –10 cm in diameter, with a maximum diameter of 100 cm. Glaciogenic features include striated marks, snicks and notches on the surface of cobble and pebble size clasts, and parallel cracks of clasts. The diamictite at the base and top of the interval is characterized by 30–50% poorly sorted clasts with traces of bedding. Laminated mudstone and siltstone with dropstones are observed in the middle and top of the interval. The upper diamictite interval of the Yutang Fm. is only 10 m thick, showing similar sedimentary features and a similar clast composition as the lower diamictite. Similar dropstone-bearing laminated mudstone also occurs at the top of the upper diamictite. It is interesting to note that both diamictite intervals are underlain by an interval consisting of grey laminated muddy

Fig. 33.2. Geological map showing outcrop of the Tielikeli area, SW Tarim Basin (modified from Ma et al. 1989).

THE TARIM BLOCK, NW CHINA

369

Stratigraphy The Neoproterozoic succession in the area (Fig. 33.5) consists of the Qiaoenblaq and Wushinanshan groups; the former overlies the metamorphic rocks of the Aksu Group unconformably. The Aksu Group is older than 962 + 12 Ma (whole-rock Rb –Sr dating) and is dominated by blueschist (Liou et al. 1989; Gao et al. 1993). The Qiaoenblaq Group, overlying the Aksu blueschists, consists of (in ascending order) the Xifangshan, Dongqiaoenblaq, Muyangtan and Dongwu formations. The Xifangshan Fm. is c. 1715 m thick and is characterized by rhythmic beds of grey-green feldspathic sandstone, feldspathic quartz sandstone and siltstone. The rhythmites exhibit sandstone with grading features in the lower part, and siltstone with horizontal and crossstratification in the upper part. The rhythmic beds range from 1 cm to 4 m in thickness, but are predominantly 50– 100 cm. They have been interpreted as gravity flow deposits or proximal turbidites (Gao et al. 1993). The Dongqiaoenblaq Fm. consists of 15 –20-m-thick grey-green conglomerate with interbeds of sandstone in the lower and massive diamictite with a maximum thickness of 70 m in the upper part. Overlying the diamictite is the Muyangtan Fm., which is c. 107 m thick and is composed of grey-green calcareous siltstone and quartz sandstone with parallel and cross-stratification. The overlying Dongwu Fm., c. 176 m thick, is composed of grey-green and grey-brownish coarse-grained feldspathic sandstone. There is a 1– 2-m-thick conglomerate bed at the base of the Dongwu Fm. The conglomerate is dominated by sub-rounded or rounded cobbles and pebbles of granite. The Wushinanshan Group consists of the Yulmeinak, Sugetblaq and Qigeblaq formations. The Yulmeinak Fm. overlies sandstone of the Dongwu Fm. above an angular unconformable contact and is composed of purplish massive diamictite with a maximum thickness of 70 m, but the diamictite gradually decreases and eventually pinches out towards the east. The overlying Sugetblaq Fm. consists of two members. The lower member is c. 728 m thick is and composed of brownish coarse-grained feldspathic sandstone with interbeds of yellow-green siltstone and sandstone in the lower part, and red, grey and brownish ferrous quartz sandstone with thin interbeds of hematite and a 15-m-thick unit of peridotite basalt in the upper part. The uppermost part of the lower member of the Sugetblaq Fm. is characterized by 79 m of dark green and brownish amygdaloidal olivine basalt. The upper member of the Sugetblaq Fm. is composed of brownish and greyish thin-bedded sandy limestone with edgewise clasts and interbedded thin siltstone in the lower part and yellow to green, thin-bedded, glauconitic feldspathic sandstone in the upper part. The overlying Qigeblaq Fm. is a characteristic interval of thick-bedded dolostone with a thickness of 186 m. This unit is overlain by early Cambrian cherts and phosphorite, which contain small shelly fossils (Gao et al. 1981; Qian & Xiao 1984; Yao et al. 2005).

Fig. 33.3. Composite stratigraphic log of the Cryogenian and Ediacaran in the Tielikeli area, SW Tarim Basin (modified from Ma et al. 1989). Key as in Figure 33.10.

cherts and silicified shale, demonstrating a similar depositional sequence.

Glaciogenic diamictites in the Aksu-Wusi area, NW Tarim Basin Neoproterozoic outcrops exposed in the Aksu-Wusi area, NW Tarim basin, are easily accessible (Figs 33.1, 33.2 & 33.4). Glaciogenic diamictite in the area was first documented in local geological reports during the 1950s and has been well studied by Gao & Qian (1985) and Gao & Chen (1993).

Description of the diamictites As in the Tielikeli area, there are two diamictite intervals interpreted as glaciogenic in the Aksu-Wushi area. The lower diamictite interval of the Dongqiaoenblaq Fm. is composed of 20–80% randomly organized sub-rounded clasts of felsite, dacite, granite, jasper, quartzite, sandstone, andesite, diabase and so on. Clasts are dominated by c. 10-cm-sized pebbles with a maximum diameter of 60 cm. The matrix is composed of feldspathic sandstone and siltstone with a mud component, showing soft deformation features and contorted bedding in some beds. Although clasts are randomly organized, and striated marks have been observed on the surface of some clasts from the diamictite of the Dongqiaoenblaq Fm., the high degree of sorting, contorted bedding and subrounded shapes of the clasts, which are similar in lithology to the clasts from the underlying conglomerate, led Gao et al.

370

M. ZHU & H. WANG

Fig. 33.4. Geological map showing the Cryogenian and Ediacaran outcrop of the Aksu-Wusi area, NW Tarim Basin (modified from Gao et al. 1985b).

(1993) to reinterpret the diamictite as having been rapidly deposited from submarine debris flows. The upper diamictite interval of the Yulmeinak Fm. is a discontinuous interval with an unconformable basal surface. The diamictite shows distinctly glaciogenic characteristics, including a large abraded surface at its base with notching, scouring and striations, which is a clear indication of ice movement or glacial pavement. The diamictite is massive and matrix-supported with 20– 30% clasts of various lithologies including granites and volcanic rocks, blueschists, sandstone and siltstone originally derived from the underlying strata. The clasts vary in size, with the largest ones at the base (maximum diameter of 6 m), and show a poor degree of sorting. Dominant clasts are subangular with a muddy or calcareous envelope, but characteristic selliform (saddle-shaped) and other irregularly shaped clasts are also observed. Striation marks, nicks and notches are common on the surface of clasts. Dropstones of various lithologies occur in laminated siltstone in the middle of the diamictite interval.

Glaciogenic diamictites in the Quruqtagh area, southern margin of the eastern Tianshan Mountains The Quruqtagh area is located at the northeastern margin of the Tarim basin and belongs to the southern margin of the Eastern Tianshan, covering an area of about 30 000 km2 (Figs 33.1– 33.3). The Quruqtagh area is customarily subdivided into three districts. The area west of Xindi is called the East Quruqtagh, the area between Xindi and Xinger is the Middle Quruqtagh, while the area east of Xinger is the West Quruqtagh. The Neoproterozoic outcrops and glaciogenic diamictites are well exposed in all three districts (Fig. 33.6) and were first reported by a Swedish geologist, E. Norin, after a joint 1927 –1935 Chinese – Swedish expedition in the area (Norin 1937). The Neoproterozoic stratigraphy of the area has been well established by local geological surveys during the 1950s and has been documented in detail by Gao & Zhu (1984). Since then, the latest Neoproterozoic successions have been investigated using various methods (e.g. Zhong & Hao 1990; Xiao et al. 2004; Duan et al. 2005; Huang et al. 2005; He et al. 2007a; Liu et al. 2007; Zhang et al. 2007; Shen et al. 2008; Kou et al. 2008; Zhu et al. 2008; Xu et al. 2009).

Stratigraphy The latest Neoproterozoic successions in the Quruqtagh area (Fig. 33.7) are known as the Quruqtagh Group, and overlie a

c. 800-m-thick carbonate unit, the Paergangtagh Fm., with an unconformable contact (Zhao et al. 1985; Gao et al. 1993). The Quruqtagh Group is subdivided (in ascending order) into nine formations: the Bayixi, Zhaobishan, Altungol, Huangyanggou, Tereeken, Zhamoketi, Yukkengol, Shuiquan and Hankalchough formations (Gao & Zhu 1984; Xiao et al. 2004; Kou et al. 2008). The best continuous outcrops are located in the Middle Quruqtagh area. The Bayixi Fm. This consists predominantly of shallow-water marine siliciclastic and volcanic rocks with several intervals of diamictite. The overall thickness of the formation ranges from 640 to 1670 m. It starts with a basal conglomeratic interval with predominant clasts of dolostone and limestone from the underlying Paergangtagh Fm. The clasts are angular to subangular in shape, with the largest clasts being 30 cm in diameter. The lower part of the Bayixi Fm. consists of volcanic rocks, slates and sandstones associated with diamictite. The middle part of the formation consists of sandstone, and siltstone interbedded with diamictite and slates. The upper part of the formation consists of two diamictite intervals of c. 200 m thickness each, separated by interbedded slates and siltstones. The Zhaobishan Fm. This is a characteristic interval with thick-

nesses ranging from 358 to 570 m, consisting predominantly of sandstones and calcareous siltstones in the lower part and slates with interbedded thin calcareous siltstones and sandstones in the upper part. The sandstones in the lower part of the formation are lithic, coarser-grained (0.2 –2 mm), better rounded and sorted and show large cross-bedding, whereas sandstones in the upper part are more fine-grained (0.1 –1 mm) and show a lower degree of textural maturity. The overall sequence of the Zhaobishan Fm. represents a deepening-upward sequence. The Altungol Fm. This is characterized by three massive diamictite

intervals associated with sandstones, siltstones and shales. The thickness of the formation increases from East Quruqtagh (764 m) to West Quruqtagh (2395 m). In the South Quruqtagh area, the formation is only 8 m thick. The basal and upper massive diamictites mark the basal and top boundaries of the Altungol Fm. The thickness of the three diamictite intervals ranges from 50 to 100 m. Associated sandstones, siltstones and shales exhibit rhythmic beddings with average bed thicknesses of 40–60 cm.

THE TARIM BLOCK, NW CHINA

371

The Huangyanggou Fm. This is a newly erected formation representing an interval between the uppermost diamictite of the Altungol Fm. and the diamictite of the Tereeken Fm. (Cao 1991; Kou et al. 2008). The interval without diamictite was originally the upper part of the Altungol Fm., and was interpreted to represent an interglacial interval between the Altungol and Tereeken glaciations (Kou et al. 2008). The base of the Huangyanggou Fm. is marked by c. 2 m of grey-purplish massive micritic dolostone, which represents a cap carbonate above the Altungol diamictite. The lower part of the formation consists of meta-andesite, clinkering breccia and silty tuff, while the upper part consists of silty mudstone and muddy siltstones with thin interbeds or lenses of limestone. A 1–3-m-thick limestone unit marks the top boundary of the Huangyanggou Fm., underlying the base of the Tereeken Fm. A c. 50-m-thick dolostone interval reported from the Altungol Fm. by Xiao et al. (2004) at the Xishankou section may correspond to the Huangyanggou Fm. The Tereeken Fm. This is characterized by grey massive diamictites with thicknesses ranging from 689 to 1845 m. The diamictite is separated into several units by thin interbeds of finely laminated silty mudstones and carbonate. The Zhamoketi Fm. This is a siliciclastic interval consisting of metre-scale rhythmites of sandstone, siltstone and mudstone, with thicknesses ranging from 297 m to 793 m. The base of the formation is marked by a 2 –10-m-thick unit of dolostone with characteristics of basal Ediacaran cap carbonates, overlying the Tereeken diamictites. The top of this formation is marked by a c. 20– 80-m-thick diabase sill. The rhythmites have been interpreted as turbidites, showing distinct structures of the Bouma sequences deposited on a continental slope (Li & Dong 1991; Kou et al. 2008). The rhythmites exhibit grading at their base with overlying laminated and wavy or cross-bedded intervals; convoluted laminations or soft-deformation have been observed within some beds. Sole marks such as flute casts, groove casts and load casts are common at the base of the rhythmites. The Yukkengol Fm. This consists of green-greyish finely laminated

siltstones and shales with thicknesses ranging from 80 to 583 m, representing condensed sequences of deeper- water facies. The Shuiquan Fm. This is characterized by laminated carbonates

and silty shales. Its variable thicknesses between 22 and 307 m was interpreted to result from erosion at the top of the formation (Gao & Zhu 1984). Centimetre-scale ribbonites, stromatolites and microbially laminated dolostone were reported in the Shuiquan Fm., suggesting sedimentation in the euphotic zone (Xiao et al. 2004). The wormiform carbonaceous fossils – filaments c. 2 mm wide and up to 40 mm long – from the Shuiquan Fm. reported by Gao et al. (1985) were reinterpreted as Vendotaenid fossils (Xiao et al. 2004). The Hankalchough Fm. This consists of light grey diamictite. Its thickness decreases eastward from 467 m to less than 10 m. A dolostone unit (1 –6 m thick) at the base of the Xishanblaq Fm. marks the top of the Hankalchough Fm. The Xishanblaq Fm. consists of black cherts and cherty phosphorites associated with volcanic rocks. Cambrian microfossils have been reported from the phosphorites (Yao et al. 2005).

Description of the diamictites

Fig. 33.5. Composite stratigraphic log of the Cryogenian and Ediacaran in the Aksu-Wusi area, NW Tarim Basin (after Gao et al. 2005b). Key as in Figure 33.10.

The four major intervals of diamictite in the Quruqtagh Group have been assigned to three glaciations (Gao & Zhu 1984). Below is a summary of these four diamictites based on detailed descriptions by Gao & Zhu (1984), Xiao et al. (2004) and Kou et al. (2008).

372

M. ZHU & H. WANG

Fig. 33.6. Geological map showing the Cryogenian and Ediacaran outcrop of the Quruqtagh area, southern margin of the Eastern Tianshan Mountains (modified from Gao & Zhu 1984).

The Bayixi diamictite. The Bayixi diamictites are distributed across only a limited area largely near the Middle Quruqtagh. The thickness of the diamictite units in the Bayixi Fm. varies from a few metres to more than 100 m. Poorly sorted and angular clasts (5 – 20%) occur in the silty and muddy matrix of the diamictites (see Supplementary Material, Fig. 1S-a). Clasts are mostly derived from older metamorphic and igneous rocks, including gneiss, granite, diabase, dolostones or marbles, and quartz; siliciclastic clasts are rare. Clast sizes are predominantly ,6 cm in diameter. Some boulders are perpendicularly or obliquely oriented relative to bedding plane. The Bayixi diamictites were interpreted as glaciomarine in origin by Gao & Zhu (1984), supported by the variety of lithic compositions, the angular and irregular shapes of the clasts, scratch marks and non-unidirectional striations on the clast surfaces. However, glacial striations were not confirmed by Xiao et al. (2004). In addition, no cap carbonate has been reported. The Altungol diamictites. Based on Kou et al. (2008), there are three

diamictite units within the newly defined Altungol Fm. However, as documented by Gao & Zhu (1984), the Altungol diamictite units appear unevenly distributed because only one unit has been observed in some sections. The Altungol diamictites are massive, have a matrix of low maturity and include clasts of quartzite, granite, cherts and quartz. The clasts are generally .1 cm with the largest being c. 15 cm in diameter (see Supplementary Material, Fig. 1S-b). Clasts with stepped fractures and glaciogenic striations have been observed (Kou et al. 2008). Gao & Zhu (1984) claimed that the Altungol diamictites have a closer affinity with the overlying Tereeken diamictites and argued that the Altungol and Tereeken diamictites jointly represent a mid-Quruqtagh ice age. However, Xiao et al. (2004) suggest the alternative interpretation that the Altungol diamictites resemble more closely the Bayixi diamictites in clast composition, degree of metamorphism and thickness of individual diamictite beds. Xiao et al. (2004) argued that if all the diamictites in the Bayixi and Altungol formations were found to be glaciogenic, it would be more likely that the Altungol and Bayixi diamictites represent pulses of a single ice age. Xiao et al.’s (2004) hypothesis is further supported by the observations of Kou et al. (2008) who confirmed the existence of an interglacial interval, the Huangyanggou Fm., separating the Altungol diamictites from the Tereeken ice age. A 2 –4-m-thick, grey-purplish carbonate unit with negative values of d13C at the base of the Huangyanggou Fm. was considered to be a cap carbonate above the Altungol diamictites (Kou et al. 2008). However, in opposition to Xiao et al.’s (2004) hypothesis, Kou et al. (2008) assigned the Altungol diamictites to a distinctly separate ice age.

The Tereeken diamictite. The Tereeken diamictite is the most con-

tinuous and distinct interval in the Quruqtagh area, showing unambiguous glaciogenic features, striated clasts and dropstones throughout. Clasts in the diamictites are dominated by carbonate, granite and other igneous rocks; marble and siliciclastic clasts are more abundant in comparison to the Bayixi and Altungol diamictites (see Supplementary Material, Fig. 1S-c). Striated clasts are common (Fig. 1S-d). Three to eight individual diamictite units can be recognized within the Tereeken Fm. (Gao & Zhu 1984). These units are intercalated with non-glaciogenic, laminated silty marine mudstones and carbonates. Several thin carbonate laminae (,10 mm thick and laterally continuous for more than 100 m) occur within the silty mudstone in the lower Tereeken Fm. at the Heishan –Zhaobishan section (Xiao et al. 2004). These laminae consist of vertically oriented, upward-growing calcite crystals, and are typically inundated by overlying siltstone. In addition, a bedded carbonate unit occurs in the lower Tereeken Fm. between two massive diamictite units, at the same section. As suggested by Xiao et al. (2004), the existence of laminated siltstones and bedded dolostones between Tereeken diamictite units indicates that there was an active hydrological system during the Tereeken ice age. A 2–10-m-thick carbonate unit above the Tereeken diamictites represents a distinct lithostratigraphic marker, and was recognized as the ‘Yukkengol limestone’, defining the boundary between the Tereeken diamictite and overlying strata (Norin 1937). The same boundary was subsequently adapted to define the boundary between the Lower and Upper Sinian Systems (Gao & Zhu 1984). The carbonate unit in the Yukkengol area has been described in detail as a cap carbonate atop the Tereeken diamictite by Xiao et al. (2004). According to observations from a section between Xidashan and Moheshan, similar to the Yukkengol area, the carbonate unit consists of three subunits (see Supplementary Material, Fig. 1S-e, f ). The lower subunit (c. 1.5 m) is a massive dolostone with angular allochthonous clasts of various compositions. The middle subunit (c. 2.5 m) consists of thinly bedded, finely laminated dolostone with thin interbeds of shales. The upper subunit (c. 1 m thick) consists of medium-bedded dolostone. It is interesting to note that the uppermost part of the Tereeken diamictite underlying the cap carbonate shows well-defined bedding and represents deposits of a deglaciation interval. The Hankalchough diamictite. The Hankalchough diamictite is composed of light grey and greenish-grey massive diamictite (see Supplementary Material, Fig. 2S-a, b, c). The clast composition differs from that of the underlying three diamictite intervals,

THE TARIM BLOCK, NW CHINA

373

consisting of angular clasts dominated by carbonate and granite. Siliciclastic clasts from underlying strata are abundant. Some carbonate boulders contain overturned stromatolite, probably derived from older stromatolitic dolostones in the Paergangtagh Group, which underlies the Quruqtagh Group (Zhao et al. 1985). The Hankalchough diamictite was originally considered to be terrestrial in origin (Gao & Zhu 1984), but has been reinterpreted as glaciomarine by Xiao et al. (2004) due to the occurrence of dropstones. A distinct 1– 5-m-thick grey-greenish or light grey, finely laminated muddy-silty dolostone or calcareous mudstone unit above the Hankalchough diamictite (see Supplementary Material, Fig. 2S-d) contains common lithic fragments and small clasts (3 –5%) with disseminated pyrite granules (5 –10 mm). It was originally interpreted as glaciolacustrine in origin (Gao & Zhu 1984) and reinterpreted as an atypical type of cap dolostone (Xiao et al. 2004).

Glaciogenic diamictites in the Guozigou-Keguqingshan area, western range of North Tianshan Mountains The Guozigou-Keguqingshan area belongs to the western range of the North Tianshan Mts. (Figs 33.1– 33.4). The well-exposed late Precambrian successions in the area (Fig. 33.8) is composed of three intervals of diamictite, which were first recognized by local geologists in 1976 and subsequently documented in detail by Wang et al. (1983) and Gao et al. (1985b) based on fieldwork in 1978 –1979. Since then, no more new information has been published.

Stratigraphy The Cryogenian and Ediacaran Kailaketik Group in the area (Fig. 33.9) is c. 700–1000 m thick and is subdivided into six formations (in ascending order): the Kulutieliekti, Tulasu, Biexibastao, Keyindi, Tarqat and Talisayi formations. The outcrop quality and thicknesses of the succession in the area vary from west to east. The lower part of the succession is well exposed in the Kuguqingshan area and the upper part of the succession is well developed in the Guozigou area. The succession is overlain disconformably by the Lingkuanggou Fm., which contains early Cambrian phosphatized small shelly fossils, and is underlain by the carbonate of the Kusongmqiek Group with unconformable contact. The Kulutieliekti Fm. This is c. 115 m thick. A basal conglomerate interval of 7.2 m is composed predominantly of limestone and dolostone clasts derived from the underlying Kusongmqiek Group. The well-sorted clasts have rounded or sub-rounded shapes with diameters ranging from 1 to 25 cm. The basal conglomerate is overlain by a 9-m-thick greenish tuff bed. The tuffaceous bed contains c. 30% volcanic clasts (0.03 –0.3 cm in diameter) and carbonate clasts (0.5 –1.5 cm in diameter). The subsequent strata are composed of dark-greyish, massive amygdaloidal basalt with a thickness of c. 90 m; the basalt occasionally shows subangular carbonate and volcanic clasts (2 –5 cm in diameter). Overlying the basalt is a 9.4-m-thick greyish, massive diamictite. The Tulasu Fm. This is composed of laminated, black, carbon-

aceous, silicified mudstone with intercalations of black siltstone with a thickness of 105 m. The siltstone layers increase up-section and form rhythmic thin mudstone –siltstone couplets. Organic microfossils are reported from this interval. Fig. 33.7. Composite stratigraphic log of the Cryogenian and Ediacaran in the Quruqtagh area, southern margin of the Eastern Tianshan Mountains (after Gao & Zhu 1984; He et al. 2007; Kou et al. 2008). Legends as in Figure 33.10.

The Biexibastao Fm. This consists of a lower member (25 m) of silt-

stone and sandstone and an upper member of diamictite (30 m). Limestone beds occur within the well-bedded siltstone and sandstone interval.

374

M. ZHU & H. WANG

Fig. 33.8. Geological map showing the Cryogenian and Ediacaran outcrop of the Guozigou-Keguqingshan area, western range of Northern Tianshan Mountains (modified from Wang et al. 1983).

The Keyindi Fm. This is a thin interval in the Kuguqing Mountains

area and becomes thicker in the Guozigou area, where it is more than 115 m thick, but the basal part is unexposed. The formation is composed of laminated, dark-greyish muddy siltstone and black silty mudstone. The Tarqat Fm. This is also composed predominantly of laminated,

muddy siltstone but with grey-greenish or purplish colour and cross-stratification in some layers. Thin lenticular limestone beds (0.7– 3 cm) occur in some sections. The Talisayi Fm. This is a distinct interval consisting of greyish or purplish, massive diamictite with a thickness ranging from 79 to 424 m. Several dark-greyish, laminated siltstone layers are recognized within the diamictite. The diamictite is overlain by Cambrian phosphorite of the Linkuanggou Fm., which contains early small shelly fossils and trilobites. A conglomerate bed at the base of the Linkuanggou Fm. and an irregular contact marks the unconformity between the Talisayi and the Linkuanggou formations.

Description of the diamictites The three diamictite intervals in the area were first interpreted by Wang et al. (1983) as representing three glaciations. These intervals were used for correlation with the Quruqtagh area. The Kulutieliekti diamictite. This diamictite consists of 40– 50%

clasts of dolostone, limestone, marbles, cherts and basalt. The diameter of these clasts varies from 1 to 25 cm, and the largest diameter reaches 35 cm. However, the diamictite lacks convincing glaciogenic features, and its glaciogenic provenance remains uncertain. The Biexibastao diamictite. This is not a continuous interval but

often changes laterally to silty dolostone. The diamictite shows distinct glaciogenic features, such as striated clasts. The diamictite contains 5 –20% clasts with average diameters ranging from 0.2 to 15 cm. The diameters of the largest clasts reach up to 100 cm. The clasts are generally sub-angular and sub-rounded, dominated by carbonate, chert and volcanic rocks. The laterally equivalent silty dolostone also contains 5% angular or subangular clasts with average diameters of 0.3–3 cm. The diameters of the largest clasts within the dolostone range up to 20 cm. Similar to the laterally equivalent diamictite, the clast composition is also dominated by carbonate. The associated mudstone, which occurs either below the diamictite or is laterally equivalent, contains 5–25% sand grains and 5% angular or sub-angular clasts with

an average diameter of 0.2–5 cm. The large clasts (up to 13 cm in diameter) exhibit characteristics of dropstone within the finely laminated mudstone, providing solid evidence of glaciogenic provenance of these diamictite, dolostone and mudstone units. The Talisayi diamictite. This contains 10% clasts with an average diameter ranging from 0.2 –15 cm. Diameters of the largest clasts range up to 100 cm. The clasts consist of predominantly carbonate, chert and quartz, and rare clasts of granite and other volcanic or metamorphic rocks. The shapes of the clasts vary from round to angular. Some clasts show striation marks, rolling cracks or lacunules on the surface. Based on the sedimentary features described above, the Biexibastao and Talisayi diamictites have been considered by Wang et al. (1983) to be marine glaciogenic diamictites.

Correlation Based on the stratigraphic descriptions of the four type areas, the successions in the Tielikeli and Aksu-Wusi areas in the Tarim Block show significant similarities. Similar to the wellstudied Cryogenian and Ediacaran successions of the Yangtze Block, only two glaciation intervals are developed, and the uppermost part of the succession is composed of thick carbonate. The C-isotope variation of the upper Sugetblaq and Qigeblaq formations in the Aksu-Wusi area is similar to that at the top Doushantuo and Dengying formations in the Yangtze Block (He et al. 2007a; Zhan et al. 2007). In particular, a major negative excursion in the lower part of the Upper Member of the Sugetblaq Fm. can be correlated with the ‘DOUNCE’ excursion at the top of the Doushantuo Fm. (Zhu et al. 2007). The chemostratigraphy supports the correlation of the Yulmeinak glaciation with the Nantuo glaciation (635 Ma), even though there is no distinct cap carbonate recorded atop the Yulmeinak diamictite. While the successions in the Quruqtagh and GuozigouKeguqingshan areas in the Tianshan Mountains are similar to each other, they show significant differences from those in the Tielikeli and Aksu-Wusi areas. At least three glaciogenic diamictites with possible cap carbonates are recorded in the area. The cap carbonates atop the Altungol and Tereeken diamictites show characteristic C-isotopic features (Xiao et al. 2004; Kou et al. 2008; Shen et al. 2008). However, C-isotope values from the cap carbonate atop the Hankalchough diamictite vary dramatically from section to section (Xiao et al. 2004), demonstrating great differences when compared with the cap carbonates atop the Altungol and Tereeken diamictites. The reason for the spatial

THE TARIM BLOCK, NW CHINA

375

variation of C-isotope values of the cap carbonate atop the Hankalchough diamictite remains problematic. Nevertheless, the distinct association of glaciogenic diamictites and cap carbonates provides strong evidence for correlation of these diamictites. In particular, the cap carbonate above the Tereeken diamictite (see Supplementary Material, Fig. 1S) shows similar sedimentary features to that of the Nantuo-Marinoan diamictite in other continents, supporting their correlation. Furthermore, C-isotope chemostratigraphy also supports a post-Nantuo or post-Cryogenian age of the Hankalchough diamictite, because the Dounce-Shuram-Wonaka excursion, which is regarded as a distinct global mid-Ediacaran event, has been recorded from the Shuiquan Fm. (Xiao et al. 2004). Detailed integrated correlations of the Cryogenian and Ediacaran successions of these four areas are given in Figure 33.10. It should be noted that biostratigraphic correlations using organicwalled microfossils have been intensively adapted for correlation of the Neoproterozoic successions in the literature. Because taxonomic problems and restricted occurrences resulted from poor preservation, biostratigraphic correlation of organic-walled microfossils remains problematic (Gao et al. 1985a; Zhong & Hao 1990). Magnetostratigraphy may support an improved correlation of the Neoproterozoic successions; however, because of complex and long-lasting tectonic processes, the original magnetic polarity can only rarely be discerned. Correlations based on previous palaeomagnetic studies (see Gao & Zhu 1984; Gao et al. 1985a, 1993) are inconsistent with other correlations and are thus not yet considered reliable.

Geochronological constraints on the diamictites

Fig. 33.9. Composite stratigraphic log of the Cryogenian and Ediacaran in the Guozigou-Keguqingshan area, western range of Northern Tianshan Mountains (after Wang et al. 1983; Gao et al. 1985a). Legends as in Figure 33.10.

Volcanic rocks and tuffaceous beds are well developed within the Cryogenian and Ediacaran sequences in the Tarim Block. Three major volcanic episodes are recorded particularly in the Quruqtagh area, and have been termed (in ascending order) the Beiyixi, Zhamoketi and Shuiquan episodes by Gao et al. (1985a). These volcanic rocks and tuffaceous beds, including basalt, diabase, andesite, rhyolite and quartz porphyry, provide potential samples for radiometric dating. Numerous radiometric ages have been published in past decades (Lu et al. 1985; Gao et al. 1993; Xu et al. 2005, 2009 and references therein). However, as these ages were based on different radiometric methods, it is difficult to adapt them for stratigraphic correlation. The most reliable SHRIMP zircon U –Pb ages from the Qurugtagh area published by Xu et al. (2009) provide constraints on the four diamictites in the area, suggesting that the Bayixi diamictite was deposited between 740+7 Ma and 725+10 Ma, the Altungol and Tereeken diamictites between 725+10 Ma and 615+6 Ma, and the Hankalchough diamictite between 615+6 Ma and c. 542 Ma. These ages indicate that the oldest (Beiyixi) glaciation in the Tarim Block is younger than 740 Ma. Although we are not sure whether the Beiyixi glaciation represents the oldest Neoproterozoic glaciation on a global scale, however, geochronological data from the magmatic rocks underlying all the lowest glaciogenic diamictites in the Tarim Block suggest that the oldest glaciation should be younger than 780 Ma. These ages were reported by Zhang et al. (2009), including a U–Pb zircon age of 773+3 Ma for the mafic dyke swarms intruding the c. 820 Ma granite in the Quruqtagh area, a U–Pb zircon age of 759+7 Ma for the mafic dyke swarms intruding the blueschist of the Aksu Group of the Aksu-Wusi region, and a U–Pb zircon age of 815+57 Ma for a mantle-sourced gneissic granite in the Tielikeli area. Nevertheless, no sedimentary evidence of glaciation overlying these 760–780 Ma magmatic rocks and underlying the Beiyixi diamictite rules out glaciations(s) between 740 Ma and 780 Ma in the Tarim Block. The reported ages also support the correlation of the Tereeken glaciation with the c. 635 Ma Nantuo glaciation, and that the Hankalchough diamictite represents a younger Ediacaran glaciation. However, whether the Hankalchough and Talisayi

376

M. ZHU & H. WANG

Fig. 33.10. Stratigraphic correlation of the Cryogenian and Ediacaran successions in the Tarim area. Key: 1, dolostone; 2, limestone; 3, conglomerate; 4, coarse-grained sandstone; 5, sandstone; 6, siltstone; 7, diamictite; 9, bedded diamictite; 8, shale and mudstone; 10, silicates or cherts; 11, phosphorite; 12, blueschist; 13, volcanic rocks; 14, dropstones.

diamictites are time-correlative to the 582 Ma Gaskiers glaciation remains poorly constrained. Based on the C-isotope data from the Shuiquan Fm. in the Qurugtagh area (Xiao et al. 2004), it is proposed here the Hankalchough diamictite is younger than 551 Ma, because the negative C-isotopic data from the Shuiquan Fm. exhibit distinct values of the DOUNCE excursion at the top of the Doushantuo Fm. of the Yangtze Block. Strata at the top of this excursion have a high-resolution zircon U – Pb age of 551 Ma (Condon et al. 2005). There is no doubt that Hankalchough and Talisayi diamictites represent an end-Ediacaran glaciation rather than a Cambrian one, because Cambrian small shelly fossils have been reported from the strata overlying the Hankalchough and Talisayi diamictites (Wang et al. 1983; Qian 1999; Yao et al. 2005).

Palaeolatitudes and palaeogeography Systematic palaeomagnetic analyses of the samples from Cryogenian and Ediacaran strata of the Aksu-Wusi, Quruqtagh and Guozigou-Keguqinshan areas indicate that the palaeolatitudes of the three areas are ,278 (Table 33.1), which supported the lowlatitude glaciations during the late Neoproterozoic put forward by the Snowball Earth hypothesis. These palaeomagnetic data, combined with new data from the mafic dyke swarms (c. 810– 820 Ma) intruding the strata below the Cryogenian successions in the Tielikeli (328N; Zhang et al. 2004) and Aksu-Wusi (43 + 68N; Chen et al. 2004) areas, collectively suggest a close palaeogeographic affinity of the Tarim and Yangtze plates with Australia and Antarctica during the late Neoproterozoic,

THE TARIM BLOCK, NW CHINA

Table 33.1. Palaeolatitude data from four areas of the Tarim Block (Gao et al. 1993; Huang et al. 2005; Zhan et al. 2007) Guozigou-Keguqinshan Tarqat Fm. Kulutieliekti Fm.

8.18 6.68 10.38

Quruqtagh Shuiquan Fm. Zhamoketi Fm. Tereeken Fm. Beiyixi Fm.

Aksu-Wusi 21.78 21.98 19.98 0.98

Sugetblaq Fm. Yulmeinak Fm.

24.38 c. 278 24.18

supporting the supercontinental configuration and subsequent rifting of Rodinia. This work benefited from discussions with many colleagues, including C. Zhou, G. Li, B. Xu, S. Xiao and G. Shields. Fieldwork was jointly carried out with G. Li, F. Zhao, X. Zhao and Z. Yin. This project is supported by the National Natural Science Foundation of China (nos 40715005 and 40930211). This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) project #512.

References Cao, R. 1991. New observation of the Sinian System in the southern Yardang Mountains, Xinjiang. Regional Geology of China, 1991, 30 – 34 (in Chinese with English abstract). Chen, Y., Xu, B., Zhan, S. & Li, Y. 2004. First mid-Neoproterozoic palaeomagnetic results from the Tarim Basin (NW China) and their geodynamic implications. Precambrian Research, 133, 271– 281. Condon, D., Zhu, M., Bowring, S., Wang, W., Yang, A. & Jin, Y. 2005. U –Pb ages from the Neoproterozoic Doushantuo Fm., China. Science, 308, 95 – 98. Duan, J. Y., Xia, D. X. & An, S. L. 2005. Deep-water sedimentation and tectono-palaeogeography of the Neoproterozoic – early Palaeozoic aulacogen in Kuruktag, Xingjiang, China. Acta Geologica Sinica, 79, 7 – 14 (in Chinese with English abstract). Gao, Z. & Chen, K. 2003. The Nanhua System of Xinjiang and some geological issue of Nanhua System in China. Geological Survey and Research, 26, 8 –14 (in Chinese with English abstract). Gao, Z. & Qian, J. 1985. Sinian glacial deposits in Xinjiang, Northwest China. Precambrian Research, 29, 143– 147. Gao, Z. & Zhu, S. 1984. Precambrian Geology in Xinjiang, China. Xinjiang People’s Publishing House, Urumuqi, China (in Chinese with English summary). Gao, Z., Wu, S., Li, Y. & Qian, J. 1981. Sinian –Cambrian stratigraphy of Aksu-Keping area, Xinjiang. Chinese Science Bulletin, 12, 741–743 (in Chinese). Gao, Z., Wang, W. et al. 1985a. The Sinian System of Xinjiang. Xinjiang People’s Publishing House, Urumuqi, China (in Chinese with English abstract). Gao, Z., Wang, W. et al. 1985b. The Sinian System on Aksu-Wushi Region, Xinjiang, China. Xinjiang People’s Publishing House, Urumuqi, China (in Chinese with English summary). Gao, Z., Chen, J., Lu, S., Peng, C. & Qin, Z. 1993. The Precambrian Geology in Northern Xinjiang (Precambrian Geology, No. 6). Geological Publishing House, Beijing (in Chinese with English summary). He, J. Y., Xu, B., Meng, X. Y., Kou, X. W., Liu, B., Wang, Y. & Mi, H. 2007a. Neoproterozoic sequence stratigraphy and correlation in Kurugtagh area, Xinjiang. Acta Petrologica Sinica, 23, 1645–1654 (in Chinese with English abstract). He, X. B., Xu, B. & Yuan, Z. Y. 2007b. C-isotope composition and correlation of the Upper Neoproterozoic in Keping area, Xingjiang. Chinese Science Bulletin, 52, 504– 511. Huang, B. C., Xu, B., Zhang, C. X., Li, Y. A. & Zhu, R. X. 2005. Palaeomagnetism of the Baiyisi volcanic rocks (c. 740 Ma) of Tarim, Northwest China: a continental fragment of Neoproterozoic Western Australia? Precambrian Research, 142, 83 –92.

377

Kou, X. W., Wang, Y., Wei, W., He, J. Y. & Xu, B. 2008. The Neoproterozoic Altungol and Huangyanggou formations in Tarim plate: recognized newly glaciation and interglaciation? Acta Petrologica Sinica, 24, 2863– 2868 (in Chinese with English abstract). Li, H. & Dong, Y. 1991. Sedimentary features of the Sinian Zhamoketi Formation in the Middle Quruqtagh area of Xinjiang. Xinjiang Geology, 9, 340–351. Liou, J. G., Graham, S. A. et al. 1989. Proterozoic blueschist belt in western China: best documented Precambrian blueschists in the world. Geology, 17, 1127–1131. Liu, B., Xu, B., Meng, X. Y., Kou, X. W., He, J. Y., Wei, W. & Mi, H. 2007. Study on the chemical index of alteration of Neoproterozoic strata in the Tarim plate and its implications. Acta Petrologica Sinica, 23, 1664–1670 (in Chinese with English abstract). Lu, S., Ma, G., Gao, Z. & Lin, W. 1985. Sinian ice ages and glacial sedimentary facies – areas in China. Precambrian Research, 29, 53– 63. Lu, S., Li, H., Zhang, C. & Niu, G. 2008. Geological and geochronological evidence for the Precambrian evolution of the Tarim craton and surrounding continental fragments. Precambrian Research, 160, 94– 107. Ma, S., Wang, Y. & Fang, X. 1989. The Sinian at north slope of western Kunlun Mountains. Xinjiang Geology, 7, 68– 79 (in Chinese with English abstract). Ma, S., Wang, Y. & Fang, X. 1991. Basic characteristics of Proterozoic Eonothem as a table cover on northern slope of western Kunlun Mountain. Xinjiang Geology, 9, 59– 71 (in Chinese with English abstract). Norin, E. 1937. Reports from the Scientific Expedition to the Northwestern Provinces of China under the Leadership of Dr. Sven Hedin, III. Geology, 1. Geology of Western Quruqtagh, Eastern Tien-Shan. Bokfo¨rlags Aktiebolaget Thule, Stockholm. Qian, Y. (ed.) 1999. Taxonomy and Biostratigraphy of Small Shelly Fossils in China. Science Press, Beijing (in Chinese with English summary). Qian, J. & Xiao, B. 1984. An Early Cambrian small shelly fauna from Aksu-Wushi region, Xinjiang. In: Professional Papers of Stratigraphy and Palaeontology, No. 13. Geological Publishing House, Beijing, 65– 90 (in Chinese with English abstract). Shen, B., Xiao, S., Kaufman, A. J., Bao, H., Zhou, C. & Wang, H. 2008. Stratification and mixing of a post-glacial Neoproterozoic ocean: evidence from carbon and sulfur isotopes in a cap dolostone from northwest China. Earth and Planetary Science Letters, 265, 209– 228. Wang, J., Cheng, S., Bai, W. & Wang, L. 1983. Glacigenous strata in the western part of the northern Tianshan Mountains. In: Precambrian Geology, No.1 of the Collected Works of Late Precambrian Glacigenous Rocks in China. Geological Publishing House, Beijing, 105– 118 (in Chinese with English summary). Wang, A. G., Zhang, C. L. & Guo, K. Y. 2004. Depositional types and its tectonic significance of lower member of Nanhuan System in North margin of West Kunlun. Journal of Stratigraphy, 28, 248– 256 (in Chinese with English abstract). Xiao, S., Bao, H. et al. 2004. The Neoproterozoic Quruqtagh Group in eastern Chinese Tianshan: evidence for a post-Marinoan glaciation. Precambrian Research, 130, 1– 26. Xu, B., Jian, P., Zheng, H., Zou, H., Zhang, L. & Liu, D. 2005. U– Pb zircon geochronology and geochemistry of Neoproterozoic volcanic rocks in the Tarim Block of northwest China: implications for the breakup of Rodinia supercontinent and Neoproterozoic glaciations. Precambrian Research, 136, 107– 123. Xu, B., Xiao, S. et al. 2009. SHRIMP zircon U– Pb age constraints on Neoproterozoic Quruqtagh diamictites in NW China. Precambrian Research, 168, 247–258. Yao, J., Xiao, S., Yin, L., Li, G. & Yuan, X. 2005. Basal Cambrian microfossils from the Yurtus and Xishanblaq formations (Tarim, north-west China): systematic revision and biostratigraphic correlation of Micrhystridium-like acritarchs from China. Palaeontology, 48, 687– 708. Zhan, S., Chen, Y., Xu, B., Wang, B. & Faure, M. 2007. Late Neoproterozoic palaeomagnetic results from the Sugetbrak Formation of the

378

M. ZHU & H. WANG

Aksu area, Tarim basin (NW China) and their implications to palaeogeographic reconstructions and the snowball Earth hypothesis. Precambrian Research, 154, 143– 158. Zhang, C. L., Shen, J. L. & Guo, K. Y. 2004. Geochemistry of the Neoproterozoic mafic dyke swarm and basalt in south of Tarim Plate and its tectonic significance. Acta Petrologica Sinica, 20, 473– 482 (in Chinese with English abstract). Zhang, C. L., Li, X. H., Li, Z. X., Lu, S. N., Ye, H. M. & Li, H. M. 2007. Neoproterozoic ultramafic –mafic– carbonatite complex and granitoids in Quruqtagh of northeastern Tarim Block, western China: geochronology, geochemistry and tectonic implications. Precambrian Research, 152, 149– 169. Zhang, C. L., Li, Z. X., Li, X. H. & Ye, H. M. 2009. Neoproterozoic mafic dyke swarms at the north margin of the Tarim Block, NW China: age, geochemistry, petrogenesis and tectonic implications. Journal of Asia Earth Sciences, 35, 167– 179.

Zhao, W., Li, C., Gao, Z. & Miao, C. 1985. Stromatolites from the Paergangtag Group at the southern slope of Mt. Kuluketag, Xinjiang. Acta Palaeontologica Sinica, 24, 71 –82. Zhong, D. & Hao, Y. (eds) 1990. Sinian to Permian Stratigraphy and Palaeontology of the Tarim Basin, Xinjiang, (I) Kuruktag Region. Nanjing University Press, Nanjing (in Chinese with English abstract). Zhu, M., Strauss, H. & Shields, G. A. 2007. From Snowball Earth to the Cambrian bioradiation: calibration of Ediacaran– Cambrian Earth history in South China. Palaeogeography, Palaeoclimatology, Palaeoecology, 254, 1 –6. Zhu, W., Zhang, Z., Shu, L., Lu, H., Su, J. & Yang, W. 2008. SHRIMP U –Pb zircon geochronology of Neoproterozoic Korla mafic dykes in the northern Tarim Block, NW China: implications for the long lasting breakup process of Rodinia. Journal of the Geological Society, London, 165, 887–890.

Chapter 34 The Hula Hula Diamictite and Katakturuk Dolomite, Arctic Alaska FRANCIS A. MACDONALD Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA 02138, USA (e-mail: [email protected]) Abstract: The Katakturuk Dolomite is a c. 2-km-thick Neoproterozoic carbonate succession (units K1– K4) exposed in the NE Brooks Range of Alaska. These strata were deposited on a south-facing (present coordinates), rifted passive margin on the North Slope subterrane (NSST) of the Arctic Alaska-Chukotka Plate (AACP). The glaciogenic Hula Hula diamictite rests below the Katakturuk Dolomite and consists of 2– 50 m of diamictite that interfingers with the underlying Mt. Copleston volcanic rocks. Unit K1 of the Katakturuk Dolomite begins with less than 10 m of dark grey, finely laminated limestone with ‘roll-up’ structures, and continues upwards with nearly 500 m of recrystallized, ooid-dominated grainstone. The Nularvik dolomite (unit K2 of the Kataktruk Dolomite) rests on unit K1 with a knife-sharp contact on a heavily silicified surface. The Nularvik dolomite is composed predominantly of laminated micro-peloids hosting tubestone stromatolites and giant wave ripples, followed by decametres of dolomatized, pseudomorphosed former aragonite crystal fans. Carbon-isotope chemostratigraphy suggests that the Hula Hula diamictite is an early Cryogenian glacial deposit, and that, despite the absence of directly underlying glacial deposits, the Nularvik dolomite is a basal Ediacaran cap carbonate. These correlations are supported by the characteristic sedimentological features in both the carbonate capping the Hula Hula diamictite and the Nularvik dolomite. Detrital zircon and Palaeozoic fauna provenance studies support the inference that much of the AACP is exotic to Laurentia; however, the pre-Mississipian relationship between the NSST and the rest of the AACP remains uncertain. Previous palaeomagnetic surveys have been hampered by pervasive Late Cretaceous overprints. Additional geological mapping, sequence stratigraphy and geochronological data are needed to correlate Neoproterozoic and Palaeozoic units across the AACP, and constrain relationships between subterranes in the AACP.

The Neoproterozoic Katakturuk Dolomite (units K1 –K4) and the overlying Cambrian to Ordovician Nanook Limestone form the backbone of the northernmost ranges of the Arctic National Wildlife Reserve of Alaska (for a field guide and an introduction to the general geology of the region, see Molenaar et al. 1987). The preMississippian stratigraphy of the NE Brooks Range has been described through geological mapping (Leffingwell 1919; Reed 1968; Reiser 1971; Sable 1977; Robinson et al. 1989), palaeontological reconnaissance (Dutro 1970) and sequence analysis (Clough & Goldhammer 2000). Recently, Macdonald et al. (2009b) conducted integrated chemo- and lithostratigraphic studies, identified two putative glacial horizons and suggested a late Neoproterozoic age for the Katakturuk Dolomite. The Katakturuk Dolomite was named by Dutro (1970) for its exposure in the Katakturuk River canyon in the Sadlerochit Mountains. Katakturuk is an English derivation of the Inupiaq word Qattaqtuuraq, which translates to ‘a wide open place’ (Clough 1989). The Katakturuk Dolomite has only been positively identified in three localities: the Sadlerochit Mountains, the Shublik Mountains and on both sides of the Hula Hula River near Kikitak Mountain (Fig. 34.1). The succession attains its greatest thickness in the Sadlerochit Mountains, where it is composed of c. 2080 m of shallow-water dolomite, and progressively thins to the south in the Shublik Mountains and at Kikitak Mountain. Exposures in the NE Brooks Range capture the shelf – slope transition, facing to the south (present coordinates). Macdonald et al. (2009b) identified the Hula Hula diamictite at the base of the Katakturuk Dolomite in the eastern Sadlerochit Mountains and along the Hula Hula River near Kikitak Mountain (Fig. 34.1). The type section of the Hula Hula diamictite is on the east side of the Hula Hula River along Eustik Creek (Figs 34.2 & 34.3a, section F621; 69825.4910 N, 144823.2160 W). At this locality, the diamictite is overlain by a 10-m-thick, dark, variably dolomitized limestone (unit K1 of the Katakturuk Dolomite) that contains microbial ‘roll-up’ structures (Hoffman et al. 1998; Pruss et al. 2010). This is one of only two limestone horizons within the Katakturuk Dolomite. In the Sadlerochit Mountains, nearly 500 m higher in the sequence, a micro-peloidal dolostone containing tubestone stromatolites, giant wave ripples and decametres of pseudomorphosed former aragonite crystal fans rests on a silicified surface. The

type section of the Nularvik dolomite (referred to as the Nularvik cap carbonate in Macdonald et al. 2009b) is on the west side of the Nularvik River (Figs 34.2 and 34.3b, section F601; 69837.5880 N, 145805.4510 W). The Nularvik dolomite is succeeded by a major transgression marked by shale and allodapic carbonate, and then an additional c. 1200 m of Ediacaran shallow-water dolomite (units K3 –K4). Exposure of the Katakturuk Dolomite is limited to the North Slope subterrane (NSST) of the Arctic Alaska-Chukotka Plate (AACP). The NSST is tied to the AACP by the Carboniferous Lisburne Group; however, the pre-Mississippian relationship between the NSST, the AACP and Laurentia remain uncertain. Dark-coloured limestones, potentially correlative to the lowermost unit of the Katakturuk Dolomite (K1), are present in the Third and Fourth Ranges of the NSST (Fig. 34.1), c. 10 km and 15 km south of the Shublik Mountains, respectively (Reiser 1971). However, the thickness of this unit is unknown as it is poorly exposed in the Third Range and structurally duplicated in the Fourth Range (Macdonald et al. 2009b). Precambrian dolomites dominated by stromatolites and coated grains, possibly correlative with the Katakturuk Dolomite, have been described on Seward Peninsula and near Snowden Mountain on the southern subterranes of the AACP (Dumoulin 1988; Dumoulin & Harris 1994), and in the Farewell Terrane (Babcock et al. 1994); however, no diamictites have been described at these localities. Terminal Neoproterozoic diamictites occur in the Upper Tindir Group of East-Central Alaska, but these were definitively deposited on Laurentia (Allison et al. 1981; Young 1982; Macdonald & Cohen 2011).

Structural framework Present exposures of the Katakturuk Dolomite in the Sadlerochit, Shublik and Kikitak Mountains (Fig. 34.1) are the product of Palaeogene north-vergent thrusting (Wallace & Hanks 1990). These structures formed during a late-stage reactivation of the mostly Mesozoic Brookian orogeny (Moore et al. 1997), and are related to the piecemeal accretion of southern Alaska (Fuis et al. 2008). Pre-Mississippian strata in the NE Brooks Range also preserves Early to Middle Devonian SE-vergent structures associated

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 379– 387. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.34

380

F. A. MACDONALD

Fig. 34.1. Geological map of the Shublik, Sadlerochit and Kikitak Mountains, and the Fourth Range, with a simplified tectonic and location map inset. Post-Ordovician geology modified from Robinson et al. (1989) and Bader & Bird (1986). Tectonic map modified and simplified from Johnston (2001) and Colpron et al. (2007). Abbreviations for terranes: AACP, Arctic Alaska-Chukotka Plate; FR, Farewell-Ruby-Hammond Terranes; NAm, Ancestral North America; LA, Late Accreted Terranes; ?, poorly exposed and of uncertain affinity.

with the Romanzof orogeny (Oldow et al. 1987). These structures are thought to be distinct from those related to the Late Devonian Ellesmerian orogeny in the Yukon and Canadian Arctic Islands (Lane 2007). Metamorphic grade generally increases from north to south with exposures of the Katakturuk Dolomite in the Shublik and Sadlerochit Mountains displaying little folding and simple block faults. In the NE Sadlerochit Mountains, the underlying Neoproterozoic rocks are tightly folded and chloritized; however, this apparent difference in deformation may be due entirely to lithology and rheology. The overlying Nanook Limestone contains conodonts with a Conodont Alteration Index (CAI) of 3.5 suggesting that at least these strata have experienced less than 300 8C (Harris et al. 1990). The Katakturuk Dolomite is heavily dolomitized and recrystallized; however, the age of the dolomitzation is unknown and, typically, primary sedimentary features are preserved. In the Kikitak Mountain area, small-scale folding is more common, and the basalts are pervasively chloritized. These modifications, coupled with incomplete exposures, make an exact determination of the thickness of the Hula Hula diamictite impossible.

Clough & Goldhammer (2000) inferred a NE– SW palaeostrandline (present coordinates) from palaeo-current data measured in tabular cross-bedded grainstones, the orientation of elongated stromatolites, and from the thickening of outer-ramp to slope facies to the south. While readily measurable elongated stromatolites are not present in the lower kilometre of the Katakturuk Dolomite, facies changes are apparent between the Sadlerochit, Shublik and Kikitak Mountains, deepening from north to south (Macdonald et al. 2009b); deepening is not demonstrable in facies changes from west to east along the ranges. Consequently, for the lower kilometre of the Katakturuk Dolomite (the Hula Hula Diamictite and units K1 –K3), Macdonald et al. (2009b) assumed an east – west palaeo-strandline. Balanced cross-sections from seismic data reveal a c. 30% Palaeogene north – south shortening in the NE Brooks Range (Molenaar et al. 1987; Moore et al. 1997). This results in a restored distance between the Sadlerochit and Shublik Mountains of c. 11 km, and an additional 10 km north –south and 50 km east –west between the Shublik and Kikitak Mountains. However, there are few constraints on the pre-Palaeogene tectonic movement between the two ranges, inhibiting confidence in threedimensional basin reconstructions.

HULA HULA DIAMICTITE AND KATAKTURUK DOLOMITE

381

Fig. 34.2. Composite carbon chemo- and lithostratigraphy of the Katakturuk Dolomite (K1– K4) and the Nanook Limestone (N1, N2) in the NE Brooks Range. Sadlerochit data from measured sections F601, F602, F607, F501 and F505; Shublik data from measured sections F513, F514, F517, F613 and F614; Kikitak data from measured sections F619 and F624 (see Figs 34.1 & 34.3 for locations). All carbonate carbon measurements in ‰ notation. Depositional distance assumes c. 30% Palaeogene shortening (Molenaar et al. 1987; Moore et al. 1997). Time scale is inferred from the detrital zircon ages in the ‘O.G.’ (Macdonald et al. 2009b), Cambrian trilobites in the Nanook Limestone and the similarity of the chemostratigraphic profile and sedimentary structures in the intervening carbonates to other Neoproterozoic age successions (see text for discussion). HH, Hula Hula diamictite; V, Mt. Copleston volcanic rocks.

Stratigraphy As much as 3 km of Neoproterozoic strata are exposed in the NE Brooks Range of Arctic Alaska. The oldest strata in the region are in map unit ‘O. G.’, which was previously mapped as the Neurokpuk Formation (Fm.) and consists of tightly folded, mixed siliciclastic and carbonate rocks (Reiser et al. 1980; Robinson et al. 1989; Macdonald et al. 2009b). The thickness of this unit is unknown and its base is not exposed. Map unit O. G. is succeeded by the Mt. Copleston volcanic rocks. Although this contact is structural in the Sadlerochit Mountains (Macdonald et al. 2009b), regionally it has been described as an unconformity (Reiser et al. 1980; Robinson et al. 1989). The Mt. Copleston volcanic rocks underlie and interfinger with the Hula Hula diamictite, which is in turn succeeded by the c. 2-km-thick Neoproterozoic Katakturuk Dolomite. Subsidence analysis suggests that these strata were accommodated by extension on the southern margin of the North Slope subterrane with units K1 – K3 of the Katakturuk Dolomite deposited on a thermally subsiding passive margin (Macdonald et al. 2009b).

Mt. Copleston volcanic rocks The Mt. Copleston volcanic rocks are rusty weathering, dark maroon to black and green tholeitic basalts with 5 mm chlorite, calcite, zeolite amygdales, and common native copper. In the western Shublik Mountains, the basalt is up to 450 m thick with metre-scale individual flows. Along the Hula Hula River, the Mt. Copleston volcanic rocks are c. 500 m thick with a true thickness difficult to determine due to structural complexities. The volcanic rocks are often greenstone, and dominated by volcaniclastic units in the upper c. 100 m. On the west flank of Kikitak Mountain, the basalts are over 100 m thick and metamorphosed to greenstone. In the eastern Sadlerochit Mountains, the basalts appear relatively low grade with minimal chlorite, 5 –10 mm long plagioclase lathes, and intact, spherical amygdales. Exposures measure up to 105 m thick with an unconformable, often faulted basal contact with the underlying unit O. G. The Mt. Copleston volcanic rocks

and the Katakturuk Dolomite are separated by the Hula Hula diamictite, which ranges in thickness from c. 2 m to 50 m.

Katakturuk Dolomite In the Sadlerochit Mountains, the Katakturuk Dolomite is over 2000 m thick, whereas in the Shublik Mountains, it is only c. 1000 m thick. This is due both to internal thinning and a basal truncation (Macdonald et al. 2009b). Generally, the Katakturuk Dolomite is composed of massive, light-grey, shallow-water dolomite with common ooids and cement. However, further to the SE, along the Hula Hula River, the Katakturuk Dolomite is composed predominantly of allodapic carbonate of unit K1 and measures only 400 m thick, with the upper units of the Katakturuk Dolomite truncated under the sub-Mississippian unconformity (Figs 34.2 & 34.3b). To highlight the major unconformities and disconformities, and departing from Robinson et al.’s (1989) lithostratigraphic subdivision, Macdonald et al. (2009b) divided the Katakturuk Dolomite into four informal units (Fig. 34.2): the Cryogenian map unit K1, the basal Ediacaran Nularvik dolomite (map unit K2), the early –middle Ediacaran map unit K3 and the late Ediacaran map unit K4. The uppermost portion of the Katakturuk Dolomite in the Sadlerochit Mountains, as described by Clough & Goldhammer (2000) has been included with the lowermost Nanook Limestone (Macdonald et al. 2009b).

Nanook Limestone The Katakturuk Dolomite is overlain by the Cambrian to Ordovician Nanook Limestone. Previous studies suggested an unconformity at this level (Clough & Goldhammer 2000; Macdonald et al. 2009b); however, this interpretation may have been complicated by cave breccias from a karstic surface higher in the succession (personal observations). The lower c. 250 m of the Nanook Limestone host bed-parallel ichnogenera and C-isotopic profiles that are consistent with an Early to Middle Cambrian age

382

F. A. MACDONALD

Fig. 34.3. (a) Geological map of the Kikitak Mountain area near the Hula Hula River. (b) Geological map of the eastern Sadlerochit Mountains along Nularvik Creek. See Figure 34.1 for locations and keys for both maps.

(Macdonald et al. 2009b). The upper Nanook Limestone contains Late Cambrian trilobites (Blodgett et al. 1986), and terminates with c. 160 m of Middle to Late Ordovician strata with the uppermost beds containing a diverse gastropod assemblage along with other molluscs, ostracods and brachiopods with a definitive Late Ordovician age (Blodgett et al. 1986) and a Siberian affinity (Blodgett et al. 2002).

Glaciogenic deposits and associated strata Hula Hula diamictite Reiser et al. (1970) described the orange weathering diamictite along the Hula Hula River as carbonate debris flows with clasts of basalt, and included these deposits with the Katakturuk Dolomite. Macdonald (2009) separated the Hula Hula diamictite from the Katakturuk Dolomite and suggested that sedimentation occurred under a glacial influence. On the east side of the Hula Hula River, along Eustik Creek, the Hula Hula diamictite is c. 50 m thick. Although incomplete exposure and structural repetition compromise the measurement of exact thicknesses, discrete thrust panels allow for confidence in the general stratigraphic relations. The lower 12 m of the diamictite is composed of cobblesized clasts of orange dolomite and green to black basalt, together with quartzite pebbles, in a green to tan siltstone, and is interfingered with at least four basaltic flows that range in thickness from 0.2 to 2 m. This lower diamictite is overlain with c. 30 m of poorly exposed, fine millimetre-laminated siltstone with rare, gravel- and cobble-sized bedding-piercing outsized clasts and multiple orange allodapic carbonate beds. The upper 5 m of the Hula Hula diamictite is a massive, clast-supported diamict with boulders of dolomite and cobbles of basalt in a calcareous silt matrix. No striated clasts have been observed. On the west side of the Hula Hula River near Kikitak Mountain (Figs 34.2 & 34.3b), the basal Hula Hula diamictite is composed of c. 10 m of diamictite with cobbles of dolomite and basalt in a siltstone matrix, and an additional c. 40 m of fine millimetrelaminated siltstone with rare outsized clasts and multiple beds of orange allodapic carbonate. The Hula Hula diamictite is also exposed in the eastern Sadlerochit Mountains along the Nularvik

Creek, where the diamictite is only 2 m thick, and is composed of angular cobbles of dolomite, quartzite and basalt in a coarse, arkose grit (Fig. 34.4).

Katakturuk Dolomite Near Kikitak Mountain and on both sides of the Hula Hula River, the basal unit of the Katakturuk Dolomite (K1) consists of c. 10 m of dark limestone, and an additional c. 75 m of rhythmite and allodopic carbonate. Near the Hula Hula River, on the north side of Eustik Creek, the basal 10 m limestone overlying the Hula Hula diamictite exhibits ‘roll-up’ microbial structures, reminiscent of the basal Rasthof Formation of northern Namibia (Hoffman et al. 1998; Pruss et al. 2010), whereas these are not present in the Sadlerochit Mountains. These shallow upwards to ,500 m of massively bedded, often silicified, grainstone and packstone with poorly defined parasequences that are littered with giant ooids (c. 5 mm in diameter). Unit K1 culminates with c. 30 m of resistant, silicified grainstone and broken beds of recrystallized black chert that weather to a distinct black and white (the lower zebra dolomite of Robinson et al. 1989).

Nularvik dolomite The Nularvik dolomite (K2) is equivalent to the upper portion of the zebra dolomite (Robinson et al. 1989; Clough & Goldhammer 2000). In the Sadlerochit Mountains, the Nularvik dolomite (K2) consists of 20– 45 m of white to buff-coloured, recrystallized, finely laminated, micro-peloidal dolomite, overlain by tens of metres of former aragonite crystal fans (Macdonald et al. 2009b). Funnel-shaped calcite and silica cements are common in the lower 15 m of unit K2. In cross-section, the funnels are less than 5 cm tall, and up to 2 cm wide, taper downward, and are commonly linked at the top along bed parallel cements. The funnels are filled with isopacous, void-filling cements. In the western Shublik Mountains, these funnel-shaped cements are laterally equivalent with tubestone stromatolite bioherms (Corsetti & Grotzinger 2005). The peculiar tubestone stromatolites are distinguished by evenly distributed, c. 1-cm-diameter, cement-filled cylindrical tubes, and span as much as 8 m of stratigraphy.

HULA HULA DIAMICTITE AND KATAKTURUK DOLOMITE

383

Fig. 34.4. C-isotope chemo- and lithostratigraphy of the Mt. Copleston volcanic rocks, the Hula Hula diamictite, and unit K1 along the Nularvik Creek in the eastern Sadlerochit Mountains (F607), and along the east side of the Hula Hula River (F619).

In the central and western Sadlerochit Mountains, giant wave ripples (Allen & Hoffman 2005) are also present in the Nularvik dolomite, at the top of the micro-peloidal dolomite. These are succeeded by breccia, and tens of metres of pseudomorphosed aragonite crystal fans, with individual fans measuring as tall as 60 cm (Clough & Goldhammer 2000; Macdonald et al. 2009b). The strata hosting the fans are dominated by grainstone and cement and are often broken, brecciated and recrystallized. Crystal fans are not present at this horizon in the Shublik Mountains. In the Sadlerochit Mountains, unit K2 is succeeded by 2 m of shale and laterally discontinuous allodapic carbonate beds. In the Shublik Mountains, these deeper water facies expand to over 100 m of shale, rhythmite and allodapic carbonate. These are in turn overlain by an additional c. 1200 m of dolomite, primarily in grainstone, biolaminate and stromatolitic facies (units K3 and K4).

deposit is a debris flow or a glacial diamictite formed of reprocessed volcanic rocks. Macdonald et al. (2009b) included this unit with the Mt. Copleston volcanic rocks rather than the Hula Hula diamictite because it lacks foreign clasts, shows no obvious evidence for a glacial origin, and because volcanic breccias occur at other horizons within the Mt. Copleston volcanic rocks. Along the Nularvik Creek in the Sadlerochit Mountains the basal contact of the Katakturuk Dolomite (unit K1) is poorly exposed, whereas near Kikitak Mountain and on both sides of the Hula Hula River, the Hula Hula diamictite is overlain with a knife-sharp contact by c. 10 m of dark limestone. The Nularvik dolomite (K2) rests above unit K1 on a heavily silicified surface with a knife-sharp contact (Fig. 34.5).

Chemostratigraphy Boundary relations with overlying and underlying non-glacial units In the eastern Sadlerochit Mountains, along the Nularvik Creek, the Hula Hula diamictite rests disconformably on pillow basalt of the Mt. Copleston volcanic rocks, with a basally erosive contact. In the Kikitak Mountain area, the Hula Hula diamictite also rests disconformably on the Mt. Copleston volcanic rocks; yet, the upper 50 m of the Mt. Copleston volcanic rocks consists of a volcaniclastic diamictite of outsized volcanic gravel and cobbles in matrix-supported volcanic grit. It is unclear if this

High-resolution, carbonate C- and O-isotope chemostratigraphy through the Katakturuk Dolomite and Nanook Limestone were reported by Macdonald et al. (2009b). In the Sadlerochit Mountains, above the Hula Hula diamictite, in the basal 20 m of K1, C-isotope values rise from – 2‰ to þ6‰ (Fig. 34.4) where they hover between þ3‰ and þ6‰. In the Kikitak Mountain area, C-isotope values rise from þ1‰ to þ8‰, and also oscillate around þ5‰ with slightly more variability. In the last parasequence of K1, values drop to 0‰. C-isotope profiles of the Nularvik dolomite display an inverted S-shaped profile with a nadir at –2‰ (Fig. 34.5). In the Sadlerochit

384

F. A. MACDONALD

Fig. 34.5. Chemo- and lithostratigraphy of the Nularvik dolomite. See Figure 34.1 for locations of measured sections. All carbonate carbon (filled) and oxygen (hollow) data are in ‰ notation.

Mountains, above the Nularvik dolomite, C-isotope values are highly variable through the cement-dominated crystal fans. In the Shublik Mountains, no crystal fans are present, and instead the transgressive sequence progresses from grainstone, to ribbonite, to variably dolomitized limestone rhythmite and shale. C-isotope values bottom out at – 3‰ in these rhythmites, then jump to þ3‰ above a sharp surface below the overlying allodapic carbonate (Fig. 34.5).

Palaeolatitude and palaeogeography The most popular model for the opening of the Arctic Ocean involves a c. 668 counterclockwise rotation of the AACP away from the Canadian Arctic islands about a pole in the Mackenzie Delta region (Carey 1955, 1958; Hamilton 1970; Grantz et al. 1979). Barring any earlier movement relative to Laurentia, this model would place the Neoproterozoic exposures in the NE Brooks Range offshore of what is now Banks Island. Lane (1997) pointed out multiple geological inconsistencies with the rotation model, including ages of deformation and deposition and proposed a model pinning Arctic Alaska to near its present position since Palaeozoic times. However, his ‘fixed’ Alaska model does not account for growing palaeontological evidence of Siberian and Baltican Palaeozoic fauna in Alaskan terranes (Blodgett et al. 2002; Dumoulin et al. 2002). There are several modified versions of the rotation model that include differential motion within the AACP (Miller et al. 2006), and pre-rotation displacement relative to North America (Sweeney 1982). The Neoproterozoic stratigraphy and Palaeozoic

palaeobiogeographic affinities of the AACP do not necessarily contradict the rotation model for the opening of the Arctic Ocean, but they do indicate that the pre-Devonian AACP was exotic to Laurentia (Macdonald et al. 2009b). These models are difficult to test directly because most palaeomagnetic studies in northern Alaska have been compromised by a pervasive Late Cretaceous overprint (Plumley et al. 1989; Stone 1989).

Geochronological constraints A minimum age constraint on the Katakturuk Dolomite is provided by Late Cambrian trilobites in the upper portion of the overlying Nanook Limestone (Blodgett et al. 1986). A lower age constraint is provided by map unit O. G., which is stratigraphically below the Mt. Copleston volcanic rocks and the Hula Hula diamictite, and contains c. 760 Ma (206Pb/207Pb LA-ICPMS) detrital zircon grains (Macdonald et al. 2009b). A coarse, diabase sill within map unit O. G., previously assumed to be coeval with the Mt. Copleston volcanic rocks, yielded a whole rock Rb –Sr isochron age of 801 + 20 Ma (Moore 1987; Clough & Goldhammer 2000); however, recent U –Pb dates of badellyite in these sills suggest they are Cretaceous in age (Macdonald 2009).

Discussion A glacial origin of the Hula Hula diamictite is indicated by the presence of bed-penetrating outsized clasts interpreted as dropstones and the association with a geochemically and

HULA HULA DIAMICTITE AND KATAKTURUK DOLOMITE

sedimentologically distinct overlying dark-coloured limestone. Although the bulk of the Mt. Copleston volcanic rocks are below the Hula Hula diamictite with a few small flows inter-fingering with the lowermost Hula Hula diamictite, it is not certain that the Mt. Copleston volcanic rocks mark the onset of glaciation, as there could have been glacial activity prior to the emplacement of the basalts that failed to leave a record. Nonetheless, a potential interpretation of the stratigraphy of the Hula Hula diamictite along Eustik Creek is that the basal c. 10 m of massive diamictite with interfingering basalts were deposited as glaciomarine deposits during the encroachment of sea ice; the middle c. 30 m of millimetre-laminated silts with occasional debris flows and rare dropstones formed under total ice cover; and the upper c. 5 m of massive diamictite represents the ice-retreat phase. Above the Hula Hula diamictite, C-isotope values rise from –2‰ to þ6‰ in the Sadlerochit Mountains and from þ1‰ to þ8‰ near Kikitak Mountain (Fig. 34.4). Although it is not clear why values are more enriched in the deeper-water sections, the positive, concave trend, and the extremely enriched values are typical of Cryogenian post-glacial carbonates (Halverson et al. 2005), and the basal negative anomaly is similar in magnitude to the basal Rasthof Formation (Yoshioka et al. 2003). An early Cryogenian age of the Hula Hula diamictite is also suggested by the presence of ‘roll-up’ microbial structures in the overlying dark limestones, which are reminiscent of the basal Rasthof cap carbonate of Northern Namibia (Hoffman et al. 1998). C-isotope profiles of the Nularvik dolomite (K2) display an inverted S-shaped profile with a nadir at –2‰ (Fig. 34.5). Normalized for thickness (and excluding isotopic values of cements), this isotopic profile is similar to that of the Ediacaran basal Doushantuo in South China (Jiang et al. 2003; Zhou & Xiao 2007), which has been dated at 635.2 + 0.6 Ma (Condon et al. 2005). The C-isotope profile is also reminiscent of slope sections of the Keilberg Fm. in northern Namibia, where underlying glacial deposits have been dated at 635 + 0.5 Ma (Hoffmann et al. 2004), and shelf sections are 3–4‰ lighter than foreslope sections (Hoffman et al. 2007). Assuming that carbon is well-mixed in the oceans and on platforms, the relatively enriched values of the Nularvik dolomite suggest it was deposited early compared to shelf sections in northern Namibia, and then truncated by exposure surfaces before seawater reached extremely negative values. Funnel-shaped calcite and silica cements in unit K2 are similar to those in the basal Ediacaran Ol cap carbonate of Mongolia (Macdonald et al. 2009a), in the Keilberg Fm. Of Northern Namibia, and in the basal Doushantuo in South China (Macdonald, unpublished data). As these are laterally equivalent with tubestone stromatolite bioherms and are reminiscent in plan view, they may be a related facies that is characteristic of basal Ediacaran cap carbonates. The crystal fans in the Nularvik dolomite are formed in grainstone with multiple exposure surfaces, pervasive cements, and broken and brecciated beds (Clough & Goldhammer 2000; Macdonald et al. 2009b). Occasionally, individual fans are tipped over on their side from the buckling of teepees. C-isotope values are highly variable through this interval. Together, these data suggest the fans were formed in a restricted, lagoonal setting with multiple exposure surfaces (Macdonald et al. 2009b). This is a very different depositional environment than the settings for sea-floor precipitate crystal fan development in other basal Ediacaran cap carbonates (Peryt et al. 1990; James et al. 2001; Hoffman & Halverson 2008). Isotopic scatter, evidence of exposure and pervasive cementing is also a common feature in other basal Ediacaran cap carbonates that were deposited in basins lacking active stretching, such as the upper portion of the Doushantuo Fm. in South China (Jiang et al. 2003) and the Jbeliat dolostone in Mauritania (Hoffman & Schrag 2002; Shields et al. 2006). This pretransgression shoaling could be a product of isostatic rebound outpacing passive subsidence and post-glacial eustatic sea-level rise. Although no late Cryogenian glacial diamictites have been identified, the Nularvik dolostone contains an isotopic profile

385

and sedimentary textures in a particular order that are both characteristic of basal Ediacaran cap carbonates globally (Allen & Hoffman 2005). The lack of glacial deposits can be attributed to poor preservation potential as the glacio-eustatic sea-level drop left the carbonate platform exposed or possibly covered with grounded ice until the post-glacial transgression. Glacial diamictites are also rare under the Keilberg cap carbonate on the Otavi platform in Northern Namibia (Hoffman & Halverson 2008). In the Sadlerochit and Shublik Mountains, a late Cryogenian glaciation may be contained in a silicified surface, with the overlying Nularvik dolomite representing the basal Ediacaran cap carbonate. The Katakturuk Dolomite has previously been correlated with the early Neoproterozoic deposits of NW Laurentia (Rainbird et al. 1996), such as the Lower Tindir Group of the Yukon-Alaska border area (Young 1982; Macdonald et al. 2010a), the pre-717.4 Ma Fifteenmile Group in the Ogilvie Mountains (Macdonald et al. 2010b), the Little Dal Group in the Mackenzie Mountains (Aitken 1981) and the pre-716.3 Ma Shaler Supergroup of Victoria Island (Young 1981; Macdonald et al. 2010b). This correlation was based in large part on palaeogeographical reconstructions that tie the AACP to northwestern Laurentia. However, the Katakturk Dolomite is younger than c. 760 Ma and likely Cryogenian to Ediacaran in age (Macdonald et al. 2009b). Using the 801 + 20 Ma Rb –Sr date, Clough & Goldhammer (2000) argued that the Mt. Copleston volcanic rocks represent a rifting episode coeval with the c. 780 Ma Gunbarrel Event (Park et al. 1995; Harlan et al. 2003), and that the Katakturuk Dolomite was deposited during the thermal subsidence stage on a passive carbonate ramp. Instead, the Mt. Copleston basalts could be correlative with the Franklin igneous event. However, NW Laurentia lacks potentially correlative Cryogenian to Ediacaran successions dominated by platformal carbonate rocks. At present, the origin and Neoproterozoic position of the NSST and AACP are unknown; however, the Neoproterozoic and Palaeozoic strata of the AACP potentially have features in common with peri-Siberian and Baltican terranes (Colpron & Nelson 2009; Macdonald et al. 2009b). Further work is necessary to test these correlations. I thank my field assistants B. Black, P. Kreycik and W. Macdonald for pushing through inclement weather and difficult terrain. I thank D. Schrag and G. Eischied for use of and help in Harvard’s Paleooceanography Laboratory. U. Bold, K. Knudson, W. Macdonald and K. Wecht are thanked for helping to prepare samples. I am grateful to P. Hoffman and the NSF Arctic program for providing financial support, and VECO polar resources for providing logistical support. I also thank the GSA for a student research grant. I would like to thank P. Hoffman, D. Jones and J. Clough for helpful discussions and comments throughout this work. Finally, I thank B. McClelland and E. Arnaud for their helpful comments that greatly improved the manuscript. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) project #512.

References Aitken, J. D. 1981. Stratigraphy and sedimentology of the Upper Proterozoic Little Dal Group, Mackenzie Mountains, Northwest Territories. In: Campbell, F. H. A. (ed.) Proterozoic Basins of Canada. Geological Survey of Canada Paper 81-10. Allen, P. A. & Hoffman, P. F. 2005. Extreme winds and waves in the aftermath of a Neoproterozoic glaciation. Nature, 433, 123– 127. Allison, C. W. A., Young, G. M., Yeo, G. M. & Delaney, G. D. 1981. Glaciogenic rocks of the Upper Tindir Group, east-central Alaska. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge. Babcock, L. E., Blodgett, R. B. & St. John, J. 1994. New Late(?) Proterozoic-age formations in the vicinity of Lone Mountain, McGrath Quadrangle, West-Central Alaska. In: Till, A. B. & Moore, T. E. (eds) Geologic Studies in Alaska by the U.S. Geological Survey, 1993. United States Government Printing Office, Washington.

386

F. A. MACDONALD

Bader, J. W. & Bird, K. J. 1986. Geologic map of the Demarcation Point, Mt. Michelson, Fluxman Island quadrangles, northeastern Alaska. U.S. Geological Survey Miscellaneous Investigations 1791, 1 sheet, scale 1:250,000. Blodgett, R. B., Clough, J. G., Dutro, J. T., Ormiston, A. R., Palmer, A. R. & Taylor, M. E. 1986. Age revisions of the Nanook Limestone and Katakturuk Dolomite, northestern Brooks Range, Alaska. In: Bartsch-Winkler, S. & Reed, K. M. (eds) Geological Studies in Alaska by the Geological Survey during 1985. US Geological Survey Circular 978. Blodgett, R. B., Rohr, D. M. & Boucot, A. J. 2002. Palaeozoic links among some Alaskan accreted terranes and Siberia based on megafossils. In: Miller, E. L., Grantz, A. & Klemperer, S. L. (eds) Tectonic Evolution of the Bering Shelf– Chukchi Sea –Arctic Margin and Adjacent Landmasses. Geological Society of America Special Paper 360, Boulder, Colorado. Carey, S. W. 1955. The orocline concept in geotectonics. Royal Society of Tasmania Proceedings, 89, 255– 288. Carey, S. W. 1958. Continental drift. In: Carey, S. W. (ed.) Continental Drift, a Symposium. University of Tasmania, Hobart, Australia. Clough, J. G. 1989. General stratigraphy of the Katakturuk Dolomite in the Sadlerochit and Shublik Mountains, Arctic National Wildlife Refuge, northeastern Alaska. Alaska Division of Geological & Geophysical Surveys, Public Data File 89-4a, 1– 11. Clough, J. G. & Goldhammer, R. K. 2000. Evolution of the Neoproterozoic Katakturuk dolomite ramp complex, northeastern Brooks Range, Alaska. In: Grotzinger, J. P. & James, N. P. (eds) Carbonate Sedimentation and Diagenesis in the Evolving Precambrian World. SEPM Special Publication, No. 67. Society of Sedimentary Geology, Tulsa, Oklahoma. Colpron, M. & Nelson, J. L. 2009. A Palaeozoic Northwest Passage: incursion of Caledonian, Baltican and Siberian terranes into eastern Panthalassa, and the early evolution of the North American Cordillera. In: Cawood, P. & Kroner, A. (eds) Accretionary Orogens through Space and Time. Geological Society, London, Special Publications, 318, 273–307. Colpron, M., Nelson, J. L. & Murphy, D. C. 2007. Northern Cordilleran terranes and their interactions through time. GSA Today, 17, 1 –7. Condon, D. J., Zhu, M., Bowring, S. A., Wang, W., Yang, A. & Jin, Y. 2005. U– Pb ages from the Neoproterozoic Doushanto Formation, China. Science, 308, 95 – 98. Corsetti, F. A. & Grotzinger, J. P. 2005. Origin and significance of tube structures in Neoproterozoic post-glacial cap carbonates: example from Noonday Dolomite, Death Valley, United States. Palaios, 20, 348– 363. Dumoulin, J. A. 1988. Stromatolite- and coated-grain-bearing carbonate rocks of the western Brooks Range. In: Galloway, J. P. & Hamilton, T. D. (eds) Geologic Studies in Alaska by the U.S. Geological Survey during 1987. United States Government Printing Office, Washington. Dumoulin, J. A. & Harris, A. G. 1994. Depositional framework and regional correlation of pre-Carboniferous metacarbonate rocks of the Snowden Mountain area, Central Brooks Range, Northern Alaska. US Geological Survey Professional Paper 1545, 1 – 55. Dumoulin, J. A., Harris, A. G., Gagiev, M., Bradley, D. C. & Repetski, J. E. 2002. Lithostratigraphic, conodont, and other faunal links between lower Palaeozoic strata in northern and central Alaska and northeastern Russia. In: Miller, E. L., Grantz, A. & Klemperer, S. L. (eds) Tectonic Evolution of the Bering Shelf – Chukchi Sea– Arctic Margin and Adjacent Landmasses. Geological Survey of America Special Paper 360, Boulder, Colorado. Dutro, J. T. 1970. Pre-Carboniferous carbonate rocks, northeastern Alaska. In: Adkison, W. L. & Brosge, M. M. (eds) Proceedings of the Geological Seminar on the North Slope of Alaska, American Association of Petroleum Geologists, Pacific Section. Los Angeles, CA. Fuis, G. S., Moore, T. E. et al. 2008. Trans-Alaska Crustal Transect and continental evolution involving subduction underplating and synchronous foreland thrusting. Geology, 36, 267–270. Grantz, A., Eittreim, S. & Dinter, D. A. 1979. Geology and tectonic development of the continental margin north of Alaska. Tectonophysics, 59, 263– 291.

Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Toward a Neoproterozoic composite carbon-isotope record. Geological Society of America Bulletin, 117, 1181– 1207. Hamilton, W. 1970. The Uralides and the motion of the Russian and Siberian Platforms. Geological Society of America Bulletin, 81, 2553– 2576. Harlan, S. S., Heaman, L. M., LeCheminant, A. N. & Premo, W. R. 2003. Gunbarrel mafic magmatic event: a key 780 Ma time marker for Rodinia plate reconstructions. Geology, 31, 1053– 1056. Harris, A. G., Lane, R. H. & Tailleur, I. L. 1990. Conodont thermal maturation patterns in Palaeozoic and Triassic rocks, Northern Alaska – geological and exploration implications. In: Grantz, A., Johnshon, L. & Sweeney, J. F. (eds) The Arctic Ocean Region, DNAG Series L. Geological Society of America, Boulder, CO. Hoffman, P. F. & Halverson, G. P. 2008. Otavi Group of the western Northern Platform, the Eastern Kaoko Zone and the western Northern Margin Zone. In: Miller, R. M. (ed.) The Geology of Namibia, vol. 2. Handbook of the Geological Survey of Namibia, Windhoek. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis; testing the limits of global change. Terra Nova, 14, 129–155. Hoffman, P. F., Kaufman, A. J. & Halverson, G. P. 1998. Comings and goings of global glaciations on a Neoproterozoic tropical platform in Namibia. GSA Today, 8, 1– 9. Hoffman, P. F., Halverson, G. P., Domack, E. W., Husson, J. M., Higgins, J. A. & Schrag, D. P. 2007. Are basal Ediacaran (635 Ma) post-glacial ‘cap dolostones’ diachronous? Earth and Planetary Science Letters, 258, 114–131. Hoffmann, K. H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. U– Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817– 820. James, N. P., Narbonne, G. M. & Kyser, T. K. 2001. Late Neoproterozoic cap carbonates; Mackenzie Mountains, northwestern Canada; precipitation and global glacial meltdown. Canadian Journal of Earth Sciences, 38, 1229–1262. Jiang, G., Kennedy, M. J. & Christie-Blick, N. 2003. Stable isotopic evidence for methane seeps in Neoproterozoic postglacial cap carbonates. Nature, 426, 822– 826. Johnston, S. T. 2001. The Great Alaskan Terrane Wreck: reconciliation of palaeomagnetic and geological data in the northern Cordillera. Earth and Planetary Science Letters, 193, 259–272. Lane, L. S. 1997. Canada Basin, Arctic Ocean: evidence against a rotational origin. Tectonics, 16, 363–387. Lane, L. S. 2007. Devonian –Carboniferous palaeogeography and orogeesis, northern Yukon and adjacent Arctic Alaska. Canadian Journal of Earth Sciences, 44, 679– 694. Leffingwell, E. 1919. The Canning River region, northern Alaska. US Geological Survey Professional Paper 109. Macdonald, F. A. 2009. Neoproterozoic stratigraphy of Alaska and Mongolia. PhD thesis, Harvard University. Macdonald, F. A. & Cohen, P. A. 2011. The Tatonduk inlier, AlaskaYukon border. In: Arnaud, E., Halverson, G. P. & Shields, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 389–396. Macdonald, F. A., Jones, D. S. & Schrag, D. P. 2009a. Stratigraphic and tectonic implications of a new glacial diamictite-cap carbonate couplet in southwestern Mongolia. Geology, 37, 123–126. Macdonald, F. A., McClelland, W. C., Schrag, D. P. & Macdonald, W. P. 2009b. Neoproterozoic glaciation on a carbonate platform margin in Arctic Alaska and the origin of the North Slope subterrane. Geological Society of America Bulletin, 121, 448–473. Macdonald, F. A., Cohen, P. A., Duda´s, F. O. & Schrag, D. P. 2010a. Early Neoproterozoic scale microfossils in the Lower Tindir Group of Alaska and the Yukon Territory. Geology, 38, 143–146. Macdonald, F. A., Schmitz, M. D. et al. 2010b. Calibrating the Cryogenian. Science, 327, 1241–1243. Miller, E. L., Toro, J. et al. 2006. New insights into Arctic palaeogeography and tectonics from U –Pb detrital zircon geochronology. Tectonics, 25, 1 –19. Molenaar, C. M., Mull, C. G. & Swauger, D. A. 1987. Geologic features of Ignek Valley and adjacent mountains, northeastern Alaska.

HULA HULA DIAMICTITE AND KATAKTURUK DOLOMITE

Centennial Field Guide Volume 1: Cordilleran Section of the Geological Society of America. Moore, T. E. 1987. Geochemistry and tectonic setting of some volcanic rocks of the Franklinian assemblage, central and eastern Brooks Range. In: Tailleur, I. & Weimer, P. (eds) Alaskan North Slope Geology. SEPM, Pacific Section, California and Alaska Geological Society, Alaska. Moore, T. E., Wallace, W. K., Mull, C. G., Adams, K. E., Plafker, G. & Nokleberg, W. J. 1997. Crustal implications of bedrock geology along the Trans-Alaska Crustal Transect (TACT) in the Brooks Range, northern Alaska. Tectonics, 102, 20 645–20 684. Oldow, J. S., Lallemant, A., Julian, F. E. & Seidensticker, C. M. 1987. Ellesmerian (?) and Brookian deformation in the Franklin Mountains, northeastern Brooks Range, Alaska, and its bearing on the origin of the Canada Basin. Geology, 15, 37 – 41. Park, J. K., Buchan, K. L. & Harlan, S. S. 1995. A proposed giant radiating dyke swarm fragmented by the separation of Laurentia and Australia based on palaeomagnetism of ca. 780 Ma mafic intrusions in western North America. Earth and Planetary Science Letters, 132, 129– 139. Peryt, T. M., Hoppe, A., Bechstadt, T., Koster, J., Pierre, C. & Richter, D. K. 1990. Late Proterozoic aragonitic cement crusts, Bambui Group, Minas Gerais, Brazil. Sedimentology, 37, 279–286. Plumley, P. W., Vance, M. S. & Milazzo, G. 1989. Structural and palaeomagnetic evidence for Tertiary bending of the Eastern Brooks Range Flexure, Alaska. In: Hillhouse, J. W. (ed.) Deep Structure and Past Kinematics of Accreted Terranes, Geophysical Monograph 50. American Geophysical Union, Washington, DC. Pruss, S. B., Bosak, T., Macdonald, F. A., McLane, M. & Hoffman, P. F. 2010. Microbial facies in a Sturtian cap carbonate, the Rasthof Formation, Otavi Group, northern Namibia. Precambrian Research, 181, 187– 198. Rainbird, R. H., Jefferson, C. W. & Young, G. M. 1996. The early Neoproterozoic sedimentary Succession B of Northwestern Laurentia: correlations and palaeogeographic significance. Geological Society of America Bulletin, 108, 454– 470. Reed, B. L. 1968. Geology of the Lake Peters area, northeastern Brooks range, Alaska. US Geological Survey Bulletin 1236, 1 –132. Reiser, H. N. 1971. Northeastern Brooks Range – a surface expression of the Prudhoe Bay Section. In: Adkison, W. L. & Brosge, M. M. (eds) Proceedings of the Geological Seminar on the North Slope of Alaska. American Association of Petroleum Geologists, Pacific Section, Los Angeles, CA. Reiser, H. N., Dutro, J. T., Brosge, W. P., Armstrong, A. K. & Detterman, R. L. 1970. Progress map, geology of the Sadlerochit

387

and Shublik Mountains, Mt. Michelson C-1, C-2, C-3, C-4 Quadrangles, Alaska. US Geological Survey Open-file report, 70-273, scale 1:63,360, 5 sheets. Reiser, H. N., Brosge, W. P., Dutro, J. T. & Detterman, R. L. 1980. Geologic map of the Demarcation Point quadrangle, Alaska, 1:250,000, map I-1133. Miscellaneous Investigation Series. U.S. Geological Survey. Robinson, M. S., Decker, J., Clough, J. G., Reifenstuhl, R. R., Dillon, J. T., Combellick, R. A. & Rawlinson, S. E. 1989. Geology of the Sadlerochit and Shublik Mountains, Alaska National Wildlife Reserve, northeast Alaska, Professional Report 100, 1:10,000 scale. State of Alaska, Department of Natural Resources, Division of Geological and Geophysical Survey. Sable, E. G. 1977. Geology of the western Romanzof Mountains, Brooks Range, northeastern Alaska: a comprehensive study of plutonic, sedimentary, and metamorphic rocks in an eastern Brooks Range area. Geological Survey, Washington, D.C., Professional Paper 897. Shields, G., Deynoux, M., Strauss, H., Paquet, H. & Nahon, D. 2006. Barite-bearing cap dolostones of the Toudeni Basin, northwest Africa: sedimentary and isotopic evidence for methane seepage after a Neoproterozoic glaciation. Precambrian Research, 153, 209–235. Stone, D. B. 1989. Palaeogeography and rotations of Arctic Alaska – an unresolved problem. In: Kissel, C. & Laj, C. (eds) Palaeomagnetic Rotations, Continental Deformation. Kluwer Academic, Norwell, MA. Sweeney, J. F. 1982. Mid-Palaeozoic travels of Arctic Alaska. Nature, 298, 647–649. Wallace, W. K. & Hanks, C. R. 1990. Structural provinces of the Northeastern Brooks Range, Arctic National Wildlife Refuge, Alaska. AAPG Bulletin, 74, 1100– 1118. Yoshioka, H., Asahara, Y., Tojo, B. & Kawakami, S. 2003. Systematic variations in C, O and Sr isotopes and elemental concentrations in Neoproterozoic carbonates in Namibia: implications for a glacial to interglacial transition. Precambrian Research, 124, 69 – 85. Young, G. M. 1981. The Amundsen Embayment, Northwest Territories: relevance to the Upper Proterozoic evolution of North America. In: Campbell, F. H. A. (ed.) Proterozoic Basins of Canada. Geological Survey of Canada Paper 81-10. Young, G. M. 1982. The late Proterozoic Tindir Group, east-central Alaska; evolution of a continental margin. Geological Society of America Bulletin, 93, 759–783. Zhou, C. & Xiao, S. 2007. Ediacaran d13C chemostatigraphy of South China. Chemical Geology, 237, 107–126.

Chapter 35 The Tatonduk inlier, Alaska –Yukon border FRANCIS A. MACDONALD* & PHOEBE A. COHEN Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA 02138, USA *Corresponding author (e-mail: [email protected]) Abstract: Glaciogenic deposits of the Rapitan and Hay Creek Groups are exposed in the Tatonduk inlier of east-central Alaska and the western Yukon. The Rapitan Group ranges in thickness from c. 50 to 700 m with Fe-formation common in the upper 10 m. In the most distal settings, the Rapitan Group is separated from the diamictite of the Hay Creek Group by over 100 m of sandstone and siltstone; however, the Hay Creek Group contains large erosive surfaces and cannibalizing breccia, and rarely preserves strata between the two glaciogenic deposits. The diamictite of the Hay Creek Group is capped by a white- to buff-coloured dolostone with pseudo-teepee structures, bed-parallel, isopachous sheet-crack cements, and a depleted C-isotope signature. Late Neoproterozoic glacial deposits in the Tatonduk inlier were formerly assigned to the Tindir Group. To simplify the nomenclature in the northwestern Canadian Cordillera, the Tindir Group was abandoned and replaced with nomenclature consistent with that of the Windermere Supergroup in the Mackenzie Mountains. The mixed lithology and anchizone-grade metamorphism distinguish the Rapitan and Hay Creek Groups in the Tatonduk inlier as attractive future targets for integrated micropalaeontology, geochemistry, palaeomagnetism and geochronology.

Cairnes (1914) referred to Precambrian stratigraphy exposed along the international border between the United States and Canada as the Tindir Group. Mertie (1930, 1933) later described the stratigraphy (and natural history of the region) in remarkable detail and divided the Tindir Group into seven units. Brabb & Churkin (1969) produced an excellent map of the geology on the Alaskan side of the border, while the Tindir Group was mapped on the Yukon side of the border by Norris (1982); however, exact correlations of specific units of the Tindir Group across the border remained ambiguous until more recent mapping and compilation (Dover 1992; Van Kooten et al. 1997) and integrated litho- and chemostratigraphy (Macdonald et al. 2010a, b, 2011; Macdonald & Roots 2010). To improve the consistency of geological maps of the Yukon and promote the synthesis of geological data, Macdonald et al. (2011) reassigned the Lower Tindir Group to the Pinguicula and Fifteenmile Groups of the Mackenzie Mountains Supergroup and the Upper Tindir Group to the Rapitan, Hay Creek and ‘upper’ groups of the Windermere Supergroup (Figs 35.1 & 2). The Upper and Lower Tindir Groups were first distinguished by Payne & Allison (1981), despite the fact that no unconformable contacts have been observed between the two (Young 1982). The Lower Tinder Group consists primarily of dolomite and shale that are commonly cut with north-trending mafic dykes (Van Kooten et al. 1997). Unlike the Lower Tindir Group, the units of the Upper Tindir Group are not cut by mafic dykes (Van Kooten et al. 1997). Young (1982) separated the Upper Tindir Group into five units in ascending stratigraphic order from unit 1 to unit 5: unit 1 consists primarily of mafic volcanic rocks; unit 2, purple mudstones and diamictite, including Fe-formation; unit 3, an additional massive diamictite; unit 4, platfomal dolomites and green to light grey shales; unit 5, grey to black shales and limestones. In the updated stratigraphic framework (Macdonald et al. 2011), unit 1 is renamed the Pleasant Creek volcanic rocks, unit 2 is the Rapitan Group, unit 3 and the ‘teepee’ dolomite at the base of unit 4 constitute the Hay Creek Group, and the rest of unit 4 and unit 5 are assigned to the ‘upper’ group. The informal upper group is a provisional name that encompasses the Sheepbed, Gametrail, Blueflower and Risky formations in the Mackenzie Mountains (e.g. Aitken 1989; Dalrymple & Narbonne 1996; MacNaughton et al. 2000, 2008) and equivalent strata in the Ogilvie Mountains (Macdonald et al. 2011). Exposures of the Windermere Supergroup in the Tatonduk inlier are typically limited to the walls of creek beds. Allison et al. (1981)

followed Mertie (1933) in designating the exposures along the Tatonduk River as the type locality of the diamictite-bearing units. This choice is due in part to the relative ease of access, but is problematic because of structural complications and the presence of a major disconformity under the diamictite in the Hay Creek Group. Owing to lack of exposure, the critical features of both diamictites are not present at one single locality. Along Pass Creek (Fig. 35.3), Fe-formation of the Rapitan Group and portions of both diamictites are exposed, but the overlying teepee dolomite is not present. Along Hard Luck Creek (Fig. 35.3), the contact between the Fe-formation and underlying diamictite can be observed, and the teepee dolomite is present, although the diamictite of the Hay Creek Group is not exposed. This lack of exposure coupled with the presence of multiple disconformities creates a significant challenge (and many uncertainties) in regional stratigraphic correlations.

Structural framework Neoproterozoic diamictites of the Windermere Supergroup are exposed through an erosional window, referred to as the Tatonduk inlier, which extends across the Canadian border into Alaska (Fig. 35.2). Present exposures are the product of a mid-Cretaceous to Palaeogene foreland fold-and-thrust belt (Norris 1972) associated with the underplating of the Yakutat Block (Fuis et al. 2008). The Tatonduk inlier protrudes from the elbow of the Ogilvie deflection, where the orientation of folds and thrusts transition from north-vergent to west-vergent due to Cenozoic activity on the Kaltag fault and associated dextral transpressional structures (Norris 1972). Anchizone-grade metamorphism is suggested by the presence of economic oil deposits in the overlying Phanerozoic strata sourced in part from Palaeozoic shales (Van Kooten et al. 1997). Late Neoproterozoic strata in the Tatonduk inlier are a northwestern continuation of the Windermere Supergroup in the Mackenzie (Aitken 1989), Wernecke (Pyle et al. 2004) and Ogilvie Mountains (Mustard & Roots 1997), for which deposition commenced with rifting of the northwestern margin of Laurentia (Stewart 1975). Neoproterozoic strata in the Tatonduk inlier have also been correlated with the Katakturuk Dolomite in the northeastern Brooks Range of Arctic Alaska (Rainbird et al. 1996); however, the Katakturuk Dolomite was likely deposited

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 389– 396. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.35

390

F. A. MACDONALD & P. A. COHEN

Ogilvie Mountains Thompson et al. 1994; Mustard & Roots, 1997

Young, 1992

Macdonald et al., 2010b; Macdonald & Roots, 2010

Coal Creek Inlier

Van Kooten et al., 1997

Palaeozoic undiff.

Bouvette Bouvette Funnel Ck., Adams, Jones Ridge Formation Formation Hillard, Jones Formation PH5 PH5 Ridge Fms.

Rapitan Gp.

4b 4a 3b 3a

Pleasant Creek/Mt. Harper volcanic rocks 1

717.43 ± 0.14 Ma

2

pCtl

pCtss

pCtr pCtbs

1 2

716.47 ± 0.24 Ma

5

4 3 2 1

Upper Harper Group

Hay Creek Group

Upper Tindir Group

“upper” group

5

Upper Tindir Group

Stratigraphy Upper Tindir Group

Windermere Supergroup

Jones Ridge & Bouvette Fms., Cambrian Undiff.

Macdonald et al., 2010a

Macdonald et al., 2011

Tatonduk Inlier

surfaces and breccias. The Hard Luck Creek Fault (HLF) marks an abrupt expansion of the stratigraphy, and can be extended to the SE into the Ogilvie Mountains where it has been named the Mt. Harper Fault, and roughly marks a former rift shoulder of the Laurentian margin (Mustard & Roots 1997). The HLF and other Precambrian faults in the region were inverted during midCretaceous to Palaeogene shortening (Brabb & Churkin 1969; Van Kooten et al. 1997).

PH4

PH4

PH3

PH3

PH2 PH1

MHVC

2 1

PH1/PH2

MHVC

unconformity

Fig. 35.1. Nomenclature chart depicting names used by different authors to describe Neoproterozoic stratigraphy in the Tatonduk and Coal Creek inliers. Dates are U –Pb CA-IDTIMS zircon ages (Macdonald et al. 2010b).

on a separate margin, as it is doubtful that the pre-Mississippian Arctic Alaska-Chukotka Plate was part of Laurentia (McClelland, 1997; Blodgett et al. 2002; Dumoulin et al. 2002). Although unconformities have not been observed in the Windermere Supergroup of the Tatonduk inlier, multiple disconformities are present in the Hay Creek Group marked by erosional

The stratigraphy of the diamictite-bearing units in the Tatonduk inlier was reviewed by Allison et al. (1981) and described in detail by Young (1982). Additional measured sections are presented in Macdonald et al. (2010a, supplementary material, 2010b, supplementary material, 2011). Diamictite of the Rapitan Group is underlain by the Pleasant Creek volcanic rocks, which are up to 200 m thick and consist chiefly of amygdaloidal pillow basalt and cherty hyaloclastic breccia, with minor tuff, shale and conglomerate. The Rapitan Group is composed primarily of finelaminated purple and red mudstone and siltstone that are sprinkled with dolomite and basaltic lonestones. The upper c. 15 m of the Rapitan Group commonly hosts a massive diamictite and Fe-formation. The Rapitan Group is overlain by the Hay Creek Group, which is composed of up to 150 m of planar laminated siltstone and sandstone with minor dolomite marl, massive diamictite, and dolomite breccia capped by a white to buff-coloured dolostone with bed-parallel cements. The uppermost dolostone of the Hay Creek Group is less than 5 m thick and rests disconformably on all of the underlying units of the Windermere Supergroup and the Pleasant Creek volcanic rocks (Fig. 35.4). The upper group is composed largely of black shale with minor allodapic carbonate.

Fig. 35.2. Location map of the NW Cordillera with Cryogenian and Ediacaran strata in grey. AACP, Arctic AlaskaChukotka Plate. Inset shows the distribution of the Mackenzie Mountains and Windermere Supergroups in the Tatonduk Inlier. Teeth on thrust faults.

THE TATONDUK INLIER, ALASKA– YUKON BORDER

391

Fig. 35.3. Geological map of exposures along the Tatonduk River and Hard Luck Creek.

Glaciogenic deposits and associated strata Rapitan Group The Rapitan Group (formerly Upper Tindir unit 2) is exposed near the international border along Pleasant Creek and in outcrops close to the Tatonduk River (Macdonald et al. 2010b). These strata are chiefly composed of fine-laminated purple and red mudstone and siltstone speckled with dolomite gravel lonestones. Scattered throughout the siltstone are thin diamictite beds (,1 m) dominated by clasts of dolomite, clastic grit and volcanic cobbles. The Rapitan Group also contains faceted clasts, boulders with striations, slumped beds, flame structures, groove casts and flute structures (Young 1982). The upper c. 15 m of the Rapitan Group (Fig. 35.5a; section T709) hosts multiple c. 10-cm-thick beds of Fe-formation, which are interbedded with a laminated diamictite with bed-penetrating, outsized clasts. The majority of clasts in the Rapitan Group consist of dolomite derived from the underlying Fifteenmile Group. Where the upper contact of the Fifteenmile Group is exposed, it is overlain by a well-sorted, imbricated carbonate matrix conglomerate with abundant dolostone clasts of variable size, interpreted as a debris flow, which is followed by the parallel-bedded siltstone and sandstone of the Hay Creek Group. The thickness of the Rapitan Group varies greatly from ,50 m near Pleasant Creek to .700 m c. 20 km to the NW (Young 1982). Palaeocurrent measurements in the interbedded siltstone suggest a west-facing margin (present coordinates) (Young 1982).

Calcareous concretions and jasper lenses are present in the Fe-formation, while foreign clasts are rare. Chemical analyses of the Fe-formation yield FeO concentrations up to 50%, high SiO2/Al2O3, and depletions in most other elements, suggesting low clastic deposition and a chemical silica contribution (Young 1982).

Hay Creek Group ( formerly units 3a, 3b and 4a of the Upper Tindir Group) In the most western sections along Pass Creek (T710), the Rapitan Group is overlain by c. 140 m of planar bedded, siltstone, sandstone and marl. These beds lack any evidence of tidal influence, lack lonestones, are commonly stacked in fining-upward Bouma sequences, and as such, are interpreted as deep-water turbidites. The turbidite beds are succeeded by an additional diamictite unit, but the contact between the two is not exposed. In section T710 the diamictite of the Hay Creek Group is 22 m thick and consists of 10 m of massive diamictite, 10 m of laminated centimetre-beds of siltstone and an additional 2 m of stratified diamictite (Fig. 35.5b). The massive diamictite is clast-dominated with boulders of dolomite, and cobbles of Fe-formation, siltstone, conglomerate and basalt in a marly pink matrix. The bedded diamictite has the same clast composition as the rock below it, but clasts are slightly smaller with no boulders, and a matrix of purple silt. Approximately 20 km to the NW of Pass Creek (section Y.S. 9), still west of the HLF, the diamictite of the Hay Creek Group is poorly

392

F. A. MACDONALD & P. A. COHEN

Fig. 35.4. Chemo- and lithostratigraphy of the ‘upper’ group in east-central Alaska. Y.S. indicates section from Young (1982). Sections have been projected along the Hard Luck Fault (i.e. to the NW) to an east– west line, with distances between sections estimated from the projection. Note the change in scale for two sections SW of the Hard Luck fault, which are approximately four times as thick as sections NE of the fault. See Figure 35.3 for section locations.

exposed, but lonestones c. 2 m in diameter are present, and the diamictite is capped by a buff-coloured dolostone (Young 1982). Along the Tatonduk River, Young (1982) described a massive, over 250-m-thick, crudely stratified, purple diamictite interbedded with minor amounts of contorted purple mudstone and lenses of chert. We assign this diamictite to the Hay Creek Group. Clasts are up to 0.6 m across, and consist predominantly of dolomite from the Fifteenmile Group, with minor limestone, basalt and chert. Faceted and striated clasts have been described at this locality (Allison et al. 1981), but have not been observed elsewhere in the Tatonduk inlier. The diamictite of the Hay Creek Group is also very thick c. 20 km due north, NE of the HLF (Van Kooten et al. 1997), where it cuts down into the underlying stratigraphy. Northeast of the HLF, the diamictite of the Hay Creek Group is either absent, or represented by a dolomite (matrix and clast) breccia. No foreign clasts have been identified with the exception of some clasts of the Pleasant Creek volcanic rocks near the base. This breccia cuts out the underlying units of the Windermere Group and rests on the Pleasant Creek volcanic rocks (Fig. 35.4). The overlying dolomite, ,5 m thick, rests disconformably on underlying units of the Hay Creek Group as well as the Pleasant Creek volcanic rocks and Rapitan Group strata. It is white to buff-coloured dolostone with bed-parallel cements (pseudoteepee structures of Young 1982). These pseudo-teepees do not show a polygonal plan-form or a concentration of cements along the broken pieces, as is typical of teepees that are of a subaerial exposure origin (Kendall & Warren 1987). Instead, cements are isopachous and bed-parallel, and beds are contorted and irregularly buckled, suggesting intraformational detachment during deposition. Similar ‘sheet-crack’ cements are present in

several basal Ediacaran cap carbonates globally (Hoffman & Macdonald 2010).

The ‘Upper’ Group The Hay Creek Group is overlain by the ‘upper’ group (formerly units 4b and 5 of the Upper Tindir Group), which consists of as much as 50 m of planar laminated siltstone, sandstone and dolomitic marl, and an additional sequence of black shale interbedded with minor organic-rich limestone (Fig. 35.4; section T707). Like the underlying units, the ‘upper’ group displays a major stratigraphic expansion to the SW ranging from 40–75 m thick in the Yukon to c. 700 m thick along the Tatonduk River in Alaska (Macdonald et al. 2010a). Both Hay Creek and the ‘upper’ group strata are consistent with a SW-facing margin (present coordinates). These ‘upper’ group units form the final clastic-carbonate cycle prior to deposition of the Cambrian sandstone (Backbone Ranges Formation) and the commencement of the miogeocline after break-up of Rodinia. This group completes Windermere Supergroup sedimentation and episodic extension.

Boundary relations with overlying and underlying non-glacial units The basal contact of the Rapitan Group was not seen; however, volcanic fragments similar in composition to the underlying Pleasant Creek volcanic rocks are common in the lower half of the massive diamictite. The uppermost exposures of the Rapitan Group commonly consist of Fe-formation. At some localities the upper contact is exposed, and overlain by a well-sorted, dolomite

THE TATONDUK INLIER, ALASKA– YUKON BORDER

(a)

(b)

Upper Unit 2 Hard Luck Creek T709

Unit 3b Pass Creek T710 v

20

(metres)

10 8

16

6

12 v

4

8

2

4 0 diamictite lithologies massive unsorted stratified imbricated conglomerate

matrix composition ironformation dolomite marl siltstone

erosional disconformity. These interpretations are consistent with the erosional disconformity and breccia in the Hay Creek Group NE of the HLF where much of the underlying stratigraphy is missing. Although the top contact is not exposed SW of the HLF, to the NE the breccia is sharply overlain by the buff-coloured dolostone that defines the top of the Hay Creek Group.

Chemostratigraphy Carbon-isotope values of the dolomite at the top of the Hay Creek Group are extremely depleted, although somewhat variable (Fig. 35.4; Macdonald et al. 2010a, supplementary data). Near the international border (section T708), values are consistently between –3‰ and –3.5‰, whereas near Hard Luck Creek (sections T701 and T709), values are more scattered, ranging from –6‰ to þ2‰. In the ‘upper’ group, C-isotope values tend to be enriched (Fig. 35.4). Carbon and Sr isotopes were reported from unit 5 of the Upper Tindir Group in Canada (Kaufman et al. 1992); however, these sections are actually correlative with the Fifteenmile Group (Macdonald et al. 2010a, b, 2011). The lowest Sr-isotope values from these sections are near 0.7064 (Macdonald et al. 2010a), a value that is typical for pre-Sturtian carbonate rocks (Halverson et al. 2007).

Palaeolatitude and palaeogeography

v

0

393

clast size boulder cobble gravel clast compostion v volcanic clastic carbonate ironformation

Fig. 35.5. (a) Stratigraphy of the Rapitan Group exposed along Hard Luck Creek, showing diamictite and its relationship to the Fe-formation. (b) Stratigraphy of the Hay Creek diamictite along Pass Creek, including diamictite and associated lithofacies. See Figure 35.3 for section locations.

clast, dolomite matrix diamictite, interpreted as a debris flow, which is followed by the parallel-bedded siltstone and sandstone with Bouma sequences of the Hay Creek Group. Young (1982) interpreted the diamictite of the Hay Creek Group along the Tatonduk River as allochthonous in the sedimentary sense, having been derived from a more proximal setting to the east of the HLF in a massive slope failure, with an unconformable basal contact. Allison et al. (1981) interpreted this contact as an

No palaeomagnetic studies have been reported from the Tatonduk inlier. Although the lithologies of Pleasant Creek volcanic rocks and the Rapitan Group (basalts and Fe-rich clastic sediments, respectively) are ideal targets, most palaeomagnetic studies in northern Alaska are compromised by a pervasive Late Cretaceous overprint (Plumley et al. 1989). However, palaeopoles on the Rapitan Group in the Mackenzie Mountains yield a palaeolatitude of 6 + 48 (Park 1997). This pole is consistent with the grand mean pole on the contemporaneous Franklin LIP (Denyszyn et al. 2009), and demonstrates that Laurentia straddled the equator during deposition of the Rapitan Group (Macdonald et al. 2010b). Although Laurentia remained at low latitudes until 615 Ma according to the controversial Long Range Dyke pole (Murthy et al. 1992; Hodych et al. 2004), Laurentia appears to have migrated to high latitudes by 590 Ma (Murthy 1971). McCausland et al. (2007) provide an excellent review of the Ediacaran palaeomagnetic data from Laurentia.

Geochronological constraints Radiometric data The NNW-trending mafic dykes that intrude the Fifteenmile Group in the Tatonduk inlier have yielded a wide range of K –Ar biotite and whole-rock ages: 532 + 11 Ma, 572 + 16 Ma, 588 + 14 Ma and 644 + 18 Ma (Van Kooten et al. 1997). These dykes have not been observed to intrude any of the Windermere Supergroup units and are possibly feeder dykes for the Pleasant Creek volcanic rocks. Because alteration can lead to argon loss (Westcott 1966), these dates are minimum age constraints for the Fifteenmile Group. Along strike to the east in the Coal Creek inlier (Fig. 35.3), Rapitan Group correlatives rest above the Mt. Harper volcanic complex (MHVC). Zircons extracted from rhyolite in member D of the MHVC were dated at 717.43 + 0.14 Ma, and zircons from a tuff within the Rapitan Group correlatives were dated at 716.47 + 0.24 Ma (U – Pb ID-TIMS; Macdonald et al. 2010b). In the Rapitan Group of the Mackenzie Mountains, a clast of leucogranite in the Rapitan Group has a U –Pb TIMS bulk zircon age of 755 + 18 Ma (Ross & Villeneuve 1997).

394

F. A. MACDONALD & P. A. COHEN

Micropalaeontology Fossils interpreted as microscopic flatworm impressions by Allison (1975), but possibly sponge spicules (Andy Knoll pers. comm.), have been described in the shale interbedded with basalt in the lower portion of the Windermere Supergroup. It is not clear if these samples were collected from the Pleasant Creek volcanic rocks, Rapitan Group or the Hay Creek Group, as they are reported in the ‘basalt and red beds’ unit of Brabb & Churkin (1969), which incorporates all three. Recent biomarker work suggests that the presence of sponge spicules in the Windermere Supergroup is not inconsistent with a Cryogenian age (Love et al. 2009). Microfossils have also previously been described in chert of Fifteenmile Group (in strata previously mismapped as unit 5 of the Upper Tindir Group; Macdonald et al. 2010a), including cyanobacterial coccoids, acritarchs such as Trachyhystrichosphaera, vase-shaped microfossils and unique, enigmatic siliceous scales (Allison 1980; Allison & Hilgert 1986; Allison & Awramik 1989). More recent studies have extracted these microfossils from the surrounding carbonate rock, and have demonstrated that they are composed of phosphate rather than silica, suggesting a green algae taxonomic affinity (Cohen et al. 2011). A pre-717 Ma for these fossils is supported by C- and Sr-isotope correlations (Macdonald et al. 2010a), the presence of dykes cutting the sections that are co-magmatic with the Pleasant Creek volcanic rocks, and

geochronology of correlative rocks in the Coal Creek inlier (Macdonald et al. 2010b, 2011).

Discussion Depositional setting The Windermere Supergroup contains two Cryogenian glaciogenic deposits (Rapitan Group and the diamictite of the Hay Creek Group) separated by c. 140 m of non-glacial strata (Fig. 35.4). A glaciomarine depositional setting for the Rapitan Group is suggested by the presence of faceted and striated clasts, bed-penetrating dropstones, and common outsized and exotic clasts, along with evidence for subaqueous slumping in the form of graded grain flows and debris flows. Young (1988) ascribed the Fe-formations of the Rapitan Group, and equivalent strata in the Tatonduk inlier, to rift-related hydrothermal activity. However, these Fe-formations are intimately associated with well-developed dropstones. Presuming these diamictites were deposited during the terminal ice retreat, and presuming an increased solubility of iron in the ocean due to low oxygen levels under long-lived sea ice (Martin 1965), Fe-formation can be attributed to an influx of oxygenated fresh water concentrated at the termini of ice streams (Kirschvink 1992). The parallel bedded sandstones and siltstones in the lower portion of the Hay Creek Group are interpreted as turbidites as

Fig. 35.6. Neoproterozoic stratigraphy in the Tatonduk inlier and correlations with the Windermere Supergroup exposed c. 500 km east in the Mackenzie Mountains of Canada. The schematic stratigraphy of the Windermere Supergroup is modified from Halverson et al. (2008). IRD mudstones are laminated fine-grained sediments with lonestones interpreted as ice rafted debris.

THE TATONDUK INLIER, ALASKA– YUKON BORDER

they lack any evidence of wave action or traction currents, and beds are stacked in fining-upward Bouma cycles. These green-grey turbidites are distinguished from the red and purple grain flows in the Rapitan Group by the lack of lonestones. Young (1982) interpreted the diamictite of the Hay Creek Group SW of the HLF as the product of a massive slope failure, and identified an unconformity at the base. Nonetheless, there is evidence of a glaciomarine influence on deposition with the presence of striated clasts (Allison et al. 1981), and a well-bedded diamictite with dropstones near the top of Hay Creek Group along Pass Creek (Fig. 35.5b). A depositional setting along the upper slope, with diamictite being the product of syn-glacial slope failure, is consistent with the interpretation that the expansion of the stratigraphy of all of the Windermere Supergroup across the HLF roughly approximates the slope – shelf transition. From outcrops along the Tatonduk River, Allison et al. (1981) cited deformation in Rapitan Group sediments near the overlying disconformity to suggest that the underlying stratigraphy was unconsolidated at the time of deposition of the upper diamictite. However, slumps and syn-sedimentary folds are common throughout the Rapitan Group (Young 1982), so this deformation in the Rapitan Group may be syn-sedimentary or the product of grounded ice during the Rapitan glaciation.

Regional correlations Another line of evidence that the upper diamictite represents a second, later glacial event is that it is overlain with a buff-coloured dolostone. The latter is considered a basal Ediacaran cap carbonate as it contains bed-parallel, isopachous sheet-crack cements (Hoffman & Macdonald 2010) and a C-isotope profile similar to that of thin basal Ediacaran cap carbonates in China (Zhou & Xiao 2007), Australia (Kennedy 1996) and Namibia (Halverson et al. 2005). Lithologically, the Rapitan Group in the Tatonduk inlier is very similar to the clast poor siltstone and Fe-formation of the Sayunei Formation in the Mackenzie Mountains (Young 1976; Yeo 1984). The diamictite of the Hay Creek Group can be correlated with the Ice Brook Formation (Fig. 35.6). This correlation is particularly attractive because the carbonate at the top of the Hay Creek Group shares sedimentological and isotopic characteristics with the Ravensthroat cap dolomite in the Mackenzie Mountains (Aitken 1991; James et al. 1999). Regional correlations of Neoproterozoic strata in NW Canada are particularly important in light of the recent geochronological constraints provided by volcanic tuffs interbedded with the Fifteenmile and Rapitan Groups in the Ogilvie Mountains (Macdonald et al. 2010b). Coupled with the robust palaeomagnetic poles in NW Canada (Evans 2000; Denyszyn et al. 2009) and the rich micropalaeontological record in the Tatonduk inlier (Allison & Arwimik 1989; Macdonald et al. 2010b), the inliers of the Ogilvie Mountains hold great promise of a calibrated record of tectonics, climate, chemistry and life in Cryogenian oceans. We are grateful to the Yukon Geological Survey, the Department of Earth and Planetary Sciences, Harvard University, P. Hoffman, A. Knoll and the NSF for support. We thank R. B. Blodgett and D. Jones for helpful discussions. We also thank the YUCH National Park office in Eagle for allowing access to the field area. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) project #512.

References Aitken, J. D. 1989. Uppermost Proterozoic formations in central Mackenzie Mountains, Northwest Territories. Geological Survey of Canada Bulletin, 368, 1 –26. Aitken, J. D. 1991. The Ice Brook Formation and Post-Rapitan, Late Proterozoic glaciation, Mackenzie Mountains, Northwest Territories. Geological Survey of Canada Bulletin, 404, 1 – 43.

395

Allison, C. W. A. 1975. Primitive fossil flatworm from Alaska: new evidence bearing on ancestry of the Metazoa. Geology, 3, 649– 653. Allison, C. W. A. 1980. Siliceous microfossils from the lower Cambrian of northwest Canada: possible source for biogenic chert. Science, 211, 53– 55. Allison, C. W. A. & Hilgert, J. W. 1986. Scale micro-fossils from the Early Cambrian of northwest Canada. Journal of Paleontology, 60, 973– 1015. Allison, C. W. A. & Awramik, S. M. 1989. Organic-walled microfossils from earliest Cambrian or latest Proterozoic Tindir Group rocks, northwest Canada. Precambrian Research, 43, 253–294. Allison, C. W. A., Young, G. M., Yeo, G. M. & Delaney, G. D. 1981. Glaciogenic rocks of the Upper Tindir Group, east-central Alaska. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 720–723. Blodgett, R. B., Rohr, D. M. & Boucot, A. J. 2002. Paleozoic links among some Alaskan accreted terranes and Siberia based on megafossils. In: Miller, E. L., Grantz, A. & Klemperer, S. L. (eds) Tectonic Evolution of the Bering Shelf – Chukchi Sea –Arctic Margin and Adjacent Landmasses. Geological Society of America Special Paper 360, Boulder, Colorado, 273– 290. Brabb, E. E. & Churkin, M. J. 1969. Geologic Map of the Charley River Quadrangle, East-Central Alaska. U.S. Geological Survey Map I-573. Cairnes, D. D. 1914. The Yukon-Alaska International Boundary Between Porcupine and Yukon Rivers. Geological Survey of Canada Memoir 67, Map 140-A. Cohen, P. A., Schopf, J. W., Butterfield, N. J., Kudryaytsev, A. & Macdonald, F. A. 2011. Phosphate biomineralization in midNeoproterozoic protists. Geology, 39, 539– 542. Dalrymple, R. W. & Narbonne, G. M. 1996. Continental slope sedimentation in the Sheepbed Formation (Neoproterozoic, Windermere Supergroup), Mackenzie Mountains, N.W.T. Canadian Journal of Earth Sciences, 33, 848–862. Denyszyn, S. W., Halls, H. C., Davis, D. W. & Evans, D. A. D. 2009. Paleomagnetism and U– Pb geochronology of Franklin dykes in High Arctic Canada and Greenland: a revised age and paleomagnetic pole for constraining block rotations in the Nares Strait region. Canadian Journal of Earth Sciences, 46, 689– 705. Dover, J. H. 1992. Geologic map and fold and thrust belt interpretation of the south eastern part of the Charley River Quadrangle, East Central Alaska. US Geological Survey Miscellaneous Investigations, Map I-1942, scale 1:100,000, 2 sheets. Dumoulin, J. A., Harris, A. G., Gagiev, M., Bradley, D. C. & Repetski, J. E. 2002. Lithostratigraphic, conodont, and other faunal links between lower Paleozoic strata in northern and central Alaska and northeastern Russia. In: Miller, E. L., Grantz, A. & Klemperer, S. L. (eds) Tectonic Evolution of the Bering Shelf – Chukchi Sea –Arctic Margin and Adjacent Landmasses. Geological Survey of America Special Paper 360, Boulder, Colorado, 291– 312. Evans, D. A. D. 2000. Stratigraphic, geochronological and paleomagnetic constraints upon the Neoproterozoic climatic paradox. American Journal of Science, 300, 347–433. Fuis, G. S., Moore, T. E. et al. 2008. Trans-Alaska Crustal Transect and continental evolution involving subduction underplating and synchronous foreland thrusting. Geology, 36, 267–270. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Toward a Neoproterozoic composite carbonisotope record. Geological Society of America Bulletin, 117, 1181–1207. Halverson, G. P., Duda´s, F. O., Maloof, A. C. & Bowring, S. A. 2007. Evolution of the 87Sr/86Sr composition of Neoproterozoic Seawater. Palaeogeography, Palaeoclimatology, Palaeoecology, 256, 103–129. Hodych, J. P., Cox, R. A. & Kosler, J. 2004. An equatorial Laurentia at 550 Ma confirmed by Grenvillian inherited zircons dated by LAM ICP-MS in the Skinner Cove volcanics of western Newfoundland: implications for inertial interchange true polar wander. Precambrian Research, 129, 93– 113. Hoffman, P. F. & Macdonald, F. A. 2010. Sheet-crack cements and early regression in Marinoan (635 Ma) cap dolostones: regional benchmarks of vanishing ice-sheets? Earth and Planetary Science Letters, 300, 374– 384.

396

F. A. MACDONALD & P. A. COHEN

James, N. P., Narbonne, G. M. & Kyser, K. T. 1999. Neoproterozoic Cap Carbonate Facies; Mackenzie Mountains, NW Canada; Abiotic Precipitation and Global Glacial Meltdown. Geological Society of America Abstracts with Programs, 31, 487. Kaufman, A. J., Knoll, A. H. & Awramik, S. M. 1992. Biostratigraphic and chemostratigraphic correlation of Neoproterozoic sedimentary successions: Upper Tindir Group, northwestern Canada, as a test case. Geology, 20, 181– 185. Kendall, C. G. S. C. & Warren, J. 1987. A review of the origin and setting of teepees and their associated fabrics. Sedimentology, 34, 1007– 1027. Kennedy, M. J. 1996. Stratigraphy, sedimentology and isotope geochemistry of Australian Neoproterozoic postglacial cap dolostones: deglaciation, d13C excursions and carbonate precipitation. Journal of Sedimentary Research, 66, 1050–1064. Kirschvink, J. L. 1992. Late Proterozoic low-latitude global glaciation: the snowball earth. In: Schopf, J. W. & Klein, C. (eds) The Proterozoic Biosphere. Cambridge University Press, Cambridge, 51 –52. Love, G. D., Fike, D. A. et al. 2009. Fossil steroids record the appearance of Demospongiae during the Cryogenian period. Nature, 457, 718– 722. Macdonald, F. A. & Roots, C. F. 2010. Upper Fifteenmile Group in the Ogilvie Mountains and correlations of early Neoproterozoic strata in the northern Cordillera. In: MacFarlane, K. E., Weston, L. H. & Blackburn, L. R. (eds) Yukon Exploration and Geology 2009. Yukon Geological Survey, Whitehorse, YT, 237–252. Macdonald, F. A., Cohen, P. A., Duda´s, F. O. & Schrag, D. P. 2010a. Early Neoproterozoic scale microfossils in the Lower Tindir Group of Alaska and the Yukon Territory. Geology, 38, 143– 146. Macdonald, F. A., Schmitz, M. D. et al. 2010b. Calibrating the Cryogenian. Science, 327, 1241– 1243. Macdonald, F. A., Smith, E. F., Strauss, J. V., Cox, G. M., Halverson, G. P. & Roots, C. F. 2011. Neoproterozoic and early Paleozoic correlations in the western Ogilvie Mountains, Yukon. In: MacFarlane, K. E., Weston, L. H. & Blackburn, L. R. (eds) Yukon Exploration and Geology 2010. Yukon Geological Survey, Whitehorse, 161– 182. MacNaughton, R. B., Narbonne, G. M. & Dalrymple, R. W. 2000. Neoproterozoic slope deposits, Mackenzie Mountains, northwestern Canada: implications for passive-margin development and Ediacaran faunal ecology. Canadian Journal of Earth Sciences, 37, 997–1020. MacNaughton, R. B., Roots, C. F. & Martel, E. 2008. Neoproterozoic– (?)Cambrian lithostratigraphy, northeast Sekwi Mountain map area, Mackenzie Mountains, Northwest Territories: new data from measured sections. Geological Society of Canada, Current Research 2008, 16, 1– 17. Martin, H. 1965. Beobachtungen zum Problem der jung-pra¨kambrischen Glazialen Ablagerungen in Su¨dwestafrika. (Observations concerning the problem of the late Precambrian glacial deposits in South West Africa.) Geologische Rundschau, 54, 115– 127. McCausland, P. J. A., Van der Voo, R. & Hall, C. M. 2007. Circum-Iapetus paleogeography of the Precambrian – Cambrian transition with a new paleomagnetic constraint from Laurentia. Precambrian Research, 156, 125– 152. McClelland, W. C. 1997. Detrital zircon studies of the Proterozoic Neruokpuk Formation, Sadlerochit and Franklin Mountains, northern Alaska. Geological Society of America Abstracts with Programs, 25, 28. Meertie, J. B. 1930. Geology of the Eagle-Circle district, Alaska. US Geological Survey Bulletin, 816, 121–122. Meertie, J. B. 1933. The Tatonduk-Nation district, Alaska. US Geological Survey Bulletin, 836-E, 345–454.

Murthy, G. S. 1971. The paleomagnetism of diabase dykes from the Grenville Province. Canadian Journal of Earth Sciences, 8, 802– 812. Murthy, G. S., Gower, C. F., Tubrett, M. & Patzold, R. 1992. Paleomagnetism of Eocambrian Long Range dykes and Double Mer Mormation from Labrador, Canada. Canadian Journal of Earth Sciences, 29, 1224–1234. Mustard, P. S. & Roots, C. F. 1997. Rift-related volcanism, sedimentation and tectonic setting of the Mount Harper Group, Ogilvie Mountains, Yukon Territory. Geological Survey of Canada Bulletin, 492. Norris, D. K. 1972. En echelon folding in the northern Coldillera of Canada. Bulletin of Canadian Petroleum Geology, 20, 634– 642. Norris, D. K. 1982. Geology, Ogilvie River, Yukon Territory, Geological Survey of Canada, Map 1526A, 1:250,000 scale. Park, J. K. 1997. Paleomagnetic evidence for low-latitude glaciation during deposition of the Neoproterozoic Rapitan Group, Mackenzie Mountains, N.W.T., Canada. Canadian Journal of Earth Sciences, 34, 34 – 49. Payne, M. W. & Allison, C. W. A. 1981. Paleozoic continental-margin sedimentation in east-central Alaska. Geology, 9, 274–279. Plumley, P. W., Vance, M. S. & Milazzo, G. 1989. Structural and paleomagnetic evidence for Tertiary bending of the Eastern Brooks Range Flexure, Alaska. In: Hillhouse, J. W. (ed.) Deep Structure and Past Kinematics of Accreted Terranes, Geophysical Monograph 50. American Geophysical Union, Washington, DC, 127– 150. Pyle, L. J., Narbonne, G. M., James, N. P., Dalrymple, R. W. & Kaufman, A. J. 2004. Integrated Ediacaran chronostratigraphy, Wernecke Mountains, northwestern Canada. Precambrian Research, 132, 1– 27. Rainbird, R. H., Jefferson, C. W. & Young, G. M. 1996. The early Neoproterozoic sedimentary Succession B of Northwestern Laurentia: correlations and paleogeographic significance. Geological Society of America Bulletin, 108, 454– 470. Ross, G. M. & Villeneuve, M. E. 1997. U –Pb geochronology of stranger stones in Neoproterozoic diamictites, Canadian Cordillera: implications for provenance and ages of deposition. Radiogenic age and isotopic studies, Report 10, Geological Survey of Canada, Current Research 1997-F, 141– 155. Stewart, J. H. 1975. Initial deposits in the Cordilleran geosyncline: evidence of a late Precambrian (,850 m.y.) continental seperation. Geological Society of America Bulletin, 83, 1345–1360. Van Kooten, G. K., Watts, A. B. et al. 1997. Alaska Division of Geological and Geophysical Surveys, Report of Investigations 96-6A, 3 sheets, scale 1:125,000, Fairbanks. Westcott, M. R. 1966. Loss of argon from biotite in a thermal metamorphism. Nature, 210, 83– 84. Yeo, G. M. 1984. The Rapitan group: relevance to the global association of Late Proterozoic glaciation and iron-formation. PhD thesis, University of Western Ontario, London, Ontario. Young, G. M. 1976. Iron-formation and glaciogenic rocks of the Rapitan Group, Northwest Territories, Canada. Precambrian Research, 3, 137– 158. Young, G. M. 1982. The late Proterozoic Tindir Group, east-central Alaska; evolution of a continental margin. Geological Society of America Bulletin, 93, 759–783. Young, G. M. 1988. Proterozoic plate tectonics, glaciation and iron-formations. Sedimentary Geology, 58, 127– 144. Zhou, C. & Xiao, S. 2007. Ediacaran d13C chemostatigraphy of South China. Chemical Geology, 237, 107– 126.

Chapter 36 Neoproterozoic glacial record in the Mackenzie Mountains, northern Canadian Cordillera P. F. HOFFMAN1,2* & G. P. HALVERSON3,4 1

Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA 02138, USA 2

School of Earth and Ocean Sciences, University of Victoria, Victoria, BC V8W 2Y2, Canada

3

School of Earth and Environmental Sciences, The University of Adelaide, North Terrace, Adelaide, SA 5005, Australia 4

Present address: Department of Earth and Plantary Sciences, McGill University, Montreal, QC H3A 2A7, Canada *Corresponding author (e-mail: [email protected]) Abstract: In the Mackenzie Mountains, an arcuate foreland thrust-fold belt of Late Cretaceous–Paleocene age in the northern Canadian Cordillera, two discrete glacial– periglacial sequences of Cryogenian age (the Rapitan Group and the Stelfox Member of the Ice Brook Fm.) are separated by c. 1.0 km of non-glacial strata. The older Rapitan diamictite occurs in an amagmatic rift basin; the younger Stelfox diamictite occurs on a passive-margin continental slope. The Rapitan Group consists of three formations. The lower Mount Berg Fm. is a complex of diamictites and conglomerates of limited extent. The middle Sayunei Fm. is a thick sequence of maroon-coloured mudrocks hosting innumerable graded layers of silt- and finegrained sandstone. It lacks wave- or traction current-generated bedforms, and is lightly sprinkled with granule aggregates (‘till pellets’) and lonestones of dolostone and rare extrabasinal granitoids. It is capped by a hematitic Fe-formation that was reworked into the disconformably overlying Shezal diamictite. The Shezal Fm. is a complex of olive-green coloured boulder diamictites with subordinate, darkgrey shales, siltstones and parallel-sided sandstones. Some of the boulders are faceted and striated, and include dolostone, quartzite, siltstone and gabbro in declining order of abundance. Diamictite terminates abruptly at the top of the Shezal Fm., which is sharply overlain by dark shales or by ,52 m of fetid, dark-grey, 13C-depleted limestone with graded bedding. The Stelfox Member is dominated by non-stratified, carbonate-clast diamictite with faceted and striated clasts, locally associated with subordinate, well-laminated shales containing till pellets and ice-rafted dropstones. It is thin or absent on the palaeocontinental shelf, but thickens seaward (southwestward) on the palaeocontinental slope. A thin clay drape separates it from a laterally continuous post-glacial ‘cap’ dolostone, which is a very pale coloured, micro- to macropeloidal dolostone with low-angle cross-laminae, giant wave ripples and local bioherms of corrugated stomatolites. In the NW, the dolostone is followed by reddish and greenish marls, followed by black shale of the Sheepbed Fm. In the SE, the dolostone is overlain by pink or grey limestones with well-developed sea-floor cements pseudomorphic after aragonite. In this area, the top of the dolostone is ferruginous and contains digitate rosettes of sea-floor barite cement, variably calcitized. The dolostone– limestone contact is perfectly conformable, and synclinal structures previously intepreted as karst features are tectonic in origin. The grand mean palaeomagnetic pole for the well-studied Franklin Large Igneous Province (c. 718 Ma) of Arctic Laurentia, coeval with the basal Rapitan Group in the Mount Harper area, Yukon Territory, places the Mackenzie Mountains firmly in the tropics, at 18 + 38N palaeolatitude, at the onset of the Rapitan glaciation. Carbon (d13C), oxygen (d18O) and strontium (87Sr/86Sr) isotopes have been measured in carbonates bracketing the Rapitan and Stelfox diamictites. Sulphur isotope data (d34S) have been obtained from carbonate-associated sulphate and barite above the younger diamictite, and calcium isotope data (d44Ca) from the younger carbonate itself. The results are broadly consistent with data from other areas. Iron isotope (d57Fe) and cerium anomaly (Ce/Ce*) values increase systematically upwards through the Sayunei Fe-formation, supporting an interpretation that deposition occurred within a redox chemocline through which the basin floor descended as a consequence of isostatic loading by the advancing Shezal ice sheet. Supplementary material: Data are available at http://www.geolsoc.org.uk/SUP18470.

Glacial marine diamictites (‘tillites’) were recognized in the northwestern Mackenzie Mountains (Fig. 36.1) and to the west in the Yukon Territory by Shell Oil Company geologists in 1958 (Ziegler 1959), during the course of economic assessment of associated Fe-formations. At the time, they were thought to be early Palaeozoic in age (Ziegler 1959). They were referred to as the Rapitan Group by Green & Godwin (1963). Correlative strata were studied in the Hayhook Lake area (Fig. 36.1), 300 km to the SE, by Upitis (1966), Gabrielse et al. (1973), Young (1976) and Eisbacher (1978), where they were recognized as being of Neoproterozoic (Hadrynian) age. Gabrielse et al. (1973) erected a type section for the Rapitan Group near Hayhook Lake at 63834’03”N, 127802’41”W. It includes the Sayunei, Shezal and Twitya formations (Fig. 36.2). Eisbacher (1978) enlarged the Rapitan Group to include the Keele Fm. because of its gradational relationship with the Twitya Fm. However, Yeo (1981) and most subsequent workers follow the original definition (Green & Godwin 1963) and limit the Rapitan Group to glaciogenic strata beneath the Twitya Fm. Additional stratigraphic and sedimentological studies of the Rapitan Group along the 400 km arc of the Mackenzie Mountains were carried out by Eisbacher (1981a, b, 1985) and Yeo (1981, 1984, 1986).

A second, thinner, glaciogenic unit was discovered .1000 m stratigraphically above the Rapitan Group by Aitken (1991a, b). He defined it as the Stelfox Member (,272 m) of the Ice Brook Fm., the type section of which is located in the Sayunei Range at 648080 0700 N, 1298000 3800 W (Aitken 1991b). It is overlain by a distinctive carbonate unit, the ‘Tepee dolostone’ of Eisbacher (1978, 1981a) and Aitken (1991b), subdivided into the Ravensthroat and Hayhook formations by James et al. (2001). The dolomite features unusual sedimentary structures, including sheet-crack cements, reverse-graded peloids, giant wave ripples, corrugated stromatolites and sea-floor cements (crystal fans) originally composed of aragonite and barite (Eisbacher 1981a; Aitken 1991b; James et al. 2001; Hoffman & Schrag 2002; Allen & Hoffman 2005; Hoffman & Macdonald 2010). It has many similarities with postglacial cap dolostones on other palaeocontinents (Hoffman et al. 2011), which allow the base of the Ediacaran Period to be recognized globally (Knoll et al. 2006). Correlatives of the Mackenzie Mountains and Windermere supergroups, including both glacial horizons, are discontinuously exposed in Wernecke and Ogilvie in central Yukon Territory (Macdonald & Roots 2009; Macdonald et al. 2010b), and in the Tatonduk inlier of east-central Alaska (Young 1982; Macdonald et al. 2010a; Macdonald & Cohen 2011).

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 397– 411. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.36

398

P. F. HOFFMAN & G. P. HALVERSON

134 oW

132 oW

platfor

Iron SNAKE RIVER Ck

mou ntain Cranswick R.

m

a

ck

Norman Wells en

zi

e

Ri

65oN

v e

r

u

t

Th

Stone Knife R. Shale Lake

M

on fr

Gayna R. P la Arctic te Red R. a

Corn Creek

126o W

128oW

130oW

ru st

Twitya R.

64oN Boomerang L. Moose Horn R.

Stelfox Mtn June Lake

Hayhook Lake

REDSTONE

RavensSekwi throat R. Brook

This figure

Windermere Supergroup Rapitan Group

PLATEAU

63oN Coates L. Thundercloud R.

Sheepbed Creek

measured sections

This chapter is a synopsis of previous work, augmented by seven new sections of the Rapitan Group, measured by P.F.H. in the Iron Creek (658030 1500 N, 1338070 3100 W: .775 m), Gayna

Group

Formation

Member

Sheepbed Hayhook

Reference Gabrielse et al. (1973) James et al. (2001)

Ravensthroat Stelfox Ice Brook

Delthore

Aitken (1991b)

Durkan Keele

Fig. 36.1. Outcrop distribution of the Rapitan Group and the Windermere Supergroup in the Mackenzie Mountains, showing locations of measured sections forming the basis of this report.

River (648490 2300 N, 1308270 1100 W: 22 m), Stone Knife River (648410 5300 N, 1298530 1900 W: 114 m), Shale Lake (648320 4100 N, 1298220 3700 W: 995 m), Boomerang Lake (638460 4600 N, 1278280 0500 : .236 m), Hayhook Lake (638340 3800 N, 1278050 3400 W: 997 m) and Ravensthroat River (638130 5700 N, 1278030 1800 W: 230 m) areas; and 13 sections of the Stelfox diamictite (all ,35 m) and its post-glacial carbonate sequence at Cranswick River (658050 5400 N, 1328260 1700 W), Arctic Red River (648560 0500 N, 1318030 3000 W), Gayna River (648490 2200 N, 1308280 2700 W), Stoneknife River (648400 3600 N, 1298530 4800 W), Shale Lake (648310 1200 N, 1298290 0900 W), Twitya River (648130 0700 N, 1288380 1100 W), Moose Horn River (638570 1300 N, 1278300 3900 W), Hayhook Lake (638340 1700 N, 1278110 4400 W), Stelfox Mountain (638360 0100 N, 1278500 4700 W; 638350 4600 N, 1278530 0400 W; 638350 6000 N, 1278530 5700 W; 638350 3700 N, 1278550 1100 W) and Ravensthroat River (638170 0500 N, 1278080 4000 W). The named locations are indicated in Figure 36.1. The coordinates refer to the base of the section.

Twitya Shezal Rapitan

Eisbacher (1978)

Sayunei

Structural framework

Mt Berg Copper Cap Coates Lake Redstone River Thundercloud Little Dal

Gabrielse et al. (1973) Jefferson (1983) Aitken (1981)

Fig. 36.2. Stratigraphic nomenclature and defining references for units associated with Cryogenian glaciogenic strata (black with inverted triangles) in the Mackenzie Mountains.

The Mackenzie Mountains are the physiographic expression of an arcuate, NE-vergent, foreland thrust –fold belt of Late Cretaceous –Paleocene age in the northern Canadian Cordillera (Aitken & Long 1978; Aitken 1982; Narbonne & Aitken 1995). Late Neoproterozoic strata of the Windermere Supergroup (Fig. 36.3) are principally exposed in the hanging wall of the Plateau Thrust system (Fig. 36.1). Their subhorizontal (Redstone Plateau) or SW-dipping attitude depends on their position above a flat or ramp in the thrust plane. Outcrop-scale strain is heterogeneous; structurally intact, low-strain sections can be located by

THE MACKENZIE MOUNTAINS, NORTHERN CANADIAN CORDILLERA

(b)

Keele

-0.5

fs

Gametrail

-100

-150

-150

Ravensthroat Ice Stelfox Brook

-0.8

Keele

-0.9 -1.0

fs fs

-250 fs

Twitya Shezal

Rapitan Group

-300

-1.1

unnamed cap

-200

Shezal

-0.7

-1.2

Sayunei -1.3 Mt Berg -1.4

-1.5

fs

-350

-200

BIF

-250 BIF

-300

-350

-400

-400

-450

-450

-1.7

-2.0

Katherine Group 780 Ma (U-Pb)

-2.3 -2.4

wacke-matrix diamictite

-550

-2.1 -600

-2.2

Tsezotene

BIF

-650

maroon

-1.9

LEGEND mud-matrix diamictite

-500

-1.8 R R

Sayunei

Little Dal Group

scaly-matrix diamictite mudstone with silty turbidites, debrites, sparse ice-rafted debris fs

R

carbonate

BIF BIF

BIF

Sayunei

MACKENZIE MTNS SUPERGROUP

TONIAN(?)

-100

fs

Sheepbed

'Unit H1'

-12

-50

-0.6

-1.6

-8

-50

-0.2

maroon

-0.4

(m)

Shezal

-0.3

(m)

Sayunei

Ingta

Risky Blueflower

63o34’38”N, 127o05’34”W

Mount Berg

-0.1

Twitya

CAMBRIAN

(km)

Vampire Backbone Ranges

Coates Lake Group

-10

Iron Creek 65o03’15”N, 133o07’31”W

Coates

-6

CRYOGENIAN

-4

WINDEREMERE SUPERGROUP

-2

EDIACARAN

(km)

Sekwi

Hayhook Lake (type sec.) Hayhook Lake (ref. sec.) 63o34’03”N, 127o02’41”W 0.0

Shezal

0

(d)

Coates?

Mackenzie Mountains composite

(c)

olive green (Little Dal-Katherine derivation)

(a)

399

reefal carbonate

sulphate mudstone evaporite

siltstone

sandstone

basaltgabbro

haematitic haematitemudstone jaspilite

flooding surface

Fig. 36.3. (a) Composite columnar section of Neoproterozoic strata in the Mackenzie Mountains (modified after Narbonne & Aitken 1995). (b) Type section of the Rapitan Group (Gabrielse et al. 1973) in the Hayhook Lake area (see Fig. 36.1 for location) as remeasured by P.F.H. (c) Detailed section of the upper Sayunei and Shezal formations 1.6 km NW of the Rapitan Group type section. Note Fe-formation (‘BIF’) at the top of the Sayunei Formation, representing ‘basin-facies’ Fe-formation (Eisbacher 1985). (d) Section of the lower Rapitan Group near Iron Creek, representative of ‘transitional-facies’ Fe-formation (Eisbacher 1985).

mapping. Long slopes are unattractive because of scree. Published 1:250 000-scale map coverage is incomplete and stratigraphically inconsistent.

An unconformity at the base of the Backbone Ranges Fm. (Early Cambrian) cuts progressively down-section from SW to NE, with the result that the Windermere Supergroup is missing to the NE of

400

P. F. HOFFMAN & G. P. HALVERSON

the Plateau Thrust system. To the SW, the glaciogenic units are buried by younger strata. Although the outcrop belt of Windermere strata is less than 30 km wide, it fortuitously preserves the outer shelf-edge and upper slope of a continental terrace developed in late Cryogenian (Keele Fm.) to early Ediacaran (Sheepbed Fm.) time (Ross 1991; Narbonne & Aitken 1995; Dalrymple & Narbonne 1996; Day et al. 2004). Continental rifting, leading to the formation of the continental terrace, controlled sedimentation during the Coates Lake and Rapitan groups (Eisbacher 1981a, 1985; Jefferson 1983), as well as during the preceding Mackenzie Mountains Supergroup (Turner & Long 2008). The Coates Lake and Rapitan basins are interpreted by Jefferson & Ruelle (1986) and Yeo (1981), respectively, as rhombochasms (‘pull-apart’ basins) associated with hypothetical strike –slip systems, systems that are not mutually compatible in orientation.

Stratigraphy Mackenzie Mountains Supergroup The Mackenzie Mountains Supergroup is a broadly conformable succession composed of sandstone, siltstone, carbonate and evaporite of mostly shallow-marine origin (Fig. 36.3a). Its upper part, the Little Dal Group, is 2 km thick in its type section (Gabrielse et al. 1973; Aitken 1981; Halverson 2006) and is dominated by carbonate with a recessive interval of gypsiferous siltstone. The carbonates are platformal in the SE and basinal in the NW, with reefal build-ups in the lower part (Aitken 1981; Turner et al. 1997). The carbonate is conformably overlain by and locally interstratified with pillow basalt, the Little Dal lavas of Aitken (1982). These are possibly the extrusive equivalents of the Tsezotene dykes and sills (Aitken 1982), which belong to the 780 Ma Gunbarrel large igneous province (Harlan et al. 2003).

Coates Lake and Rapitan groups The Coates Lake Group is an assemblage of sandstone, carbonateclast conglomerate, gypsiferous siltstone and carbonate, including fetid basinal limestone with turbidites and ‘debrites’ (i.e. coarsegrained mass-flow deposits) (Ruelle 1982; Jefferson & Ruelle 1986). The Rapitan Group (Eisbacher 1978, 1981a) includes the Sayunei (sigh-YOU-knee) Fm., composed of maroon coloured, subaqueously deposited, fine-grained clastics with subordinate debrites and lonestones, and the overlying Shezal (shiz-ALL) Fm., a stack of mostly olive-coloured, polymictic ‘diamictites’ (i.e. massive, foliated or bedded wackestone with randomly dispersed, matrix-supported pebbles and boulders, characteristically faceted and striated) with thin interbeds of dark shale and sandstone. Locally, an older diamictite complex (Mount Berg Fm.) occurs below the Sayunei Fm. (Yeo 1981). The Rapitan Group onlaps tilted Coates Lake and Little Dal group strata unconformably (Eisbacher 1978, 1981a, 1985). Basinal facies of the Coates Lake Group are black, pyritic and organic-rich; those of the Rapitan Group are maroon, hematitic and organic-poor.

Twitya, Keele and Ice Brook formations A major marine transgression followed the Rapitan glaciation, providing accommodation for 330–765 m of dark grey shale, siltstone and fine-grained sandstone of the Twitya Fm. (Eisbacher 1978, 1981a; Aitken 1982). Resting sharply upon boulder diamictite of the upper Shezal Fm., at the base of the Twitya Fm., are 0– 52 m of dark-grey, thin-bedded limestone (Eisbacher 1978, 1981a; Aitken 1982), which are absent in the Rapitan Group type section (Hayhook Lake) and thickest at Shale Lake (Fig. 36.1). Above the Twitya Fm. lie 220 –600 m of cyclic, shallow-marine sandstone, siltstone, limestone and dolomitized limestone of the

Keele Fm. (Day et al. 2004). A thin, jasper-pebble conglomerate occurs widely near its base. Glendonites, interpreted as pseudomorphs after ikaite (CaCO3.6H2O), occur close to the same horizon (James et al. 2005). The outer edge of the Keele platform is exposed near Shale Lake and on Stelfox Mountain (Fig. 36.1), where a SW-dipping ‘breakaway scarp’ (Aitken 1991b, p. 26) separates shallow-marine Keele strata on the footwall from hotel-size megaclast breccia (Durkan Member), turbiditic siltstone (Delthore Member) and glaciogenic diamictite (Stelfox Member) of the Ice Brook Fm. on the hanging wall (Aitken 1991b; Shen et al. 2008). The Stelfox diamictite is buttressed against the palaeoscarp and its feather-edge steps across the trace of the palaeoscarp on Stelfox Mountain (at 638350 5800 N, 1278520 0200 W) onto the Keele Formation of the footwall (Fig. 36.4b). The Ravensthroat Fm. passes across the fault line without displacement, proving that movement on the palaeoscarp ended before the glacial termination. Eisbacher (1981a) inferred that the Durkan megabreccia was triggered by glacioeustatic fall, but this interpretation was rejected by Aitken (1991b), who argued that pore-fluid overpressures would not develop in unconsolidated sediments unless the base-level fall was catastrophically rapid. He (Aitken 1991b) insisted that only the Stelfox Member was glacial in origin. P.F.H. therefore prefers the name Stelfox glaciation over Ice Brook glaciation, despite the latter’s priority in the literature (Kaufman et al. 1997).

Ravensthroat, Hayhook and Sheepbed formations As with the Rapitan glaciation, major marine flooding followed by Stelfox diamictite deposition, providing accommodation on the Keele platform for the formation of a transgressive, post-glacial carbonate (Raventhroat/Hayhook formations) and 446 –562 m of black shale and minor siltstone of the Sheepbed Fm. (Pyle et al. 2004; Shen et al. 2008). On the continental slope SW of the Keele platform, the Sheepbed Fm. reaches 1050 m in thickness at Sekwi Brook (Fig. 36.1) and includes significantly more siltstone, deposited by both turbidity flows and unidirectional (NW-directed) contour currents (Dalrymple & Narbonne 1996). The contourites are estimated to have formed in water depths of 1.0– 1.5 km and imply that the slope was open to the world ocean in a gulf .100 km wide (Dalrymple & Narbonne 1996).

Glaciogenic and associated strata Rapitan Group The remeasured type section of the Rapitan Group (Gabrielse et al. 1973) is shown in Figure 36.3b. The Sayunei Fm. (487 m) consists of maroon-coloured mudstone with innumerable graded beds of siltstone and/or fine-grained sandstone, some with outsized clasts of dolostone. Granule- to small pebble-sized lonestones are sprinkled throughout the mudstone. The clasts are commonly rimmed by chlorite, formed by reaction between the dolostone and the host mudstone. Some clasts are aggregates of sand or granules, interpreted as ‘till pellets’ (Ovenshine 1970) by Young (1976) and Eisbacher (1981a). Bedforms attributable to wave action are absent, as is evidence of bottom traction currents. Greenish mudstone beds occur, but none are dark grey or black, in stark contrast to the basinal facies of the underlying Coates Lake Group. In the Hayhook Lake area (Fig. 36.1), the top of the Sayunei Fm. is marked by 10–16 m of brick-red, hematitic mudstone and hematite jaspilite (Fig. 36.3c). This interval is parallel-laminated and lacks evidence for bottom traction currents. Within the Fe-rich formation are lonestones and thin dolostone-clast debrites (Young 1976). The former include rounded cobbles and small boulders of porphyritic quartz monzonite (i.e. two-feldspar

THE MACKENZIE MOUNTAINS, NORTHERN CANADIAN CORDILLERA

401

(a) o

155-210

NW

SE

50

50

G

Sheepbed Formation

G

o

200-220

R

G

20

198-210 R

G

R

o

Hayhook fm

30

190-200o

o

215

G

240o 190-210o

R

R

190o

195o

190o 195o

230o

210-230o

0

Ravensthroat fm

10

Sheepbed Fm

40

Ice Brook Fm (Stelfox Mb)

-10 aeolian(?)

-20 Keele Formation

40 30 20 10 0 -10 -20 -30

-30 Keele Fm

150-225

-40

o

o

200

-40

o

200

-50 Cranswick R. 65o05’54”N, 132o26’17”W

(b)

W

Arctic Red R. 64o56’05”N, 131o03’30”W

-50 Gayna R. 64o49’22”N, 130o28’27”W

red / green marly shale

RG

50

Stone Knife R. 64o40’36”N, 129o53’48”W

Moose Horn R. Ravensthroat R. 63o57’13”N, 63o17’05”N, 127o30’39”W 127o08’40”W

E 50

current ripples (flow azimuth) sea-floor aragonite cement

40

sea-floor barite cement giant wave ripples (crestline)

limestone dolostone (secondary) quartz sandstone quartz-sericite siltstone

30

Twitya R. 64o13’07”N, 128o38’11”W

LEGEND

black shale allodapic limestone peloidal dolostone (primary) carbonate-clast diamictite

40

Shale Lake 64o31’12”N, 129o29’09”W

corrugate stromatolite sheet-crack cement

30

crossbedding (flow azimuth)

Hayhook formation

20

20

Sheepbed Formation

10 170-205

o

275-280

o

295-300o

235o

Ravensthroat formation

0

10 0 -10

en, 1991b)

-10 Stelfox Member Ice Brook Formation

scarp” (Aitk

-20

Delthore Member Ice Brook Formation

-40

Keele Formation

-30

“breakaway

-30

-20

-40 -50

-50 Stelfox Mt. 63o36’08”N, 127o57’43”W

Stelfox Mt. 63o35’37”N, 127o55’11”W

Stelfox Mt. 63o36’00”N, 127o53’57”W

Stelfox Mt. 63o35’46”N, 127o53’04”W

Stelfox Mt. 63o36’01”N, 127o50’47”W

granitoid with 5–15% quartz). The Fe-formation is overlain disconformably by maroon-coloured diamictite (basal Shezal Fm.), which contains rounded and angular clasts from the lower unit. In some sections, the Fe-formation is missing and was probably removed by erosion at the base of the Shezal Fm.

Moose Horn R. 63o57’13”N, 127o30’39”W

Hayhook Lake 63o34’17”N, 127o11’44”W

Fig. 36.4. Stratigraphic relations based on columnar sections spanning the Stelfox Member and its post-glacial carbonate sequence (Raventhroat and Hayhook formations) in lines oriented parallel (a) and transverse (b) to the depositional strike of the Neoproterozoic passive margin. The ‘breakaway scarp’ (Aitken 1991b) marks the outer edge of the Keele Formation shelf. Megaclast breccia of the Durkan Member (not shown) conformably underlies the Delthore Member and both abut the ‘breakaway scarp’ at a buttress unconformity.

In the Iron Creek area (Fig. 36.3d), an aggregate thickness of 100– 120 m of hematite jaspilite with pea-sized jasper nodules (miniconcretions) is sandwiched between units of poorly stratified, boulder diamictite with subordinate beds and lenses of tabular crossbedded, medium-grained, well-rounded, monocrystalline, quartz

402

P. F. HOFFMAN & G. P. HALVERSON

sandstone (Yeo 1981, 1986; Klein & Beukes 1993). Outsized clasts and debrites occur within the jaspilite intervals, as do thin seams and lenses of Fe-carbonate (siderite). The Hayhook Lake area and Iron Creek sections represent respectively the ‘basinal’ and ‘transitional’ (more proximal) facies of Fe-formation distinguished in the Rapitan Group by Eisbacher (1985). The Shezal Fm. diamictite is mostly olive-green or grey in colour, except for the basal part, which is maroon and clearly derived from the underlying Sayunei Fm. (Fig. 36.3c). The clasts consist of dolomite (Little Dal Group), commonly stromatolitic, quartzite (Katherine Group), siltstone (Coates Lake Group and Tsezotene Fm.), basalt and gabbro. Extrabasinal clasts are rare. Faceted and striated clasts, particularly siltstone, are abundant and unequivocally glacial in origin (Young 1976; Eisbacher 1981a). Diamictite occurs in crudely tabular, poorly stratified, matrix-supported bodies up to 90 m thick. The matrix consists variably of wackestone, structureless mudstone, or mudstone with anastomosing (‘scaly‘) foliation (Fig. 36.3c) that is subparallel to bedding but unrelated to the axis of maximum tectonic compression. The Shezal Fm. (320 m thick) is unusually well-exposed in a river canyon 1.6 km NW of the type area (Fig. 36.3c), where diamictite units are separated by subordinate, recessive intervals of well-stratified siltstone, graded sandstone and dark-grey shale. A 6-m-thick duplex structure with thrusting directed toward the NE occurs in the reference section between –320 and –330 m (Fig. 36.3c). In this section, the tops of the diamictite bodies tend to be more sharply defined than their bases. No post-glacial carbonate is present in the type section of the Rapitan Group (Fig. 36.3b), but elsewhere the basal Twitya Fm. consists of thin graded beds of dark-grey, allodapic limestone (Eisbacher 1978, 1981a; Aitken 1982). The carbonate reaches a thickness of 52.5 m at Shale Lake. At Stone Knife River (Fig. 36.1), where 40 m of limestone is exposed, its lower part is hummocky cross-stratified. Near Corn Creek (Fig. 36.1), bordering the Wernecke Mountains, up to 300 m of massive dolomite (Mount Profeit Fm.) is laterally equivalent to the lower Twitya post-glacial limestone (Eisbacher 1981a).

Stelfox Member (Ice Brook Fm.) Whereas the Rapitan Group is up to 1500 m thick (Yeo 1981), the Stelfox Member is less than 40 m thick in most sections (Fig. 36.4) and the thickest, 308 m (Atiken 1991a), is also the deepest-water section of Stelfox diamictite in the Mackenzie Mountains. Basinward thickening could account for the modest thickness of the Stelfox diamictites, which are only preserved in shelf and upperslope settings, in contrast to the deep rift-basin setting of the Rapitan Group. At Shale Lake (Fig. 36.4a), the Stelfox diamictite is directly underlain by 5–9 m of well-sorted, well-rounded, mediumgrained, quartz sandstone of possible aeolian origin. Large-scale tabular cross-bedding indicates transport to the SSW (c. 2058 azimuth, n ¼ 2), consistent with easterly palaeowinds (see below). The presence of aeolianite beneath the diamictite would suggest that glacioeustasy preceded or outranked glacioisostasy. The Stelfox Member is typically recessive and masked by scree from the overlying Ravensthroat Fm., which along with its modest thickness accounts for its delayed recognition. The dominant lithology is massive to weakly stratified diamictite, in which rounded clasts of tan dolostone and grey limestone are supported by a matrix of brownish-tan carbonate-rich mudrock with dispersed grains of quartz sand and granules. The clasts are derived from the underlying Keele Fm. and the absence of quartzite clasts suggests that carbonates were the only components of the Keele Fm. that were strongly lithified at the time of glaciation. Beds and lenses of quartz-chert sandstone, micaceous wacke and polymictic debrite are subordinate components of the Stelfox Member. At Cranswick River section (Fig. 36.4a), the diamictite

has a terrigenous matrix and the carbonate clasts, freed in float, are visibly faceted and striated. Ice-rafted dropstones and ‘till pellets’ (Ovenshine 1970) are well-developed in laminated facies (Aitken 1991a, b).

Ravensthroat– Hayhook post-glacial carbonate couplet The Stelfox diamictite or its equivalent erosion surface is everywhere sharply overlain by a transgressive, 10–15-m-thick, cap dolostone (Fig. 36.4), previously known informally as the ‘Tepee dolostone’ (Eisbacher 1981a, 1985; Aitken 1991b). The pale, incandescent, yellowish-grey scree profusely generated by the informally defined Ravensthroat Fm. (James et al. 2001) could not be more unlike the jet-black feathers of its avian namesake. East of 1308W longitude, the dolostone is overlain by up to 15 m of ledge-forming grey or pink limestone (Fig. 36.4), informally defined as the Hayhook Fm. (James et al. 2001). West of 1308W longitude, it is overlain by an ascending sequence of red, green and grey shales with thin beds of current-rippled limestone (Fig. 36.4). This tricoloured sequence is remarkably similar to that overlying the homologous postglacial cap dolostones (Hambrey & Spencer 1987; Halverson et al. 2004) in East Greenland (e.g. Kap Weber, Andre´e Land) and East Svalbard (e.g. Ditlovtoppen, Ny Friesland). The Ravensthroat Fm. consists of micro- and macropeloidal dolostone with ubiquitous, low-angle, cross-lamination defined by normal and reverse-graded laminae (James et al. 2001). The peloids are structureless, sub-spherical aggregates of dolomicrite up to 3 mm in diameter. They commonly rest on abraded facets. Giant wave ripples (cf. Allen & Hoffman 2005) are well developed in many sections (Fig. 36.4). They are trochoidal in profile, with sharp crests and lobate troughs, and individual ripple trains aggrade for up to 1.4 m through bidirectional accretion of peloidal laminae. Their synoptic relief, crest to trough, is up to 40 cm. The structures are different in form and origin from peritidal ‘tepees’ (Assereto & Kendall 1977; Kendall & Warren 1987); they are not brecciated, lack syndepositional cements, and their crestlines are linear and parallel, not polygonal, in plan view (Eisbacher 1981a; Aitken 1991b; James et al. 2001). Their crestlines generally trend NE, at a high angle to the Keele shelf edge (Fig. 36.5). Decametric bioherms composed of corrugated stromatolites (James et al. 2001) oriented subparallel to the crests of giant wave ripples occur in some sections (Fig. 36.4). The corrugated

130o

1 3

1

128o

65o

Keele 4

shelf o

64

slope 50 km

132o

5

5

130o

1

Fig. 36.5. Mean orientations of crestlines (double bars) of giant wave ripples in the Ravensthroat Fm. at different locations. Numerals are the numbers of independent ripple trains measured at each location.

THE MACKENZIE MOUNTAINS, NORTHERN CANADIAN CORDILLERA

stromatolites are closely similar to those that are intergradational with ‘tubestone’ stromatolites in the homologous Keilberg cap dolostone in northern Namibia. Evidence of subaerial exposure is absent, but there can be no doubt that the Ravensthroat Fm. was deposited above wave-base and within the euphotic zone. In virtually all sections from Shale Lake to Ravensthroat River (Fig. 36.1), a distance of nearly 200 km, the uppermost 4– 10 cm of the Ravensthroat Fm. contains a remarkable digitate crust composed of sea-floor barite (BaSO4) cement, variably pseudomorphosed by calcite. At Shale Lake, the barite crust is developed directly on a train of giant wave ripples. Because of pseudomorphic replacement in some sections, the barite cement was formerly misinterpreted as an aragonite cement (Aitken 1991b, fig. 8) or as microdigitate stromatolites (James et al. 2001, fig. 10). The digits consist of rosettes or bundles of bladed crystals whose morphology is quite distinct from the prismatic (pseudohexagonal) habit of the former aragonite cements in the overlying Hayhook Fm. (James et al. 2001). The digits preserve growth lamellae, 1– 2 mm thick, and typically tilt to the SW due to preferential growth in the seaward direction. The spaces between the digits are filled by laminated, peloidal, ferroan dolomite, the dark brown colour of which contrasts with the pale colouration of the rest of the Ravensthroat Fm. The borders of the barite digits are ragged due to the growth of vertically standing barite crystals that lap out onto the upper surfaces of individual laminae of infilling peloidal dolomite. This provides textural proof that the barite digits formed simultaneously with the peloidal infills. This interpretation agrees with that of barite cements at the homologous post-glacial dolostone– limestone transition in central Australia (Kennedy 1996). The tips of the barite digits, marking the Ravensthroat – Hayhook transition, are typically ‘colonized’ by slender prismatic fans of pseudomorphosed aragonite, while at the same level the peloidal ferroan dolostone changes to calcimicrite, initially with some reworked dolostone. Locally, the barite crust was dissolved away instead of pseudomorphically replaced, and the barite cement layer is represented by a collapse microbreccia of ferroan dolostone clasts, overlain by calcimicrite of the basal Hayhook Fm.. The local occurrence of this microbreccia contributed to the misinterpretation of the contact as a karstic unconformity (James et al. 2001, fig. 9B). Numerous mesoscopic folds at the Ravensthroat –Hayhook contact, previously interpreted as karst channels (James et al. 2001, fig. 9A), are consistently oriented parallel to tectonic cleavage and regional folds. They are not karst channels because the primary stratification in both units is everywhere parallel to the contact. We interpret them as lobate-cuspate folds (Ramsay & Huber 1987) of tectonic origin and Late Cretaceous –Paleocene age, related to bedding-parallel shortening at the rheological interface between the stiff Ravensthroat dolostone and the weak Hayhook limestone. Contrary to James et al. (2001), we find no evidence of karstic unconformity, subaerial exposure or isostatic adjustment at the Ravensthroat –Hayhook contact. The Hayhook Fm. is 0 –15 m thick (Fig. 36.4) and consists of flaggy micrite with pseudomorphosed crystal fans of aragonite cement (James et al. 2001). The cement fraction generally increases upward and reaches nearly 100% in some sections, notably Shale Lake (Fig. 36.1). The detailed interplay between cement growth, burial by micrite and renewed cement development indicates precipitation at a free-face on the sea-floor.

Boundary relations with overlying and underlying non-glacial units Rapitan Group A low-angle unconformity separates the glaciogenic Rapitan Group from various units of the underlying Coates Lake and

403

Little Dal groups (Gabrielse et al. 1973; Eisbacher 1978, 1981a, 1985; Aitken 1982; Jefferson & Ruelle 1986). In the Hayhook Lake area (Fig. 36.1), the Coates Lake Group is onlapped by the Sayunei Fm. along the unconformity (Eisbacher 1978, 1981a). The presence of submarine carbonate-clast talus breccias in the lower Rapitan Group (beneath and within the lower Sayunei Fm.) on the northeastern margin of the outcrop belt near Shale Lake (648320 4100 N, 1298220 3700 W) suggests a master normal fault on that side of the basin. It may have been structurally inverted during orogenic contraction to form a part of the Plateau Thrust system. The maroon colour of the Sayunei stems in part from its derivation from the Coates Lake Group, which is dominated by reddish siltstones of the Redstone River Fm. Similarly, the olivegreen colouration of the overlying Shezal diamictites reflects its derivation from the grey carbonates and greenish basalts of the Little Dal Group. A knife-sharp, marine, flooding surface separates the Shezal Fm. from fetid, dark-grey, flaggy limestone and shale of the lower Twitya Fm. (Eisbacher 1978, 1981a). Where the Rapitan Group is absent, the Twitya transgresses older units disconformably (Narbonne & Aitken 1995).

Stelfox Member A low-angle unconformity (Fig. 36.4a) separates the Stelfox diamictite from shelf carbonates and clastics of the underlying Keele Fm. (Day et al. 2004). West of the palaeoscarp at Stelfox Mountain (Fig. 36.4b), a sharp disconformity separates the Stelfox diamictite from parallel-laminated, greenish-grey siltstone of the underlying Delthore Member (Aitken 1991b). As the Delthore Member was deposited below storm wave-base, glacial erosion may be required to account for the sharp disconformity. A major marine flooding surface separates the Stelfox Member, or the Keele Fm. where the Stelfox is absent, from the overlying Ravensthroat Fm. Where the basal contact of the dolostone is not covered by its own scree (e.g. Cranswick River, Arctic Red River, Stoneknife River), 10–20 cm of clay separate it from the underlying diamictite. The top of the Hayhook limestone is a conformable marine flooding surface overlain by organic-rich black shale of the Sheepbed Fm. (Pyle et al. 2004; Shen et al. 2008).

Chemostratigraphy Carbon isotopes Carbon-isotopic records have been published for the Little Dal and Coates Lake groups (Halverson 2006), the Twitya post-glacial limestone and Keele Fm. (Narbonne et al. 1994; Kaufman et al. 1997; Hoffman & Schrag 2002), and the Ravensthroat –Hayhook post-glacial carbonate (James et al. 2001). The Little Dal Group is enriched in 13C relative to the PDB standard, except for a depleted interval in the Upper Carbonate Fm. The overlying Coates Lake Group begins with a deep negative excursion to –8‰ in d13C, followed by a positive excursion to þ8‰ before settling back to modestly enriched values. Values of d13C in the Twitya post-glacial limestone increase up-section from – 2.5‰ at the base. Carbonates of the Keele Fm. are strongly enriched,  þ 8‰, but plunge to – 8‰ in the last 30 m of section beneath the Stelfox diamictite. The Ravensthroat Fm. is also depleted in 13C (Fig. 36.6) and values decline up-section. The Ravensthroat–Hayhook contact coincides with a –2‰ step function. This is broadly consistent with lowtemperature equilibrium fractionation between dolomite and calcite (Friedman & O’Neil 1977). Within the Hayhook Fm., d13C rises by c. 1‰ up-section, with no systematic difference in values between micro-drilled micrite and sea-floor cement (Fig. 36.6).

404

P. F. HOFFMAN & G. P. HALVERSON

(a) -20

-18

-16

-14

(per mil V-PDB) -8 -12 -10

-6

-4

-2

0

25

20

15

An alternative, less consequential, interpretation of the d18O data is that the micrite was preferentially altered by diagenetic fluids on account of its finer grain size compared with the cements, and therefore its greater surface-to-volume ratio. The d18O data are more depleted in 18O and display more scatter in the second section (Fig. 36.6b), at Ravensthroat River. This section illustrates the ease with which d18O, but not d13C, was diagenetically altered.

10 Sheepbed Fm

5

Hayhook fm

Strontium isotopes

Ravensthroat fm

0 (m)

Stelfox Mb

(b) 30

25

20

δ O δ C 18

13

15

calcite (micrite) calcite (cement)

10

dolopelmicrite James et al. (2001)

5

-20

-18

-16

-14

-8 -12 -10 (per mil V-PDB)

-6

-4

-2

0

0 (m)

Fig. 36.6. C- and O-isotope data from dolomite of the Ravensthroat Fm. and from coexisting micrite and sea-floor cement of the Hayhook formation at (a) Shale Lake and (b) Ravensthroat River.

Oxygen isotopes In order to test the possibility of low-temperature dolomite – calcite equilibrium fractionation, which would imply a sea-floor origin for the Ravensthroat dolomite, and also for the possibility that the Hayhook sea-floor cements and associated micrites might retain a record of bottom and surface waters, respectively, we plot unpublished O-isotope data for two sections that we sampled in detail (Fig. 36.6). Exceptional petrographic preservation of the sea-floor cements motivated this effort. At Shale Lake, the Ravensthroat-Hayhook contact coincides with a –6‰ step function in d18O, taking the micritic component to be representative of the surface waters, from which the Ravensthroat was also derived (Fig. 36.6a). The magnitude of the step function in d18O, roughly 3 that in d13C, is compatible with a lowtemperature equilibrium fractionation between dolomite and calcite (Friedman & O’Neil 1977). The sea-floor cements are consistently more enriched in 18O than the micritic component of the Hayhook (Fig. 36.6a), consistent with lower temperatures and/or higher salinities of the bottom waters, from which the cements were precipitated, relative to the surface waters. The cements become more enriched in 18O up-section (Fig. 36.6a), which could indicate an increase in the temperature and/or salinity gradient, or alternatively an increase in water depth, with time. The cements exhibit considerably more scatter in d18O than does the micrite, which could reflect the difficulty in micro-drilling cement without contamination from the micrite that fills in between the slender prisms of former aragonite and even their originally hollow interiors. Accordingly, the reduction in scatter towards the top of the Hayhook reflects the more massive and continuous nature of the uppermost cements.

New and previously published (Kaufman et al. 1993, 1997) 87 Sr/86Sr data from Sr-rich (300 –3000 ppm), low-Mn (Mn/ Sr , 0.1) limestones bracketing the Rapitan and Stelfox glaciations in the Mackenzie Mountains are given in Halverson et al. (2007). Non-radiogenic values of 0.70550 –0.70622 (n ¼ 10) are observed in the Little Dal Group, compared with 0.70644– 0.70669 (n ¼ 5) in the Coates Lake Group. No data are currently available for the post-Rapitan limestone (basal Twitya Fm.), but limestones directly beneath the Stelfox diamictite (uppermost Keele Fm.) have ratios of 0.70718–0.70720 (n ¼ 3). Sea-floor cements from the Hayhook post-glacial limestone are marginally less radiogenic at 0.70714– 0.70716 (n ¼ 4) than the Coates Lake Group, whereas the coexisting micrites are more radiogenic, with minimum values of 0.70751–0.70792 (n ¼ 5), which is consistent with preferential alteration. All reported Hayhook values are from samples with Sr concentrations .600 ppm and Mn/Sr , 0.1.

Sulphur isotopes and iron speciation Sulphur isotopes have been measured from carbonate-associated sulphate in the Ravensthroat and Hayhook formations, as well as from the barite sea-floor cement at the top of the Ravensthroat Fm. (M.T. Hurtgen, pers. comm. 2008). The d34SCAS and d34Sbarite values are mutually consistent and agree with values from correlative strata in Namibia (Hurtgen et al. 2006), but the data have yet to be published. Sulphide S isotopes and Fe-speciation (ratio of highly reactive Fe to total Fe) have been studied in organic-rich black shales of the Sheepbed Fm. (Fig. 36.3) in order to infer changes in the oxygenation of the atmosphere and the deep ocean, respectively, in the aftermath of the Stelfox glaciation (Shen et al. 2008). Large variability (.35‰) in d34Ssulphide is observed from bottom to top, indicating that sulphate was not limiting (i.e. .2 mM). This implies higher rates of oxidative weathering (the major source of marine sulphate) than before the glaciation (Hurtgen et al. 2002). There is a dramatic shift in the Fe-speciation ratio (FeHR/FeT), however, between 140 and 160 m above the base of the lithologically homogeneous black shale sequence that is 420 m thick in total. The average value below 140 m is 0.43 (n ¼ 32), indicating bottom water anoxia, whereas above 160 m it is 0.11 (n ¼ 27), which is consistent with oxygenated bottom waters (Raiswell & Canfield 1998; Lyons & Severmann 2006; Canfield et al. 2007).

Calcium isotopes Calcium isotopes (d44Ca) were measured in limestone of the lower Twitya Fm., and in dolostone and limestone of the Ravensthroat and Hayhook formations, respectively (Silva-Tamayo et al. 2010). The data from the first are consistent with mid-Cryogenian postglacial carbonates in Brazil, and the latter with basal Ediacaran post-glacial sequences in Brazil and Namibia (Silva-Tamayo et al. 2010). Compared with the younger sequences, the older d44Ca profiles appear basally truncated, similar to their sequence stratigraphies and d13C profiles (Hoffman & Schrag 2002). The

THE MACKENZIE MOUNTAINS, NORTHERN CANADIAN CORDILLERA

younger sequences exhibit paired negative and positive d44Ca excursions that span twice the entire Phanerozoic range of variability (Farkas et al. 2007). The paired anomaly in the lower Hayhook limestone is preceeded (Ravensthroat Fm.) and followed (upper Hayhook Fm.) by relatively stable values lying well within the Phanerozoic range (Silva-Tamayo et al. 2010). Equilibrium isotope fractionation during carbonate precipitation maintains seawater roughly 1.0‰ heavier than the sources and sinks of Ca in the ocean. Thus, imbalance between Ca input and output will cause deviations in the d44Ca of seawater, and therefore in marine carbonate. The magnitude of the equilibrium fractionation is expected to vary with temperature and carbonate precipitation rate, but relatively modest variability in d44Ca (n ¼ 10) within the Ravensthroat Fm. (Silva-Tamayo et al. 2010), deposited during global deglaciation when changes in temperature and precipitation rate may have been large, implies that those dependencies were either small or counteracting at that time. The paired negative and positive excursions of 1.0‰ apiece in the lower Hayhook limestone and correlatives (n ¼ 31) represents first, a large excess of Ca input over output, followed by the reverse, a large excess of Ca output over input (Silva-Tamayo et al. 2010).

Synglacial hematite jaspilite (Fe-formation), Rapitan Group REE geochemistry of the Sayunei jaspilite (Klein & Buekes, 1993) is described in Hoffman et al. (2011). Planavsky et al. (2010) present elemental data (P, Fe, Mn and Al) for jaspilite samples from the Sayunei Fm. (n ¼ 41) and correlative strata (Upper Tindir Group) in the Tatonduk inlier. In common with other Cryogenian glaciogenic jaspilites, the samples have much higher phosphate contents (average P/Fe ¼ 1.44%, n ¼ 44) than Palaeoproterozoic or Archaean Fe-formations (Planavsky et al. 2010). Iron isotopes (d57Fe) and accompanying rare-earth elements (REE), including redox-sensitive cerium anomalies (Ce/Ce*), have been measured through the 16.4-m-thick interval of hematitic mudstone and hematite-jaspilite (banded Fe-formation) at the top of the Sayunei Fm. in the Hayhook Lake area (Fig. 36.7). The

405

d57Fe values increase systematically upsection from – 0.7‰ to 1.2‰. The largest rise coincides with the transition from hematitic mudstone to hematite-jaspilite, which also corresponds to the highest Fe concentration (c. 58 wt% Fe2O3). The increase in d57Fe is accompanied by a rise in Ce/Ce* (Fig. 36.7).

Palaeontology Neoproterozoic palaeontology of the Mackenzie Mountains is summarized in Narbonne & Aitken (1995). When combined with contemporary work in the Wernecke Mountains (Narbonne & Hofmann 1987) and ongoing stratigraphic and palaeontological studies in correlative strata of the Tatonduk inlier, a pattern of change over time emerges that is broadly consistent with other areas (Macdonald et al. 2010a, b; Macdonald & Cohen 2011). A number of eukaryotic crown groups, both algal and protistan, had evolved and diversified before the Rapitan glaciation. The record of these groups is cryptic from shortly before the Rapitan glaciation until long after the Stelfox glaciation. Soft-bodied, Ediacara-type macrofossils, representing benthic polypoid and frond-like organisms, occur in the upper Ediacaran Blueflower Fm. (Fig. 36.1), where they coexist with infaunal burrows (simple, meandering and subordinate patterned) that extend into the overlying Risky Fm. (Narbonne & Aitken 1990; Narbonne 1994). In the same area, near Sekwi Brook (Fig. 36.1), simple radially symmetric body fossils and rare trace fossils occur in the middle Sheepbed Fm. (Narbonne & Aitken 1990; Narbonne 1994). The Stelfox glaciation is not recognized at Sekwi Brook, but at Shale Lake (Fig. 36.1), Fe-speciation data indicate that bottom waters became oxygenated just below the middle Sheepbed Fm. (Shen et al. 2008). An assemblage of simple, centimetric annuli and discs, thought to be biogenic, occur in the upper Twitya Fm. (Fig. 36.3a) of the Sayunei Range (Hofmann et al. 1990), south of Twitya River (Fig. 36.1). As possible sponge-grade metazoan fossils pre-dating the terminal Cryogenian glaciation, the ‘Twitya-discs’ have recently been joined by calcareous structures at different horizons

Fig. 36.7. Fe isotopes (d57Fe), wt% Fe2O3 and Ce anomaly data (Ce/Ce*) from the Sayunei Fe-formation in the Hayhook Lake reference section. Data tabulated in Halverson et al. (2011).

406

P. F. HOFFMAN & G. P. HALVERSON

in South Australia (Wallace & Woon 2008; Maloof et al. 2010), and a sponge biomarker in the Oman Salt Basin (Love et al. 2009).

Other characteristics The Crest deposit at Iron Creek (Fig. 36.3d), near Snake River, is a hematite –jaspilite interval 120 m thick within glaciogenic diamictites of the Rapitan Group (Yeo 1986; Klein & Beukes 1993). The deposit contains 5.6 billion tonnes of ‘ore’ averaging 47.2% Fe and regional reserves are estimated to exceed 18.6 billion tonnes (Yeo 1986). Because of remoteness and difficult terrain, the deposit has never been economic.

Palaeolatitude and palaeogeography Palaeomagnetic data from the Mackenzie Mountains relating to the Rapitan and Stelfox glaciations have been reviewed (Evans 2000; Evans & Raub 2011). Morris (1977) recognized three natural remanent magnetic components in the Rapitan Group, of which the one inferred to be the youngest (Cretaceous?) gives a high palaeolatitude and the other two give low palaeolatitudes. A larger data set obtained by Park (1997) gives broadly similar results, but only after anomalous declinations are corrected for large vertical-axis rotations of presumed structural origin. The high-inclination poles were acquired during tectonic folding and the low-inclination poles before that time, although Morris (1977) and Park (1997) disagreed on which of the two clusters of low-inclination poles was older. The palaeolatitudes of their inferred oldest components are 08 + 28 (Morris 1977) and 06 + 78 (Park 1997). Neither result is definitively syndepositional, but the near-equatorial palaeolatitudes agree with more recent palaeomagnetic results from mid-Neoproterozoic (800 –740 Ma) sedimentary rocks in western Laurentia (Weil et al. 2004, 2006). Shallow mafic igneous rocks provide more reliable palaeomagnetic results than sedimentary rocks because they are strongly magnetized, resist low-temperature remagnetization, have not been compacted (no flattening of palaeomagnetic inclination), and provide baked-contact tests of the primary age of remanent magnetization, which can be directly dated as the crystallization age of the rock. The Franklin Large Igneous Province (LIP) comprises comagmatic mafic dykes, sills and lavas extending collectively for 2500 km across Arctic Laurentia. The grand mean palaeopole (8.48N, 163.88E, A95 ¼ 2.88, n ¼ 78 sites) for the Franklin LIP (Denyszyn et al. 2009) places the Mount Harper area and the Mackenzie Mountains, respectively, at 21+3 and 18 + 38N palaeolatitude at the time of magmatism, which spanned the onset of the Rapitan glaciation (see below). These results for the Rapitan glaciation are more reliable than those from the Rapitan Group itself. There are no direct palaeomagnetic constraints on the Stelfox glaciation. McMechan (2000b) cites Park (1994) as giving a palaeolatitude of c. 358 for the Stelfox glaciation, but this was based on palaeopoles from the Risky Fm., which is c. 2 km stratigraphically above the Stelfox Member (Fig. 36.3a). Park (1994) estimated the Risky Fm. to be younger than the 615 Ma Long Range Dykes of Newfoundland and Labrador (Kamo et al. 1989; Kamo & Gower 1994). Although more results are needed for better statistics, the preliminary palaeopole for the Long Range Dykes (19.08N, 355.38E, A95 ¼ 17.48, n ¼ 5 dykes) places the Mackenzie Mountains near the palaeoequator at 615 Ma (McCausland et al. 2007).

Geochronological constraints In the Mount Harper area of the Ogilvie Mountains, 310 km west of Corn Creek (Fig. 36.1), up to 120 m of massive and bedded

diamictite, correlative with the lower Rapitan Group, was deposited during the waning stages of bimodal volcanism in an active rift basin (Mustard & Roots 1997; Macdonald et al. 2010b). The presence of faceted and striated clasts, bed-penetrating dropstones, common outsized and exotic clasts, and glacial push structures (Hart River inlier) within a marine succession support a glacial marine origin for the diamictite at an ice-sheet groundingline. A quartz-phyric rhyolite within the volcanic pile, stratigraphically beneath the diamictite, and a felsic tuff within diamictite, c. 60 m above the base of the glacial –periglacial sequence, give U –Pb zircon dates of 717.43 + 0.14 and 716.47 + 0.24 Ma (2s), respectively (Macdonald et al. 2010b). These dates are indistinguishable in age from mafic dykes and sills of the Franklin Large Igneous Province (Heaman et al. 1992; Denyszyn et al. 2009; Macdonald et al. 2010b), which straddled the palaeoequator at the time of their emplacement. The beginning of glacial marine sedimentation in the Mount Harper rift basin at 717 Ma is the best indicator of the onset of Rapitan glaciation in the northern Canadian Cordillera (Macdonald et al. 2010b). These data are consistent with maximum age constraints obtained from the Mackenzie Mountains and Windermere Supergroup, including a weighted mean 207Pb/206Pb date of 779 + 2.3 Ma (Harlan et al. 2003) obtained for two baddeleyite analyses from a gabbro sill (Carcajou Canyon Gabbro) intruding the Tsezotene Fm. (Fig. 36.3), and a weighted average 207Pb/206Pb date of 755 + 18 Ma obtained for two concordant but low-Pb zircon grains from a single leucogranite ‘dropstone’ within turbidites close to the base of the Sayunei Fm. near Shale Lake (Ross & Villeneuve 1997). The Stelfox glaciation has no independent radiometric age constraint. Mafic tuffs in the upper Keele Fm. (Fig. 36.3a) contain only xenocrystic zircons. The Stelfox glaciation is assumed, largely on the basis of its post-glacial carbonate sequence, to be correlative with the terminal Cryogenian glaciation (Knoll et al. 2006), which terminated in 635 Ma.

Discussion Palaeoenvironmental interpretation of the Rapitan Group The Rapitan Group is widely interpreted as a glacial marine succession because of the occurrence of faceted and striated clasts, sand aggregates (‘till pellets‘), erratic dropstones with impactrelated structures, and soft-sediment deformation horizons (e.g. push structures, scaly matrix in particular diamictite sheets) of apparent glacitectonic origin (Young 1976; Eisbacher 1981a, 1985; Yeo 1981; Aitken 1982; Macdonald et al. 2010b). Eisbacher (1985) presents a comprehensive facies model ranging from fluctuating grounding-line diamictites (Shezal Fm.) through transitional facies to distal proglacial turbidites and fine-grained suspension fallout accumulated in structurally controlled, deepwater basins (Sayunei Fm.). The distal nature of the Sayunei turbidites, characterized by Bouma C-D sequences, is emphasized (Yeo 1981; Eisbacher 1985) and the source of suspended sediment is inferred to be turbid meltwater plumes emanating from discharge sites at ice grounding-lines at the basin margin (Young 1976; Yeo 1981; Klein & Beukes 1993). Eisbacher (1985, p. 242, fig. 8) suggests that the paucity of dropstones and diamictite, respectively, in the Sayunei Fm. reflect a suppression of icebergs and a buttressing of ice grounding-lines by a ‘stable sea-ice cover’ (Dowdeswell et al. 2000). Such a permanent ice cover would account for the low organic content and deep maroon colour of the Sayunei mudrocks, in contrast to the basinal shale and limestone of the underlying non-glacial Copper Cap Fm., which are mostly fetid and black in colour. Removal of the permanent ice cover by climate warming would have allowed icebergs and ice grounding-lines to advance basinward at the time of the Sayunei-Shezal transition (Eisbacher 1985).

THE MACKENZIE MOUNTAINS, NORTHERN CANADIAN CORDILLERA

Palaeomagnetic data place the Rapitan rift basins in palaeolatitudes comparable to the present southern Red Sea Rift basin (Evans & Raub 2011). At the Last Glacial Maximum (LGM, c. 20 ka), arguably as severe a glaciation as any in the Phanerozoic eon, the lowest nearby glacial moraines, in the Ethiopian Highlands, are 3750 m above sea level (Umer et al. 2004). Air temperatures at that altitude are on average 26 8C colder than the same latitude at sea level, reached by Rapitan glaciers. This gives some sense of the contrast between the Rapitan glaciation and the LGM.

Origin of ‘transitional-facies’ Fe-formation A local hydrothermal source of Fe for transitional-facies Fe-formation (Fig. 36.3d) was favoured by Yeo (1981) and Young (1988, 2002). However, when compared with Neoarchaean and Palaeoproterozoic Fe-formations, the rare-earth element (REE) chemistry of the Rapitan Fe-formation is ‘much less distinctly influenced by hydrothermal input’, suggesting that such input was ‘highly diluted by ocean waters at Rapitan time’ (Klein & Beukes 1993). Moreover, where submarine volcanism was demonstratively active during the Rapitan glaciation (Mount Harper area), Fe-formation is absent. Where Fe-formation is present, in the Tatonduk inlier (Macdonald & Cohen 2011) and the Mackenzie Mountains, interbedded volcanic rocks are nonexistent (Macdonald et al. 2010b). Global Fe-speciation data suggest that Cryogenian and Ediacaran deep waters were ferruginous, rather than euxinic (Canfield et al. 2008). This raises the question why Fe-formation (hematite jaspilite) was only deposited during glacial times, and possibly only during the ‘Sturtian’ glaciation (Hoffman et al. 2011). If the Rapitan rift-basins were covered by thick permanent marine ice during fine-grained sedimentation of the Sayunei Fm. (Eisbacher 1985), primary productivity would have crashed, limiting bacterial sulphate reduction and allowing high concentrations of Fe in acidic waters in the absence of hydrothermal sources (Mikucki et al. 2009). The Crest Fe deposit at Iron Creek (Fig. 36.3d) is situated within a sequence of transitional-facies diamictites (Eisbacher 1985), where local oxidizing power capable of titrating dissolved Fe would have been provided by subglacial meltwater discharges at basin-margin ice grounding-lines, assuming air bubbles in the meteoric ice (compressed snow) contained oxygen. Low primary productivity related to the permanent ice cover as well as low seawater pH would account for elevated P/Fe ratios in the Fe-formation at Iron Creek and perhaps other synglacial Cryogenian Fe-formations (Planavsky et al. 2010). The explanation for the high P/Fe ratio favoured by Planavsky et al. (2010) involves ‘unprecedented continental P fluxes during postglacial and interglacial time periods, given the extraordinary extent and duration of Cryogenian ice cover and the high levels of P delivery expected from glaciated catchments’. The problem with their explanation is that six of the seven Cryogenian Fe-formations in their database were deposited before the termination of the first major Cryogenian glaciation.

Origin and stratigraphic localization of ‘basin-facies’ Fe-formation Why does Fe-formation in the basin facies (Eisbacher 1985) occur at the top of the Sayunei Fm., disconformably beneath the Shezal diamictite (Fig. 36.3b,c)? If we accept Eisbacher’s (1985) explanation that the advance of the Shezal grounding-line was triggered by the removal of permanent sea ice over the basin, two factors would have contributed to the deposition of Fe-formation. First, as the ice cover thinned and finally disappeared, anoxic and oxygenic photosynthesis could have precipitated Fe2O3-precursor from anoxic Fe(II)-rich basin waters. Second, after the ice cover was removed, air –sea gas exchange and wind-driven mixing

407

would have quickly oxygenated the surface waters, leading to abiotic Fe(III) precipitation. After the Sayunei basin became ice-free, the water vapour content of air blowing across the basin would have increased due to evaporation. This would likely have caused a positive change in the mass-balance of downwind ice sheets, due to increased precipitation and decreased melting on account of fog and clouds. On a ‘modern Snowball Earth’ (i.e. present geography with complete sea-ice cover, Voigt & Marotzke 2009), the Red Sea Rift would be isolated from the Indian Ocean because of the roughly 1.0 km fall in global mean sea level (glacioeustasy). As tropical sea-surface temperatures rose to melting point, sea ice on the Red Sea basin would disappear, while the Indian Ocean would remain ice-bound because of sea – glacial flow from the south (Warren et al. 2002; Goodman & Pierrehumbert 2003). Accordingly, the Shezal diamictite would represent the time interval between the opening of the Sayunei basin ‘oasis’ and overall glacial termination. Once illuminated, primary productivity in the basin, now crowded with icebergs, would be stimulated. Fe-isotope and Ce-anomaly data (Fig. 36.7) shed additional light on the basin-facies Fe-formation near Hayhook Lake (Fig. 36.3c). Controls on Fe-isotope variation in natural systems are not well known, and data from ancient rocks are still quite limited. Laboratory experiments indicate that broadly similar equilibrium and kinetic isotope fractionations are associated with both abiotic and biological Fe-oxidation pathways (Johnson & Beard 2006). Consider the d57Fe trend in the Fe-formation (Fig. 36.7) not in terms of a secular change in seawater Fe-isotope composition, but as a change in net Fe-isotope fractionation (D57Fehem-Fe2þ) between hematite (or ferric-oxyhydroxide precursor) and dissolved Fe(II) across a redox chemocline in the water column (Halverson et al. 2011). This could come about either because the Fe-isotope fractionation varied as a function of the Fe(II) concentration, assumed to increase with depth across the chemocline, or because of progressive oxidation of Fe(II) as it upwelled across the chemocline. Progressive oxidation would drive the isotopic composition of the dissolved Fe(II) to more 57Fe-depleted values, lowering the d57Fe of the hematite produced. In either case, the d57Fe of hematite should decrease from the base of the chemocline to the top. Accordingly, the upward increase in d57Fe observed within the Fe-formation (Fig. 36.7) suggests an increase in water depth with time, which is consistent with more reducing conditions up-section inferred from the Ce/Ce* data and with transgression inferred on sedimentological grounds over the same stratigraphic interval (Klein & Beukes 1993). As the Sayunei Fe-formation is directly overlain by the Shezal diamictite (Fig. 36.3), the rise in relative sea level is logically attributed to lithospheric downwarping under the load of the advancing Shezal ice sheet (Halverson et al. 2011).

Palaeoenvironmental interpretation of the Stelfox (Ice Brook) glaciation The Stelfox Member of the Ice Brook Fm. is a subaqueous unit interpreted as glaciomarine because of the presence of faceted and striated clasts in diamictite, sand and granule aggregates (‘till pellets’), dropstones with impact-related structures in finely laminated mudrocks, extremely angular quartz grains derived from well-rounded quartz arenites of the Keele Fm., and diamictite in a shallow-shelf setting where large-scale mass-flows are less likely to occur (Aitken 1991a, b). In discussing the Stelfox-equivalent Vreeland Diamictites in northeastern British Columbia, McMechan (2000a, b) described mud-rich diamictites and related fine-grained terrigenous facies as resembling deposits associated with temperate, wet-based glaciers in southern Alaska, in contrast to those of cold-based glaciers in Antarctica. She argued that they are inconsistent with the existence of a Cryogenian ‘Snowball’ Earth, assuming that its

408

P. F. HOFFMAN & G. P. HALVERSON

mean-annual tropical surface temperature was close to that of present-day Antarctica. She noted the prevalence of biogenic facies (e.g. diatomaceous ooze) beneath Antarctic ice shelves, on the Antarctic continental shelf and adjacent to the outlets of fast-flowing ice streams, implying a near-absence of fine-grained terrigenous input from meltwater plumes (Domack 1988; Anderson et al. 1991). Citing the same authors, she posited that significant meltwater flow requires surface melting and therefore summer air temperatures above 0 8C, which do not occur on Antarctica or on a ‘Snowball’ Earth (McMechan 2000b). Much has been learned about the Antarctic glacial regime since the studies cited by McMechan (2000a, b). More than 150 subglacial lakes have been found, some at the origins of fast-flowing ice streams, and they are not stagnant but are subject to active hydrodynamic exchange (Siegert et al. 2005; Wingham et al. 2006; Bell et al. 2007). Modelling suggests that basal melting occurs under most of the interior of the East Antarctic Ice Sheet and all of its outlet ice streams (Pattyn 2010). Recent collapse of the Larsen B ice shelf revealed the former existence of both terrigenous and biogenic sedimentation beneath the ice shelf (Domack et al. 2005; Damiani & Giorgetti 2008). Springtime diatom blooms on the East Antarctic margin may be triggered, not inhibited, by Fe-rich sub-glacial meltwater disharges, from which terrigenous mud subsequently settles (Leventer et al. 2006). There is no doubt that accumulation rates for muds are lower in polar than in temperate settings, but sedimentation rates in the Cryogenian are as yet unknown in the absence of biogenic facies, varve chronologies, or data on meteoritic dust content. We agree that ‘any model for Neoproterozoic glaciation should include a significant period of more moderate glacial conditions prior to the end of glaciation’ (McMechan 2000b), but we maintain that even the most extreme glaciation must experience more moderate conditions at its termination, especially in tropical marine environments, and that this is fully consistent with the hysteresis loop predicted by the Snowball Earth hypothesis (Hoffman 2009).

Calcium isotope record of the Stelfox glaciation and its aftermath The Snowball Earth hypothesis (Hoffman & Schrag 2002) provides a potential explanation for the paired negative and positive d44Ca excursions and for their stratigraphic location in the lower Hayhook Fm. (Silva-Tamayo et al. 2010). During a snowball glaciation, a steady rise in CO2 drives carbonate dissolution in the ocean (a source of Ca) and precludes carbonate deposition (Hoffman & Schrag 2002, fig. 10). This is consistent with the absence of primary carbonate in the relevant synglacial strata. The d44Ca of dissolved Ca in snowball brine should therefore fall towards the value of the Ca input. The first carbonate that precipitates from the brine will be depleted in 44Ca by c. 1.0‰ compared with normal carbonates. In the snowball aftermath, Ca input to the ocean from erosion is high, but rapid ocean warming and the more leisurely drawdown of CO2, through silicate weathering, drive excess Ca output (carbonate deposition) over input, raising d44Ca. During deglaciation, however, ocean surface waters are flooded by glacial meltwater, marine and terrestrial, and are thus isolated from the deep brine by an exceptionally stable density stratification (Shields 2005). Consequently, the 44Ca-depleted snowball signature is not observed in the Ravensthroat Fm., which was deposited in surface waters during the snowball meltdown (James et al. 2001). As melting neared completion, the flux of meltwater waned and ocean mixing by winds and tides resumed. The negative d44Ca excursion in the lower Hayhook Fm. (Silva-Tamayo et al. 2010) records the mixing of 44Ca-depleted snowball brine and the meltwater lid. It does not record the generation of the isotope anomaly in dissolved Ca of the brine, but rather its delayed expression as precipitated carbonate when the brine was mixed into the meltwater lid, which was saturated with

carbonate from dissolution of carbonate rock ‘flour’, exposed by deglaciation (Fairchild 1993). Once the ocean was mixed, the postglacial excess of Ca output over input drove the positive d44Ca excursion. Silva-Tamayo et al. (2010), who assume that the stratigraphic expression of the negative d44Ca excursion corresponds to its time of formation, attribute the negative excursion to Ca influx from weathering immediately after deglaciation. Their model requires that Ca input greatly exceeded output, which appears unlikely at a time of ocean warming (raising saturation) and mixing (facilitating degassing), when the surface ocean is known to have been critically oversaturated from the widespread occurrence of syndeglacial cap dolostones (Hoffman et al. 2007).

Correlation and palaeogeography The post-glacial carbonate sequences following the older and younger Cryogenian glaciations are stratigraphically, lithologically and isotopically distinct (Kennedy et al. 1998; Hoffman & Schrag 2002). They provide a surprisingly successful basis for distinguishing the glaciations globally (Knoll et al. 2006; Hoffman et al. 2011). As cosmopolitans, however, they are less well suited for palaeogeographic reconstruction. In a wide orogen like the North American Cordillera, the distinction between Neoproterozoic successions that are indigenous to Laurentia and those originating elsewhere is fundamental (Johnston 2008; Hildebrand 2009). All exposed Neoproterozoic rocks in the Mackenzie Mountains (Fig. 36.1) were thrust northeastward relative to cratonic Laurentia in the Upper Cretaceous –Paleocene. Sequence and chemostratigraphic comparison of successions deposited before, between and after the Cryogenian glaciations implies that the Tatonduk, Coal Creek, Hart River and Wernecke inliers of the Yukon Territory expose a Neoproterozoic succession continuous with that of the Mackenzie Mountains (Macdonald & Roots 2009; Macdonald et al. 2010a, b; Macdonald & Cohen 2011). The co-occurrence of 718– 716 Ma mafic magmatism in the Mount Harper area (Coal Creek inlier) and on cratonic Laurentia (Franklin LIP) therefore implies that the Neoproterozoic succession of the Ogilvie and Mackenzie Mountains formed on the margin of Laurentia. In contrast, the lithostratigraphically alien Katakturuk Dolomite (Ediacaran) in the northeastern Brooks Range, Arctic Alaska, is unlikely to have had a Laurentian affinity before Siluro-Devonian time, at the earliest (Macdonald et al. 2009). Accordingly, the stratigraphically underlying Hula Hula diamictite represents a Cryogenian glaciation of Chukotka-Arctic Alaska, not Laurentia, and the Mount Copleston volcanics within and beneath the diamictite are predicted to be asynchronous with the Franklin LIP. What about Neoproterozoic successions in the rest of the North American Cordillera? They host named glacial horizons in British Columbia (Toby and Vreeland), Idaho (Scout Mountain), Utah (Dutch Peak and Mineral Fork), California (Surprise and Wildrose) and Sonora (Mina el Mezquite). Conventionally, all have been assumed to originate at the rifted margin of Laurentia (Stewart 1972). In the Cordilleran collisional model (Johnston 2008; Hildebrand 2009), all except possibly the Mineral Fork originated elsewhere. In that model, they collided with the Laurentian margin in the Late Cretaceous as part of a composite ribbon continent named ‘Rubia’ (Hildebrand 2009) or ‘Saybia’ (Johnston 2008). Targeted detrital zircon dating of the various glacial horizons, and their host successions, should be an effective means of testing the Rubia/Saybia hypothesis. From the perspective of the Mackenzie Mountains, the sequence stratigraphy and lithology of Neoproterozoic successions assigned to the Windermere Supergroup in British Columbia (Ross et al. 1989) and to the South are not similar. The differences may reflect facies changes, as conventionally assumed, but it seems fair to conclude that Neoproterozoic strata of the North American Cordillera provide no grounds at present for rejecting the Rubia/Saybia hypothesis. To date, all

THE MACKENZIE MOUNTAINS, NORTHERN CANADIAN CORDILLERA

that we do know is that palaeomagnetic data from the Franklin LIP place the then northward-facing rifted margin of Laurentia in the Mackenzie Mountains at 18 + 38N latitude at the onset of the Rapitan glaciation, and possibly closer to the palaeo-equator during the Stelfox glaciation. Fieldwork in the Mackenzie Mountains was licensed by the Aurora Research Institute and supported by a grant from the Astrobiology Institute of the US National Aeronautics and Space Administration (NASA), and by grants EAR-9905495 and EAR-0417422 from the US National Science Foundation (NSF) to P.F.H. C. A. Ferguson, M. T. Hurtgen, F. A. Macdonald, A. C. Maloof, Y. Shen and A. V. Turchyn participated in the fieldwork. C and O isotopes were analysed at the Laboratory for Geochemical Oceanography at Harvard University, directed by D. P. Schrag. Sr isotope data were obtained at the Geochronology Laboratory at the Massachusetts Institute of Technology, directed by S. A. Bowring. Fe isotopes were measured at the Laboratoire des Me´canismes et Transferts en Ge´ologie (LMTG) at the Universite´ P. Sabatier, Toulouse, France, directed by F. Poitrasson. P.F.H. acknowledges discussions with J. D. Aitken, R. W. Dalrymple, E. W. Domack, G. H. Eisbacher, H. Gabrielse, L. M. Heaman, H. J. Hofmann, N. P. James, C. W. Jefferson, A. J. Kaufman, T. Kurt Kyser, M. E. McMechan, W. A. Morris, D. C. Murphy, G. M. Narbonne, J. K. Park, G. M. Ross, G. M. Yeo and G. M. Young. Comments by E. Arnaud, G. Narbonne, E. C. Turner and G. Yeo substantially improved the manuscript. This represents a contribution of the IUGS- and UNESCOfunded IGCP (International Geoscience Programme) project #512.

References Aitken, J. D. 1981. Stratigraphy and sedimentology of the Upper Proterozoic Little Dal Group, Mackenzie Mountains, Northwest Territories. In: Campbell, F. H. A. (ed.) Proterozoic Basins of Canada. Geological Survey of Canada Paper, 81 – 10, 47– 71. Aitken, J. D. 1982. Precambrian of the Mackenzie fold belt — a stratigraphic and tectonic overview. In: Hutchinson, R. W., Spence, C. D. & Franklin, J. M. (eds) Precambrian Sulfide Deposits. Geological Association of Canada Special Paper, 25, 149– 161. Aitken, J. D. 1991a. Two late Proterozoic glaciations, Mackenzie Mountains, northwestern Canada. Geology, 19, 445– 448. Aitken, J. D. 1991b. The Ice Brook Formation and post-Raptian, Late Proterozoic glaciation, Mackenzie Mountains, Northwest Territories. Geological Survey of Canada Bulletin, 404. Aitken, J. D. & Long, D. G. F. 1978. Mackenzie tectonic arc – reflection of early basin configuration? Geology, 6, 626–629. Allen, P. A. & Hoffman, P. F. 2005. Extreme winds and waves in the aftermath of a Neoproterozoic glaciation. Nature, 433, 123–127. Anderson, J. B., Kennedy, D. S., Smith, M. J. & Domack, E. W. 1991. Sedimentary facies associated with Antarctica’s floating ice masses. In: Anderson, J. B. & Ashley, G. M. (eds) Glacial Marine Sedimentation: Paleoclimatic Significance. Geological Society of America Special Paper, 261, 1– 25. Assereto, R. L. A. M. & Kendall, C. G. St. C. 1977. Nature, origin and classification of peritidal tepee structures and related breccas. Sedimentology, 24, 153– 210. Bell, R. E., Studinger, M., Shuman, C. A., Fahnestock, M. A. & Joughin, I. 2007. Large subglacial lakes in East Antarctica at the onset of fast-flowing ice streams. Nature, 445, 904– 907. Canfield, D. E., Poulton, S. W. & Narbonne, G. M. 2007. LateNeoproterozoic deep-ocean oxygenation and the rise of animal life. Science, 315, 92 – 95. Canfield, D. E., Poulton, S. W., Knoll, A. H., Narbonne, G. M., Ross, G., Goldberg, T. & Strauss, H. 2008. Ferruginous conditions dominated later Neoproterozoic deep-water chemistry. Science, 321, 949– 952. Dalrymple, R. W. & Narbonne, G. M. 1996. Continental slope sedimentation in the Sheepbed Formation (Neoproterozoic, Windermere Supergroup), Mackenzie Mountains, N.W.T. Canadian Journal of Earth Sciences, 33, 848–862. Damiani, D. & Giorgetti, G. 2008. Provenance of glacial-marine sediments under the McMurdo/Ross Ice Shelf (Windless Bight, Antarctica): Heavy minerals and geochemical data. Palaeogeography, Palaeoclimatology, Palaeoecology, 260, 262– 283.

409

Day, E. S., James, N. P., Narbonne, G. M. & Dalrymple, R. W. 2004. A sedimentary prelude to Marinoan glaciation, Cryogenian (Middle Neoproterozoic) Keele Formation, Mackenzie Mountains, northwestern Canada. Precambrian Research, 133, 223– 247. Denyszyn, S. W., Halls, H. C., Davis, D. W. & Evans, D. A. D. 2009. Paleomagnetism and U– Pb geochronology of Franklin dykes in High Arctic Canada and Greenland: a revised age and paleomagnetic pole constraining block rotations in the Nares Strait region. Canadian Journal of Earth Sciences, 46, 689– 705. Domack, E. W. 1988. Biogenic facies in the Antarctic glacimarine environment: basis for a polar glacimarine summary. Palaeogeography, Palaeoclimatology, Palaeoecology, 63, 357– 372. Domack, E. W., Duran, D. et al. 2005. Stability of the Larsen B ice shelf on the Antarctic Peninsula during the Holocene epoch. Nature, 436, 681– 685. Dowdeswell, J. A., Whittington, J. A., Jennings, A. E., Andrews, J. T., Mackensen, A. & Marienfield, P. 2000. An origin for laminated glacimarine sediments through sea-ice build-up and suppressed iceberg rafting. Sedimentology, 47, 557– 576. Eisbacher, G. H. 1978. Re-definition and Subdivision of the Rapitan Group, Mackenzie Mountains. Geological Survey of Canada Paper 77– 35. Eisbacher, G. H. 1981a. Sedimentary tectonics and glacial record in the Windermere Supergroup, Mackenzie Mountains, northwestern Canada. Geological Survey of Canada Paper, 80– 27. Eisbacher, G. H. 1981b. Late Precambrian tillites of the northern YukonNorthwest Territories region, Canada. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 724–727. Eisbacher, G. H. 1985. Late Proterozoic rifting, glacial sedimentation and sedimentary cycles in the light of Windermere deposition, western Canada. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 231– 254. Evans, D. A. D. 2000. Stratigraphic, geochronological and Paleomagnetic constraints upon the Neoproterozoic climatic paradox. American Journal of Science, 300, 347–433. Evans, D. A. D. & Raub, T. D. 2011. Neoproterozoic glacial Palaeolatitudes: a global update. In: Arnaud, E., Halverson, G. P. & Shields, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 93 –112. Fairchild, I. J. 1993. Balmy shores and ice wastes: the paradox of carbonates associated with glacial deposits in Neoproterozoic times. Sedimentology Review, 1, 1 –16. Farkas, J., Bo¨hm, F. et al. 2007. Calcium isotope record of Phanerozoic oceans: implications for chemical evolution of seawater and its causative mechanisms. Geochimica et Cosmochimica Acta, 71, 5117–5134. Friedman, I. & O’Neil, J. R. 1977. Chapter KK. Compilation of stable isotope fractionation factors of geochemical interest. In: Fleischer, M. (ed.) Data of Geochemistry 6th edn. United States Geological Survey Professional Paper 440– KK, Washington, DC. Gabrielse, H., Blusson, S. L. & Roddick, J. A. 1973. Geology of the Flat River, Glacier Lake and Wrigley Lake map-areas, District of Mackenzie and Yukon Territory. Geological Survey of Canada Memoir 366. Goodman, J. & Pierrehumbert, R. T. 2003. Glacial flow of floating marine ice in ‘Snowball Earth’. Journal of Geophysical Research, 108, 3308, doi: 10.1029/2002JC001471. Green, L. H. & Godwin, C. I. 1963. Snake River area: mineral industry of Yukon Territory and southwestern District of Mackenzie, 1962. Geological Survey of Canada Paper 63 –38, 15 – 18. Halverson, G. P. 2006. A Neoproterozoic chronology. In: Xiao, S. & Kaufman, A. J. (eds) Neoproterozoic Geobiology and Paleobiology. Springer, Dordrecht, 231– 271. Halverson, G. P., Maloof, A. C. & Hoffman, P. F. 2004. The Marinoan glaciation (Neoproterozoic) in northeast Svalbard. Basin Research, 16, 297– 324. ¨ ., Maloof, A. C. & Bowring, S. A. 2007. Halverson, G. P., Duda´s, F. O Evolution of the 87Sr/86Sr composition of Neoproterozoic seawater. Palaeogeography, Palaeoclimatology, Palaeoecology, 256, 103– 129.

410

P. F. HOFFMAN & G. P. HALVERSON

Halverson, G. P., Poitrasson, F., Hoffman, P. F., Nedelec, A., Montel, J.-M. & Kirby, J. 2011. Fe isotope and trace element geochemistry of the Neoproterozoic syn-glacial Raptian iron formation. Earth and Planetary Science Letters, doi: 10.1016/j.epsl.2011.06.21. Hambrey, M. J. & Spencer, A. M. 1987. Late Precambrian Glaciation of Central East Greenland. Meddelelser om Grønland, Geoscience 19. Harlan, S. S., Heaman, L., LeCheminant, A. N. & Premo, W. R. 2003. Gunbarrel mafic magmatic event: a key 780 Ma time marker for Rodinia plate reconstructions. Geology, 31, 1053–1056. Heaman, L. M., LeCheminant, A. N. & Rainbird, R. H. 1992. Nature and timing of Franklin igneous events, Canada: implications for a Late Proterozoic mantle plume and the break-up of Laurentia. Earth and Planetary Science Letters, 109, 117–131. Hildebrand, R. S. 2009. Did westward subduction cause CretaceousTertiary orogeny in the North American Cordillera? Geological Society of America, Special Paper 457. Hoffman, P. F. 2009. Pan-glacial – a third state in the climate system. Geology Today, 25, 107– 114. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129–155. Hoffman, P. F. & Macdonald, F. A. 2010. Sheet-crack cements and early regression in Marinoan (635 Ma) cap dolostones: regional benchmarks of vanishing ice-sheets? Earth and Planetary Science Letters, 300, 374–384. Hoffman, P. F., Halverson, G. P., Domack, E. W., Husson, J. M., Higgins, J. A. & Schrag, D. P. 2007. Are basal Ediacaran (635 Ma) post-glacial ‘cap dolostones’ diachronous? Earth and Planetary Science Letters, 258, 114–131. Hoffman, P. F., Macdonald, F. A. & Halverson, G. P. 2011. Chemical sediments associated with Neoproterozoic glaciation: iron formation, cap carbonate, barite and phosphorite. In: Arnaud, E., Halverson, G. P. & Shields, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 67 –80. Hofmann, H. J., Narbonne, G. M. & Aitken, J. D. 1990. Ediacaran remains from intertillite beds in northwestern Canada. Geology, 18, 1199– 1202. Hurtgen, M. T., Arthur, M. A., Suits, N. S. & Kaufman, A. J. 2002. The sulfur isotopic composition of Neoproterozoic seawater sulfate: implications for a snowball Earth? Earth and Planetary Science Letters, 203, 413–429. Hurtgen, M. T., Halverson, G. P., Arthur, M. A. & Hoffman, P. F. 2006. Sulfur cycling in the aftermath of a 635-Ma snowball glaciation: evidence for a syn-glacial sulfidic deep ocean. Earth and Planetary Science Letters, 245, 551–570. James, N. P., Narbonne, G. M. & Kyser, T. K. 2001. Late Neoproterozoic cap carbonates: Mackenzie Mountains, northwestern Canada: precipitation and global glacial meltdown. Canadian Journal of Earth Science, 38, 1229– 1262. James, N. P., Narbonne, G. M., Dalrymple, R. W. & Kyser, T. K. 2005. Glendonites in Neoproterozoic low-latitude, interglacial, sedimentary rocks, northwest Canada: insights into Cryogenian ocean and Precambrian cold-water carbonates. Geology, 33, 9 – 12. Jefferson, C. W. 1983. The Upper Proterozoic Redstone Copper Belt, Mackenzie Mountains, Northwest Territories. PhD thesis, University of Western Ontario, London, Ontario, 445. Jefferson, C. W. & Ruelle, J. C. L. 1986. The Late Proterozoic Redstone Copper Belt, Mackenzie Mountains, Northwest Territories. In: Morin, J. A. (ed.) Mineral Deposits of Northern Cordillera. Canadian Institute of Mining and Metallurgy Special Volume, 37, 154– 168. Johnson, C. M. & Beard, B. L. 2006. Fe isotopes: an emerging technique for understanding modern and ancient biogeochemical cycles. GSA Today, 16, 4– 10. Johnston, S. T. 2008. The Cordilleran ribbon continent of North America. Annual Reviews of Earth and Planetary Sciences, 36, 495– 530. Kamo, S. L. & Gower, C. F. 1994. Note: U– Pb baddeleyite dating clarifies age of characteristic paleomagnetic remanence of Long Range dykes, southeastern Labrador. Atlantic Geology, 30, 259– 262. Kamo, S. L., Gower, C. F. & Krogh, T. E. 1989. Birthplace for the Iapetus Ocean? A precise U –Pb zircon and baddeleyite age for the Long Range dikes, southeast Labrador. Geology, 17, 602–605.

Kaufman, A. J., Jacobsen, S. B. & Knoll, A. H. 1993. The Vendian record of Sr and C isotopic variations in seawater: implications for tectonics and paleoclimate. Earth and Planetary Science Letters, 120, 409– 430. Kaufman, A. J., Knoll, A. H. & Narbonne, G. M. 1997. Isotopes, ice ages and terminal Proterozoic earth history. Proceedings of the National Academy of Sciences (USA), 94, 6600– 6605. Kendall, C. G. St. C. & Warren, J. 1987. A review of the origin and setting of tepees and their associated fabrics. Sedimentology, 34, 1007– 1027. Kennedy, M. J. 1996. Stratigraphy, sedimentology and isotopic geochemistry of Australian Neoproterozoic postglacial cap dolostones: deglaciation, d13C excursions and carbonate precipitation. Journal of Sedimentary Research, 66, 1050– 1064. Kennedy, M. J., Runnegar, B., Prave, A. R., Hoffmann, K.-H. & Arthur, M. A. 1998. Two or four Neoproterozoic glaciations? Geology, 26, 1059–1063. Klein, C. & Beukes, N. J. 1993. Sedimentology and geochemistry of the glaciogenic Late Proterozoic Rapitan Fe-formation in Canada. Economic Geology, 88, 542–565. Knoll, A. H., Walter, M. R., Narbonne, G. M. & Christie-Blick, N. 2006. The Ediacaran Period: a new addition to the geologic time scale. Lethaia, 39, 13 – 30. Leventer, A., Domack, E. et al. 2006. Marine sediment record from the East Antarctic margin reveals dynamics of ice sheet recession. GSA Today, 16, 4– 10. Love, G. D., Grosjean, E. et al. 2009. Fossil steroids record the appearance of Demospongiae during the Cryogenian period. Nature, 457, 718– 721, doi: 10.1038/nature07673. Lyons, T. W. & Severmann, S. 2006. A critical look at iron paleoredox proxies based on new insights from modern euxinic marine basins. Geochimica et Cosmochimica Acta, 70, 5698– 5722. Macdonald, F. A. & Roots, C. F. 2009. Upper Fifteenmile Group in the Ogilvie Mountains and correlations of early Neoproterozoic strata in the northern Cordillera. In: MacFarlane, K. E., Weston, L. H. & Blackburn, L. R. (eds) Yukon Exploration and Geology 2009, Yukon Geological Survey, Whitehorse, 237–252. Macdonald, F. A. & Cohen, P. A. 2011. The Tatonduk inlier, Alaska-Yukon border. In: Arnaud, E., Halverson, G. P. & Shields, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 389–396. Macdonald, F. A., Mcclelland, W. C., Schrag, D. P. & Macdonald, W. P. 2009. Neoproterozoic glaciation on a carbonate platform margin in Arctic Alaska and the origin of the North Slope subterrane. Geological Society of America Bulletin, 121, 448–473. ¨ . & Schrag, D. P. 2010a. Macdonald, F. A., Cohen, P. A., Duda´s, F. O Early Neoproterozoic siliceous scale microfossils in the Lower Tindir Group of Alaska and the Yukon Territory. Geology, 38, 143– 146. Macdonald, F. A., Schmitz, M. D. et al. 2010b. Calibrating the Cryogenian. Science, 327, 1241– 1243. Maloof, A. C., Rose, C. V. et al. 2010. Possible animal-body fossils in pre-Marinoan limestones from South Australia. Nature Geoscience, 3, 653– 659. McCausland, P. J. A., Van der Voo, R. & Hall, C. M. 2007. Circum-Iapetus paleogeography of the Precambrian– Cambrian transition with a new paleomagnetic constraint from Laurentia. Precambrian Research, 156, 125– 152. McMechan, M. E. 2000a. Vreeland Diamictites – Neoproterozoic glaciogenic slope deposits, Rocky Mountains, northeast British Columbia. Bulletin of Canadian Petroleum Geology, 48, 246– 261. McMechan, M. E. 2000b. Reply to discussion. Vreeland Diamictites – Neoproterozoic glaciogenic slope deposits, Rocky Mountains, northeast British Columbia. Bulletin of Canadian Petroleum Geology, 48, 364– 366. Mikucki, J. A., Pearson, A. et al. 2009. A contemporary microbially maintained subglacial ferrous ‘ocean’. Science, 324, 397– 400. Morris, W. A. 1977. Paleolatitude of upper Precambrian glaciogenic Rapitan Group and the use of tillites as chronostratigraphic marker horizons. Geology, 5, 85 – 88. Mustard, P. S. & Roots, C. F. 1997. Rift-related volcanism, sedimentation and tectonic setting of the Mount Harper Group, Ogilvie Mountains, Yukon Territory. Geological Survey of Canada Bulletin, 492.

THE MACKENZIE MOUNTAINS, NORTHERN CANADIAN CORDILLERA

Narbonne, G. M. 1994. New Ediacaran fossils from the Mackenzie Mountains, northwestern Canada. Journal of Palaeontology, 68, 411– 416. Narbonne, G. M. & Hofmann, H. J. 1987. Ediacaran biota of the Wernecke Mountains, Yukon, Canada. Palaeontology, 30, 647– 676. Narbonne, G. M. & Aitken, J. D. 1990. Ediacaran fossils from the Sekwi Brook area, Mackenzie Mountains, northwestern Canada. Palaeontology, 33, 945– 980. Narbonne, G. M. & Aitken, J. D. 1995. Neoproterozoic of the Mackenzie Mountains, northwestern Canada. Precambrian Research, 73, 101– 121. Narbonne, G. M., Kaufman, A. J. & Knoll, A. H. 1994. Integrated chemostratigraphy and biostratigraphy of the Windermere Supergroup, northwestern Canada: implications for Neoproterozoic correlations and early evolution of animals. Geological Society of America Bulletin, 106, 1281– 1292. Ovenshine, T. A. 1970. Observations of ice rafting in Glacier Bay, Alaska, and the identification of ancient ice-rafted deposits. Geological Society of America Bulletin, 81, 891–894. Park, J. K. 1994. Palaeomagnetic constraints on the position of Laurentia from middle Neoproterozoic to Early Cambrian times. Precambrian Research, 69, 95 – 112. Park, J. K. 1997. Paleomagnetic evidence for low-latitude glaciation during deposition of the Neoproterozoic Rapitan Group, Mackenzie Mountains, N.W.T., Canada. Canadian Journal of Earth Sciences, 34, 34 – 49. Pattyn, F. 2010. Antarctic subglacial conditions inferred from a hybrid ice sheet/ice stream model. Earth and Planetary Science Letters, 295, 451– 461. Planavksy, N. J., Rouxel, O. J., Bekker, A., Lalonde, S. V., Konhauser, K. O., Reinhard, C. T. & Lyons, T. W. 2010. The evolution of the marine phosphate reservoir. Nature, 467, 1088– 1090. Pyle, L. J., Narbonne, G. M., James, N. P., Dalrymple, R. W. & Kaufman, A. J. 2004. Integrated Ediacaran chronostratigraphy, Wernecke Mountains, northwestern Canada. Precambrian Research, 132, 1– 27. Raiswell, R. & Canfield, D. E. 1998. Sources of iron for pyrite formation in marine sediments. American Journal of Science, 298, 219– 245. Ramsay, J. G. & Huber, M. I. 1987. The Techniques of Modern Structural Geology, Vol. 2. Academic Press, New York. Ross, G. M. 1991. Tectonic setting of the Windermere Supergroup revisited. Geology, 19, 1125– 1128. Ross, G. M., Villeneuve, M. E. 1997. U– Pb geochronology of stranger stones in Neoproterozoic diamictites, Canadian Cordillera: implications for provenance and ages of deposition. In: Radiogenic Age and Isotopic Studies: Report 10. Geological Survey of Canada Current Research, 1997-F, 141– 155. Ross, G. M., McMechan, M. E. & Hein, F. J. 1989. Proterozoic history: the birth of the miogeocline. In: Rickets, B. D. (ed.) Western Canada Sedimentary Basin: A Case History. Canadian Society of Petroleum Geologists, Calgary, 79 – 104. Ruelle, J. C. L. 1982. Depositional environments and genesis of stratiform copper deposits of the Redstone Copper Belt, Mackenzie Mountains, N.W.T. In: Hutchinson, R. W., Spence, C. D. & Franklin, J. M. (eds) Precambrian Sulfide Deposits. Geological Association of Canada Special Paper, 25, 701–737. Shen, Y., Zhang, T. & Hoffman, P. F. 2008. On the co-evolution of Ediacaran oceans and animals. Proceedings of the National Academy of Sciences USA, 105, 7376–7381. Shields, G. A. 2005. Neoproterozoic cap carbonates: a critical appraisal of existing models and the plumeworld hypothesis. Terra Nova, 17, 299 – 310. Siegert, M. J., Carter, S., Tabacco, I., Popov, S. & Blankenship, D. 2005. A revised inventory of Antarctic sub-glacial lakes. Antarctic Science, 17, 453– 460.

411

Silva-Tamayo, J. C., Na¨gler, T. F. et al. 2010. Global Ca isotope variations in c. 0.7 Ga old post-glacial carbonate successions. Terra Nova, 22, 188– 194. Stewart, J. H. 1972. Initial deposits in the Cordilleran Geosyncline: evidence of a Late Precambrian (,850 m.y.) continental separation. Geological Society of America Bulletin, 83, 1345– 1360. Turner, E. C. & Long, D. G. F. 2008. Basin architecture and syndepositional fault activity during deposition of the Neoproterozoic Mackenzie Mountains supergroup, Northwestern Territories, Canada. Canadian Journal of Earth Sciences, 45, 1159– 1184. Turner, E. C., James, N. P. & Narbonne, G. M. 1997. Growth dynamics of Neoproterozoic calcimicrobial reefs, Mackenzie Mountains, northwest Canada. Journal of Sedimentary Research, 67, 437– 450. Umer, M., Kebede, S. & Osmaston, H. 2004. Quaternary glacial activity on the Ethiopian mountains. In: Ehlers, J. & Gibbard, P. L. (eds) Quaternary Glaciations – Extent and Chronology, Part III: South America, Asia, Africa, Australia, Antarctica. Elsevier, Amsterdam, 171–174. Upitis, U. 1966. The Rapitan Group, southwestern Mackenzie Mountains, Northwest Territories. MSc thesis, McGill University, Montreal. Voigt, A. & Marotzke, J. 2009. The transition from the present-day climate to a modern Snowball Earth. Climate Dynamics, doi: 10.1007/s00382-009-0633-5. Wallace, M. W. & Woon, E. 2008. Giant Cryopgenian reefs as windows into pre-Ediacaran life. In: Gallagher, S. J. & Wallace, M. W. (eds) Selwyn Symposium 2008: Neoproterozoic Extreme Climates and the Origin of Early Metazoan Life. Geological Society of Australia, Extended Abstracts, 91, 17 – 22. Warren, S. G., Brandt, R. E., Grenfell, T. C. & McKay, C. P. 2002. Snowball Earth: ice thickness on the tropical ocean. Journal of Geophysical Research, 107(C10), 3167, doi: 10.1029/2001JC001123. Weil, A. B., Geismann, J. W. & Van der Voo, R. 2004. Paleomagnetism of the Neoproterozoic Chuar Group, Grand Canyon Supergroup, Arizona: implications for Laurentia’s Neoproterozoic APWP and Rodinia break-up. Precambrian Research, 129, 71 –92. Weil, A. B., Geismann, J. W. & Ashby, J. M. 2006. A new paleomagnetic pole for the Uinta Mountain supergroup, Central Rocky Mountain States, USA. Precambrian Research, 147, 234– 259. Wingham, D. J., Siegert, M. J., Shepherd, A. & Muir, A. S. 2006. Rapid discharge connects Antarctic subglacial lakes. Nature, 440, 1033–1036. Yeo, G. M. 1981. The Late Proterozoic Rapitan glaciation in the northern Cordillera. In: Campbell, F. H. A. (ed.) Proterozoic Basins of Canada. Geological Survey of Canada Paper, 81 –10, 25 –46. Yeo, G. M. 1984. The Rapitan Group: relevance to the global association of Late Proterozoic glaciation and Fe-formation. PhD thesis, University of Western Ontario, London (Ontario). Yeo, G. M. 1986. Fe-formation in the late Proterozoic Rapitan Group, Yukon and Northwest Territories. In: Morin, J. A. (ed.) Mineral Deposits of Northern Cordillera. Canadian Institute of Mining and Metallurgy, Special Volume, 37, 142– 153. Young, G. M. 1976. Fe-formation and glaciogenic rocks of the Rapitan Group, Northwest Territories, Canada. Precambrian Research, 3, 137– 158. Young, G. M. 1982. The late Proterozoic Tindir Group, east-central Alaska: evolution of a continental margin. Geological Society of America Bulletin, 93, 759–783. Young, G. M. 1988. Proterozoic plate tectonics, glaciation and Fe-formations. Sedimentary Geology, 58, 127–144. Young, G. M. 2002. Stratigraphic and tectonic settings of Proterozoic glaciogenic rocks and banded Fe-formations: relevance to the snowball Earth debate. Journal of African Earth Sciences, 35, 451– 466. Ziegler, P. 1959. Fru¨hpala¨ozoische Tillite im o¨stlichen YukonTerritorium (Kanada). Eclogae Geologicae Helvetiae, 52, 735– 741.

Chapter 37 The record of Neoproterozoic glaciations in the Windermere Supergroup, southern Canadian Cordillera MARK D. SMITH1, EMMANUELLE ARNAUD2*, R.W.C. ARNOTT1 & GERALD M. ROSS3 1

Department of Earth Sciences, University of Ottawa, Ottawa, Ontario, K1N 6N5, Canada

2

School of Environmental Sciences, University of Guelph, Guelph, Ontario, N1G 2W1, Canada 3

Kupa’a Farm, P.O.Box 458, Kula, Hawai’i, 96790, USA

*Corresponding author (e-mail: [email protected]) Abstract: The Neoproterozoic Windermere Supergroup (WSG) is exposed over an area of 35 000 km2 in the southern Canadian Cordillera, and consists primarily of deep-marine meta-sedimentary rocks interpreted to have been deposited during rifting and subsequent post-rift thermal relaxation. The main exposures of the WSG occur within thrust panels and structural culminations of the eastern Cordilleran orogen. Within the thick stratigraphic succession (c. 9 km) are three units of glaciogenic affinity: Toby, Vreeland and Old Fort Point (OFP) formations. The Toby Formation (Fm.) is composed of up to 2500 m of diamictite, interbedded with conglomerate, sandstone, mudstone, carbonate and mafic volcanic rocks. The Vreeland Formation ranges from 350 m to 2000 m in thickness and consists of diamictite, interbedded with mudstone, sandstone and conglomerate. The OFP ranges from 60 to 450 m in thickness and consists of a distinctive threefold stratigraphic package of basal siltstone grading upward into limestone –siltstone rhythmite, organic-rich mudstone and an overlying heterolithic unit of diamictite, breccia, conglomerate, sandstone, siltstone to mudstone and limestone. A locally prominent unconformity marks the base of the OFP upper member. Both the Toby and Vreeland formations represent remobilized glacially derived marine sediments deposited by sediment-gravity flows. Deposition of the Toby Fm. was fault-controlled during an active tectonic phase (rifting), whereas the Vreeland Fm. accumulated during the subsequent quiescent phase (post-rift) with limited structural control. The OFP is interpreted to be a regionally extensive deep-marine post-glacial marker temporally associated with the glaciogenic Vreeland Fm. Although direct geochronological ages for WSG units in southwestern Canada are generally absent, highprecision radiometric ages of underlying and overlying igneous events constrain the relative maximum and minimum timing of deposition from c. 740– 728 Ma to c. 569 Ma. At the base of the WSG succession, the Toby Fm. may be as young as c. 685 Ma based on ages obtained from potential stratigraphic correlatives in the USA. There is no direct age constraint for the deposition of the Vreeland Fm.; its minimum timing is based on its stratigraphic relationship with the post-glacial OFP. The middle member of the OFP was precisely dated at 607.8 + 4.7 Ma using the Re–Os method, placing it in the Ediacaran Period. Published geochemical and stable isotopic data are similarly limited for all three units with only some d34Spy values available from one section of the OFP. Recent work has focused on detailed sedimentological and stratigraphic studies of the Toby and OFP formations with future efforts being directed towards integrated geochemical and isotopic research. Additional geochronological constraints are needed to refine palaeogeographical models and strengthen regional correlations with other North American Cordilleran glaciogenic units.

The record of Neoproterozoic glaciogenic sedimentation in the southern Canadian Cordillera (Fig. 37.1) is preserved in three units within the Windermere Supergroup (WSG): a glacial interval at the base of the succession is inferred from the diamictite-bearing Toby Fm., with another event higher in the succession inferred from the diamictite-bearing Vreeland Fm. (glacial) and Old Fort Point Fm. (post-glacial) (Fig. 37.2) (Ross et al. 1989, 1995).

including mapping of the Toby Fm. and related structures in the Paradise Mine (Atkinson 1975), Delphine Creek (Root 1987), Mount Forster (Bennett 1986), Creston/Salmo (Glover & Price 1976) and Crawford Bay/Columbia Point areas (Lis & Price 1976). Ongoing investigations are focusing on the relationship between Toby Fm. facies variability and structural trends as well as isotopic analyses of associated strata.

Toby Fm.

Vreeland Fm.

The Toby Fm. (also known as the Toby Conglomerate) is best exposed in the Purcell Mountains of southeastern British Columbia. The variability of its lithofacies has precluded the identification of a type section; the formation was named by Walker (1926) after well-exposed outcrops located along Toby Creek, west of Invermere, British Columbia (508300 N, 1168020 W) (Fig. 37.1, A). This area has been extensively modified by Proterozoic, Devonian and Mesozoic deformation and the structural context of these rocks is often complicated (e.g. Root 1987). Previous studies of the Toby Fm. are relatively few, with primary emphasis on regional mapping relative to Proterozoicand Mesozoic-age structures (Walker 1926; Leech 1954; Aalto 1971; Reesor 1973; Root 1987; Pope 1990; Warren 1997). The glacial origin of the Toby Fm. was first proposed based on a more detailed study of facies types and variability, petrography, as well as trends in clast characteristics in the Toby Fm. itself (Aalto 1971). Additional mapping of specific regions has followed,

The Vreeland Fm. has received little study (due largely to inaccessibility) despite its impressive exposures and thickness in the Pine and Monkman Pass regions, central British Columbia (from 548050 N to 578000 N) (Fig. 37.1). It was originally mapped by Slind & Perkins (1966) as an unnamed Precambrian conglomerate –schist. Later, it was briefly examined by Stelck et al. (1978), but was not really described until regional mapping conducted by McMechan (1987) and McMechan & Thompson (1985, 1995a, b). It was initially referred to as a diamictite within the Misinchinka Group or the Mount Vreeland –Paksumo Pass diamictites (McMechan 1990), until Hein & McMechan (1994) referred to it as the Vreeland Fm. McMechan (2000) provided the first detailed stratigraphic description and palaeogeographical interpretation in the Monkman Pass area. No type section has been defined for the Vreeland Fm., although the Mount Vreeland sections (548350 N, 1218320 W) in the Snake Indian thrust sheet (Fig. 37.1,B) and the Paksumo Pass section (548200 N,

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 413– 423. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.37

414

M. D. SMITH ET AL.

Fig. 37.1. Generalized geological map showing outcrop distribution of Windermere Supergroup in western Canada and specific outcrop localities (A– K) discussed in the text (modified from Ross et al. 1995). A comprehensive list of outcrop locations for the Toby Fm. can be found in Aalto (1971) and Reesor (1973). Outcrop localities for the Vreeland Fm. are from McMechan (2000) and for the OFP from Ross & Murphy (1988) and Smith (2009). Mtns, Mountains.

1208490 W) in the Wapiti Pass thrust sheet (Fig. 37.1,C) of the Monkman Pass area are thought to be representative exposures (McMechan 2000).

Old Fort Point Fm. The Old Fort Point Fm. (OFP) is the preferred name to describe the unique stratigraphic marker in the WSG that is exposed locally over the entire outcrop region of southwestern Canada (Fig. 37.1,D – K). Walcott (1910) was the first to describe the OFP from the Lake Louise area of Alberta using its distinctive purple and green colours and lithology (siltstone, limestone, breccia) to separate the Precambrian Corral Creek (lower conglomerate to sandstone) and Hector (upper slate) formations. The name and type section (528520 1600 N, 1188030 4000 W) come from outcrops at a prominent landmark near Jasper, Alberta

(Charlesworth et al. 1967). A more representative and better exposed section crops out c. 15 km west of Jasper along the major highway (528520 2100 N, 1188170 3500 W). The regionally extensive nature of the OFP has led to various names: (i) basal Hector Fm. (Walcott 1910) or Mount Temple and Taylor Lake members (e.g. Gussow 1957) in the Lake Louise area, Alberta; (ii) Old Fort Point Fm. (e.g. Weiner 1966; Charlesworth et al. 1967) in the Jasper area, Alberta; (iii) Middle Miette Marker (e.g. McDonough 1989) in the Mount Robson area, British Columbia; (iv) Kaza Group marker (e.g. Ross & Murphy 1988) in the Cariboo Mountains, British Columbia; (v) Baird Brook Division (e.g. Kubli 1990) in the Dogtooth Range of the Purcell Mountains, British Columbia; (vi) Comedy Creek unit (Grasby & Brown 1993), Selkirk Mountains, British Columbia; (vii) upper and lower markers (e.g. Warren 1997) in the Purcell Mountains, British Columbia (Fig. 37.1). Although there were early attempts to correlate between certain locations (Charlesworth et al. 1967;

WINDERMERE SUPERGROUP

415

Fig. 37.2. Comparative stratigraphic columns for the Windermere Supergroup from the Monkman Pass area and the Cariboo-Purcell-Rocky Mountains (modified from Ross et al. 1995). Nomenclature of rift and post-rift strata (Kaza-Cariboo, Horsethief Creek and Miette groups refer to similar strata, depending on location within the region). Relative stratigraphic position of the glaciogenic Toby, Vreeland and Old Fort Point formations and proposed correlation between the basal Framstead and Old Fort Point formations are shown. Ages shown are discussed in the text and, in stratigraphic order, are Malton Gneiss (McDonough & Parrish 1991); rift-related volcanic rocks from Idaho (Lund et al. 2003; Fanning & Link 2004); Old Fort Point Fm. (Kendall et al. 2004); and Hamill Group volcanic rocks (Colpron et al. 2002). Bracketed ages indicate that the ages are not from rocks of the southern Canadian Cordillera, but rather interpreted correlatives of the Irene Fm.

Aitken 1969), it was Ross & Murphy (1988) who first recognized the basin-wide correlation of the OFP (and equivalents), its palaeogeographical significance and importance in Neoproterozoic event stratigraphy. Subsequent work by Ross et al. (1995) examined the S-isotopic evolution of the OFP and WSG strata in the Cariboo Mountains, British Columbia (Fig. 37.1). A regionally comprehensive study of the OFP integrating sedimentology, stratigraphy, geochemistry and stable isotopes has recently been completed (Smith 2009).

Structural framework and basin setting The WSG forms part of a long arcuate belt of semicontinuous outcrops in western North America that extends from the Yukon – Alaska border region to northwestern Mexico (e.g. Ross et al. 1989). In southwestern Canada, exposures crop out over an area of 35 000 km2 in a series of thrust sheets in the Main Ranges of the western Fold-and-Thrust Belt and the Omineca Belt of the southern Canadian Cordillera (Fig. 37.1). Major structures affecting WSG strata in the region include the Southern Rocky Mountain Trench (Fig. 37.1) and a broad gently north-plunging structure in the Purcell Mountains (the Purcell Anticlinorium). Proterozoicage structures that affected WSG deposition include regionally significant transfer faults (e.g Mount Forster and Moyie) and associated uplifted blocks (e.g. Windermere High and Montania; Reesor 1973; Lis & Price 1976; Root 1987; Warren 1997). All WSG strata were subjected to Mesozoic orogenic deformation that resulted in a wide range of structural complexity and metamorphic grade between different structural panels. Extensive tracts of sub-greenschist to greenschist grade exist in the western Main Ranges of the Rocky Mountains (Lake Louise, Jasper, northern Cariboo Mountains and eastern Purcells Mountains), whereas higher-grade metamorphic rocks (biotite to upper amphibolite)

are found in the southern Cariboo and western Purcell Mountains, as well as the Shuswap Complex (e.g. Simony et al. 1980; Ross et al. 1995). Areas of low metamorphic grade and relatively uncomplicated tectonic deformation show exceptional preservation of primary sedimentary textures and structures. This allowed the development of a consistent internal stratigraphy of the WSG and reconstruction of the Windermere basin in southwestern Canada (Ross & Murphy 1988; Ross et al. 1989, 1995; Ross 2000). The WSG is thought to represent deposition in two tectonic phases (e.g. Stewart 1972; Ross 1991). The first phase was characterized by rifting during the break-up of the supercontinent Rodinia, whereas the second phase occurred during post-rift thermal relaxation. At the base of the WSG, the Toby Fm. together with the Irene Fm. volcanic rocks are interpreted to have accumulated during the rifting phase (Aalto 1971; Glover & Price 1976; Root 1987). The variable thickness of the Toby Fm. is attributed to syn-sedimentary fault-controlled deposition (Aalto 1971; Root 1987). Based on the nature of Proterozoic-age faults, Root (1987) suggested that the basin underwent two episodes of extension; one trending approximately east –west, resulting in a series of NNE and NNW faults and another trending NW –SE, resulting in a series of NE– SW normal faults. The Vreeland Fm. and OFP both accumulated during the postrift phase with possible local structural control (McMechan 2000; Smith 2009). During this later phase, the basin is thought to be an elongate NW-flowing turbidite system (Ross et al. 1989, 1995; Ross 1991), based on consistent palaeocurrent data and bimodal pattern (.2.6 Ga and 1.9–1.75 Ga) of U –Pb detrital zircon provenance (Ross & Bowring 1990; Ross & Parrish 1991). Basement clasts from the Vreeland Fm. yielded ages between 1865 to 1842 Ma (Ross & Villeneuve 1997), suggesting derivation from a source area to the north and NE (McMechan 2000). Although there is broad agreement about these two tectonic phases, the exact tectonic setting of the Windermere basin and

416

M. D. SMITH ET AL.

its evolution remains controversial. Some argue that the upper part of the WSG in the southern Canadian Cordillera accumulated on a continental passive margin with a substantial Proto-Pacific Ocean (Ross 1991; Ross et al. 1995; Dalrymple & Narbonne 1996). Others prefer an intracontinental rift and restricted ocean basin setting for the southern WSG, and a late Neoproterozoic –early Cambrian inception for the Proto-Pacific Ocean (Colpron et al. 2002). The controversy and uncertainty is in part due to the presence of two intervals of rift-related igneous rocks within the southern Canadian Cordillera (Bond & Kominz 1984; Ross 1991; Colpron et al. 2002), diachronous dates for the timing of rifting along the whole of Laurentia’s Pacific margin (see discussion in Lund et al. 2003), and tectonic deformation that currently precludes certain identification of a western Windermere basin margin.

Stratigraphy In the southern Canadian Cordillera, the WSG is a thick (c. 9 km), unconformity-bounded succession of predominantly coarsegrained deep-marine siliciclastic rocks with subordinate carbonate and mafic volcanic rocks (Fig. 37.2). Older units associated with the rift phase (e.g. Toby and Irene formations) are more limited in areal extent and tend to show greater lateral facies variations compared with units associated with the post-rift phase (e.g. Ross et al. 1989, 1995; Ross 1991; Warren 1997). At the base of the WSG, the Toby Fm. unconformably overlies shallow-water deposits of the Mount Nelson Fm. (Neoproterozoic/Mesoproterozoic?, Root 1987; Ross & Villeneuve 2003) or deep-marine sedimentary rocks of the Mesoproterozoic Purcell Supergroup (Aalto 1971; Root 1987). The Toby Fm. is associated with localized rift-related mafic volcanic rocks of the Irene Fm. and a laterally discontinuous carbonate horizon at its top (e.g. Root 1987; Warren 1997). These basal units of the WSG are largely restricted to the Purcell Mountains region (Fig. 37.1). Potential high-grade metamorphic equivalents of the Toby (e.g. Simony et al. 1980; Murphy et al. 1991) and Irene volcanic rocks (e.g. Simony et al. 1980; Sevigny 1988) may exist in the Shuswap Complex and northern Cariboo Mountains, although the stratigraphic control is poor (Fig. 37.1). The nomenclature of WSG succession that overlies the Toby Fm. in the southern Canadian Cordillera is rather complex: a direct result of correlation challenges across thrust-bound panels with varying levels of metamorphism, general absence of reliable stratigraphic markers and poor geochronological constraints. The Horsethief Creek Group (Purcell Mountains), Kaza and lower/ middle Cariboo groups (Cariboo Mountains), Miette Group (Rocky Mountains) and Misinchinka Group (Rocky Mountains) all refer to parts of the same c. 5 km deep-marine turbidite system exposed over 35 000 km2 (e.g. Ross et al. 1989) (Figs 37.1 & 37.2). Lithologically, the thick package is dominated by coarsegrained arkosic sandstone and granule conglomerate interbedded with mudstone-dominated intervals or minor carbonate. The Vreeland Fm. of the Misinchinka Group is a local exception, being composed of up to 2000 m of diamictite interbedded with mudstone, sandstone and conglomerate (McMechan 2000). Mapping of the Vreeland Fm. has shown that the diamictite units undergo a lateral facies change into typical deep-marine coarsegrained turbidites of the WSG in both a westward (McMechan 1987) and southward direction (McMechan & Thompson 1995a). Throughout much of the basin, the overall monotony of the thick-bedded, immature turbidites of the Horsethief Creek, Miette and Kaza groups is interrupted by the fine-grained, regionally widespread OFP (and correlative units) (Ross & Murphy 1988; Ross et al. 1995). Nowhere is the OFP observed to be in contact with diamictite of the Vreeland Fm.; rather, the OFP overlies WSG turbiditic strata interpreted as deep-marine lateral equivalents. The OFP is overlain by coarse-grained basin-floor turbidite deposits that form a kilometre-scale shoaling upwards trend

to mudstone-dominated upper slope units and shelf platform carbonates (Ross et al. 1989). A significant regional unconformity marks the end of WSG deposition and the beginning of a latest Neoproterozoic to Cambrian rift– drift succession with the clastic and volcanic sediments of the upper Cariboo, Hamill and Gog groups.

Glaciogenic deposits and associated strata Toby Fm. The Toby Fm. consists primarily of diamictite interbedded with conglomerate, sandstone, dolomitized or recrystallized limestone and mudstone (Fig. 37.3) (Aalto 1971). Its thickness is highly variable, ranging from 0 to 2500 m. Lithofacies, in general, and sedimentary characteristics of diamictite, in particular (including clast concentration, clast size, matrix and clast lithology, lateral continuity and bed thickness) are highly variable over short distances, a fact that has been stressed by all workers in this region. The following descriptions rely heavily upon the regional study of Aalto (1971), with additional observations based on more recent fieldwork by E. Arnaud and K. Root. Diamictite is predominantly massive, matrix- or clast-supported with sub-angular to sub-rounded clasts up to 2 m, floating in a muddy sandstone or sandy mudstone matrix. Clast lithology includes quartzite, volcanic greenstone, slate, dolomite and chert derived from the underlying Mount Nelson Fm. as well as rare extrabasinal granite (Leech 1954; Reesor 1973; Loveridge et al. 1981). The diamictite matrix exhibits a similar lithological variability. Tectonic overprinting typically obscures clast fabric in many outcrops. Where relatively unaltered, clasts show no preferred orientation. Diamictite units vary in thickness from several metres to tens of metres. In outcrops with good lateral exposure, basal contacts are relatively planar, and conformable or erosional. Conglomerate occurs in lenticular or planar units (centimetre to metre scale) interbedded with diamictite and sandstone. It is distinguished from the diamictite by its coarse sandstone matrix, and consistently high clast concentration (65 –80%). It is generally unorganized and massive with occasional subtle coarse-tail grading. Sandstone varies from coarse- to fine-grained, moderately to poorly sorted, and occurs in planar beds or lenticular units (centimetre to metre scale). It is most commonly massive or laminated, although some sandstone units exhibit ripple crosslamination, climbing ripples or normal grading. Dolomitized or recrystallized limestone (centimetre to metre scale thickness) with varying amounts of quartz grains is also observed within and at the top of the Toby Fm. at various sites within the region. These carbonate strata are planar –tabular and can be laterally persistent over hundreds of metres, although their overall distribution is patchy. Finally, mudstone is relatively common throughout the Toby Fm. It is massive or laminated and occurs in relatively planar tabular units (centimetre to metre scale thickness). Outsized clasts are found within these mudstone units, and some are seen to depress or puncture underlying laminations (Aalto 1971). In terms of associated strata, the underlying Mount Nelson and Dutch Creek formations consist of thick packages of carbonate and siliciclastic strata (kilometre scale; Walker 1926; Reesor 1973; Root 1987). Slate predominates in the Dutch Creek Fm., whereas thick packages of white quartzite and dolomite interbedded with intervals of argillite, siltstone and conglomerate characterize the Mount Nelson Fm. In the Purcell Mountains, the Toby Fm. is overlain by the Horsethief Creek Group, which consists of a thick package (,100–2000 þ m) of mudstone, sandstone, conglomerate, calcareous mudstone and dolostone (Walker 1926; Root 1987; Kubli 1990; Warren 1997). Further to the south, the Toby Fm. is interbedded and overlain by volcanic greenstone of the Irene Fm., and coarse- to fine-grained clastic and carbonate lithofacies of the Monk Fm. (Aalto 1971).

WINDERMERE SUPERGROUP

417

Vreeland Fm. Description of the Vreeland Fm. is based entirely on regional mapping studies in the Pine and Monkman Pass areas (McMechan & Thompson 1985, 1995a, b; McMechan 1987) and geological descriptions from the Snake Indian and Wapiti Pass thrust sheets (Fig. 37.1) (McMechan 1990, 2000). The Vreeland Fm. comprises a thick succession of diamictite interbedded with mudstone, sandstone and conglomerate (Fig. 37.4). Diamictite units are generally massive (c. 98%) with a sandstone or siltstone to mudstone matrix

Fig. 37.3. Stratigraphic columns of the Toby Fm. from the Mount Brewer (c. 508230 N, 1168140 W) and Paradise Mine localities (c. 508280 N, 1168180 W) with lithofacies representative of other outcrops; however, their lateral continuity, stratigraphic thickness and stratigraphic distribution vary widely across the Purcell Mountains. Note the different scales used in each log. Facies codes used: (first letters) D, diamictite; G, conglomerate; S, sandstone; F, fine-grained facies; (second letters in diamictite and conglomerate) m, matrix supported; c, clast-supported; (third letter in any lithofacies) m, massive; l, laminated; d, deformed; r, rippled; h, horizontally bedded; s, stratified.

Fig. 37.4. Stratigraphic columns of the Vreeland Fm. near Mount Vreeland in the Snake Indian thrust sheet, and Paksumo Pass in the Wapiti thrust sheet, modified from McMechan (2000). St, cross-trough bedded sandstone; Fld, laminated fine-grained facies with lonestones. See Figure 37.3 caption for other facies codes.

418

M. D. SMITH ET AL.

and form laterally extensive (up to 2 km) tabular sheets that range from a few metres to .40 m thick. The basal contact of massive diamictite units is sharp and locally erosive. Subtle normal grading is observed locally in diamictite in addition to rare stratified diamictite defined by clast layers, or sandstone/siltstone stringers and lenses. Clasts are lithologically diverse and consist of both intra-basinal (mudstone, sandstone, carbonate) and extra-basinal (felsic or mafic plutonic, volcanic, quartzite) varieties. Differing provenance trends are apparent in the two main areas based on clast lithology. Interstratified mudstone units are up to 38 m thick with rare sandstone interbeds, and common lonestones. Sandstone units are sharp-based and range from very fine- to very coarse-grained and very thin- to thick-bedded. Sedimentary structures include scours, cross-stratification, graded bedding and intraclast rip-ups. Conglomerate and sandstone units commonly form lenticular channellike deposits. Conglomerate is matrix- and clast-supported with a sandstone matrix, which contrasts with the typical siltstone to mudstone matrix of the diamictite. Clast lithologies are similarly diverse, rounded to angular, and range up to boulder size. In the Snake Indian thrust sheet, the top of the Vreeland Fm. is marked by a pyrite-rich zone up to 30 cm thick and locally the diamictites are overlain by a thin (few metres), parallel laminated limestone. The conformably overlying Framstead Fm. is mudstone-dominated, but locally is composed of sandstone to conglomerate or limestone. Locally, large (up to 650 m long) carbonate olistoliths occur on, or just above the basal contact with the Vreeland Fm., and also near its upper contact. These olisoliths occur as discrete, randomly oriented blocks that lack internal deformation. Lithologies are predominantly tan-weathered shallow-water carbonates and include laminated dolomite with rare layers of teepee structures and bladed calcite, stromatolitic dolomite, sandy to pebbly dolomite, dolomite conglomerate, and quartzose fenestral and vuggy dolomite.

Old Fort Point Fm. The following sedimentological and stratigraphic description stems mostly from a recent detailed regional study (Smith 2009), but also incorporates work of earlier authors (Fig. 37.1,D –K) (Walcott 1910; Weiner 1966; Charlesworth et al. 1967; Aitken 1969; Murphy 1986; Pell & Simony 1987; Ross & Murphy 1988; McDonough 1989; Deschesne 1990; Kubli 1990; Grasby & Brown 1993; Ross et al. 1995; Warren 1997; Ross & Ferguson 2003a, b). The OFP comprises three distinctive lithological members, which despite variable metamorphic grade, form a consistent stratigraphic relationship throughout the southern Canadian Cordillera (Fig. 37.5). The lower member is a purple, green, grey or red-brown finegrained unit that ranges from 50 to 125 m thick. The basal portion is composed of siltstone to mudstone that grades upward into rhythmic couplets of limestone –siltstone. Beds typically range from ,1 cm to 10 cm in thickness and exhibit common tractional sedimentary structures including planar lamination, lenticular bedding, ripple cross-lamination, normal grading and minor scours. Subordinate lithofacies include very fine- to fine-grained sandstone interstratified with uncommon limestone-clast breccia beds. Palaeocurrents measured from 3D current-ripples generally indicate flow toward the SW to NW. The middle member ranges from 2 to 15 m in thickness and consists of a dark grey organic-rich mudstone/pelite. The basal contact with the underlying lower member is usually gradational over a few metres. The unit is characterized by alternating siltstone and mudstone laminae with subtle normal grading. Isolated, thinbedded dark-grey massive or planar-laminated limestone and very fine ripple-stratified sandstone occur locally. This member is regionally extensive. Its fissile nature typically results in poor or covered exposures, although notable exceptions occur (e.g. Ross et al. 1995; Kendall et al. 2004).

Fig. 37.5. Stratigraphic columns of the Old Fort Point Fm., simplified from Smith (2009). At the Geikie Siding section in Jasper National Park, the regionally widespread lower two members are exposed and overlain sharply by a thinly developed upper member. The Re– Os isochron age was obtained from the middle member at this section (Kendall et al. 2004). At the Upper Boomerang section on the Lake Louise Ski Resort, only a thickly developed Upper Member is exposed, with complete erosion of the lower two members. Str, stratified (ripple cross-laminated and cross-bedded). See Figure 37.3 for other facies codes.

In contrast to the lower two members, the upper member is lithologically diverse and highly variable in thickness (,0.5–165 m). The basal contact with underlying strata is always sharp and, locally, it completely erodes the lower two members. Lithologies include diamictite, breccia, conglomerate, mudstone, siltstone, sandstone, quartz arenite, calcareous arenite, arenaceous limestone and limestone. Diamictite and breccia to conglomerate are

WINDERMERE SUPERGROUP

generally sharp, commonly erosively based, massive beds with mudstone/siltstone or well-sorted, coarse quartz-rich sandstone matrix. Bed thickness (,0.5 to .10 m), clast size (centimetre to metre) and clast lithologies are variable. Clasts include fragments of the OFP members and shallow-marine carbonates, some with rare bladed calcite crystals. Mudstone and siltstone exhibit planar laminations, micro-scours, subtle normal grading, large single chlorite flakes and rare scours. This lithofacies is locally thick (c. 100 m) and monotonous with rare interbeds of diamictite, breccia, conglomerate or sandstone. Sandstone in the upper member consists of a range of textures: fine- to very coarsegrained, poor- to well-sorted and mineralogical maturity (immature to supermature). The well-sorted, mature quartz arenites are interbedded with dark, organic-rich limestone beds and these units can exhibit a range of compositions between the two end members (e.g. calcareous arenites or arenaceous limestones). Sedimentary structures include massive beds, planar lamination and cross-stratification. Rare palaeocurrent measurements from the upper member are consistently towards the SW to NW. Softsediment deformation features such as load structures, convolute or contorted bedding and ductile folding are common in both fine- and coarse-grained facies of the upper member.

Boundary relations with overlying and underlying non-glacial units Toby Fm. The Toby Fm. rests unconformably on various members of the Mount Nelson Fm. or the underlying Dutch Creek Fm. In outcrop, the unconformity is commonly subtle (,108), although a distinctive angular unconformity with tilted underlying Mount Nelson strata is evident in some places. The upper contact of the Toby Fm. with the Horsethief Creek Group is conformable and gradational throughout the Purcell Mountains and typically established based on the loss of diamictite. To the west and south of the Purcell Mountains, the Toby Fm. is conformably overlain by pillow lavas of the Irene Fm. as well as conglomerate and diamictite facies of the lower Monk Fm. (Aalto 1971). Considering there are localized lenses of conglomerate and diamictite in the Irene Fm., and that the upper contact of the Toby Fm. is defined in the Purcell Mountains with the loss of diamictite facies, the upper contact of the Toby Fm. may actually occur several tens of metres above the Irene volcanic rocks (Aalto 1971).

419

of the Vreeland Fm. diamictite. The contact is commonly marked by a distinctive change in colour (e.g. grey to purple) and lithology (appearance of lower member siltstones). However, at some locations the lower and middle members have been eroded and the upper member forms a sharp unconformity over older WSG strata. The top of the OFP is conformably overlain by younger WSG strata. In many cases the contact is identified by a sharp change in grain size and/or framework mineralogy (e.g. OFP quartz arenite to WSG arkosic sandstone). The challenge arises where OFP mudstone units are overlain directly by a younger mudstone unit of the WSG. The contact is taken to coincide with the termination of organic-rich limestone beds or the appearance of common sandstone interbeds. Field observations can sometimes be confirmed by a decrease in gamma-ray counts or correlative geochemical trends (e.g. decrease in total organic carbon (TOC), Mo) (Smith 2009).

Chemostratigraphy The only published geochemical or isotopic data from the three units is the S-isotopic study of Ross et al. (1995) on pyrite in the post-rift strata of the WSG, Cariboo Mountains, British Columbia (Fig. 37.1,E). A total of 170 samples were analysed for d34Spy and these showed a broad range of isotopic values (c. 50‰) (Ross et al. 1995). As part of that study, a suite of 36 samples were collected from a measured section that includes the OFP. A strong correlation between lithology (and inferred sedimentation rate) and pyrite isotopic composition was reported (Ross et al. 1995). For example, the most negative d34Spy values ( –31.9‰ to –25.3‰) corresponded to siltstone and mudstone of the OFP, whereas the interval in the underlying Kaza Group with the highest sandstone –mudstone ratio was the most positive (around þ11.9‰ to þ14.5‰) (Ross et al. 1995). New isotopic data (d18Ocarb, d13Ccarb, d13Corg, d34Spy) from OFP sections and lithogeochemical (major, trace and rare-earthelement analyses) data from a regional study of WSG mudstones have recently been obtained (Smith 2009). Preliminary analysis shows lower member limestone units exhibiting negative d13Ccarb isotopic values and the middle member characterized by distinctive chemical (e.g. TOC, Mo, V/Cr) and isotopic (e.g. d13Corg) signatures.

Other characteristics

In the western Snake Indian thrust sheet, the basal contact of the Vreeland Fm. is not exposed (McMechan 2000). In the eastern Wapiti Pass thrust sheet, the basal diamictite appears to interfinger with the upper 150 m of the fine-grained facies of the underlying Paksumo Fm. (McMechan & Thompson 1995b). The Vreeland Fm. is conformably overlain by the Framstead Fm., and in the western exposures by a carbonate olistolith-bearing unit, whereas in the eastern exposures it is overlain by a sandstone unit (McMechan 2000). The carbonate olistolith-bearing unit correlates from west to east where it overlies the lowermost Framstead Fm. sandstone. This suggests at the western Snake Indian thrust sheet locality that either local erosion of the sandstone unit occurred or it is actually a lateral equivalent of diamictite of the Vreeland Fm. (McMechan 1990).

Economic deposits associated with the glaciogenic strata are limited. Local copper mineralization has been observed below, or in the basal parts of the OFP (Ross 2000; Smith 2009). The only description of possible Precambrian fossils from glaciogenic strata are some problematic discoid structures at the base of the OFP near Arnica Lake, west of Banff, Alberta (Fig. 37.1). The structures are c. 3 mm in size, lentil-shaped with either hypo- or epirelief, have concentric wrinkles and a five-point star-like figure inside the concentric pattern (Hofmann 1971; Gussow 1973; Smith 2009). The specimens are attributed to Chuaria circularis (Hofmann 1971; Gussow 1973), but they lack any carbonaceous material that would strengthen a biogenic origin and certainly require further detailed work. Ediacaran fauna (including Namacalathus and Cloudina assemblages similar to those found in the Nama Group of Namibia) have otherwise been reported from the uppermost part of the WSG (Hofmann & Mountjoy 2001 and references therein).

Old Fort Point Fm.

Palaeolatitude and palaeogeography

The OFP conformably and gradationally overlies deep-marine turbiditic strata of the WSG interpreted to be lateral facies equivalent

There are no published palaeolatitute data for the WSG in southwestern Canada. A near-equatorial palaeogeographical position

Vreeland Fm.

420

M. D. SMITH ET AL.

is based on palaeomagnetic data obtained from broadly correlative WSG strata in the Mackenzie Mountains, northern Canadian Cordillera (Park 1997).

Geochronological constraints The predominantly siliciclastic nature of the WSG in the southern Canadian Cordillera has led to a paucity of high-precision geochronological constraints. Currently, the local maximum depositional age of the WSG in the region is constrained by a U– Pb zircon date of 736 þ 23/ –17 Ma [mean square weighted deviation (MSWD) 2.36, four-point TIMS regression] from orthogneiss basement rocks of the Malton Gneiss Complex (McDonough & Parrish 1991). This age corresponds well with U –Pb zircon dates obtained from WSG basement rocks elsewhere in the Canadian Cordillera, such as 740 + 36 Ma (Parrish & Scammell 1988) and 728 þ 8/– 7 Ma (Evenchick et al. 1984; see Ross et al. 1995 for a full review). A U –Pb zircon age of 569.6 + 5.3 Ma (three-point 207Pb/206Pb weighted average) from the syn-rift felsic volcanic rocks of the Hamill Group that unconformably overlie the WSG constrains the minimum timing of deposition (Colpron et al. 2002). No direct ages have been obtained from the Toby Fm. or Irene volcanic rocks in southern British Columbia. Mafic metavolcanic rocks of the Huckleberry Fm. in northeastern Washington, interpreted to be Irene Fm. correlatives were imprecisely dated with a preliminary age of 762 + 44 Ma (MSWD 0.06) from a threepoint Sm –Nd isochron of one whole-rock and two pyroxene separates (Devlin et al. 1988). Other possible treatments of the Sm – Nd data produced even larger error estimates: 795 + 115 Ma (MSWD 6.71) from all the data excluding one pyroxene separate; 674 + 212 Ma (MSWD 1.0) from a seven-point whole-rock isochron; 719 + 200 Ma (MSWD 1.0) from two whole-rock and two pyroxene separates (Devlin et al. 1988). More recent work on possible correlative rift-related volcanic rock units in Idaho has obtained precise SHRIMP U –Pb zircon ages of 685 + 7 Ma, 684 + 4 Ma (Lund et al. 2003) and 709 + 5 Ma (Fanning & Link 2004). A comparable U – Pb zircon age of 688.9 þ 9.5/ –6.2 Ma was obtained from syn-rift felsic volcaniclastics of the Gataga Volcanics, northern British Columbia (Ferri et al. 1999). In contrast, high-precision U –Pb (ID-TIMS) dates were obtained from possible correlative strata in the Central Ogilvie Mountains, NW Canada (716.47 + 0.24 Ma; Macdonald et al. 2010). No absolute age constraints are available for the Vreeland Fm. Its relative age is based on regional mapping and lithostratigraphic correlations within the WSG (McMechan 1987, 1990, 2000). Kendall et al. (2004) obtained two comparable five-point Re –Os isochron ages from organic-rich black mudstone of the OFP middle member in the Jasper area, Alberta (Fig. 37.1): an imprecise 634 + 57 Ma (MSWD 65) using the conventional inverse aqua regia digestion method, and a more precise 607.8 + 4.7 Ma (MSWD 1.2) using the CrO3-H2SO4 digestion technique that is thought to best reflect depositional age of the OFP.

Discussion Toby Fm. The variable level of erosion below the sub-Toby unconformity suggests a period of significant erosion and uplift preceded WSG sedimentation. The depositional setting of the Toby Fm. itself has been controversial, specifically regarding the extent of glacial influence on deposition. Most lithologies record remobilization of sediment by sediment-gravity flows. Some authors have favoured a glacial setting for these reworked deposits based on the regional extent of the Toby Fm., the presence (albeit

localized) of extrabasinal clasts, striated clasts, clast clusters within diamictites and outsized clasts interpreted as far-travelled and ice-rafted glacial debris (Aalto 1971; Warren 1997). Others favoured the resedimentation of locally derived material formed along a fault-bounded margin undergoing extension with limited or no glacial influence (Walker 1926; Reesor 1973; Root 1987). The predominance of intrabasinal clasts, the mounting evidence for structural controls on facies type and variability (ongoing research), in addition to the localized and rare occurrence of glacial indicators, suggest that tectonic activity imparted a primary control on deposition with localized glacial influence. Carbonate strata within and at the top of the Toby Fm. have received limited attention to date. Preliminary results from ongoing work suggest that primary carbonate was accumulating on topographic highs while coarse-grained clastics accumulated in fault-bounded basins. Stable isotopic analyses of these carbonates have so far yielded limited information, with current research focused on assessing the effects of metamorphism on isotopic signatures.

Vreeland Fm. The Vreeland Fm. is interpreted to have been deposited in a midslope, glacially influenced marine setting (McMechan 2000). The diamictite units are interpreted to be resedimented sedimentgravity flow deposits (e.g. unconfined debris flows) of glaciogenic debris or the result of rainout processes from fine-grained suspended sediment plumes and ice-rafted debris (McMechan 2000). The argument for a glacial origin is based on large, sub-angular extrabasinal basement clasts in the diamictite and the occurrence of lonestones interpreted as dropstones in the mudstone units of the Snake Indian thrust sheet (McMechan 2000). The absence of dropstones in the Wapiti Pass thrust sheet Vreeland sediments suggests that there was little to no ice-rafted material at this palaeogeographical location (McMechan 2000). The considerable strike length (c. 400 km) of diamictite suggests glacial erosion over a large area, whereas its thickness (c. 2000 m) suggests active subsidence accompanied by local faulting (McMechan 2000). In the Snake Indian thrust sheet, diamictite units are overlain locally by a discontinuous thin-laminated grey limestone (cap carbonate?) deposited during post-glacial sea-level rise and concomitant shutdown of coarse clastic sediment flux to the basin (McMechan 1990, 2000). The large shallow-marine carbonate olistoliths in the Framstead Fm. suggest continued uplift or faulting causing instability and downslope movement of parts of the carbonate shelf edge (McMechan 2000), although simple highstand shedding during post-glacial eustastic rise could also be a potential delivery mechanism.

Old Fort Point Fm. The OFP was deposited along a deep-marine basin-slope to basinfloor transect and overlies coarse-grained turbidites of the WSG. The OFP is interpreted to be related to an Ediacaran-aged postglacial eustatic rise and a shutdown of the supply of coarse, immature siliciclastic sediment into the basin (Ross & Murphy 1988). Fine-grained, mostly turbiditic strata of the lower and middle members are regionally uniform in thickness and lithofacies, suggesting synchronous, basin-wide deep-marine deposition during transgression and highstand (Ross & Murphy 1988; Ross et al. 1995; Smith 2009). Conversely, the variable lithology and thickness of the upper member appears to be controlled by a combination of more local factors, including palaeogeographical location, topography along the slope, erosional mass wasting, structural activity and/or relative sea-level changes (Smith 2009). Deposition of the upper member is interpreted to have coincided with a fall of a relative

WINDERMERE SUPERGROUP

sea level caused, at least in part, by regional uplift (Smith 2009). In western North America, other Ediacaran-aged uplift or erosion features have been observed, suggesting some type of renewed extensional activity during the post-rift phase (e.g. Warren 1997; Fedo & Cooper 2001; Clapham & Corsetti 2005). Where observable, the locally sharp, basal contact is interpreted as a sequence boundary (possibly overprinting the basal surface of forced regression) within the OFP (Smith 2009). Related erosional mass wasting, submarine canyon incision and/or syn-sedimentary growth faults controlled initial sediment transport fairways (lowstand) that were subsequently filled during a renewed relative sea-level rise (transgression and highstand) (Smith 2009). The quartz-rich lithologies of the upper member compared to the arkosic sandstone of other WSG turbidites reflect winnowing and maturation from residence on the shelf during the earlier transgression and highstand (Ross 2000; Smith 2009).

Regional correlations The Toby, Vreeland and OFP formations have been correlated with other units along the North American Cordillera based on lithostratigraphic similarities (see Lund et al. 2003 or Colpron et al. 2002 for most recent reviews). The Toby Fm. has been correlated with the Kingston Peak Fm. (Death Valley), the Edwardsburg and Pocatello formations (Idaho), the Shedroof Conglomerate (NW Washington) and the Sayunei and Shezal formations (Mackenzie Mountains, NW Canada) (Gabrielse & Campbell 1991; Link et al. 1993; Ross et al. 1995). The Vreeland Fm. has been correlated to the similar diamictite-bearing Toobally (Pigage & MacNaughton 2004) and Icebrook (Aitken 1991; James et al. 2001) formations in the northern Canadian Cordillera. The Mount Lloyd George diamictite in north-central British Columbia (Eisbacher 1981a, b) is of uncertain age and may correlate with either the Toby or the Vreeland formations. The OFP overlies turbidites of the WSG that are interpreted to be correlative with the Vreeland Fm. (McMechan 1990, 2000; Ross et al. 1995). The OFP lower member rhythmic limestone – siltstone couplets are interpreted as a deep-marine equivalent of the thin limestone overlying the Vreeland Fm. (McMechan 2000) and of the cap-carbonate succession overlying the Ice Brook Fm. (Aitken 1991; James et al. 2001). The organic-rich OFP middle member potentially correlates with the pyrite-rich horizon at the top of the Vreeland Fm. (McMechan 2000) and represents maximum flooding conditions and anoxic (?) bottom-water conditions during earliest post-glacial highstand. The correlative link between the glaciogenic Ice Brook/Vreeland formations and the post-glacial OFP is based primarily on stratigraphic relationships and on the consistent stratigraphic presence of associated carbonates and carbonate olistoliths possessing distinctive bladed calcite crystals (Ross et al. 1995). The basal glaciogenic event in the southern Canadian Cordillera is represented by deposition of the Toby Fm., which is thought to be younger than c. 740– 728 Ma (see Ross et al. 1995 for full review). Various data treatments of the Sm –Nd ages (c. 795– 674 Ma) from correlative volcanic rocks (Devlin et al. 1988) are considered too imprecise to be useful. Current precise ages from potentially equivalent strata in Idaho and NW Canada provide better absolute timing constraints on the glacial episodes because they were obtained from volcanic rocks intercalated with diamictite units, as opposed to unconformably underlying basement rocks. However, further geochronological studies are needed to determine the timing of glacial conditions in the southern Canadian Cordillera relative to the timing of glacial conditions in Idaho and NW Canada. The disparity in radiometric age constraints for rift-related magmatic rocks along the Cordilleran margin suggests a diachronous, protracted history of crustal extension and magmatism and thus underscores the need for caution when correlating Cordilleran glacial deposits without precise

421

geochronological constraints (Lund et al. 2003; Fanning & Link 2004). The timing of the younger glaciogenic event is constrained by the c. 608 Ma OFP (Kendall et al. 2004), which provides a relative minimum age for Vreeland diamictite deposition based on current regional stratigraphic correlations. M. D. S.’s PhD research on the OFP is a product of the Windermere Consortium, initiated by R. W. C. A. and G. M. R. and jointly funded by the Natural Sciences and Engineering Research Council of Canada and industry partners (Anadarko Petroleum, Canadian Natural Resources Ltd, Devon Canada Ltd, Encana Corp., Husky Energy Corp., Nexen Inc. and Shell). M. D. S. has received additional funding from an Ontario Graduate Scholarship (OGS), a Canadian Society of Petroleum Geologists Scholarship and two Grants-In-Aid from the American Association of Petroleum Geologists. E. A.’s work on the Toby Fm. is funded by the Natural Sciences and Engineering Research Council of Canada. Parks Canada and B. C. Parks are thanked for granting research and collections permits to M. D. S. Critical reviews and detailed comments by Brian Kendall and Paul Link improved the content and clarity of the manuscript. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) project #512.

References Aalto, K. R. 1971. Glacial marine sedimentation and stratigraphy of the Toby conglomerate (upper Proterozoic), southeastern British Columbia, northwestern Idaho and northeastern Washington. Canadian Journal of Earth Sciences, 8, 753– 787. Aitken, J. D. 1969. Documentation of the sub-Cambrian unconformity, Rocky Mountains main ranges, Alberta. Canadian Journal of Earth Sciences, 6, 193–200. Aitken, J. D. 1991. Two late Proterozoic glaciations, Mackenzie Mountains, northwestern Canada. Geology, 19, 445– 448. Atkinson, S. J. 1975. Surface geology of the Paradise Basin (82K/8W). In: Geology in British Columbia. British Columbia Ministry of Mines and Petroleum Resources 1975: G7– G12. Bennett, S. 1986. Geology of the Mount Forster map area. British Columbia Ministry of Energy, Mines and Petroleum Resources, Preliminary Map 62. Bond, G. C. & Kominz, M. A. 1984. Construction of tectonic subsidence curves for the early Paleozoic miogeocline, southern Canadian Rocky Mountains-implications for subsidence mechanisms, age of breakup, and crustal thinning. Geological Society of America Bulletin, 95, 155– 173. Charlesworth, H. A. K., Weiner, J. L. et al. 1967. Precambrian Geology of the Jasper Region. Research Council of Alberta Bulletin 23. Clapham, M. E. & Corsetti, F. A. 2005. Deep valley incision in the terminal Neoproterozoic (Ediacaran) Johnnie Fm., eastern California, USA: tectonically or glacially driven? Precambrian Research, 141, 154– 164. Colpron, M., Logan, J. M. & Mortensen, J. K. 2002. U– Pb zircon age constraint for late Neoproterozoic rifting and initiation of the lower Paleozoic passive margin of western Laurentia. Canadian Journal of Earth Sciences, 39, 133– 143. Dalrymple, R. W. & Narbonne, G. M. 1996. Continental slope sedimentation in the Sheepbed Fm. (Neoproterozoic, Windermere Supergroup), Mackenzie Mountains, N.W.T. Canadian Journal of Earth Sciences, 33, 848–862. Dechesne, R. G. 1990. Geology of the Ptarmigan Creek map area (east half) and adjacent regions, Main Ranges, Rocky Mountains, British Columbia. Geological Survey of Canada Paper 90-1D, 81 – 89. Devlin, W. J., Brueckner, H. K. & Bond, G. C. 1988. New isotopic data and a preliminary age for volcanics near the base of the Windermere Supergroup, northeastern Washington, U.S.A. Canadian Journal of Earth Sciences, 25, 1906–1911. Eisbacher, G. H. 1981a. Sedimentary tectonics and glacial record in the Windermere Supergroup, Mackenzie Mountains, northwestern Canada. Geological Survey of Canada Paper 80-27. Eisbacher, G. H. 1981b. The Late Precambrian Mount Lloyd George diamictites, northern British Columbia. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 728–730.

422

M. D. SMITH ET AL.

Evenchick, C. A., Parrish, R. R. & Gabrielse, H. 1984. Precambrian gneiss and Late Proterozoic sedimentation in north-central British Columbia. Geology, 12, 233– 237. Fanning, C. M. & Link, P. K. 2004. U –Pb SHRIMP ages of Neoproterozoic (Sturtian) glaciogenic Pocatello Fm., southeastern Idaho. Geology, 32, 881–884. Fedo, C. M. & Cooper, J. D. 2001. Sedimentology and sequence stratigraphy of Neoproterozoic and Cambrian units across a craton-margin hinge zone, southeastern California, and implications for the early evolution of the Cordilleran margin. Sedimentary Geology, 141– 142, 501– 522. Ferri, F., Rees, C. J., Nelson, J. L. & Legun, A. S. 1999. Geology and mineral deposits of the northern Kechika Trough between Gataga River and the 60th parallel. British Columbia Ministry of Energy and Mines Bulletin 107. Gabrielse, H. & Campbell, R. B. 1991. Upper Proterozoic assemblages. In: Gabrielse, H. & Yorath, C. J. (eds) Geology of the Cordilleran Orogen. Geological Survey of Canada, Geology of Canada, 4, 125– 150. Glover, J. K. & Price, R. A. 1976. Stratigraphy and structure of the Windermere Supergroup, southern Kootenay Arc, British Columbia. Geological Survey of Canada Paper 76-1B, 21 –23. Grasby, S. E. & Brown, R. L. 1993. New correlations of the Hadrynian Windermere Supergroup in the northern Selkirk Mountains, British Columbia. Geological Survey of Canada Paper 93-1A, 199– 206. Gussow, W. C. 1957. Cambrian and Precambrian geology of southern Alberta. Alberta Society of Petroleum Geologists Guidebook, Seventh Annual Field Conference, 7 – 19. Gussow, W. C. 1973. Chuaria sp. cf. C. circularis Walcott from the Precambrian Hector Fm., Banff National Park, Alberta, Canada. Journal of Paleontology, 47, 1108– 1112. Hein, F. J. & McMechan, M. E. 1994. Proterozoic – Lower Cambrian strata of the Western Canada Sedimentary Basin. In: Mossop, G. D. & Shetsen, I. (comps) Geological Atlas of the Western Canada Sedimentary Basin. Canadian Society of Petroleum Geologists, Calgary, 57 –68. Hofmann, H. J. 1971. Precambrian fossils, pseudofossils and Problematica in Canada. Geological Survey of Canada Bulletin 189. Hofmann, H. J. & Mountjoy, E. W. 2001. Namacalathus-Cloudina assemblage in Neoproterozoic Miette Group (Byng Fm.), British Columbia; Canada’s oldest shelly fossils. Geology, 13, 819– 821. James, N. P., Narbonne, G. M. & Kyser, T. K. 2001. Late Neoproterozoic cap carbonates; Mackenzie Mountains, northwestern Canada; precipitation and global glacial meltdown. Canadian Journal of Earth Sciences, 38, 1229– 1262. Kendall, B. S., Creaser, R. A., Ross, G. M. & Selby, D. 2004. Constraints on the timing of Marinoan ‘Snowball Earth’ glaciation by 187 Re– 187Os dating of a Neoproterozoic, post-glacial black shale in Western Canada. Earth and Planetary Science Letters, 222, 729– 740. Kubli, T. E. 1990. Geology of the Dogtooth Range, northern Purcell Mountains, British Columbia. PhD thesis, University of Calgary. Leech, G. B. 1954. Canal Flats, British Columbia. Geological Survey of Canada Paper 54-7. Link, P. K., Christie-Blick, N. et al. 1993. Middle and Late Proterozoic stratified rocks of the western United States Cordillera, Colorado Plateau, and Basin and Range Province. In: Reed, J., Sims, P., Houston, R. S., Rankin, D. W., Link, P. K., Van Schmus, W. R. & Bickford, M. E. (eds) Precambrian: Conterminous United States. Geological Society of America Decade of North American Geology Series, C-3, 474–690. Lis, M. G. & Price, R. A. 1976. Large-scale block faulting during deposition of the Windermere Supergroup (Hadrynian) in southeastern British Columbia. Geological Survey of Canada Paper 76-1A, 135– 136. Loveridge, W. D., Leech, G. B., Stevens, R. D. & Sullivan, R. W. 1981. Zircon and monazite age of a granitic clast in Toby Conglomerate (Windermere Supergroup), Canal Flats, British Columbia. Geological Survey of Canada Paper 81-1C, 131 – 134. Lund, K., Aleinikoff, J. N., Evans, K. V. & Fanning, C. M. 2003. SHRIMP U –Pb geochronology of Neoproterozoic Windermere

Supergroup, central Idaho; implications for rifting of western Laurentia and synchroneity of Sturtian glacial deposits. Geological Society of America Bulletin, 115, 349– 372. Macdonald, F. A., Schmitz, M. D. et al. 2010. Calibrating the Cryogenian. Science, 327, 1241– 1243. McDonough, M. R. 1989. The structural geology and strain history of the northern Selwyn Range, Rocky Mountains, near Valemount, British Columbia. PhD thesis, University of Calgary. McDonough, M. R. & Parrish, R. R. 1991. Proterozoic gneisses of the Malton Complex, near Valemount, British Columbia; U– Pb ages and Nd isotopic signatures. Canadian Journal of Earth Sciences, 28, 1202– 1216. McMechan, M. E. 1987. Stratigraphy and structure of the Mount Selwyn area, Rocky Mountains, northeastern British Columbia. Geological Survey of Canada Paper 85-28. McMechan, M. E. 1990. Upper Proterozoic to Middle Cambrian history of the Peace River Arch; evidence from the Rocky Mountains. Bulletin of Canadian Petroleum Geology, 38A, 36 – 44. McMechan, M. E. 2000. Vreeland diamictites; Neoproterozoic glaciogenic slope deposits, Rocky Mountains, Northeast British Columbia. Bulletin of Canadian Petroleum Geology, 48, 246– 261. McMechan, M. E. & Thompson, R. I. 1985. Geology of southeast Monkman Pass area (93I/SW), British Columbia. Geological Survey of Canada Open File 1150. McMechan, M. E. & Thompson, R. I. 1995a. Geology, Ovington Creek, west of sixth meridian, British Columbia. Geological Survey of Canada Map 1873A. McMechan, M. E. & Thompson, R. I. 1995b. Geology, Wapiti Pass, west of sixth meridian, British Columbia. Geological Survey of Canada Map 1872A. Murphy, D. C. 1986. Stratigraphy and structure of the east-central Cariboo Mountains, British Columbia, and implications for the geological evolution of the southeastern Canadian Cordillera. PhD thesis, Carleton University. Murphy, D. C., Walker, R. T. & Parrish, R. R. 1991. Age and geological setting of Gold Creek Gneiss, crystalline basement of the Windermere Supergroup, Cariboo Mountains, British Columbia. Canadian Journal of Earth Sciences, 28, 1217– 1231. Park, J. K. 1997. Paleomagnetic evidence for low-latitude glaciation during deposition of the Neoproterozoic Rapitan Group, Mackenzie Mountains, N.W.T., Canada. Canadian Journal of Earth Sciences, 34, 34 – 49. Parrish, R. R. & Scammell, R. J. 1988. The age of the Mount Copeland syenite gneiss and its metamorphic zircons, Monashee Complex, southeastern British Columbia. Geological Survey of Canada Paper 88-2, 21 –28. Pell, J. & Simony, P. S. 1987. New correlations of Hadrynian strata, south-central British Columbia. Canadian Journal of Earth Sciences, 24, 302– 313. Pigage, L. C. & MacNaughton, R. B. 2004. Reconnaissance geology of northern Toobally Lake (95D/8), southeast Yukon. In: Emond, D. S. & Lewis, L. L. (eds) Yukon Exploration and Geology 2003. Yukon Geological Survey, 199–219. Pope, A. 1990. Geology and mineral deposits of the Toby-Horsethief Creek map area, northern Purcell Mountains, southeast British Columbia (82K). Ministry of Energy, Mines and Petroleum Resources, Mineral Resources Division Open File Report 1990-26, 1 –53. Reesor, J. E. 1973. Geology of the Lardeau Map-area, east-half, British Columbia. Geological Survey of Canada Memoir 369. Root, K. G. 1987. Geology of the Delphine Creek area, southeastern British Columbia: implications for the Proterozoic and Paleozoic development of the Cordilleran divergent margin. PhD thesis, University of Calgary. Ross, G. M. 1991. Tectonic setting of the Windermere Supergroup revisited. Geology, 19, 1125– 1128. Ross, G. M. 2000. The Neoproterozoic Windermere Supergroup: an on-land continental margin turbidite system. Geological Survey of Canada Open File 3932. Ross, G. M. & Murphy, D. C. 1988. Transgressive stratigraphy, anoxia and regional correlations within the late Precambrian Windermere grit of the southern Canadian Cordillera. Geology, 16, 139– 143.

WINDERMERE SUPERGROUP

Ross, G. M. & Bowring, S. A. 1990. Detrital zircon geochronology of the Windermere Supergroup and the tectonic assembly of the southern Canadian Cordillera. Journal of Geology, 98, 879– 893. Ross, G. M. & Parrish, R. R. 1991. Detrital zircon geochronology of metasedimentary rocks in the southern Omineca Belt, Canadian Cordillera. Canadian Journal of Earth Sciences, 28, 1254– 1270. Ross, G. M. & Villeneuve, M. E. 1997. U– Pb geochronology of stranger stones in Neoproterozoic diamictites, Canadian Cordillera; implications for provenance and ages of deposition. Geological Survey of Canada Paper 1997-F, 141– 155. Ross, G. M. & Villeneuve, M. E. 2003. Provenance of the Mesoproterozoic (1.45 Ga) Belt basin (western North America): another piece in the pre-Rodinia paleogeographic puzzle. Geological Society of America Bulletin, 115, 1191– 1217. Ross, G. M. & Ferguson, C. R. 2003a. Geology and structure crosssections, Eddy, British Columbia. Geological Survey of Canada, ‘A’ series map 1967A. Ross, G. M. & Ferguson, C. R. 2003b. Geology and structure crosssections, Lanezi Lake, British Columbia. Geological Survey of Canada, ‘A’ series map 2001A. Ross, G. M., McMechan, M. E. & Hein, F. J. 1989. Proterozoic History: the Birth of the Miogeocline. In: Ricketts, B. D. (ed.) The Western Canadian Sedimentary Basin. Canadian Society of Petroleum Geologists, Calgary, 79 – 104. Ross, G. M., Bloch, J. D. & Krouse, H. R. 1995. Neoproterozoic strata of the southern Canadian Cordillera and the isotopic evolution of seawater sulfate. Precambrian Research, 73, 71– 99. Sevigny, J. H. 1988. Geochemistry of late Proterozoic amphibolites and ultramafic rocks, southeastern Canadian Cordillera. Canadian Journal of Earth Sciences, 25, 1323– 1337.

423

Simony, P. S., Ghent, E. D., Craw, D., Mitchell, W. & Robbins, D. B. 1980. Structural and metamorphic evolution of the northeast flank of the Shuswap Complex, southern Canoe River area, British Columbia. In: Coney, P. J. (ed.) Cordilleran Core Complexes. Geological Society of America Memoir 153. Slind, O. L. & Perkins, G. D. 1966. Lower Paleozoic and Proterozoic sediments of the Rocky Mountains between Jasper, Alberta and Pine River, British Columbia. Bulletin of Canadian Petroleum Geology, 14, 442– 468. Smith, M. D. 2009. Stratigraphic and geochemical evolution of the Old Fort Point Fm., southern Canadian Cordillera: the deep-marine perspective of Ediacaran post-glacial environmental change. PhD thesis, University of Ottawa. Stelck, C. R., Burwash, R. A. & Stelck, D. R. 1978. The Vreeland High; a Cordilleran expression of the Peace River Arch. Bulletin of Canadian Petroleum Geology, 26, 87 –104. Stewart, J. H. 1972. Initial deposits in the Cordilleran geosyncline; evidence of a Late Precambrian (,850 m.y.) continental separation. Geological Society of America Bulletin, 83, 1345– 1360. Walcott, C. D. 1910. Pre-Cambrian rocks of the Bow River Valley, Alberta, Canada. Smithsonian Miscellaneous Collections, 53, 423–431. Walker, J. F. 1926. Geology and Mineral Deposits of the Windermere Map-area, British Columbia. Geological Survey of Canada Memoir 148. Warren, M. J. 1997. Crustal extension and subsequent crustal thickening along the Cordilleran rifted margin of ancestral North America, western Purcell Mountains, southeastern British Columbia. PhD thesis, Queen’s University. Available at: http://www.collectionscanada. ca/obj/s4/f2/dsk3/ftp04/nq22501.pdf. Weiner, J. L. 1966. The Old Fort Point Fm., Jasper, Alberta. PhD thesis, University of Alberta.

Chapter 38 Neoproterozoic strata of southeastern Idaho and Utah: record of Cryogenian rifting and glaciation PAUL KARL LINK1* & NICHOLAS CHRISTIE-BLICK2 1

Department of Geosciences, Idaho State University, Pocatello, ID 83209-8072, USA

2

Department of Earth and Environmental Sciences and Lamont-Doherty Earth Observatory of Columbia University, Palisades, NY 10964-8000, USA *Corresponding author (e-mail: [email protected]) Abstract: Neoproterozoic strata in southeastern Idaho and Utah include the ,766 Ma Uinta Mountain Group and Big Cottonwood Formation (Fm.) deposited in an east-trending rift basin and, to the west, the lower part of a westward-thickening rift to passivemargin succession that initiated c. 720 Ma. The latter contains a lower diamictite and volcanic succession, with a complex stratigraphic interval of Cryogenian marine glacial deposits (Pocatello and Mineral Fork formations and correlatives). This is overlain by a mostly terrigenous succession of ,667 Ma strata assigned to the upper member of the Pocatello Fm. and Brigham Group in southeastern Idaho, to the Kelley Canyon Fm. and Brigham Group in northern and western Utah, and to the McCoy Creek Group and Prospect Mountain Quartzite in adjacent Nevada. Although the Brigham Group and correlative deposits contain no direct evidence for glaciation, widely developed, though stratigraphically restricted, incised valleys, with erosional relief from a few metres to as much as 160 m, are inferred to represent subsequent times of Cryogenian glacially lowered sea level. Overall interpretations of the stratigraphy and sedimentology of these rocks have changed little in the past 10–15 years. The most important recent advances relate to U– Pb geochronology. In strata that lie unconformably below demonstrable glacial deposits, the lower Uinta Mountain Group (formerly thought to be c. 900 Ma) contains populations of detrital zircons as young as 766 + 5 Ma. Cryogenian magmatism north of the Snake River Plain in central Idaho is recognized near House Mountain, east of Boise at c. 725 + 5 Ma, in the Pioneer Mountains Core Complex at about 695 Ma, and in central and east-central Idaho at 685–650 Ma. Clasts interpreted to be from the rift-related Bannock Volcanic Member of the Pocatello Fm. are dated at 717 + 4 Ma and 701 + 4 Ma. The overlying diamictite-bearing Scout Mountain Member contains a mafic lapilli tuff near the base (686 + 4 Ma) and a reworked fallout tuff near the top (667 + 5 Ma). Strongly negative C-isotope data have been obtained from some of the carbonate rocks, although the latter constitute only a small fraction of the succession. Palaeomagnetic data are available only for the Uinta Mountain Group, and suggest an equatorial palaeolatitude.

Neoproterozoic glaciogenic rocks, locally in excess of 1 km thick, are exposed widely but discontinuously in southeastern Idaho, and northern and western Utah (Fig. 38.1). The deposits are assigned to a plethora of local formal and informal stratigraphic units, for historical reasons and because of their varied expression. Names used for commonly correlated diamictite-bearing units include Pocatello Fm. in Idaho, and Mineral Fork Fm., Sheeprock Group, formation of Perry Canyon, Trout Creek sequence (units 3 and 5), and Horse Canyon Fm. in Utah (Fig. 38.2). Crittenden et al. (1971, 1983) and Link et al. (1993, 1994) provide regional reviews of diamictites and their interpretation. Important papers for specific locations include Blackwelder (1932), Crittenden et al. (1952), Ojakangas & Matsch (1980), Blick (1981), ChristieBlick (1982, 1983a, 1983b, 1985, 1997), Christie-Blick & Link (1988), Christie-Blick & Levy (1989) and Yonkee et al. (2000a) for northern and west-central Utah; Ludlum (1942), Trimble (1976), Link (1981, 1983, 1987; Link et al. 2005) for the Pocatello area of Idaho; and Misch & Hazzard (1962), Bick (1966) and Rodgers (1994) for westernmost Utah and eastern Nevada. Type localities are described by Misch & Hazzard (1962; Trout Creek sequence and McCoy Creek Group), Bick (1966; Horse Canyon Fm.), Crittenden et al. (1971; Kelley Canyon Fm., Brigham Group and Pocatello Fm.), Christie-Blick (1982, 1983a; Mineral Fork Fm. and Sheeprock Group), Crittenden et al. (1983; Pocatello Fm.), Link (1983, 1987; Pocatello Fm.) and Link et al. (1985; Brigham Group). The stratigraphy and sedimentology of these strata were worked out primarily in the 1970s and early 1980s. The correlations first synthesized by Crittenden et al. (1971) were based on regional field mapping, and have largely stood the test of time as summarized in Link et al. (1993). More recent research has focused on U – Pb geochronology (Fanning & Link 2004, 2008) and geochemistry

(Smith et al. 1994; Young 2002; Lorentz et al. 2004; Corsetti et al. 2007; Dehler et al. 2007).

Structural framework The rocks are thought to have accumulated in rift-related basins associated with the development of a passive continental margin in western North America between c. 665 Ma and c. 520 Ma, and in part atop erosional topography with as much as 900 m of local relief (Stewart 1972; Stewart & Suczek 1977; Bond et al. 1983, 1985; Christie-Blick & Levy 1989; Levy & Christie-Blick 1991a; Ross 1991; Christie-Blick 1983a, 1997). They crop out today within and along the eastern flank of the late Jurassic to early Cenozoic Cordilleran thrust-and-fold belt, and across the eastern edge of the late Cenozoic Basin and Range extensional province (Fig. 38.1; Armstrong & Oriel 1965; Armstrong 1968; Levy & Christie-Blick 1989; Allmendinger 1992; Wernicke 1992; DeCelles 2004; DeCelles & Coogan 2006). The structurally lowest thrust sheets encompassing thick glacial and associated deposits at the present level of exposure (generally on the eastern side of the thrust belt) belong to the Willard –Paris – Putnam system in northern Utah and southeastern Idaho, and to the Tintic – Sheeprock –Canyon Range system of west-central Utah (Levy & Christie-Blick 1989; DeCelles 2004; Fig. 38.1). For example, the Pocatello Fm. in the Bannock Range of southeastern Idaho and the formation of Perry Canyon near Ogden in the northern Wasatch Range of Utah represent a 200 km strike-parallel outcrop band within the Putnam-Paris thrust sheet. Although the rocks are generally foliated, with pervasive development of chlorite and locally biotite in the greenschist facies, detrital zircon geochronology and C-isotope studies have been successfully

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 425– 436. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.38

426

P. K. LINK & N. CHRISTIE-BLICK

Fig. 38.1. Map showing areas of outcrop of Neoproterozoic rocks of Idaho and Utah, with key locations shown (after Link et al. 1994). Inset map is present geography. Main map is palinspastic reconstruction after Levy & Christie-Blick (1989).

Windermere Supergroup Neoproterozoic C Age Cryogenian Ediacaran (Ma)

conducted (Fanning & Link 2004, 2008; Lorentz et al. 2004; Corsetti et al. 2007). One or more generations of Neogene extensional faults cut all of these ranges. In many cases, therefore, rocks transported eastward during Cretaceous thrusting have been translated westward during Neogene extension (Levy & Christie-Blick 1989). Detailed field studies suggest that primary sedimentary features and correlations are discernable through the deformation. In the structurally highest hinterland thrust sheet of the Deep Creek Range, close to the Utah – Nevada state line, garnet and staurolite grade rocks are present in the Trout Creek sequence (Nelson 1966; Rodgers 1994). East of there, but in the same thrust Allochthon

Deep Creek SE Idaho N. Wasatch Sheeprock Range

Prospect Mountain v ? McCoy Creek Gp 600 ?

700

Camelback Mountain

Prospect Mountain

Browns Hole Fm 580 v Mutual Fm Inkom Fm Caddy Canyon Fm

Tintic Fm

Uinta Mtns Tintic

‘Mutual’ Fm

v 667 Poca- d d d d Mineral tello v Fm of v Sheepd Fork Fm Perry rock v 686 Fm d Canyon d Group 1000

Sumidouro

250

One of the earliest isotopic studies of a post-glacial cap carbonate was performed in the Ireceˆ Basin (Una Group) by Torquato & Misi (1977), although at the time the unit was considered lacustrine in origin. These authors analysed 85 carbonate samples for C and O isotopes through two complete sections of the Una Group. The results reveal a d13C trend from consistently negative values in the red argillaceous dolostone (Unit C, c. –6‰) immediately above diamictite of the Bebedouro Fm. to values c. 0‰ up-section (Units B and B1) and continuing to highly positive values (þ8‰, Unit A1). The negative values in the dolostone overlying the Bebedouro Fm. diamictite are similar to post-glacial trends worldwide and may support the glacial interpretation for these deposits. This trend has been reproduced by Misi & Veizer (1998), who analysed well-preserved micro-samples that were first characterized by petrographic and geochemical techniques. In addition, these authors determined 87Sr/86Sr ratios on the best-preserved carbonates. The least radiogenic 87Sr/86Sr value of 0.70780 from Unit B is consistent with a sedimentation age between 750 and 600 Ma (cf. Jacobsen & Kaufman 1999) (Fig. 48.7).

Arrependido Sa˜o Francisco Basin

1000 arkosic sandstone D 250

Fig. 48.4. Stratigraphic units of the Vazante Group.

´ gua Fm. (200 – 1300 m) is composed of interThe Olhos d’A bedded limestones that become organic-rich towards the top. The upper section consists of c. 50 m of limestone –pelite intercalations that are overlain by laminated limestones (40 m thick) culminating with a 30-m-thick layer of organic-rich limestone. The lower section of laminated limestones is in sharp contact with the diamictites of the Palestina Fm., which is composed of metagreywackes with clasts up to 30 cm in diameter. The clasts are of gneiss, granite, quartz, quartzite, phyllite and metacarbonate supported in a quartz –sericite matrix. Lenses of Fe-cemented quartzite unconformably overlie the siliciclastic sediments of the Sima˜o Dias Group (Fig. 48.6). The Palestina Fm. is interpreted as a glaciogenic deposit. At a lower stratigraphic level, the equivalent Jueteˆ and Ribeiro´polis formations (D’el Rey Silva 1999) are also interpreted as glaciogenic and may correlate with the Bebedouro (Una Group) and Jequitaı´

The carbonates of the Bambuı´ Group in the Sa˜o Francisco Basin have been the subject of many chemostratigraphic studies (e.g. Iyer et al. 1995; Kiang 1997; Kawashita 1998; Santos et al. 2000, 2004; Kaufman et al. 2001). The most detailed chemostratigraphic study developed in the Bambuı´ Group thus far comes from high-resolution profiles from the Serra do Ramalho area, western Bahia (Powis 2006; Misi et al. 2007). Chemostratigraphic studies have also been conducted by Iyer et al. (1995), Kaufman et al. (2001) and Vieira et al. (2007) in the Sete Lagoas area of the Sa˜o Francisco Basin. As in the Ireceˆ Basin, the d13C composition of carbonate rocks in the Serra do Ramalho and in the Sete Lagoas area increase up-section from c. –4‰ at the base of the Sete Lagoas Fm. overlying glacial diamictite of the Jequitaı´ Fm. (assumed to be correlative to the Bebedouro Fm.) and rising steadily up to c. þ14‰ near the top of the succession in Lagoa do Jacare´ Fm. The best-preserved samples (Mn/Sr  0.02) show a narrow range of 87Sr/86Sr values between 0.7074 and 0.7075 (Fig. 48.7).

Brasilia Fold Belt (Vazante Group) Petrographic investigations showed that some of the Vazante carbonates were affected by dolomitization, but examination of thin sections shows that these rocks have very fine-grained and fabric-retentive dolomicrites (Azmy et al. 2001, 2006). Three generations of secondary dolomitic cements have been observed,

˜ O FRANCISCO CRATON NEOPROTEROZOIC SUCCESSIONS OF THE SA

515

Fig. 48.5. Simplified geological map of the Sergipano Belt, modified from D’el Rey Silva (1999). Basement domes: 1, Itabaiana; 2, Sima˜o Dias; 3, Jirau do Ponciano. Main thrust and strike–slip fault: MF, Macurure´ fault; BMJF, Belo Monte-Jeremoabo fault; SMAF, Sa˜o Miguel do Aleixo fault; IF, Itaporanga˜ fault.

including fibrous, equant and late fracture-filling phases (cf. Azmy et al. 2001). No significant increase in crystal size was observed associated with dolomitization, suggesting that the original sediment did not suffer from extensive and/or repeated meteoric alteration. The trace element compositions of the Vazante carbonates, particularly those of the fabric retentive micrites, suggest significant geochemical preservation (Azmy et al. 2001, 2006). The C-isotope profiles reveal a negative shift up to 8‰ in dolomite beds above the dropstone-laden shale of the Serra do Poc¸o Verde Fm. (Fig. 48.8) and near the base of the Lapa Fm. (cf. Azmy et al. 2006; Misi et al. 2007). The least radiogenic Sr-isotope values associated with the negative d13C shifts are 0.7061 in the Serra do Poc¸o Verde Fm. and 0.7068 in the Lapa Fm. (Fig. 48.8).

Sergipano Fold Belt (Vaza Barris Group) Chemostratigraphic studies have been performed on some carbonate units of the Vaza Barris Group by Kawashita (1998) and by Sial et al. (2000, 2003, 2005, 2006a, b). The first author determined least-altered 87Sr/86Sr ratios of 0.7078 in black limestones of the ´ gua Fm. In addition Sial et al. (2000) report d13C low Olhos d’A values of – 4.7‰ immediately above diamictite of the Palestina Fm. These numbers increase up-section to values as high as ´ gua Fm. Sial et al. (2006a) also þ10‰ in the upper Olhos d’A report negative excursions as low as – 5% in the lower carbonates of the Acaua˜ Fm. (equivalent to the Jacoca Fm.; Fig. 48.9) immediately above the diamictites of the Ribeiro´polis Fm., suggesting that at least two important glaciogenic events are represented in the Vaza Barris Group. The results support a possible correlation of

Fig. 48.6. Summary stratigraphy of the Itabaiana dome area, based on D’el Rey Silva (1999). Megasequences: LSM, lower siliciclastic; LCM, lower carbonate (megasequences of Cycle I); USM, upper siliciclastic; UCM, upper carbonate (megasequences of Cycle II). Formations: JU, Jueteˆ A, Acaua˜; I, Itabaiana; R, Ribeiro´polis; JC, Jacoca; LP, LaegartoPalmares; Ja, Jacare´; FP, Frei Paulo; P, ´ gua Formation. Palestina; OA, Olthos d’A

516

A. MISI ET AL.

Table 48.2. Correlations between intracratonic and passive-margin basins

Mesoproterozoic Units (overthrusted terrain)

Vazante group (Formations) Serra da Lapa

Bambuí group (Formations)

Una group (Units) (Form.)

Três Marias

Morro do Calcário

Serra sa Saudade

Serra do Poço Verde

Lagoa do Jacaré

A1

Serra de Santa Helena

A

Serra da Garrote Lagamar Rocinha

Santo Antônio do Bonito

Sergipano Fold Belt Passive-margin Basin

Intracratonic Basins

Sete Lagoas 3

B1

Sete Lagoas 2

B

Sete Lagoas 1

C

Jequitaí

Vaza barrís/miaba gr. (Formations)

–

–

–

–

Salitre Formation

Brasiliano Fold Belt Passive-margin Basin

Bebedouro

Olhos d’Água Palestina Simão Dias Group Jacoca Ribeirópolis

the Vaza Barris Group with the carbonate platform successions of the Bambuı´ and Una Groups, as sustained by some authors (e.g. Misi 1979; Teixeira & Figueiredo 1991; Trompette 1994; D’el Rey Silva 1999). The diamictites of the Ribeiro´polis Fm. may be equivalents of the Bebedouro and the Jequitaı´ formations (respectively, from the Una and Bambuı´ groups), but no equivalent glacial deposits correlating with the Palestina Fm. have been clearly demonstrated from these epicontinental successions (Fig. 48.9 & Table 48.2).

Geochronological constraints Despite the absence of reliable absolute geochronological age estimates in the platform carbonate successions of the Sa˜o Francisco craton, there is a general agreement in considering a middle Cryogenian age for the Bambuı´ and Una groups. The least radiogenic Sr isotope values of 0.7074– 0.7077 and S (from stratform barite), with a mean value of c. þ30 per mil CDT in the Una Group, indicate that sedimentation likely occurred between c. 780 Ma and 600 Ma (Misi & Veizer 1998; Misi et al. 1999). More recently, Babinski & Kaufman (2003) and Babinski et al. (2007) determined new Pb/Pb carbonate ages from wellpreserved post-glacial sea-floor cements in the Sete Lagoas Fm. of the Bambuı´ Group, suggest a depositional age of 740 + 22 Ma (MSDW ¼ 0.66). Notably, almost the same age (730 Ma, U –Pb in zircon) has been recorded by Brito-Neves (pers comm.) in acidic tuffs intercalated in the Ribeiro´polis diamictites of the Vaza Barris/Miaba Group, in the Sergipano Belt. According to D’el Rey Silva (1999), sedimentation in the Sergipano Belt lasted from c. 1.0 Ga (U –Pb zircon in metavolcanics, Brito Neves et al. 1993) to possibly 0.65 Ga. On the other hand, the global stratigraphic correlation of the Vazante Group in the Sa˜o Francisco Basin has been a challenging issue. The resetting of the radiogenic clock by the major Brası´lian –Pan African orogeny (c. 550– 600 Ma) (Trindade et al. 2004) and the scarcity of volcanic ash layers makes absolute

Fig. 48.7. Proposed lithostratigraphic and chemostratigraphic correlation between the Bambuı´ and Una Groups, respectively in the Sa˜o Francisco Basin (Serra do Ramalho area) and in the Ireceˆ Basin (Misi & Veizer 1998; Powis 2006; Misi et al. 2007).

˜ O FRANCISCO CRATON NEOPROTEROZOIC SUCCESSIONS OF THE SA

517

the basal portion of the Lagamar Fm. with a maximum age of 1.95 Ga. consistent with the c. 1.35 Ga Re – Os ages obtained by Geboy (2006) from organic shales of the same unit.

Mineralization Mineral deposits associated with Late Precambrian glacial and carbonate sediments of the Sa˜o Francisco craton include the following:

Fig. 48.8. Chemostratigraphy of the Vazante Group.

geochronological age control of this succession problematic. The abundance of Conophyton metula Kirichenko in the Lagamar Fm. (Fig. 48.4) suggests an age estimate between 1350 Ma and 950 Ma for the sedimentation of this part of the Vazante Group (Cloud & Dardenne 1973). Nevertheless, Misi et al. (1997) and Sanches et al. (2007) analysed carbonate fluorapatite and finegrained organic-rich limestone (micrite) from the lower Rocinha Formation for their Sr-isotope composition. 87Sr/86Sr values range from 0.70760 to 0.70790, similar to those reported for the Bambuı´, Una and Vaza Barris/Miaba Groups. All seven samples had high total Sr (.1300 ppm) and low Mn– Sr (0.1 –0.01), indicating a high degree of preservation of the original seawater signal. Sm –Nd studies of provenance (Pimentel et al. 2001) on finegrained sediments of the Bambuı´, Vazante and Paranoa´ groups have furnished TDM model age intervals of 1.9 –1.4 Ga, 2.3– 1.7 Ga and 2.3 –1.9 Ga for the respective groups. According to these authors, the provenance results suggest that the Vazante Group constitutes a transitional sequence between the Paranoa´ and Bambuı´ groups. Recent Re – Os investigations of organic-rich shales of the upper Vazante Group (Geboy 2006; Azmy et al. 2008) suggest a significantly older, Mesoproterozoic age for the Morro do Calca´rio and Serra do Garrote units, with ages of c. 1 Ga and older. However, U –Pb determinations on detrital zircons (Dardenne et al. 2003, 2005; Dardenne 2007a; Rodrigues 2008; Rodrigues et al. 2008) indicated maximum depositional ages for the Santo Antonio do Bonito Fm. (diamictites at the basal section of the Vazante Group, Fig. 48.4) of 988 + 15 Ma. Rodrigues et al. (2008) and Rodrigues (2008) also report a significant Neoproterozoic zircon population from the Rocinha Fm. in which the youngest grains have been dated c. 925 Ma. Nonetheless, in the same study the authors identified provenance patterns for the conglomerates at

† Fe – Mn deposits near Porteirinha (Minas Gerais state), which are intimately associated with glacial sedimentation of the Macau´bas Group in the Arac¸uaı´ belt, and likely related to exhalative processes (Dardenne & Schobbenhaus 2001). † Phosphorite deposits interbedded with carbonates in the lower Vazante Group at Coromandel, Rocinha and Lagamar (Dardenne et al. 1986, 1997; Da Rocha-Arau´jo et al. 1992; Sanches et al. 2007), at the base of the Sete Lagoas Fm. in the Bambuı´ Group at Campos Belos (Dardenne et al. 1986), in the lower half of the Serra da Saudade Fm. in the Bambuı´ Group at Cedro do Abaete´ (Dardenne et al. 1986; Lima et al. 2005), and in stromatolitic beds of the Salitre Fm. in the Ireceˆ basin (Misi & Kyle 1994). † Fluorite, barite and Pb –Zn deposits associated with the upper part of the first carbonate megacycle of the Bambuı´ Group near Janua´ria-Itacarambi, Montalvaˆnia and Serra do Ramalho (Dardenne 1979; Misi 1979; Dardenne & Freitas-Silva 1999). † Pb– Zn sulphide deposits associated with brecciated dolomite in the middle Vazante Group at Morro Agudo mine (Hitzman et al. 1995; Misi et al. 1998, 1999; Dardenne & Freitas-Silva 1999; Bettencourt et al. 2001; Monteiro 2002). † Zn silicate deposits associated with major fault structures at the Vazante mine, Minas Gerais (Dardenne & Freitas-Silva 1999; Bettencourt et al. 2001; Monteiro 2002; Misi et al. 2005).

Discussion Broadly speaking, the Neoproterozoic Era experienced at least three widespread ice ages, commonly referred to as the Gaskiers, Marinoan and Sturtian events, which are radiometrically constrained to be c. 585 Ma (Newfoundland: Bowring et al. 2003; Halverson et al. 2005), c. 630 Ma (South China: Condon et al. 2005; Zhang et al. 2005) and c. 750 Ma (Namibia: Hoffman et al. 1996), respectively. Some researchers suggest that these singular glacial events may have been diachronous, spanning over tens of millions of years, or alternatively split into several discrete ice ages (e.g. Kaufman et al. 1997; Kendall et al. 2006). In most cases, the glacial diamictites are overlain by texturally and isotopically enigmatic cap carbonates with strong negative C-isotope anomalies (e.g. Kaufman et al. 1997; Hoffman et al. 1998b), which have also been recorded in syn- and immediately pre-glacial carbonates (e.g. Kaufman et al. 1997; Kennedy et al. 2001; Halverson et al. 2002). In contrast, S-isotope compositions in the cap carbonates preserve remarkably positive d34S anomalies (e.g. Hurtgen et al. 2002). In all the studied Neoproterozoic successions from the Sa˜o Francisco craton, at least one horizon of glacial diamictite (generally recognized on the basis of faceted and striated clasts of heterogeneous composition in a fine-grained matrix) has been recognized. Most often, these glacial horizons occur at the base of the succession. These ice age deposits are overlain by cap carbonates, although in some cases a thin siliciclastic wedge, perhaps recording glacial outwash facies during meltback (Bailey & Peters 1998), separates the diamictite from carbonate lithofacies. Chemostratigraphic studies (including C, S and Sr isotopes) and Pb/Pb dating of carbonates from these glaciogenic successions provide tools for possible correlations with Neoproterozoic counterparts around the world.

518

A. MISI ET AL.

Fig. 48.9. Chemostratigraphy of the Vaza Barris/Miaba Groups (modified from Sial et al. 2000).

Perhaps of greatest interest are the correlation of Bambuı´, Una, Vazante and Vaza Barris/Miaba group deposits on the Sa˜o Francisco Craton with Neoproterozoic equivalents on the Congo Craton in Namibia. If we accept that throughout much of the Mesoproterozoic and Neoproterozoic eras these cratons were joined, the broadly equivalent sedimentary successions should reveal striking similarities. However, whereas the well-studied Otavi Group, in Namibia, hosts at least two cap carbonates atop glacial sediments (Kaufman et al. 1991; Hoffmann & Prave 1996; Hoffman et al. 1998b), only one ice age deposit is currently recognized in the Bambuı´ and equivalent Una groups. Diamictites from these units containing basement clasts in a fine-grained Fe-rich mudstone and overlained by a cap-carbonate lithofacies with centimetre-scale sea-floor precipitates, are strikingly similar to those of the younger Maieberg Fm. cap carbonate in Namibia. The Sete Lagoas precipitates are preserved in limestone with negative d13C values and high Sr abundances. Sr-isotope compositions of these lithologies are consistently around 0.7073, which is an exact match with the precipitate interval in the Maieberg Fm., supporting this intracontinental correlation. Notably, Pb/Pb dating of the Sete Lagoas precipitates yielded an age of 740 + 22 Ma (11-point isochron; MSWD ¼ 0.66; Babinski et al. 2007). Acceptance of this date and the correlation with the precipitates in the Maieberg Fm. associates both of these post-glacial deposits with the Sturtian ice age. This assignment is at odds with the view of the Maieberg Fm. as the archetypical Marinoan cap carbonate (Hoffman & Schrag 2002), and the c. 630 Ma age assigned to the underlying Ghaub diamictite by the correlation of this unit with an ash-bearing diamictite in the deformed Swakop Group of central Namibia (Hoffmann et al. 2004). Independent chemostratigraphic evidence for the equivalence of the dated Swakop Group

diamictite with the Otavi Group is missing and we regard it is possible that these glacial deposits instead correlate with possibly younger glacial deposits of the Witvlei Group on the Kalahari Craton in southern Namibia (Kaufman et al. 1997; Saylor et al. 1998). On the western margin of the Sao Francisco Craton in Brazil, the basal glacial deposit of the Vazante Group (Santo Antonio do Bonito Fm.) has long been considered as an equivalent to the basal diamictite of the Bambuı´ Group beneath the Sete Lagoas Fm. Overlying strata in both successions are variably mineralized (phosphate and Pb– Zn) and contain thick carbonate intervals. However, C-isotope compositions of the carbonates differ significantly. The d13C profile for Vazante carbonates is generally near zero, and shows only minor excursions to negative values that are interpreted to reflect cap-carbonate lithofacies (Azmy et al. 2001, 2006). In contrast, the strongly negative d13C values of the Sete Lagoas cap carbonate are followed by remarkably positive values (upwards to þ16‰) in overlying strata (Misi et al. 2007). Thus the equivalence of these successions is suspect, although Sr-isotope compositions of lower Vazante phosphorites are quite similar to those in well-preserved limestones of the Bambuı´ and Una groups (Misi & Veizer 1998). Further complicating the interbasinal correlation is the presence of a second Vazante Group diamictite (D II) in the Serra do Poc¸o Verde Fm. (Fig. 48.2: Olcott et al. 2005; Azmy et al. 2006), and the possibility that the overlying Lapa Fm. was deposited during oceanic transgression after a third possible glacial event (Brody et al. 2004; Azmy et al. 2006). It has been suggested previously that the shale and marly carbonate of the Lapa Fm. may be equivalent to the Rasthof Fm. in northern Namibia (the older of the two Otavi Group cap carbonates). This assignment, however, was based on comparable

˜ O FRANCISCO CRATON NEOPROTEROZOIC SUCCESSIONS OF THE SA

negative d13C excursions in both units, and on a single Sr-isotope value (c. 0.7068) from a dolomite cement in the basal Lapa Formation (Azmy et al. 2001, 2006) that matches with the lowest values recorded in Sr-rich Rasthof limestones. However, the Sr-isotope equivalence of these units is problematic and, furthermore, new Re –Os age constraints for shale of the Lapa and older organic-rich lithologies in the Vazante Group suggest that this succession may be far older than previously considered. Re –Os age determinations on Lapa Fm. shale suggest an age of close to one billion years (Azmy et al. 2008), and similar measurements of organic-rich shale in the underlying Serra do Poc¸o Verde and Serra do Garrote formations yield ages of c. 1.1 and 1.3 Ga (Geboy 2006; Geboy et al. 2006, 2009). A Mesoproterozoic age for this succession is consistent with the relative invariance of C-isotope trends in the Vazante Group, which are comparable with the Paranoa´ Group stratigraphically beneath Bambuı´ sediments in central and eastern Brazil (Santos et al. 2000). This older age is similarly suggested from Nd-isotope studies of detrital grains in the Vazante Group (Pimentel et al. 2001) and the appearance of Conophyton metula Kirichenko stromatolites (Cloud & Dardenne 1973), which are also known from the Paranoa´ Group. The Re –Os ages provided by two different laboratories would push the known record of Late Precambrian glaciation into the Mesoproterozoic Era, some 500 million years earlier than previously considered. Further tests of the Re – Os ages, as well as time-series elemental, isotopic and biomarker studies (cf. Olcott et al. 2005) of the Vazante Group, which are currently in progress, should shed light on the depositional environment of the glaciogenic Vazante Group and provide better resolution of these controversial age estimates. During the last ten years CNPq has provided funds for research on the Neoproterozoic basins of the Sa˜o Francisco Craton. Currently, A. M. receives support from project no. 486416/2006-2. A. J. K. acknowledges the NSF and NASA for funding for work in the Proterozoic successions of Brazil, as well as the Deutsche Forschungsgemeinschaft and Westfa¨lische Wilhelms-Universita¨t Mu¨nster for sabbatical funding in 2007–2008. A. N. S acknowledges support from the CNPq (grants PROSUL/CNPq 490136/2006-0 and 475657/2006-3) and PRONEX APQ-047-1.07/06 that have defrayed costs with fieldwork in Sergipe and Bahia. We are grateful to the directors of Votorantim Metais for supporting fieldwork at the Vazante area. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Alkmim, F. F., Brito Neves, B. B. & Alves, J. A. C. 1993. Arcabouc¸o Tectoˆnico do Cra´ton do Sa˜o Francisco: uma revisa˜o. In: Dominguez, J. M. L. & Misi, A. (eds) O Craton do Sa˜o Francisco SBG, SGM, CNPq, 45 –62. Alkmim, F. F., Chemale, F., Jr. & Endo, I. 1996. A deformac¸a˜o das coberturas proterozo´icas do Craton do Sa˜o Francisco e o seu significado tectoˆnico. Revista da Escola de Minas de Ouro Preto, 49, 22 – 38. Amaral, G. & Kawashita, K. 1967. Determinac¸o˜es da idade do Grupo Bambuı´ pelo me´todo Rb– Sr. 21st Congresso Brasileiro de Geologia. Curitiba, Anais, 214– 217. Azmy, K., Veizer, J., Misi, A., Oliveira, T. F., Sanches, A. L. & Dardenne, M. A. 2001. Dolomitization and isotope stratigraphy of the Vazante Formation, Sa˜o Francisco Basin, Brazil. Precambrian Research, 112, 303– 329. Azmy, K., Kaufman, A. J., Misi, A. & Oliveira, T. F. 2006. Isotope stratigraphy of the Lapa Formation, Sa˜o Francisco Basin, Brazil: implications for Late Neoproterozoic glacial events in South America. Precambrian Research, 149, 231– 248. Azmy, K., Kendall, B., Creaser, R. A., Heaman, L. & de Oliveira, T. F. 2008. Global correlation of the Vazante Group Sa˜o Francisco Basin, Brazil: Re– Os and U– Pb radiometric age constraints. Precambrian Research, 164, 160172.

519

Babinski, M. & Kaufman, A. J. 2003. First direct dating of a Neoproterozoic post-glaciogenic cap carbonate. IV South American Symposium on Isotope Geology, Short Papers 1, 321– 323. Babinski, M., Vieira, L. C. & Trindade, R. I. F. 2007. Direct dating of the Sete Lagoas cap carbonate (Bambuı´ Group, Brazil) and implications for the Neoproteozoic glacial events. Terra Nova, 19, 401– 406. Bailey, C. M. & Peters, S. E. 1998. Glacially influenced sedimentation in the late Neoproterozoic Mechum River Formation, Blue Ridge province, Virginia. Geology, 26, 623– 626. Bettencourt, J. S., Monteiro, L. V. S., Bello, R. M. S., Oliveira, T. F. & Juliani, C. 2001. Metalogeˆnese do Zinco e chumbo na regia˜o de Vazante-Paracatu, Minas Gerais. In: Pinto, C. P. & Martins-Neto, M. A. (eds) Bacia do Sa˜o Francisco: Geologia e recursos minerais. SBG-Nu´cleo de Minas Gerais, Belo Horizonte, 161– 198. Bowring, S., Myrow, P., Landing, E., Ramezani, J. & Grotzinger, J. 2003. Geochronological constraints on terminal Proterozoic events and the rise of metazoans. Geophysical Research Abstracts (EGS, Nice), 5, 219. Brasier, M. D. & Shields, G. 2000. Neoproterozoic chemostratigraphy and correlation of the Port Askaig glaciation, Dalradian Supergroup of Scotland. Journal of the Geological Society of London, 157, 909– 914. Brito Neves, B. B., Van Schmus, W. R., Babinski, M. & Sabin, T. 1993. O evento de magmatismo de 1,0 Ga nas faixas mo´ veis ao norte do Craa´ton Sa˜o Francisco. Sociedade Brasileira de Geologia, Simpo´sio o Cra´ton do Sa˜o Francisco 2, Salvador, Anais, 243– 245. Brito-Neves, B. B., Campos-Neto, M. C. & Fuck, R. A. 1999. From Rodinia to Western Gondwana: an approach to the Brasiliano– Pan African Cycle and orogenic collage. Episodes, 22, 155– 166. Brody, K. B., Kaufman, A. J., Eigenbrode, J .L. & Cody, G. D. 2004. Biomarker geochemistry of a post-glacial Neoproterozoic succession in Brazil (Abstract), Geological Society of America, Abstracts with Programs, 36, 477. Campos-Neto, M. C. 1984a. Litoestratigrafia, relac¸o˜es estratigra´ficas e evoluc¸a˜o paleogeogra´fica dos grupos Canastra e Paranoa´ (regia˜o Vazante-Lagamar, MG). Revista Brasileira de Geocieˆncias, 14, 81– 91. Campos Neto, M. C. 1984b. Geometria e fases de dobramentos brasilianos superpostos no oeste de Minas Gerais. Revista Brasileira de Geocieˆncia, 14, 60– 68. Chang, H. K., Miranda, F. P. & Alkmim, F. F. 1988. Considerac¸o˜es sobre a evoluc¸a˜o tectoˆnica da Bacia do Sa˜o Francisco. In: 35th Congresso Brasileiro de Geologia (Bele´m), Anais, 5, 2076–2090. Cloud, P. E. & Dardenne, M. A. 1973. Proterozoic age of the Bambuı´ Group in Brazil. Geological Society of America Bulletin, 84, 1673–1676. Condie, K. C. 2002. The supercontinent cycle: are there two patterns of cyclicity? Journal of African Earth Sciences, 35, 179– 183. Cordani, U. G., Brito-Neves, B. B. & D’Agrella-Filho, M. S. 2003. From Rodinia to Gondwana: a review of the available evidence from South Ame´rica. Gondwana Research, 6, 275– 283. Condon, D., Zhu, M., Bowring, S., Wang, W., Yang, A. & Jin, Y. 2005. U–Pb ages from the Neoproterozoic Doushantuo Formation, China. Science, 308, 95– 97. Cozzi, A., Allen, P. & Grotzinger, J. 2004. Understanding carbonate ramp dynamics using d13C profiles: examples from the Neoproterozoic Buah Formation of Oman. Terra Nova, 16, 62 –67. DA ROCHA ARAUJO, P. R., FLICOTEAUX, R., PARRON, C. & TROMPETTE, R. 1992. Phosphorites of Rocinha mine, Patos de Minas (Minas Gerais Brasil): genesis and evolution of a Middle Proterozoic Deposit tectonized by the Brasiliano Orogeny. Economic Geology, 87, 332– 351. Dardenne, M. A. 1978a. Zonac¸a˜o tectoˆnica na borda ocidental do Cra´ton Sa˜o Francisco. 30th Congresso Brasileiro de Geologia, Recife, SBG, 1, 299– 308. Dardenne, M. A. 1978b. Sı´ntese sobre a estratigrafia do Grupo Bambuı´ no Brasil Central. 30th Congresso Brasileiro de Geologia, SBG, 2, 597– 610. Dardenne, M. A. 1979. Les mine´ralisations de plomb, zinc, fluor du Prote´rozoı¨que Supe´rieur dans le Bre´sil Central. The`se de Doctorat d’E´tat, Universite´ de Paris VI. Coromandel. Dissertac¸a˜o de Mestrado.

520

A. MISI ET AL.

Dardenne, M. A. 1981. Os grupos Paranoa´ e Bambuı´ na Faixa Dobrada Brası´lia. An. Simp. Cra´ton Sa˜o Francisco e suas Faixas Marginais, Salvador, SBG, 140–157. Dardenne, M. A. 2000. The Brası´lia Fold Belt. In: Cordani, U. G., Milani, E. J., Thomaz-Filho, A. & Campos, D. A. (eds) Tectonic Evolution of South America. International Geological Congress, 31, Rio de Janeiro, 231– 263. Dardenne, M. A. 2001. Lithostratigraphic sedimentary sequence of the Vazante Group. In: Misi, A. & Teixeira, J. B. G. (eds) Proterozoic Base Metal Deposits of Africa and South America. Proceedings of the 1st Field Workshop IGCP 450, CNPq/UNESCO/IUGS, Belo Horizonte and Paracatu (MG), Brazil, 48 –50. Dardenne, M. A. 2006. A glaciac¸a˜o neoproterozo´ica do Grupo Vazante na Faixa Brası´lia: discussa˜o e alternativas. Congresso Brasileiro de Geologia, 43, SBG, Aracaju, Resumos. Dardenne, M. A. 2007a. The pseudo-diamictites of the Vazante Group. Symposium IGCP 478, Stellenbosch, Abstracts. Dardenne, M. A. 2007b. Lithostratigraphy of the Vazante and Bambuı´ groups in the Sa˜o Francisco Craton and the Brası´lia Fold Belt. IGCP 478, Stellenbosch, Abstracts. Dardenne, M. A. & Freitas-Silva, F. H. 1999. Modelos gene´ticos dos depo´sitos Pb-Zn nos grupos Bambuı´ e Vazante. Workshop: Depo´sitos Minerais Brasileiros de Metais-Base. Salvador, CAPES-PADCTADIMB, 86– 93. Dardenne, M. A. & Schobbenhaus, C. 2001. Metalogeˆnese do Brasil. Editora UnB, Brası´lia, 392. Dardenne, M. A., Trompette, R., Magalha˜es, L. F. & Soares, L. A. 1986. Proterozoic and Cambrian phosphorites-regional review: Brazil. In: Cook, P. J. & Shergold, J. H. (eds) Phosphate Deposit of the World. Proterozoic and Cambrian Phosphorites. Cambridge University Press, 116–131. Dardenne, M. A., Freitas-Silva, F. H., Nogueira, G. M. S. & Souza, J. F. C. 1997. Depo´sitos de fosfato de Rocinha e Lagamar, Minas Gerais. In: Schobbenhaus, C., Queiroz, E. T. & Coelho, C. E. S. (eds) Principais Depo´sitos Minerais do Brasil. DNPM/ CPRM, IVC, 113– 122. Dardenne, M. A., Pimentel, M. M. & Alvarenga, C. J. S. 2003. Provenance of conglomerates of the Bambuı´, Jequitaı´, Vazante and Ibia´ groups, implications for the evolution of the Brası´lia Belt. In: Sociedade Brasileira de Geologia, Simpo´sı´o Nacional de Estudos Tectoˆnicos, 9, International Symposium on Tectonics, 3, Armac¸a˜o de Bu´zios, 2003. Boletim de Resumos, 47–50. D’el-Rey Silva, L. J. H. 1995. Tectonic evolution of the Sergipano Belt, northeastern Brazil. Revista Brasileira de Geocieˆncias, 25, 315– 332. D’el Rey Silva, L. J. H. 1999. Basin infilling in the southern-central part of the Sergipano Belt (NE Brazil) and implications for the evolution of Pan-African/Brasiliano cratons and Neoproterozoic sedimentary cover. Journal of South American Earth Sciences, 12, 453– 470. Derry, L. A., Keto, L. S., Jacobsen, S. B., Knoll, A. H. & Swett, K. 1989. Sr isotopic variations of upper Proterozoic carbonates from East Greenland and Svalbard. Geochimica Cosmochimica Acta, 53, 2331– 2339. Fairchild, I. J., Marshall, J. D. & Bertrand-Safati, J. 1990. Stratigraphic shifts in carbon isotopes from Proterozoic stromatolitic carbonates (Mauritania): influence of primary mineralogy and diagenesis. American Journal of Science, 290A, 46 –79. Fairchild, T. R., Schopf, J. W. et al. 1996. Recent discoveries of Proterozoic microfossils in south-central Brazil. Precambrian Research, 80, 125– 152. Fuck, R. A., Pimentel, M. M. & Silva, J. H. D. 1994. Compartimentac¸a˜o tectoˆnica na porc¸a˜o oriental da Provı´ncia Tocantins. Anais 388 Congresso Brasileiro de Geologia, Camboriu´, SBG, 1, 215– 216. Geboy, N. J. 2006. Rhenium– osmium age determinations of glaciogenic shales from the Mesoproterozoic Vazante Formation, Brazil. MS thesis, University of Maryland, 99. Geboy, N. J., Kaufman, A. J. & Walker, R. J. 2006. A stable isotope and Re –Os study of organic-rich mudstones across a Proterozoic glacial cycle in Brazil. Geological Society of America Annual Meeting (Philadelphia), Abstracts, 38, 125. Geboy, N. J., Kaufman, A. J. et al. 2009. Evidence of Mesoproterozoic Ice Ages in Brazil. Internal report, unpublished.

Guimara˜es, J. T. 1996. A Formac¸a˜o Bebedouro no Estado da Bahia; faciologia, estratigrafia e ambientes de sedimentac¸a˜o. Dissertac¸a˜o (Mestrado em Geologia), Instituto de Geocieˆncias, Universidade Federal da Bahia (UFBA), Brasil. Guimara˜es, J. T., Misi, A., Pedreira, A. J. & Dominguez, J. M. L. 2011. The Bebedouro Formation, Una Group, Bahia (Brazil). In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 503–508. Halverson, G. P., Hoffman, P. F., Schrag, D. P. & Kaufman, A. J. 2002. A major perturbation of the carbon cycle before the Ghaub glaciation (Neoproterozoic) in Namibia: prelude to snowball Earth. Geophysics, Geochemistry, Geosystems, 3, doi: 10.1029/2001G C000244. Halverson, G. P., Hoffman, P. F. & Schrag, D. P. 2005. Toward a Neoproterozoic composite carbon isotope record. Geological Society of America Bulletin, 117, 1181– 1207. Hitzmann, M. W., Thorman, C. H., Romagna, G., Oliveira, T. F., Dardenne, M. A. & Drew, L. J. 1995. The Morro Agudo Zn –Pb deposit, Minas Gerais, Brazil: a proterozoic Irish-type carbonatehosted SEDEX – replacement deposit. Annual Meeting, New Orleans, GSA, Abstract, 408. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129–155. Hoffman, P. F., Hawkins, D. P., Isachsen, C. E. & Bowring, S. A. 1996. Precise U–Pb zircon ages for early Damaran magmatism in the Summas Mountains and Welwitschia inlier, northern Damara belt, Namibia. Geological Survey of Namibia Communications, 11, 47 – 52. Hoffman, P. F., Kaufman, A. J., Halverson, G. P. & Schrag, D. P. 1998a. A Neoproterozoic snowball Earth. Science, 281, 1342– 1346. Hoffman, P. F., Kaufman, A. J. & Halverson, G. P. 1998b. Comings and goings of global glaciation on a Neoproterozoic tropical platform in Namibia. GSA Today, 8, 1– 9. Hoffmann, K.-H. & Prave, A. R. 1996. A preliminary note on a revised subdivision and regional correlation of the Otavi Group based on glaciogenic diamictites and associated cap dolostones. Communications of the Geological Survey of Namibia, 11, 77 – 82. Hoffmann, K.-H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. A U– Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: Constraints on Marinoan glaciation. Geology, 32, 817– 820. Hurtgen, M. T., Arthur, M. A., Suits, N. S. & Kaufman, A. J. 2002. The sulfur isotopic composition of Neoproterozoic seawater sulfate: implications for a snowball earth? Earth Planet. Sci. Lett. 203, 413– 430. Hyde, W. T., Crowley, T. J., Baum, S. K. & Peltier, W. R. 2000. Neoproterozoic ‘snowball Earth’ simulations with a coupled climate/icesheet model. Nature, 405, 425– 429. Iyer, S. S., Babinski, M., Krouse, H. R. & Chemale, F., Jr. 1995. Highly 13 C enriched carbonate and organic matter in the Neoproterozoic sediments of the Bambuı´ Group, Brazil. Precambrian Research, 73, 271– 282. Jacobsen, S. B. & Kaufman, A. J. 1999. The Sr, C and O isotope evolution of Neoproterozoic seawater. Chemical Geology, 161, 37– 57. Karfunkel, J. & Hoppe, A. 1988. Late Proterozoic glaciation in Central-Eastern Brazil: synthesis and model. Paleogeography, Paleoclimatology, Paleoecology, 65, 1– 21. Kaufman, A. J. & Knoll, A. H. 1995. Neoproterozic variations in the Cisotopic composition of seawater: stratigraphic and biogeochemical implications. Precambrian Research, 73, 27 – 49. Kaufman, A. J., Hayes, J. M., Knoll, A. H. & Germs, G. J. B. 1991. Isotopic compositions of carbonates and organic carbon from upper Proterozoic successions in Namibia: stratigraphic variation and the effect of diagenesis and metamorphism. Precambrian Research, 49, 301– 327. Kaufman, A. J., Knoll, A. H. & Narbonne, G. M. 1997. Isotopes, ice ages, and terminal Proterozoic earth history. Proceedings of the National Academy of Science USA, 94, 6600–6605. Kaufman, A. J., Varni, M. A., Misi, A. & Brito-Neves, B. B. 2001. Anomalous d34S signatures in trace sulfate from a potential cap

˜ O FRANCISCO CRATON NEOPROTEROZOIC SUCCESSIONS OF THE SA

carbonate in the Neoproterozoic Bambuı´ Group, Brazil. In: Misi, A. & Teixeira, J. B. G. (organizers) Proterozoic Sediment-Hosted Base Metal Deposits of Western Gondwana. I Field Workshop, Beolo Horizonte and Paracatu (Minas Gerais), Brazil, 62 –65. Kawashita, K. 1998. Rochas carbona´ticas neoproterozo´icas da Ame´rica do Sul: idades e infereˆncias quimioestratigra´ficas. Full professor thesis, University of Sa˜o Paulo, Brazil, 126. Kiang, C. H. 1997. Iso´topos esta´veis (C, H, O) e 87Sr/86Sr: implicac¸o˜es na estratigrafia e na paleo-circulac¸a˜o de fluidos na Bacia do Sa˜o Francisco. Full Professor Thesis, UNESP, Sa˜o Paulo, Brazil. Knoll, A. H., Hayes, J. M., Kaufman, A. J., Swett, K. & Lambert, L. B. 1986. Secular variation in carbon isotope ratios from upper Proterozoic successions of Svalbard and East Greenland. Nature, 321, 832– 838. Kendall, B., Creaser, R. A. & Selby, D. 2006. Re– Os geochronology of post-glacial black shales in Australia: constraints on the timing of ‘Sturtian’ glaciation. Geology, 34, 729–732. Kennedy, M. J., Christie-Blick, N. & Sohl, L. E. 2001. Are Proterozoic cap carbonates and isotopic excursions a record of gas hydrate destabilization following Earth’s coldest intervals? Geology, 29, 443– 446. Lima, O. N. B., Uhlein, A. & Britto, W. 2005. Geologia dos depo´sitos fosfa´ticos do Grupo Bambuı´ na Serra da Saudade, Cedro do Abaete´, Minas Gerais. In: Short Papers-III Simpo´sio sobre o Craton do Sa˜o Francisco, Salvador, SBG, 320–323. Madalosso, A. 1979. Stratigraphy and sedimentation of the Bambuı´ Group in Paracatu region, minas Gerais, Brazil. MA thesis, University of Missouri, USA, 127. Marini, O. J., Fuck, R. A., Danni, J. C., Dardenne, M. A., Loguercio, S. O. C. & Ramalho, R. 1984. As faixas de dobramentos Brası´lia, Uruac¸u e Paraguai-Araguaia e o Macic¸o Mediano de Goia´s. In: Schobbenhaus, C., Campos, D. A., Derze, G. R. & Asmus, H. E. (eds) Geologia do Brasil. Texto Explicativo do Mapa Geolo´gico do ´ rea Oceaˆnica Adjacente Incluindo Depo´sitos Minerais. Brasil e da A Ministe´rio da Minas e Energias/Departamento Nacional da Produc¸a˜o Mineral, Brası´lia, 251– 303. Misi, A. 1978. Ciclos de sedimentac¸a˜o e mineralizac¸o˜es de Pb –Zn nas sequeˆncias Bambui (Supergrupo Sa˜o Francisco), Estado de Bahia. XXX Congresso Brasileiro de Geologia. Sociedade Brasileira de Geologia, Anais, 4, 2548–2561. Misi, A. 1979. O Grupo Bambuı´ no Estado da Bahia. In: Inda, H. V. (ed.) Geologia e Recursos Minerais do Estato da Bahia. Textos Ba´sicos, SME/CPM, Salvador, 1, 120– 154. Misi, A. 1999. Um modelo de evoluc¸a˜o metalogene´tica para os depo´sitos de zinco e chumbo hospedados em sedimentos proterozo´icos de cobertura do Craton do Sa˜o Francisco. Full Professror Thesis, Federal University of Bahia (Brazil). Misi, A. 2001. Estratigrafia isoto´pica das sequeˆncias do Supergrupo Sa˜o Francisco, coberturas neoproterozo´icas do Cra´ton do Sa˜o Francisco. Idade e correlac¸o˜es. In: Pinto, C. P. & Martins-Neto, M. A. (eds) Bacia do Sa˜o Francisco, Geologia e Recursos Naturais, SBG/ MG, Belo Horizonte, 67– 92. Misi, A. & Kyle, J. R. 1994. Upper Proterozoic carbonate stratigraphy, diagenesis, and stromatolitic phosphorite formation, Ireceˆ Basin, Bahia, Brazil. Journal of Sedimentary Research, 64, 299– 310. Misi, A. & Souto, P. G. 1975. Controle estratigra´fico das mineralizac¸o˜es de Pb –Zn –F –Ba do Grupo Bambui, parte leste da Chapada de Ireceˆ (Bahia). Revista Brasileria de Geoscieˆncias, 5, 30 – 45. Misi, A. & Veizer, J. 1998. Neoproterozoic carbonate sequences of the Una Group, Ireceˆ Basin, Brasil: chemostratigraphy, age and correlations. Precambrian Research, 89, 87 –100. Misi, A., Veizer, J., Kawashita, K. & Dardenne, M. A. 1997. The age of the Neoproterozoic carbonate platform sedimentation based on 87 Sr/86Sr determinations, Bambuı´ and Una Groups, Brazil. I South American Symposium on Isotope Geology, Campos do Jorda˜o, Sa˜o Paulo, Brazil. Extented Abstracts, 199– 200. Misi, A., Iyer, S. S. S. et al. 1999. Geological and isotopic constraints on the metallogenic evolution of the Proterozoic sediment-hosted Pb – Zn (Ag) deposits of Brazil. Gondwana Research, 2, 47 – 65. Misi, A., Iyer, S. S. S. et al. 2005. Sediment-hosted lead– zinc deposits of the Neoproterozoic Bambui Group and correlative sequences, Sa˜o

521

Francisco Craton, Brazil: a review and a possible metallogenic evolution model. Ore Geology Reviews, 3, 263–304. Misi, A., Kaufman, A. J. et al. 2007. Chemostratigraphic correlation of Neoproterozoic successions in South America. Chemical Geology, 237, 161–185. Montes, A. S. L. 1997. O contexto estratigra´fico e sedimentolo´gico da Formac¸a˜o Bebedouro na Bahia. Um possı´vel portador de diamantes. MSc thesis, University of Brası´lia, Brazil. Monteiro, L. V. S. 2002. Modelamento metalogene´tico dos depo´sitos de Vazante, Fagundes e Ambro´sia associados ao Grupo Vazante, Minas Gerais. Tese Doutorado, Universidade de Sa˜o PauloUSP, 317. Olcott, A. N., Sessions, A. L., Corsetti, F. A., Kaufman, A. J. & de Oliviera, T. F. 2005. Direct evidence for significant primary production during widespread Neoproterozoic glaciation. Science, 310, 471–474. Pedrosa-Soares, A. C., Dardenne, M. A., Hasui, Y., Castro, F. D. C., Carvalho, M. V. A. & Reis, A. C. 1994. Mapa Golo´gico do Estdo de Minas gerais e Nota Explicativa. Secretaria de Recursos Minerais, Hı´dricos e Energe´ticos, Companhia Mineradora de Minas Gerais (COMIG), mapas e texto, 97. Pimentel, M. M., Dardenne, M. A., Viana, M. G., Costa, S. M., Gioia, L., Junges, S. & Seer, H. J. 2001. Nd isotopes and the provenance of sediments from the Neoproterozoic Brası´lia Belt, Central Brazil: geodynamic implications. Journal of South American Earth Sciences, 14, 571– 585. Porada, H. 1989. Pan-African rifting and orogenesis in southern to equatorial Africa and Eastern Brazil. Precambrian Research, 44, 103– 136. Powis, K. 2006. Stable isotope geochemistry of the Neoproterozoic Bambuı´ Group at Serra do Ramalho, Bahia, Brazil. PhD thesis, Ottawa-Carleton Geoscience Center, University of Ottawa, Canada, 197. Rodrigues, J. B. 2007. Estudos de provenieˆncia de sedimentos da regia˜o Centro-Oeste da Faixa Brası´lia-uma abordagem geocronolo´gica. Exame de Qualificac¸a˜o Doutorado, Universidade de Brası´lia-UnB, Brası´lia, 50. Rodrigues, J. B. 2008. Provenieˆncia de sedimentos dos grupos Canastra, lbia´, Vazante e Bambuı´, Um estudo de zirco˜es detrı´ticos e ldades Modelo Sm –Nd. Dr. Thesis, University of Brasilia, Brasilia DF. Rodrigues, J. B., Pimentel, M. M., Buhn, B., Dardenne, M. A. & Alvarenga, C. J. S. 2008. Provenance of the Vazante Group– Preliminary data. VI South American Symposium on Isotope Geology, Bariloche, Argentina. Extented Abstracts, CD-ROM. Sanches, A. L., Misi, A., Kaufman, A. J. & Azmy, K. 2007. As sucesso˜es carbona´ticas neoproterozo´icas do Craton do Sa˜o Francisco e os depo´sitos de fosfato: correlac¸o˜es e fosfogeˆnese, Revista Brasileira de Geocieˆncias, 37, 1034–1046. Santos, R. V., de Alvarenga, C. J. S., Dardenne, M. A., Sial, A. N. & Ferreira, V. F. 2000. Carbon and oxygen isotopes across Meso-Neoproterozoic limestones from Central Brazil: Bambuı´ and Paranoa´ groups. Precambrian Research, 104, 107–122. Santos, R. V., Alvarenga, C. J. S. et al. 2004. Carbon isotopes of Mesoproterozoic-Neoproterozoic sequences from Southern Sa˜o Francisco Craton and Arac¸uaı´ Belt, Brazil: palaeogeographic implications. Journal of South American Earth Sciences, 18, 27 –39. Saylor, B. Z., Kaufman, A. J., Grotzinger, J. P. & Urban, F. 1998. A composite reference section for terminal Proterozoic strata of southern Namibia. Journal of Sedimentary Research, 68, 1223– 1235. Sial, A. N., Ferreira, V. P., Almeida, A. R., Romano, A. W., Parente, C., Da Costa, M. L. & Santos, V. H. 2000. Carbon isotope fluctuations in Precambrian carbonate sequences of several localities in Brazil. Anais Academia Brasileira de Cieˆncias, 72, 540– 557. Sial, A. N., Ferreira, V. P., Moura, C. V. A. & Santos, V. H. 2003. C-, O- and Sr-isotope stratigraphy of the Sturtian Jacoca and Olho D’Agua Formations, state of Sergipe, Northeastern Brazil. Short papers. IV South American Symposium on Isotope Geology. Salvador, Bahia, 394–397. Sial, A. N., Ferreira, V. P. et al. 2005. Two Neoproterozoic cap carbonates in the states of Sergipe and Bahia, Northeastern Brazil: C- and Sr-isotopes and mercury as paleoclimatic tracer. X Congresso

522

A. MISI ET AL.

Brasileiro de Geoquı´mica e II Simpo´sio de Geoquı´mica de Paı´ses do Mercosul, Porto de Galinhas. Short paper CD-ROM. Sial, A. N., Ferreira, V. P. et al. 2006a. Chemostratigraphy of two Neoproterozoic cap carbonates from the Sergipano belt (northeastern Brazil). Short Papers, V South American Symposium on Isotope Geology (V SSAGI). Punta del Este, Uruguay, 314–317. Sial, A. N., Ferreira, V. P. et al. 2006b. C- and Sr-isotopes and mercury as paleoclimatic tracer in two Neoproterozoic cap carbonates in northeastern Brazil. Snowball Earth Conference. Ascona, Suic¸a, 100– 101. Sial, A. N., Dardenne, M. A. et al. 2009. The Sa˜o Francisco Paleocontinent. In: Neoproterozoic-Cambrian Tectonics, Global Change and Evolution: A Focus on Southwestern Gondwana. Developments in Precambrian Geology, 16, 31– 69. So¨fner, B. 1973. Observac¸o˜es sobre a estratigrafia do Pre´-Cambriano da Chapada Diamantina Sudeste e da a´rea contı´gua. In: SBG, 27th Congresso Brasileiro de Geologia. Aracaju´, Anais, 1, 23 –33. Souza, J. C. F. 1997. Litoestratigrafia e sedimentologia da Formac¸a˜o Vazante na regia˜o de Coromandel. MSc thesis, Universidade de Brası´lia. Teixeira, L. B., Martins, M. & Braun, O. P. G. 1993. Evoluc¸a˜o geolo´gica da Bacia Sa˜o Francisco com base em sı´smica de reflexa˜o e me´todos potenciais. II Simpo´sio sobre o Craton do Sa˜o Francisco. SBG, Nu´cleo Bahia, Anais, 179–181. Teixeira, W. & Figueiredo, M. C. H. 1991. An outline of Early Proterozoic crustal evolution in the Sa˜o Francisco Craton, Brazil: a review. Precambrian Research, 53, 1 –22. Thomaz-Filho, A., Kawashita, K. & Cordani, U. G. 1998. A origem do Grupo Bambuı´ no contexto da evoluc¸a˜o geotectoˆnica e de idades radiome´tricas. Anais Academia Brasileira de Cieˆncias, 70, 527– 548. Torquato, J. R. F. & Misi, A. 1977. Medidas isoto´picas de carbono e oxigeˆnio em carbonatos do Grupo Bambui na regia˜o centronorte do Estado da Bahia. Revista Brasileira de Geocieˆncias, 7, 14 –24.

Trindade, R. I. F., D’Agrella-Filho, M., Babinski, M. & Brito Neves, B. B. 2004. Paleomagnetism and geochronology of the Bebedouro cap carbonate: evidence for continental-scale Cambrian remagnetization in the Sa˜o Francisco craton, Brazil. Precambrian Research, 128, 83 – 103. Trompette, R. 1994. Geology of Western Gondwana (2000– 500 Ma). A.A. Balkema, Rotterdam, 350. Uhlein, A., Alvarenga, C. J. S., Trompette, R., Dupont, H. S. J. B., Egydio-Silva, M., Cukrov, N. & Lima, O. N. B. 2004. Glaciac¸a˜o neoproterozo´ica sobre o Craton do Sa˜o Francisco e faixas dobradas adjacentes. In: Mantesso-Neto, V., Bartorelli, A., Carneiro, D. R. C. & Brito Neves, B. B. (eds) Geologia do Continente Sul-Americano: Evoluc¸a˜o da obra de Fernando Fla´vio de Almeida. Beca Ed., Sa˜o Paulo, 539– 553. Uhlein, A., de Alvarenga, C. J. S., Dardenne, M. A. & Trompette, R. R. 2011. The glaciogenic Jequitaı´ Formation, southeastern Brazil. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 541–546. Valeriano, C. M., Dardenne, M. A., Fonseca, M. A., Simo˜es, L. S. A. & Seer, H. J. 2004. A Evoluc¸a˜o Tectoˆnica da Faixa Brası´lia. In: Mantesso-Neto, V., Bartorelli, A., Carneiro, D. R. C. & Brito Neves, B. B. (eds) Geologia do Continente Sul-Americano: Evoluc¸a˜o da obra de Fernando Fla´vio de Almeida. Beca Edta, Sa˜o Paulo, 575–593. Viera, L. C., Almeida, R. P., Trindade, R. I. F., Nogueira, A. C. R. & Janikian, L. 2007. A Formac¸a˜o Sete Lagoas em sua a´rea tipo: fa´cies, estratigrafia e sistemas deposicionais. Revista Brasileira de Geoscieˆncias, 37, 1020– 1033. Young, G. M. 1995. Are Neoproterozoic glacial deposits preserved on the margins of Laurentia related to the fragmentation of two supercontinents? Geology, 23, 153– 156. Zhang, S., Jiang, G., Zhang, J., Song, B., Kennedy, M. J. & ChristieBlick, N. 2005. U– Pb sensitive high-resolution ion microprobe ages from the Doushantuo Formation in south China: constraints on late Neoproterozoic glaciations: Geology, 33, 473–476.

Chapter 49 The Neoproterozoic Macau´bas Group, Arac¸uaı´ orogen, SE Brazil A. C. PEDROSA-SOARES1*, MARLY BABINSKI2, CARLOS NOCE1, MAXIMILIANO MARTINS1, ´ UCIA QUEIROGA1 & FRANCISCO VILELA1 GLA 1

Universidade Federal de Minas Gerais, Instituto de Geocieˆncias-CPMTC, Campus Pampulha, 31270-901 Belo Horizonte, MG, Brazil

2

Universidade de Sa˜o Paulo, Instituto de Geocieˆncias, Rua do Lago 562, Cidade Universita´ria 05508-080 Sa˜o Paulo, SP, Brazil *Corresponding author (e-mail: [email protected]) Abstract: The Neoproterozoic Macau´bas Group records accumulation in the precursor basin of the Arac¸uaı´ orogen, located on the eastern margin of the Sa˜o Francisco craton (SE Brazil). The Macau´bas basin evolved from a late Tonian continental rift to a passive margin that lasted at least until c. 660 Ma. It was orogenically inverted during the late Neoproterozoic Brasiliano event. The Macau´bas Group includes the pre-glacial Mata˜o, Duas Barras and Rio Peixe Bravo formations, the glaciogenic Serra do Catuni, Nova Aurora and Lower Chapada Acaua˜ formations, and the post-glacial Upper Chapada Acaua˜ and Ribeira˜o da Folha formations. In the central sector of the Arac¸uaı´ orogen, the oldest glaciogenic unit of the group, the Serra do Catuni Formation, overlies the sandstone-conglomerate package of the Duas Barras Formation. The Serra do Catuni Formation consists of massive diamictite with minor sandstone and rare pelite, deposited mostly in a proximal glaciomarine environment. This unit passes upward and eastward into the Lower Chapada Acaua˜ Formation, a thick succession of stratified diamictite, graded sandstone, pelite, transitional basalt and rare carbonate. This distal glaciomarine unit is covered by the diamictite-free Upper Chapada Acaua˜ Formation, which passes eastward into the Ribeira˜o da Folha Formation, which includes fine-grained turbidite, pelite and ocean-floor rocks. In the northern sector of the Arac¸uaı´ orogen, the sandstone-pelite succession of the pre-glacial Rio Peixe Bravo Formation is covered by the Nova Aurora Formation, the glaciomarine unit rich in diamictite and Ferich diamictite, with minor graded sandstone and rare pelite. The Nova Aurora Formation is covered by the sandstone-pelite package of the Upper Chapada Acaua˜ Formation. The pre-glacial and glaciogenic successions record the continental rift to transitional stages of the Macau´bas basin. The post-glacial succession represents proximal and distal passive margin to ocean floor environments. The Serra do Catuni Formation seems to be a proximal glaciomarine equivalent of the Jequitaı´ glacio-terrestrial deposits located on the Sa˜o Francisco craton.

The Macau´bas Group represents the precursor basin of the Neoproterozoic Arac¸uaı´ orogen, located in southeastern Brazil (Fig. 49.1a). This orogen extends from the eastern border of the Sa˜o Francisco craton to the Atlantic Ocean (Pedrosa-Soares et al. 2001, 2008). Since the beginning of the 1930s the name Macau´bas has appeared in the international geological literature to refer to a sedimentary unit that includes Precambrian glaciogenic rocks (Moraes & Guimara˜es 1931). The first field studies were published by Moraes (1929), who described the succession of ‘conglomeratic phyllites, phyllites, quartzites and mica schists cropping out in the Macau´bas River valley and Catuni Ridge’, located to the north of the city of Diamantina (Fig. 49.2). Moraes & Guimara˜es (1930) named that succession the Macau´bas Fm. and interpreted the ‘conglomeratic phyllites’ as metamorphosed glaciogenic rocks. In 1932, Moraes noted the very extensive distribution of the Macau´bas Fm. after identifying a thick pile of ‘highly metamorphosed conglomeratic phyllonites’ associated with quartzite, schist and rare carbonate layers along the western valleys of the Jequitinhonha and Rio Pardo rivers, and many of their tributaries (Fig. 49.2). Moraes (1932) was also the first to suggest a correlation between the Macau´bas Fm. and the glaciogenic Jequitaı´ Fm. located on the Sa˜o Francisco craton (Fig. 49.1a). Since Scho¨ll (1972) and Pflug & Scho¨ll (1975), the designation Macau´bas Group (instead of formation) has been used. Almeida (1977) considered the Macau´bas Group as the main fill of the ‘Alpine-type geosyncline’ that became the Arac¸uaı´ fold belt after the Brasiliano orogenic event in late Neoproterozoic time. Based on its first definition (Moraes 1929, 1932; Moraes & Guimara˜es 1930, 1931), the name Macau´bas Group is here exclusively applied to refer to the extensive unit with metadiamictite-bearing formations that underwent regional deformation and metamorphism within the Arac¸uaı´ orogen (Figs 49.2 & 49.3). This group

was thrust over the pelite-carbonate cratonic cover (Bambuı´ Group) and does not show any direct field relation with the Jequitaı´ and Carrancas diamictites located on the southern Sa˜o Francisco craton (Figs 49.1a & 49.2), although their correlation is often assumed (Uhlein et al. 2011). Several names such as conglomeratic phyllite, conglomeratic phyllonite, paraconglomerate, conglomeratic greywacke, mixtite and tillite were applied to the Macau´bas metadiamictites until the 1970s, reflecting disagreement (and misunderstandings) over the glaciogenic interpretation, evolution of geological knowledge, and standardization of nomenclature. Indeed, the diamictitebearing units preserved on the Sa˜o Francisco craton record the best evidence of glaciation, despite the relatively small area over which they occur. The Jequitaı´ Fm. includes glacio-terrestrial sediments and structures preserved from orogenic deformation (e.g. Isotta et al. 1969; Viveiros & Walde 1976; Walde 1976; RochaCampos & Hasui 1981; Gravenor & Monteiro 1983; Karfunkel & Hoppe 1988; Martins-Neto et al. 1999; Uhlein et al. 1999, 2011; Karfunkel et al. 2002; Martins-Neto & Hercos 2002), and has a maximum sedimentation age around 880 Ma (U – Pb LA-ICPMS, detrital zircon; Rodrigues 2008). The Carrancas diamictite occurs in a few outcrops that lack clear evidence of glacial origin, but it is overlain by the c. 740 Ma carbonate typical of post-glacial rocks within the cratonic cover (Scho¨ll 1972; Babinski et al. 2007). On the other hand, most of the Macau´bas metadiamictites represent sub-aqueous deposits that underwent strong deformation and metamorphism, hampering the recognition of solid evidence of glaciation. Nevertheless, the areas of occurrence, regional stratigraphy, Neoproterozoic age and glaciogenic nature of the Macau´bas metadiamictites seem to be definitively demonstrated by the data and correlations presented by Hettich (1975, 1977), Karfunkel & Karfunkel (1976, 1977), Hettich & Karfunkel (1978), Viveiros

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 523– 534. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.49

524

A. C. PEDROSA-SOARES ET AL.

Structural framework

Fig. 49.1. (a) Location of the Arac¸uaı´ orogen in relation to the Sa˜o Francisco craton, showing the distribution of the Macau´bas Group and Neoproterozoic cratonic covers. (b) The Arac¸uaı´ –West Congo orogenic system in relation to the Sa˜o Francisco-Congo craton.

et al. (1979), Karfunkel et al. (1984, 1985), Karfunkel & Hoppe (1988), Moura˜o & Pedrosa-Soares (1992), Pedrosa-Soares et al. (1992, 1998, 2000, 2008), Grossi-Sad et al. (1997a), Noce et al. (1997), Uhlein et al. (1998, 1999, 2007), Babinski et al. (2005, 2011), Gradim et al. (2005) and Martins (2006). In this chapter, we present an updated synthesis on the Macau´bas Group with special emphasis on the glaciogenic units that are considered to represent at least one Cryogenian glacial event.

The correlation between the Arac¸uaı´ orogen and its counterpart located in southwestern Africa, the West Congo belt (Fig. 49.1b), is well-established in the geological literature (e.g. Porada 1989; Pedrosa-Soares et al. 1992, 2008; Trompette 1994; Cordani et al. 2003; Alkmim et al. 2006). The Arac¸uaı´ – West Congo orogenic system, as well as its precursor basin, developed inside an embayment of the Sa˜o Francisco –Congo craton (Fig. 49.1b), characterizing a Neoproterozoic confined orogen (Pedrosa-Soares et al. 2001, 2008; Rogers & Santosh 2004). That precursor basin evolved from a Tonian continental rift to an inland-sea basin (a large gulf) that was partially floored by Cryogenian oceanic crust along its southern sector, and was inverted during the late Neoproterozoic Brasiliano –Pan African orogenesis, resulting in the Arac¸uaı´ and West Congo belts (Pedrosa-Soares et al. 2001, 2007, 2008; Tack et al. 2001; Alkmim et al. 2006, 2007). This system was split up in two quite different but complementary counterparts, after the Atlantic Ocean opening in Cretaceous time (Pedrosa-Soares et al. 2008). The West Congo belt inherited the thick package of bimodal volcanic rocks of the early Tonian continental rift, as well as rift and passive margin successions from Tonian to late Cryogenian age, but no Neoproterozoic ophiolite sliver (Tack et al. 2001; Frimmel et al. 2006; Pedrosa-Soares et al. 2008). The Arac¸uaı´ orogen inherited the basin sector, lacking in rift-related magmatic rocks, but including the remnants of Neoproterozoic oceanic lithosphere and the extensive Macau´bas Group that includes rift, transitional and passive margin sediments from late Tonian to late Cryogenian age (Pedrosa-Soares et al. 2007, 2008). Neoproterozoic diamictites occur in both counterparts of the Arac¸uaı´ –West Congo orogenic system; those located in the West Congo belt are reviewed in Tait et al. (2011). The Macau´bas Group is the most complete record of the precursor basin within the Arac¸uaı´ orogen because it shows evidence of the rift and passive margin stages. Normal faults and clastic dykes associated with the sedimentation of the lower Macau´bas Group, as well as late Tonian mafic dykes and anorogenic granites, the main evidence of the rift-related extensional tectonics (Karfunkel et al. 1976, 1977, 1984; Machado et al. 1989; Guimara˜es et al. 1993; Grossi-Sad et al. 1997a; Uhlein et al. 1988, 1999; MartinsNeto et al. 2001; Martins-Neto & Hercos 2002; Knauer et al. 2006; Martins 2006; Martins et al. 2008; Silva et al. 2008). Transitional mafic volcanism occurred in the late continental rift stage (Gradim et al. 2005). Transgressive sedimentary successions free of diamictite and ophiolite slivers record the passive margin stage and associated oceanic spreading that lasted at least until c. 660 Ma (Pedrosa-Soares et al. 1998, 2001, 2008; Queiroga et al. 2007). The Macau´bas Group was deformed, metamorphosed and intruded by orogenic granites during the Brasiliano orogeny that lasted from c. 630 Ma (onset of the pre-collisional stage) to c. 490 Ma (end of the post-collisional stage) in the Arac¸uaı´ orogen (Pedrosa-Soares et al. 2001, 2008; Silva et al. 2005). The Macau´bas Group occurs in the west-verging fold-and-thrust belt that characterizes the external (western) tectonic domain of the Arac¸uaı´ orogen (Figs 49.1 & 49.2). Along that belt, the major structures of the first deformational phase are thrust faults and ductile shear zones, some of them representing the tectonic inversion of normal faults related to the rift stage of the Macau´bas basin, and west-verging tight asymmetric folds. These structures are associated with an east-dipping regional foliation and a stretching lineation (very well marked by stretched clasts in metadiamictites). The kinematic indicators, such as fold vergence, S-C foliation and asymmetric sigmoidal features (e.g. stretched sigma clasts in diamictites) of this first deformational phase, record tectonic transport towards the Sa˜o Francisco craton. The regional metamorphism related to the first deformational phase increases from west to east so that the Macau´bas Group records the chlorite, biotite, garnet, staurolite, kyanite and sillimanite zones of an intermediate

´ BAS GROUP NEOPROTEROZOIC MACAU

525

Fig. 49.2. Simplified geological map of the Arac¸uaı´ orogen (modified from Pedrosa-Soares et al. 2007, 2008). Section A– B shows structural and metamorphic features of the Macau´bas Group.

526

A. C. PEDROSA-SOARES ET AL.

Fig. 49.3. Sketched map showing the distribution of the different formations of the Macau´bas Group in the external (western) tectonic domain of the Arac¸uaı´ orogen, along with the regional stratigraphic scheme (modified from Pedrosa-Soares et al. 2007).

pressure metamorphic regime (Dossin & Dardenne 1984a; Pedrosa-Soares et al. 1984, 1992, 1993, 1996, 2001, 2008; Alkmim et al. 2006; Uhlein & Chaves 1989; Grossi-Sad et al. 1997a; Uhlein et al. 1998; Marshak et al. 2006). A west-dipping crenulation cleavage associated with asymmetric folds and normal faults characterizes the second deformational phase (Pedrosa-Soares et al. 1993; Moura˜o & Grossi-Sad 1997; Oliveira et al. 1997a; Pedrosa-Soares & Grossi-Sad 1997;

Tupinamba´ & Grossi-Sad 1997; Alkmim et al. 2006). These structures have been associated with the gravitational collapse of the Arac¸uaı´ orogen (Marshak et al. 2006). Two NE- and NW-trending, high-angle-dip regional fracture systems associated with open megafolds (flexures) characterize the third deformational phase (Pedrosa-Soares et al. 1992; Grossi-Sad et al. 1997a), which may also be related to the gravitational collapse of the orogen.

´ BAS GROUP NEOPROTEROZOIC MACAU

Fig. 49.4. General stratigraphic scheme for the central sector (17–188S) of the Macau´bas Group (not to scale; estimated thicknesses are referred to in the text and Fig. 49.5; modified from Pedrosa-Soares et al. 2008).

In fact, all primary rocks of the Macau´bas Group were modified by regional deformation and metamorphism of variable intensity. However, the general low metamorphic grade (greenschist to intermediate amphibolite facies) and the preservation of sedimentary and volcanic structures in low strain zones allow the interpretation of the pre-metamorphic lithology and other features, such that the protoliths can be confidently described.

Stratigraphy The regional stratigraphic scheme adopted here resulted from the Espinhac¸o Project (a 1:100 000 scale geological mapping programme that largely covered the Macau´bas Group; Grossi-Sad et al. 1997a; Noce et al. 1997), with modifications inserted by Lima et al. (2002), Pedrosa-Soares et al. (2007, 2008) and Martins et al. (2008). The Macau´bas Group is subdivided into seven formations, from oldest to youngest Mata˜o, Duas Barras, Rio Peixe Bravo, Serra do Catuni, Nova Aurora, Chapada Acaua˜ and Ribeira˜o da Folha (Fig. 49.3). Two uplifted basement blocks divide the distribution of the Macau´bas Group into three sectors with variably complete stratigraphic records: the northern sector, related to the Porteirinha block; the central sector, located between 178 and 188S; and the southern sector, related to the Guanha˜es block (Fig. 49.3). The most complete package of the Macau´bas Group occurs in the central sector where the lowermost units are the diamictite-free Mata˜o and Duas Barras formations that are covered by the most proximal and oldest diamictite-bearing unit of the group, the Serra do Catuni Fm. (Figs 49.3 & 49.4). This formation passes upward and eastward into the diamictite-bearing Lower Chapada Acaua˜ Fm., which, in turn, is covered by the diamictite-free Upper Chapada Acaua˜ Fm. This unit passes eastwards into the diamictite-free Ribeira˜o da Folha Fm. (Fig. 49.4). In the northern sector, only the Serra do Catuni Fm. occurs between the Sa˜o Francisco craton boundary and the western edge of the Porteirinha block (Fig. 49.3). To the east of the Porteirinha block the diamictite-free Rio Peixe Bravo Fm. is overlain by the diamictite-bearing Nova Aurora Fm., which is partially capped by the Upper Chapada Acaua˜ Fm. In the southern sector, the stratigraphic record of the Macau´bas Group is very incomplete. The Serra do Catuni Formation, locally overlain by strata that probably correlate to the Chapada Acaua˜ Formation, occurs along a narrow zone in the western thrust front of the Arac¸uaı´ orogen (Figs 49.2 & 49.3). To the east

527

of the Guanha˜es block, only the Ribeira˜o da Folha Fm. occurs (Figs 49.3 & 49.4). The Mata˜o, Duas Barras and Rio Peixe Bravo formations represent the oldest deposits of the Macau´bas Group (Figs 49.3 & 49.4). The Mata˜o Fm. consists of breccia and conglomerate rich in sandstone pebbles and cobbles, covered by sandstone with conglomerate lenses. Its maximum thickness is 200 m. An erosive unconformity and normal faults define the contact between the Mata˜o Fm. and the basement (locally represented by eolian sandstone of the Statherian Espinhac¸o Supergroup). The Mata˜o Fm. records sedimentation under unstable tectonic conditions related to the early rift stage of the Macau´bas basin (Martins 2006; Martins et al. 2008). The Duas Barras Fm. consists of sandstone and conglomeratic sandstone with variable contents of mica, feldspar, iron oxide and/or lithic fragments, quartz sandstone, conglomerate and rare pelite. It shows fluvial and shallow marine sedimentary structures and bimodal (NW –SE and SE –NW) palaeocurrent sets. The maximum thickness is approximately 100 m (Noce 1997a; Grossi-Sad et al. 1997b; Martins 2006). The Rio Peixe Bravo Fm. consists of micaceous, ferruginous and/or feldspathic sandstone, pelite locally rich in hematite and/ or graphite, and rare conglomerate, with a maximum thickness of around 700 m (Viveiros et al. 1979; Moura˜o & Grossi-Sad 1997; Oliveira et al. 1997a; Roque et al. 1997; Knauer et al. 2006). The Duas Barras and Rio Peixe Bravo formations, which lack evidence of glacial influence, represent fluvial to marine sedimentation during the continental rift stage of the Macau´bas basin (Noce et al. 1997; Uhlein et al. 1998, 2007; Martins-Neto et al. 2001; Martins 2006; Pedrosa-Soares et al. 2007, 2008). The Serra do Catuni Fm., the oldest diamictitic unit of the Macau´bas Group, is a package of massive diamictite with minor sandstone and rare pelitic rhythmite lenses (Figs 49.3 & 49.4). This formation is locally rich in chaotic boulders, and contains faceted and striated flat-iron-shaped cobbles and pebbles (Karfunkel & Karfunkel 1976, 1977; Karfunkel et al. 1984; Karfunkel & Hoppe 1988; Pedrosa-Soares et al. 1992, 2008; Noce et al. 1997; Uhlein et al. 1999, 2007; Martins-Neto et al. 2001; Martins-Neto & Hercos 2002; Martins 2006). The Nova Aurora Fm. mainly comprises stratified diamictite, ferruginous diamictite-bearing and diamictitic Fe formation, with minor intercalations of sandstone with fining-up graded bedding and rare pelite (Viveiros et al. 1979; Pedrosa-Soares et al. 1992; Moura˜o & Grossi-Sad 1997; Noce et al. 1997; Oliveira et al. 1997a; Pedrosa-Soares & Oliveira 1997; Roque et al. 1997; Uhlein et al. 1998, 1999; Drumond 2000). The Chapada Acaua˜ Fm. includes a lower diamictitic unit and an upper diamictite-free unit (Figs 49.3 & 49.4). The Lower Chapada Acaua˜ Fm. essentially consists of stratified diamictite, graded sandstone, pelite and mafic volcanic rocks (Hettich 1977; Moura˜o & Pedrosa-Soares 1992; Pedrosa-Soares et al. 1992, 2007, 2008; Moura˜o & Grossi-Sad 1997; Oliveira et al. 1997a, b; Tupinamba´ & Grossi-Sad 1997; Uhlein et al. 1999; Gradim et al. 2005; Martins 2006). At least one thick carbonate lens (c. 20 m) occurs intercalated in a pelite layer with isolated outsized clasts at the top of the Lower Chapada Acaua˜ Fm. (Pedrosa-Soares & GrossiSad 1997). The mafic volcanic rocks, metamorphosed to greenschists, show pillow structures and other features of sub-aqueous flows (Schrank et al. 1978; Chula et al. 1996; Uhlein et al. 1998; Gradim et al. 2005). They have tholeiitic basalt protoliths with a dominant within-plate signature (Gradim et al. 2005), Sm –Nd TDM model age of c. 1.5 Ga, and inherited zircon grains with detrital features and U – Pb SHRIMP ages from Archaean to Late Mesoproterozoic (Babinski et al. 2005). However, samples with oceanic signature and slightly positive 1Nd(900 Ma) of þ0.23, together with the inherited zircon grains, suggest a transitional mafic magma that migrated through a thinned continental crust (Babinski et al. 2005; Gradim et al. 2005). Accordingly, these greenschists provide strong evidence of volcanism in an extensional marine basin floored by thin continental crust, during the

528

A. C. PEDROSA-SOARES ET AL.

transitional phase from the late rift to early passive margin stages of the Macau´bas basin. Both the Nova Aurora and Lower Chapada Acaua˜ formations show cyclic deposition from coarse- to fine-grained sediments, fining-up graded bedding, erosional contacts between cycles and load structures (Moura˜o & Pedrosa-Soares 1992; Pedrosa-Soares et al. 1992; Noce et al. 1997; Pedrosa-Soares & Oliveira 1997; Uhlein et al. 1999, 2007; Martins-Neto et al. 2001). The features of these formations, as well as of the Serra do Catuni Fm., are described and interpreted in subsequent sections. The Upper Chapada Acaua˜ Fm. is a succession of interbedded sandstone and pelite, free of diamictite, interpreted as a postglacial unit deposited in a shelf environment during the passive margin stage of the Macau´bas basin (Pedrosa-Soares et al. 1992; Moura˜o & Grossi-Sad 1997; Noce et al. 1997; Oliveira et al. 1997a; Pedrosa-Soares & Grossi-Sad 1997; Martins-Neto et al. 2001). The Ribeira˜o da Folha Fm. is free of diamictite and includes distal passive margin and ocean floor deposits (Figs 49.2, 49.3 & 49.4). The proximal part of the Ribeira˜o da Folha Fm. is a succession of fine-grained turbidites and very rare limestone lenses. It overlies the diamictite-turbidite unit of the Lower Chapada Acaua˜ Fm. and is interpreted as a deep-water correlative of the shelf sandstone-pelite pile of the Upper Chapada Acaua˜ Fm. (Pedrosa-Soares et al. 1992, 2007, 2008; Pedrosa-Soares & Grossi-Sad 1997). The distal part of the Ribeira˜o da Folha Formation is locally associated with thrust slices of oceanic metamafic and meta-ultramafic rocks, constituting a tectonically dismembered ophiolite complex (Pedrosa-Soares et al. 1992, 1998, 2001, 2008; Queiroga et al. 2007). The distal Ribeira˜o da Folha Fm. is rich in peraluminous and carbonaceous pelites with intercalations of volcanic-exhalative deposits (sulphidebearing cherts, massive sulphide bodies, and banded iron formations of oxide, silicate and sulphide types) and mafic volcanic rocks with ocean-floor geochemical and isotopic (1Nd(660 Ma) of c. þ4) signatures (Pedrosa-Soares et al. 1998, 2001, 2008; Queiroga et al. 2006, 2007).

Glaciogenic deposits and associated strata

Fig. 49.5. Composite stratigraphic columns for the Serra do Catuni, Nova Aurora and Lower Chapada Acaua˜ formations (Macau´bas Group). The Serra do Catuni column is based on sections described by Martins (2006) along the Macau´bas River valley (Fig. 49.2). The Nova Aurora column synthesizes field and drill holes data from the region limited by a white rectangle in Figure 49.3. The column for the Lower Chapada Acaua˜ Fm. is based on field data from the eastern escarpments of the Acaua˜ Plateau (Fig. 49.3) and medium valley of the Arac¸uaı´ River (Fig. 49.2), as well as on data from the Macau´bas River valley.

The Macau´bas Group includes three diamictite-bearing units that have been interpreted as glaciogenic deposits: the Serra do Catuni, Nova Aurora and Lower Chapada Acaua˜ formations (Figs 49.3, 49.4 & 49.5). The rock-assemblage included in the Serra do Catuni Fm. was mapped and/or studied in detail by Karfunkel & Karfunkel (1976, 1977), Dossin & Dardenne (1984a), Karfunkel et al. (1984), Pedrosa-Soares et al. (1984, 1992, 2007, 2008), Uhlein & Chaves (1989), Fogac¸a (1997), Grossi-Sad et al. (1997b), Guimara˜es (1997), Guimara˜es et al. (1997), Knauer & Grossi-Sad (1997), Moura˜o et al. (1997), Noce (1997a, b), Oliveira et al. (1997b), Tupinamba´ & Grossi-Sad (1997), Uhlein et al. (1998, 1999, 2007), Knauer et al. (2006) and Martins (2006). The Serra do Catuni Fm. is a very extensive and rather homogeneous unit that persistently occurs for over than 400 km along the western border of the Arac¸uaı´ orogen (Fig. 49.3). This formation is rich in massive diamictite, the thickness of which ranges from a few tens of metres up to 300 m (Fig. 49.5). The following description is based on sections in the Macau´bas River valley (Fig. 49.2), where the Serra do Catuni Fm. is reasonably well preserved from deformation and erosion. The Serra do Catuni Fm. overlies the Duas Barras Formation and contains a 250-m-thick pile of massive diamictite with lenses of massive sandstone (Martins 2006). The diamictite has an abundant (60 –80% in volume), poorly sorted, muddysandstone matrix composed of detrital quartz, K-feldspar and carbonate, with a metamorphic foliation marked mainly by finegrained muscovite. The clasts, ranging in size from granules to boulders, are also poorly sorted in texture and composition

´ BAS GROUP NEOPROTEROZOIC MACAU

(milky quartz, sandstone, granitoid, carbonate and mafic rock). Faceted and striated flat-iron-shaped cobbles and pebbles can also be found in the Serra do Catuni diamictite. Lenses of massive sandstone up to 2 m thick appear mainly in the upper part of the section. This sandstone is poorly sorted and consists of quartz with minor K-feldspar, carbonate and iron oxides (Martins 2006). The rock-assemblage included in the Nova Aurora Fm. was mapped and/or studied in detail by Viveiros et al. (1979), Pedrosa-Soares et al. (1984, 1992, 2000, 2007, 2008), Moura˜o & Grossi-Sad (1997), Oliveira et al. (1997a), Pedrosa-Soares & Oliveira (1997), Roque et al. (1997), Uhlein et al. (1998, 1999), Drumond (2000) and Knauer et al. (2006). The Nova Aurora Fm. is well exposed in road cuts to the west of Salinas town (Fig. 49.3), and part of its stratigraphic pile is also known in drill holes for iron ore prospecting (Viveiros et al. 1979; Uhlein et al. 1999; Vilela 2010). Viveiros et al. (1979) estimated the minimum thickness for the Nova Aurora Fm. to be c. 1200 m based on detailed sections and data from drill holes. Again, the strong regional deformation hampers any accurate thickness evaluation for the whole unit, which has independently been estimated to be thicker than 2 km (e.g. Grossi-Sad et al. 1997a; Uhlein et al. 1998, 1999). The following description is based mainly on detailed studies carried out by Viveiros et al. (1979), Pedrosa-Soares et al. (1984, 1992), Pedrosa-Soares & Oliveira (1997) and Vilela (2010) in the region west of Salinas (white rectangle in Fig. 49.3). The Nova Aurora Fm. is essentially a package of diamictites with minor intercalations of graded sandstone and scarce pelite, with the most striking feature being the great amount of Fe-rich diamictite. The lower Nova Aurora Fm. is dominated by Fe-rich diamictite followed by non-ferruginous diamictite (Fig. 49.5). Despite the Fe content, the diamictites vary considerably in grain size, matrix composition and clast/ matrix ratio. They are metamorphosed to greenschist facies and their clasts are mainly of milky to blue quartz, metasandstone, carbonate, granitoid, gneiss and metapelite. The Fe-rich diamictite package consists of ferruginous diamictite (hematite þ magnetite .5% in volume) and diamictitic Fe formation (Fe .15 wt%), with minor intercalations of sulphide-bearing diamictite, ferruginous and non-ferruginous sandstone, hematite – magnetite schist and pelite with scattered outsized clasts. The foliated matrix of the Fe-rich diamictite consists of quartz, hematite, magnetite, muscovite, biotite and chlorite. The amount of magnetite and the total Fe content (in excess of 40% in places) increases in shear zones. A thin carbonate intercalation, recently found in a drill hole, occurs in the upper section of the Fe-rich diamictite. Layers of sulphide-bearing diamictite occur at the base and top of the Fe-rich diamictite package. The sulphides are mainly pyrite and pyrrhotite. The upper package consists of stratified diamictite with minor layers and lenses of fining-up graded sandstone and thin intercalations of pelite. This diamictite has a foliated matrix composed mainly of quartz, biotite, muscovite, garnet and carbonate. The arenaceous intercalations vary in composition from immature (micaceous, feldspathic, ferruginous) to quartz sandstone, and commonly show load structures. Erosional contacts can be observed between diamictite and sandstone intercalations. The rock-assemblage included in the Lower Chapada Acaua˜ Fm. was mapped and/or studied in detail by Hettich (1975, 1977), Karfunkel & Karfunkel (1976, 1977), Schrank et al. (1978), Uhlein & Chaves (1989), Moura˜o & Pedrosa-Soares (1992), Pedrosa-Soares et al. (1992, 1993, 2007, 2008), Chula et al. (1996), Grossi-Sad et al. (1997b), Moura˜o & Grossi-Sad (1997), Noce (1997b), Oliveira et al. (1997a), Pedrosa-Soares & GrossiSad (1997), Tupinamba´ & Grossi-Sad (1997), Uhlein et al. (1998, 1999, 2007), Gradim et al. (2005) and Martins (2006). The Lower Chapada Acaua˜ Fm. is well exposed along the escarpments of the Acaua˜ Plateau (Fig 49.3) and waterways in

529

the southwestern Arac¸uaı´ and Jequitinhonha valleys (Fig. 49.2). Detailed sections carried out along creeks of the medium valley of the Arac¸uaı´ River suggest a local minimum thickness of 800 m for the Lower Chapada Acaua˜ Fm. (Moura˜o & PedrosaSoares 1992; Pedrosa-Soares & Grossi-Sad 1997). However, the strong regional deformation hampers any accurate thickness evaluation for the whole unit, which seems to be thicker than 2 km (e.g. Grossi-Sad et al. 1997a; Uhlein et al. 1998, 1999). The following description is based on detailed studies carried out in the middle valley of the Arac¸uaı´ River, the type-area of the Lower Chapada Acaua˜ Fm. (Moura˜o & Pedrosa-Soares 1992; Pedrosa-Soares et al. 1992; Pedrosa-Soares & Grossi-Sad 1997), and in the Macau´bas River valley (Martins 2006). The Lower Chapada Acaua˜ Fm. has many vertical and lateral variations of lithofacies. It is essentially composed of fining-up cycles of stratified diamictite, graded sandstone and pelitic rhythmite (Fig. 49.5). The diamictite layers show great variations in clast/ matrix ratios and the foliated matrix consists mainly of quartz, feldspar, muscovite, chlorite, biotite and garnet (according to the metamorphic zone). The clasts vary from granules to boulders and are composed of milky and blue quartz, metasandstone, carbonate, metapelite, granitoid, gneiss and meta-mafic rock. The diamictite layers envelope lenses of cross-laminated sandstone, ranging in length from decimetres to a few metres and up to 1 m thick. At the top of the cycles the diamictite gives way to sandstone and pelite. Erosional contacts between cycles are common. The sandstone layers and lenses show fining-up graded bedding and vary in composition from wacke-rich in carbonate + feldspar + lithoclasts + micas to quartz sandstone. Conglomerate + conglomeratic sandstone can be present at the base of the sandstone layers and lenses. The pelitic rhythmite layers locally show isolated outsized clasts, mainly cobbles and boulders. These layers have mineralogical compositions similar to the diamictite matrix, with variable contents of metamorphic minerals, such as quartz, micas and garnet, in alternating laminae and bands.

Boundary relations with overlying and underlying non-glacial units The Serra do Catuni Fn. was thrust over the Neoproterozoic pelitecarbonate cratonic cover (Fig. 49.2), outlining a clear limit between the Arac¸uaı´ orogen and Sa˜o Francisco craton (Almeida 1977; Uhlein et al. 1998; Alkmim et al. 2006; Pedrosa-Soares et al. 2008). A regional disconformity occurs between the diamictitefree Duas Barras Fm. and the overlying Serra do Catuni Fm., the oldest diamictitic unit of the Macau´bas Group (Karfunkel & Hoppe 1988; Grossi-Sad et al. 1997a; Uhlein et al. 1999, 2007; Martins-Neto et al. 2001; Martins 2006). Locally, the Serra do Catuni diamictite shows erosional contact with the Duas Barras Fm., but normal and tectonic (sheared) contacts are also observed (Grossi-Sad et al. 1997b; Noce 1997a; Martins 2006). The Serra do Catuni Fm. gradually passes upwards and eastwards to the interlayered succession of diamictite, sandstone, pelite and mafic volcanic rocks that characterizes the Lower Chapada Acaua˜ Fm. (Fig. 49.4). Intraformational erosive contacts between diamictite and sand-pelite turbidites of both the Lower Chapada Acaua˜ and Nova Aurora formations can be locally observed (Pedrosa-Soares et al. 1992, 2008; Pedrosa-Soares & Oliveira 1997; Pedrosa-Soares & Grossi-Sad 1997; Uhlein et al. 1998, 1999). In the Arac¸uaı´ River valley and eastern escarpments of the Acaua˜ Plateau (Figs 49.2 & 49.3), the Lower Chapada Acaua˜ Fm. gradually passes upwards to the diamictite-free package of the Upper Chapada Acaua˜ Fm. To the east, this sandstonepelite unit transitions to the fine-grained turbidite package of

530

A. C. PEDROSA-SOARES ET AL.

the proximal Ribeira˜o da Folha Fm. (Pedrosa-Soares et al. 1992, 2008; Pedrosa-Soares & Grossi-Sad 1997; Uhlein et al. 1999, 2007). In the northern sector of the Macau´bas Group, the diamictitefree Upper Chapada Acaua˜ Formation conformably covers the Nova Aurora Formation (Viveiros et al. 1979; Roque et al. 1997; Moura˜o & Grossi-Sad 1997; Oliveira et al. 1997b; PedrosaSoares & Oliveira 1997).

Chemostratigraphy No chemostratigraphic study is available for the Macau´bas Group, but there are isotopic data for older carbonate units located in the Arac¸uaı´ orogen and for carbonate clasts extracted from the Serra do Catuni diamictite. Santos et al. (2004) presented C-isotopic data for metadolomite and metamarlstone samples of the Domingas Fm., formerly interpreted as the oldest unit of the Macau´bas Group (Grossi-Sad et al. 1997b; Noce 1997a). In fact, the shallow marine dolomite-pelite package of the Domingas Fm. is unconformably covered by a fluvial sandstone-conglomerate succession of the Duas Barras Fm. (age , 900 + 21 Ma, see ‘Geochronological constraints’) and contains the stromatolite Conophyton metula Kirichenko that suggests a Mesoproterozoic age (Scho¨ll 1976). The stromatolitic dolomite samples of the Domingas Formation yielded d13CPDB values from þ0.4 to þ0.7‰, similar to other Mesoproterozoic dolostones in the region (the Rio Pardo Grande Fm.; Santos et al. 2004). Other metadolomite samples of the Domingas Fm. yielded d13CPDB values from þ0.2 to –1.9‰ and one sample of metamarlstone is as low as –4.0‰ (Santos et al. 2004). Martins (2006) reported d13CPDB values ranging – 7.2 to 0‰ obtained on 14 carbonate pebbles extracted from the Serra do Catuni diamictite. These data are similar to those obtained from carbonate clasts of the Jequitaı´ diamictite (d13CPDB from – 0.6 to –3.1‰) and from some samples of the Domingas Fm. by Santos et al. (2004).

Other characteristics Manganese deposits have been mined in the southern and central sectors of the Macau´bas Group. They consist of manganese oxides formed by Cenozoic lateritic alteration and supergene concentration on pelitic horizons of the Chapada Acaua˜ Fm. (Dossin & Dardenne 1984b). The thick carbonate lens of the Lower Chapada Acaua˜ Fm. (Figs 49.4 & 49.5) has been sporadically mined (Pedrosa-Soares & Grossi-Sad 1997). The lower part of the Nova Aurora Fm. is rich in ferruginous diamictite and diamictitic Fe formation that form deposits up to 200 m thick, with Fe content up to 40 wt% (Viveiros et al. 1979; Vilela 2010). These deposits are widely capped by thick lateritic soils on the plateaus that characterize the northern sector of the Macau´bas Group. However, new aeromagnetic data suggest significant Fe-rich diamictite deposits that are the target of current exploration. Martins (2006) suggested that the Serra do Catuni diamictite is the source for diamonds found in alluvial deposits along watercourses of the Macau´bas River valley (Fig. 49.2).

Palaeolatitude and palaeogeography No palaeomagnetic data are available for the Macau´bas Group. However, palaeomagnetic data and palaeogeographic settings for the Sa˜o Francisco–Congo craton in late Mesoproterozoic and Neoproterozoic times have been discussed in several papers (e.g. Cordani et al. 2003; D’Agrella-Filho et al. 2004).

Geochronological constraints Although there are abundant data from detrital zircon grains extracted from rocks of the Mata˜o, Duas Barras, Serra do Catuni and Chapada Acaua˜ formations, and from intrusions related to distinct extensional stages of the precursor basin of the Arac¸uaı´ orogen, these data yield few useful constraints on the timing of glacial sedimentation in the Macau´bas Group. The best maximum age constraint on the onset of glacial sedimentation in the Macau´bas Group comes from a detrital study of the Duas Barras Fm., the lowermost stratigraphic unit, where the youngest zircon has an age of 900 + 21 Ma (Babinski et al. 2011). The youngest concordant zircon grain in the Upper Chapada Acaua˜ Fm. provides a maximum depositional age of 864 + 30 Ma (Pedrosa-Soares et al. 2000). The Pedro Lessa mafic dykes and Salto da Divisa granite (Fig. 49.2) are considered to be related to the continental rift stage of the precursor basin of the Arac¸uaı´ orogen (Pedrosa-Soares et al. 2008; Silva et al. 2008). The meta-dolerite dyke swarm of Pedro Lessa was dated at 906 + 2 Ma (zircon and baddeleyite, U –Pb TIMS; Machado et al. 1989). The fluorite-bearing anorogenic Salto da Divisa granite yielded a zircon U –Pb SHRIMP age of 875 + 9 Ma (Silva et al. 2008). The maximum age of deep sea sedimentation for the distal Ribeira˜o da Folha Fm. is constrained by the age of 660 + 29 Ma (U –Pb LA-ICPMS), obtained for the magmatic crystallization of euhedral zircon crystals extracted from meta-plagiogranite patchs found in ophiolitic meta-mafic rocks (Queiroga et al. 2007). The entire Macau´bas must be older than the regional metamorphism, which is dated at c. 580 Ma (Pedrosa-Soares et al. 2001, 2008; Silva et al. 2005).

Discussion The Macau´bas Group includes three widespread diamictitic units, the Serra do Catuni, Nova Aurora and Lower Chapada Acaua˜ formations (Figs 49.3, 49.4 & 49.5). The Serra do Catuni Fm. has been interpreted as a glaciogenic unit deposited mostly in a proximal marine environment based on multiple observations. First, it comprises massive diamictite occurring for over 400 km along the proximal region of the Macau´bas basin (Fig. 49.3). Second, it contains zones rich in chaotic boulders, faceted and striated cobbles and pebbles, and only scarce finer-grained sediments. Third, it is correlated with the glacio-terrestrial deposits of the Jequitaı´ Fm. located on the little deformed Sa˜o Francisco cratonic cover sequence to the west. The Nova Aurora and Lower Chapada Acaua˜ formations also occur in extensive areas, mainly to the east of the Serra do Catuni Fm., along a more distal region of the Macau´bas basin (Fig. 49.3). They essentially consist of stratified diamictite, sandstone and pelite, and show sedimentary features such as cyclic deposition from coarse- to fine-grained sediments, graded bedding, erosional contacts between cycles and load structures that are evidence of debris flow and turbiditic sedimentation in a distal glaciomarine environment. The clear relationship between the Serra do Catuni and Lower Chapada Acaua˜ formations also points to a transition from proximal to distal glaciomarine conditions, respectively (Figs 49.4 & 49.5). The Nova Aurora Fm., which overlies the pre-glacial Rio Peixe Bravo Fm., can be interpreted as a distal correlative of the Serra do Catuni Fm. and, to some extent, a lateral equivalent of the Lower Chapada Acaua˜ Fm. Based on the occurrence of glaciogenic diamictites, the Macau´bas Group is subdivided into a pre-glacial diamictite-free succession (Mata˜o, Duas Barras and Rio Peixe Bravo formations), a glaciogenic succession (Serra do Catuni, Nova Aurora and Lower Chapada Acaua˜ formations) and a post-glacial succession (Upper Chapada Acaua˜ and Ribeira˜o da Folha formations). These successions are related to different evolutionary stages of

´ BAS GROUP NEOPROTEROZOIC MACAU

the Macau´bas basin (Fig. 49.3). The pre-glacial and glaciogenic successions have been ascribed to the continental rift to transitional stages of the Macau´bas basin based on multiple lines of evidence: † the predominance of immature sandstone-conglomeratic sedimentation with bimodal (NW – SE and SE –NW) palaeocurrent sets in the pre-glacial succession; † syn-sedimentary normal faulting; † clastic dykes related to the rudite-sandstone sedimentation of the pre-glacial succession and Serra do Catuni Fm.; † the huge ferruginous diamictite deposits of the Nova Aurora Fm.; † the transitional geochemical and isotopic signatures of the Chapada Acaua˜ submarine basalts that also carry inherited zircon grains from continental sources. The post-glacial succession represents the passive margin stage of the Macau´bas basin, when sea-floor spreading continued at least until c. 660 Ma. The onset of a rifting stage in the region is constrained by the U –Pb ages of the Pedro Lessa mafic dykes (c. 906 Ma; Fig. 49.2) and Salto da Divisa anorogenic granites (c. 875 Ma; see ‘Geochronological constraints’). Indeed, these anorogenic magmatic episodes provide solid evidence that an active continental rift developed during the late Tonian (900 –850 Ma) in this region (Fig. 49.2), and it is certain that at least the pre-glacial strata were deposited during this rift stage. The Serra do Catuni Fm. and the cratonic Jequitaı´ Fm. have also been assigned to the continental rift stage (Pedrosa-Soares et al. 1992, 2008; Uhlein et al. 1998, 1999; Martins-Neto et al. 2001; Martins-Neto & Hercos 2002). However, if the regional disconformity between the pre-glacial and glaciogenic successions represents a significant hiatus (Fig. 49.3; see ‘Boundary relations’), the Serra do Catuni Fm. could be early Cryogenian in age (Uhlein et al. 2007). With regards to the age of glaciation, the cratonic Jequitaı´ and Carrancas diamictites located in the southern sector of the nearby Sa˜o Francisco craton (Fig. 49.1a) should be considered. Although the Macau´bas, Jequitaı´ and Carrancas diamictites show no direct field relation, they have been correlated with each other since the 1930s, owing to their shared glacial origin. The maximum sedimentation ages for the Jequitaı´ Fm. (c. 880 Ma; Rodrigues 2008) and the oldest Macau´bas diamictites (c. 900 Ma; see ‘Geochronological constraints’) are similar, but they do not provide any precise information on the glaciation age. On the other hand, the Carrancas diamictite can be correlated to an early Cryogenian glacial event because it is covered by the c. 740 Ma cap carbonate of the cratonic cover (Babinski et al. 2007, 2011). In order to fully understand the timing and nature of glaciation on the Sa˜o Francisco craton, the entire Arac¸uaı´ –West Congo orogenic system and its precursor basin should also be considered (Fig. 49.1b). Because these now rifted margins were juxtaposed prior to full opening of the Macau´bas basin, the Brazilian and African counterparts of the precursor basin must have lain in similar palaeoclimatic zones (e.g. Cordani et al. 2003; Pisarevsky et al. 2003). In the West Congo belt, the Lower Mixtite Fm. is the diamictite unit that has been related to the early Cryogenian glacial event (Trompette 1994; Tack et al. 2001; Frimmel et al. 2006). Pedrosa-Soares et al. (2008) correlated the Lower Mixtite and Lower Chapada Acaua˜ formations because they include glaciomarine diamictites with intercalations of transitional submarine basalts, representing a late rift stage. These authors also correlated the Lower Chapada Acaua˜ and Lower Mixtite formations with the Carrancas diamictite, and ascribed them to the early Cryogenian glacial event. The Upper Mixtite Fm. in the West Congo Belt, which is ascribed to the late Cryogenian glacial event (Frimmel et al. 2006), does not appear to have a counterpart in the Macau´bas Group (Pedrosa-Soares et al. 2008).

531

In conclusion, it seems that the Macau´bas Group records at least the early Cryogenian glacial event. The Serra do Catuni Fm. and its probable correlative, the Jequitaı´ Fm., were likely deposited during an early phase of that glacial event. The Lower Chapada Acaua˜ Fm. and its probable equivalents, the Carrancas diamictite and the Lower Mixtite Fm. then represents a late phase of the early Cryogenian glacial event. Indeed, the very extensive and thick diamictitic package of the Macau´bas Group might contain equivalents of all the Neoproterozoic glaciogenic units preserved on the Sa˜o Francisco craton. The challenging question is how many discrete glacial epochs do they represent? The authors acknowledge financial support provided by CNPq (Conselho Nacional de Desenvolvimento Cientı´fico e Tecnolo´gico), FINEP/CT-Infra (Financiadora de Estudos e Projetos), FAPEMIG (Fundac¸a˜o de Amparo a` Pesquisa de Minas Gerais) and FAPESP (Fundac¸a˜o de Amparo a` Pesquisa do Estado de Sa˜o Paulo). We also thank W. Altermann, M. A. Martins-Neto and an anonymous reviewer, as well as Emmanuelle Arnaud, for corrections and suggestions. We dedicate this work to our beloved friend and coauthor – the late Carlos Maurı´cio Noce – who contributed very significantly to understanding the Macau´bas basin. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Alkmim, F. F., Marshak, S., Pedrosa-Soares, A. C., Peres, G. G., Cruz, S. C. & Whittington, A. 2006. Kinematic evolution of the Arac¸uaı´-West Congo orogen in Brazil and Africa: nutcracker tectonics during the Neoproterozoic assembly of Gondwana. Precambrian Research, 149, 43 –63. Alkmim, F. F., Pedrosa-Soares, A. C., Noce, C. M. & Cruz, S. C. P. 2007. Sobre a evoluc¸a˜o tectoˆnica do Oro´geno Arac¸uaı´-Congo Ocidental. Geonomos, 15, 35 – 43. Almeida, F. F. M. 1977. O Cra´ton do Sa˜o Francisco. Revista Brasileira de Geocieˆncias, 7, 349–364. Babinski, M., Gradim, R. J., Pedrosa-Soares, A. C., Alkmim, F. F., Noce, C. M. & Liu, D. 2005. Geocronologia U– Pb (SHRIMP) e Sm–Nd de xistos verdes basa´lticos do Oro´geno Arac¸uaı´: implicac¸o˜es para a idade do Grupo Macau´bas. Revista Brasileira de Geocieˆncias, 35 (4-supplement), 77– 81. Babinski, M., Vieira, L. C. & Trindade, R. I. F. 2007. Direct dating of the Sete Lagoas cap carbonate (Bambuı´ Group, Brazil) and implications for the Neoproterozoic glacial events. Terra Nova, 19, 401 –406. Babinski, M., Pedrosa-Soares, A. C., Trindade, R. I. F., Martins, M., Noce, C. M. & Liu, D. 2011. Neoproterozoic glacial deposits from the Arac¸uaı´ orogen, Brazil: age, provenance and correlations with the Sa˜o Francisco craton and West Congo belt. Gondwana Research, doi: 10.1016/j.gr.2011.04.008. Chula, A. M., Knauer, L. G. & Almeida-Abreu, P. A. 1996. Estratigrafia do Supergrupo Espinhac¸o na regia˜o de Planalto de Minas, Diamantina, MG. Geonomos, 3, 69 –81. Cordani, U. G., Brito-Neves, B. B., D’agrella-Filho, M. S. & Trindade, R. I. F. 2003. Tearing-up Rodinia: the Neoproterozoic paleogeography of South American cratonic fragments. Terra Nova, 15, 343– 349. D’Agrella-Filho, M. S., Pacca, I. I. G., Trindade, R. I. F., Teixeira, W., Raposo, M. I. B. & Onstott, T. C. 2004. Paleomagnetism and 40 Ar39Ar ages of mafic dykes from Salvador, Brazil: new constraints on the Sa˜o Francisco craton APW path between 1080 and 1010 Ma. Precambrian Research, 132, 55– 77. Dossin, I. A. & Dardenne, M. A. 1984a. Geologia da borda ocidental da Serra do Cipo´, Minas Gerais. In: Congresso Brasileiro de Geologia 33, Rio de Janeiro, Anais, 7. Sociedade Brasileira de Geologia, 3104–3117. Dossin, I. A. & Dardenne, M. A. 1984b. Os depo´sitos supergeˆnicos de manganeˆs da borda ocidental da Serra do Cipo´, MG (Quadrı´cula Inhame). In: Congresso Brasileiro de Geologia 33, Rio de Janeiro, Anais, 3. Sociedade Brasileira de Geologia, 1129–1143. Drumond, J. B. V. 2000. Folhas Cordeiros, Belo Campo e Curral de Dentro. CODEMIG, Projeto Leste, Belo Horizonte.

532

A. C. PEDROSA-SOARES ET AL.

Fogac¸a, A. C. C. 1997. Geologia da Folha Diamantina. In: Grossi-Sad, J. H., Lobato, L. M., Pedrosa-Soares, A. C. & Soares-Filho, B. S. (eds) Projeto Espinhac¸o. CODEMIG, Belo Horizonte, 1575– 1666. Frimmel, H. E., Tack, L., Basei, M., Nutman, A. & Boven, A. 2006. Provenance and chemostratigraphy of the Neoproterozoic West Congolian Group in the Democratic Republic of Congo. Journal of African Earth Sciences, 46, 221–239. Gradim, R. J., Alkmim, F. F., Pedrosa-Soares, A. C., Babinski, M. & Noce, C. M. 2005. Xistos verdes do Alto Arac¸uaı´, Minas Gerais: vulcanismo ba´sico do rifte neoproterozo´ico Macau´bas. Revista Brasileira de Geocieˆncias, 35 (4-suplemento), 59 –69. Gravenor, C. P. & Monteiro, R. 1983. Ice-thrust features in the Proterozoic Macau´bas Group, Jequitaı´ area, Minas Gerais, Brazil. Journal of Geology, 91, 113– 116. Grossi-Sad, J. H., Lobato, L. M., Pedrosa-Soares, A. C. & Soares-Filho, B. S. 1997a. Projeto Espinhac¸o. CODEMIG, Belo Horizonte. Grossi-Sad, J. H., Roque, N. C., Knauer, L. G., Noce, C. M. & Fonseca, E. 1997b. Geologia da Folha Carbonita. In: Grossi-Sad, J. H., Lobato, L. M., Pedrosa-Soares, A. C. & Soares-Filho, B. S. (eds) Projeto Espinhac¸o. CODEMIG, Belo Horizonte, 1251–1372. Guimara˜es, M. L. V. 1997. Geologia da Folha Botumirim. In: GrossiSad, J. H., Lobato, L. M., Pedrosa-Soares, A. C. & Soares-Filho, B. S. (eds) Projeto Espinhac¸o. CODEMIG, Belo Horizonte, 543– 610. Guimara˜es, M. L., Crocco-Rodrigues, F. A., Abreu, F. R., Oliveira, O. A. B. & Greco, F. M. 1993. Geologia do bloco Itacambira-Monte Azul entre Barroca˜o e Porteirinha, MG. Boletim do Nu´cleo Minas Gerais-Sociedade Brasileira de Geologia, 12, 74– 78. Guimara˜es, M. L. V., Grossi-Sad, J. H. & Fonseca, E. 1997. Geologia da Folha Francisco Sa´. In: Grossi-Sad, J. H., Lobato, L. M., PedrosaSoares, A. C. & Soares-Filho, B. S. (eds) Projeto Espinhac¸o. CODEMIG, Belo Horizonte, 223– 313. Hettich, M. 1975. Zur stratigraphie und Genese des Macau´bas nordlich der Serra Negra, Espinhac¸o-Zone, Minas Gerais, Brasilien. Geologisches Jahrbuch, 14, 47 –85. Hettich, M. 1977. A glaciac¸a˜o proterozo´ica no centro-norte de Minas Gerais. Revista Brasileira de Geocieˆncias, 7, 87– 101. Hettich, M. & Karfunkel, J. 1978. Um esker, um varvito e seixos estriados no Grupo Macau´bas, norte de Minas Gerais. Revista Escola de Minas de Ouro Preto, 34, 5– 8. Isotta, C. A. L., Rocha-Campos, A. C. & Yoshida, R. 1969. Striated pavement of the Upper Precambrian glaciation in Brazil. Nature, 222, 466– 468. Karfunkel, J. & Karfunkel, B. 1976. Estudos petro-faciolo´gicos do Grupo Macau´bas na porc¸a˜o mediana da Serra do Espinhac¸o, Minas Gerais. In: Congresso Brasileiro de Geologia 29, Ouro Preto, Anais, 2. Sociedade Brasileira de Geologia, 179 – 188. Karfunkel, J. & Karfunkel, B. 1977. Fazielle Entwicklung der mittleren Espinhac¸o-Zone mit besonderer Beru¨cksichtigung des TillitProblems (Minas Gerais, Brasilien). Geologisches Jahrbuch, 24, 3 –91. Karfunkel, J. & Hoppe, A. 1988. Late Precambrian glaciation in centraleastern Brazil: synthesis and model. Palaeogeography, Palaeoclimatology, Palaeoecology, 65, 1– 21. Karfunkel, J., Moreira, P. C., Ribeiro, M. C. & Franco, A. L. 1984. Aspectos gene´ticos e deposicionais do Grupo Macau´bas na regia˜o da barragem do Parau´na e sua importaˆncia na contribuic¸a˜o para um modelo paleogeogra´fico e geotectoˆnico. In: Congresso Brasileiro de Geologia 33, Rio de Janeiro, Anais, 7. Sociedade Brasileira de Geologia, 3091– 3103. Karfunkel, J., Pedrosa-Soares, A. C. & Dossin, I. A. 1985. O Grupo Macau´bas em Minas Gerais: revisa˜o dos conhecimentos. Boletim do Nu´cleo Minas Gerais-Sociedade Brasileira de Geologia, 5, 45 –59. ´ gua Fria e Karfunkel, J., Hoppe, A. & Noce, C. M. 2002. Serra da A vizinhancas, MG: vestı´gios de glaciac¸a˜o neoproterozo´ica. In: Schobbenhaus, C., Campos, D. A., Queiroz, E. T., Winge, M. & Berbert-Born, M. (eds) Sı´tios Geolo´gicos e Paleontolo´gicos do Brasil, 1. DNPM, Brası´lia, Brazil, 165– 173.

Knauer, L. G. & Grossi-Sad, J. H. 1997. Geologia da Folha Presidente Kubtischek. In: Grossi-Sad, J. H., Lobato, L. M., Pedrosa-Soares, A. C. & Soares-Filho, B. S. (eds) Projeto Espinhac¸o. CODEMIG, Belo Horizonte, 1902– 2056. Knauer, L. G., Silva, L. L., Souza, F. B. B., Silva, L. R. & Carmo, R. C. 2006. Folha Monte Azul 1:100.000. CPRM, Programa Geologia do Brasil, Brası´lia. Lima, S. A. A., Martins-Neto, M. A., Pedrosa-Soares, A. C., Cordani, U. G. & Nutman, A. 2002. A Formac¸a˜o Salinas na a´rea-tipo, NE de Minas Gerais: uma proposta de revisa˜o da estratigrafia da Faixa Arac¸uaı´ com base em evideˆncias sedimentares, metamo´rficas e idades U– Pb SHRIMP. Revista Brasileira de Geocieˆncias, 32, 491– 500. Machado, N., Schrank, A., Abreu, F. R., Knauer, L. G. & Almeida-Abreu, P. A. 1989. Resultados preliminares da geocronologia U– Pb na Serra do Espinhac¸o Meridional. Boletim do Nu´cleo Minas Gerais-Sociedade Brasileira de Geologia, 10, 171–174. Marshak, S., Alkmim, F. F., Whittington, A. & Pedrosa-Soares, A. C. 2006. Extensional collapse in the Neoproterozoic Arac¸uaı´ orogen, eastern Brazil: a setting for reactivation of asymmetric crenulation cleavage. Journal of Structural Geology, 28, 129–147. Martins, M. S. 2006. Geologia dos diamantes e carbonados aluvionares da bacia do Rio Macau´bas, MG. PhD thesis, Instituto de Geocieˆncias, Universidade Federal de Minas Gerais, Belo Horizonte. Martins, M. S., Karfunkel, J., Noce, C. M., Babinski, M., PedrosaSoares, A. C., Sial, A. N. & Lyu, D. 2008. A sequ¨eˆncia pre´-glacial do Grupo Macau´bas na a´rea-tipo e o registro da abertura do rifte Arac¸uaı´. Revista Brasileira de Geocieˆncias, 38, 768– 779. Martins-Neto, M. A. & Hercos, C. M. 2002. Sedimentation and tectonic setting of Early Neoproterozoic glacial deposits in southeastern Brazil. In: Altermann, W. & Corcoran, P. L. (eds) Precambrian Sedimentary Environments: A Modern Approach to Ancient Depositional Systems. International Association of Sedimentologists, Special Publications, 33, 383– 403. Martins-Neto, M. A., Gomes, N. S., Hercos, C. M. & Reis, L. A. 1999. Fa´cies gla´cio-continentais (outwash plain) na Megassequ¨encia ´ gua Fria, MG) e seu contexto geotectoˆnico. Macau´bas (Serra da A Revista Brasileira de Geocieˆncias, 29, 179–188. Martins-Neto, M. A., Pedrosa-Soares, A. C. & Lima, S. A. A. 2001. Tectono-sedimentary evolution of sedimentary basins from Late Paleoproterozoic to Late Neoproterozoic in the Sa˜o Francisco craton and Arac¸uaı´ fold belt, eastern Brazil. Sedimentary Geology, 141/142, 343–370. Moraes, L. J. 1929. Geologia da regia˜o diamantina de Minas Gerais. In: Relato´rio Anual do Diretor 1928, Rio de Janeiro. Servic¸o Geolo´gico e Mineralo´gico, 29 – 34. ´ rea ocupada pela Formac¸a˜o Macau´bas no Moraes, L. J. 1932. A norte de Minas Gerais. Anais Academia Brasileira de Cieˆncias, 4, 111– 114. Moraes, L. J. & Guimara˜es, D. 1930. Geologia da regia˜o diamantı´fera do norte de Minas Gerais. Anais Academia Brasileira de Cieˆncias, 2, 153– 186. Moraes, L. J. & Guimara˜es, D. 1931. The diamond-bearing region of Northern Minas Gerais, Brazil. Economic Geology, 26, 502– 530. Moura˜o, M. A. A. & Pedrosa-Soares, A. C. 1992. Evideˆncias de sedimentac¸a˜o turbidı´tica no Grupo Macau´bas, Faixa Arac¸uaı´. Revista Escola de Minas de Ouro Preto, 45, 94– 96. Moura˜o, M. A. A. & Grossi-Sad, J. H. 1997. Geologia da Folha Padre Carvalho. In: Grossi-Sad, J. H., Lobato, L. M., Pedrosa-Soares, A. C. & Soares-Filho, B. S. (eds) Projeto Espinhac¸o. CODEMIG, Belo Horizonte, 316– 418. Moura˜o, M. A. A., Grossi-Sad, J. H. & Fonseca, E. 1997. Geologia da Folha Janau´ba. In: Grossi-Sad, J. H., Lobato, L. M., PedrosaSoares, A. C. & Soares-Filho, B. S. (eds) Projeto Espinhac¸o. CODEMIG, Belo Horizonte, 11 – 123. Noce, C. M. 1997a. Geologia da Folha Curimataı´. In: Grossi-Sad, J. H., Lobato, L. M., Pedrosa-Soares, A. C. & Soares-Filho, B. S. (eds) Projeto Espinhac¸o. CODEMIG, Belo Horizonte, 1207– 1250. Noce, C. M. 1997b. Geologia da Folha Itacambira. In: Grossi-Sad, J. H., Lobato, L. M., Pedrosa-Soares, A. C. & Soares-Filho, B. S. (eds) Projeto Espinhac¸o. CODEMIG, Belo Horizonte, 853–926.

´ BAS GROUP NEOPROTEROZOIC MACAU

Noce, C. M., Pedrosa-Soares, A. C. et al. 1997. Nova subdivisa˜o estratigra´fica regional do Grupo Macau´bas na Faixa Arac¸uaı´: o registro de uma bacia neoproterozo´ica. Boletim do Nu´cleo Minas GeraisSociedade Brasileira de Geologia, 14, 29 –31. Oliveira, M. J. R., Grossi-Sad, J. H., Romano, A. W. & Lobato, L. M. 1997a. Geologia da Folha Gra˜o Mogol. In: Grossi-Sad, J. H., Lobato, L. M., Pedrosa-Soares, A. C. & Soares-Filho, B. S. (eds) Projeto Espinhac¸o. CODEMIG, Belo Horizonte, 611–642. Oliveira, M. J. R., Fogac¸a, A. C. C. & Fonseca, E. 1997b. Geologia da Folha Baldim. In: Grossi-Sad, J. H., Lobato, L. M., PedrosaSoares, A. C. & Soares-Filho, B. S. (eds) Projeto Espinhac¸o. CODEMIG, Belo Horizonte, 2437– 2531. Pedrosa-Soares, A. C. & Grossi-Sad, J. H. 1997. Geologia da Folha Minas Novas. In: Grossi-Sad, J. H., Lobato, L. M., PedrosaSoares, A. C. & Soares-Filho, B. S. (eds) Projeto Espinhac¸o. CODEMIG, Belo Horizonte, 925– 1053. Pedrosa-Soares, A. C. & Oliveira, M. J. R. 1997. Geologia da Folha Salinas. In: Grossi-Sad, J. H., Lobato, L. M., Pedrosa-Soares, A. C. & Soares-Filho, B. S. (eds) Projeto Espinhac¸o. CODEMIG, Belo Horizonte, 419– 541. Pedrosa-Soares, A. C., Leonardos, O. H. & Correia-Neves, J. M. 1984. Aspectos metamo´rficos de sequ¨eˆncias supracrustais da Faixa Arac¸uaı´ em Minas Gerais. In: Congresso Brasileiro de Geologia 33, Rio de Janeiro, Anais, 6. Sociedade Brasileira de Geologia, 3056– 3065. Pedrosa-Soares, A. C., Noce, C. M., Vidal, P., Monteiro, R. & Leonardos, O. H. 1992. Toward a new tectonic model for the Late Proterzoic Arac¸uaı´ (SE Brazil) — West Congolian (SW Africa) Belt. Journal of South American Earth Sciences, 6, 33 –47. Pedrosa-Soares, A. C., Baars, F. J., Lobato, L. M., Magni, M. C. V. & Faria, L. F. 1993. Arquitetura tectono-metamo´rfica do setor central da Faixa Arac¸uai. Boletim do Nu´cleo Minas Gerais-Sociedade Brasileira de Geologia, 12, 176– 182. Pedrosa-Soares, A. C., Leonardos, O. H., Ferreira, J. C. & Reis, L. B. 1996. Duplo regime metamo´rfico na Faixa Arac¸uaı´: reinterpretac¸a˜o a` luz de novos dados. In: Congresso Brasileiro de Geologia 39, Salvador, Anais, 6. Sociedade Brasileira de Geologia, 5– 8. Pedrosa-Soares, A. C., Vidal, P., Leonardos, O. H. & Brito-Neves, B. B. 1998. Neoproterozoic oceanic remnants in eastern Brazil: further evidence and refutation of an exclusively ensialic evolution for the Arac¸uaı´-West Congo orogen. Geology, 26, 519– 522. Pedrosa-Soares, A. C., Cordani, U. & Nutman, A. 2000. Constraining the age of Neoproterozoic glaciation in eastern Brazil: first U –Pb (SHRIMP) data from detrital zircons. Revista Brasileira de Geocieˆncias, 30, 58 – 61. Pedrosa-Soares, A. C., Noce, C. M., Wiedemann, C. M. & Pinto, C. P. 2001. The Arac¸uaı´-West Congo orogen in Brazil: an overview of a confined orogen formed during Gondwanland assembly. Precambrian Research, 110, 307–323. Pedrosa-Soares, A. C., Noce, C. M., Alkmim, F. F., Silva, L. C., Babinski, M., Cordani, U. & Castan˜eda, C. 2007. Oro´geno Arac¸uaı´: sı´ntese do conhecimento 30 anos apo´s Almeida 1977. Geonomos, 15, 1– 16. Pedrosa-Soares, A. C., Alkmim, F. F., Tack, L., Noce, C. M., Babinski, M., Silva, L. C. & Martins-Neto, M. A. 2008. Similarities and differences between the Brazilian and African counterparts of the Neoproterozoic Arac¸uaı´-West Congo orogen. In: Pankhurst, R. J., Trouw, R. A. J., Brito Neves, B. B. & De Wit, M. J. (eds) West Gondwana: Pre-Cenozoic Correlations Across the South Atlantic Region. Geological Society, London, Special Publications, 294, 153– 172. Pflug, R. & Scho¨ll, W. U. 1975. Proterozoic glaciation in Eastern Brazil: a review. Geologische Rundschau, 64, 287– 299. Pisarevsky, S. A., Wingate, M. T. D., Powell, C. Mca., Johnson, S. & Evans, D. A. D. 2003. Models of Rodinia assembly and fragmentation. In: Yoshida, M., Windley, B. & Dasgupta, S. (eds) Proterozoic East Gondwana: Supercontinent Assembly and Breakup. Geological Society, London, Special Publications, 206, 35–55. Porada, H. 1989. Pan-African rifting and orogenesis in southern to equatorial Africa and Eastern Brazil. Precambrian Research, 44, 103–136.

533

Queiroga, G. N., Pedrosa-Soares, A. C., Que´me´neur, J. & Castan˜eda, C. 2006. A unidade metassedimentar do ofiolito de Ribeira˜o da Folha, Oro´geno Arac¸uaı´, Minas Gerais: petrografia, geotermobarometria e calcografia. Geonomos, 14, 9 –12. Queiroga, G. N., Pedrosa-Soares, A. C. et al. 2007. Age of the Ribeira˜o da Folha ophiolite, Arac¸uaı´ Orogen: the U– Pb zircon dating of a plagiogranite. Geonomos, 15, 61 – 65. Rocha-Campos, A. C. & Hasui, Y. 1981. Tillites of the Macau´bas Group (Proterozoic) in central Minas Gerais and southern Bahia, Brazil. In: Hambrey, M. J. & Harland, W. B. (eds) Earths’s Pre-Pleistocene Glacial Record. Cambridge University Press, 933– 939. Rodrigues, J. B. 2008. Provenieˆncia de sedimentos dos grupos Canastra, Ibia´, Vazante e Bambuı´: um estudo de zirco˜es detrı´ticos e idadesmodelo Sm –Nd. PhD thesis, Instituto de Geocieˆncias, Universidade de Brası´lia, Brazil. Rogers, J. W. & Santosh, M. 2004. Continents and Supercontinents. Oxford University Press. Roque, N. C., Grossi-Sad, J. H., Noce, C. M. & Fonseca, E. 1997. Geologia da Folha Rio Pardo de Minas. In: Grossi-Sad, J. H., Lobato, L. M., Pedrosa-Soares, A. C. & Soares-Filho, B. S. (eds) Projeto Espinhac¸o. CODEMIG, Belo Horizonte, 126–221. Santos, R. V., Alvarenga, C. et al. 2004. Carbon isotopes of Mesoproterozoic –Neoproterozoic sequences from southern Sa˜o Francisco craton and Arac¸uaı´ Belt: paleogeographic implications. Journal of South American Earth Sciences, 18, 27 –39. Scho¨ll, W. U. 1972. Der su¨dwestliche Randbereich der Espinhac¸o Zone, Minas Gerais, Brasilien. Geologische Rundschau, 61, 201– 216. Scho¨ll, W. U. 1976. Estromato´litos Conophyton em dolomitos do Grupo Macau´bas. In: Congresso Brasileiro de Geologia 29, Ouro Preto, Anais, 2. Sociedade Brasileira de Geologia, 67– 73. Schrank, A., Dourado, B. V. & Biondi, J. C. 1978. Estudo preliminar dos metavulcanitos do Grupo Macau´bas na regia˜o do Alto Jequitinhonha, MG. In: Congresso Brasileiro de Geologia 30, Recife, Anais, 3. Sociedade Brasileira de Geologia, 1323–1335. Silva, L. C., Mcnaughton, N. J., Armstrong, R., Hartmann, L. & Fletcher, I. 2005. The Neoproterozoic Mantiqueira Province and its African connections. Precambrian Research, 136, 203– 240. Silva, L. C., Pedrosa-Soares, A. C. & Teixeira, L. R. 2008. Tonian rift-related, A-type continental plutonism in the Arac¸uaı´ orogen, Eastern Brazil: new evidences for the breakup stage of the Sa˜o Francisco– Congo Paleocontinent. Gondwana Research, 13, 527– 537. Tack, L., Wingate, M. T .D., Lie´geois, J. P., Fernandez-Alonso, M. & Deblond, A. 2001. Early Neoproterozoic magmatism (1000–910 Ma) of the Zadinian and Mayumbian Groups (Bas-Congo): onset of Rodinian rifting at the western edge of the Congo craton. Precambrian Research, 110, 277–306. Tait, J., Delpomdor, F., Pre´at, A., Straathof, G., Kanda, V. & Tack, L. 2011. Neoproterozoic Sequences of the West Congo and Lindi/ Ubangi Supergroups in the Congo Craton, central Africa. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 185–194. Trompette, R. 1994. Geology of Western Gondwana (2000 –500 Ma). Pan-African–Brasiliano aggregation of South America and Africa. A. A. Balkema, Rotterdam. Tupinamba´, M. & Grossi-Sad, J. H. 1997. Geologia da Folha Rio Vermelho. In: Grossi-Sad, J. H., Lobato, L. M., Pedrosa-Soares, A. C. & Soares-Filho, B. S. (eds) Projeto Espinhac¸o. CODEMIG, Belo Horizonte, 1667–1806. Uhlein, A. & Chaves, M. S. C. 1989. Geologia da borda nordeste da Serra do Espinhac¸o (regia˜o de Mendanha a Sa˜o Gonc¸alo do Rio Preto, MG). Boletim do Nu´cleo Minas Gerais-Sociedade Brasileira de Geologia, 10, 175– 179. Uhlein, A., Trompette, R. & Egydio-Silva, M. 1998. Proterozoic rifting and closure, SE border of the Sa˜o Francisco craton, Brazil. Journal of South American Earth Sciences, 11, 191– 203. Uhlein, A., Trompette, R. & Alvarenga, C. 1999. Neoproterozoic glacial and gravitational sedimentation on a continental rifted margin: the Jequitaı´-Macau´bas sequence (Minas Gerais, Brazil). Journal of South American Earth Sciences, 12, 435– 451.

534

A. C. PEDROSA-SOARES ET AL.

Uhlein, A., Trompette, R., Egydio-Silva, M. & Vauchez, A. 2007. A glaciac¸a˜o Sturtiana (750 Ma), a estrutura do rifte Macau´bas-Santo Onofre e a estratigrafia do Grupo Macau´bas, Faixa Arac¸uaı´. Geonomos, 15, 45 – 60. Uhlein, A., de Alvarenga, C. J. S., Dardenne, M. A. & Trompette, R. R. 2011. The Glaciogenic Jequitaı´ Formation, Southeastern Brazil. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 541– 546. Vilela, F. T. 2010. Caracterizac¸a˜o de metadiamictitos ferruginosos da Formac¸a˜o Nova Aurora (Grupo Macau´bas, Oro´geno Arac¸uail) a

oeste de Salinas, ME. MSc thesis, Instituto de Geoscieˆncias, Universidade Federal de Minas Gerais, Belo Horizonte. Viveiros, J. F. M. & Walde, D. 1976. Geologia da Serra do Cabral, Minas Gerais, Brasil. Mu¨nsterische Forschungshefte Geologie und Palaeontologie, 38/39, 15 –27. Viveiros, J. F. M., Sa´, E. L., Vilela, O. V., Santos, O. M., Moreira, J. M. P., Holder-Neto, F. & Vieira, V. S. 1979. Geologia dos vales dos rios Peixe Bravo e Alto Vacaria, norte de Minas Gerais. Boletim do Nu´cleo Minas Gerais-Sociedade Brasileira de Geologia, 1, 75– 87. Walde, D. 1976. Neue Hinweise fu¨r eine proterozoische Vereisung in Ostbrasilien. Mu¨nsterische Forschungshefte Geologie und Palaeontologie, 38/39, 47 –59.

Chapter 50 Moema laminites: a newly recognized Neoproterozoic (?) glaciogenic unit, Sa˜o Francisco Basin, Brazil ANTONIO C. ROCHA-CAMPOS1*, BENJAMIN B. DE BRITO NEVES1, MARLY BABINSKI1, PAULO R. DOS ˆ NIA M.B. DE OLIVEIRA1 & A. ROMANO2 SANTOS1, SO 1

Instituto de Geocieˆncias, Universidade Sa˜o Paulo, Rua do Lago, 562, Cidade Universita´ria, Sa˜o Paulo, SP, Brazil 2

Instituto de Geocieˆncias, Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazil *Corresponding author (e-mail: [email protected])

Abstract: A recently identified diamictite and silt-clay laminite, which discordantly overlie the Archaean basement and underlie the Neoproterozoic Bambui Group, have been informally named as the Moema Laminites. They are preserved at the southwestern margin of the Sa˜o Francisco Basin in southeastern Brazil, and are widely distributed in the central-western Minas Gerais state. They crop out discontinuously over at least 140 km along a north–south direction. The nomenclature of and stratigraphic relationships between the Moema Laminites and other isolated Neoproterozic occurrences of similar rocks are in a state of flux. Two exceptionally good exposures of the Moema Laminites show good evidence for deposition under glacial conditions. At the Formiga locality, a single glacial advance is registered by a deformation tillite, while overlying laminite records deposition in a post-glacial, probably marine basin following deglaciation. At the SAFFRAN quarry, striations on a bedding plane may have been caused by floating seaice that just touched the bottom of the basin. Much additional work is needed to establish relationships between the Moema Laminites and other similar occurrences. If these and Moema Laminites are shown to be Cryogenian glacial deposits, the area covered by the Cryogenian glaciations in the Sa˜o Francisco basin is much larger than formerly believed.

The name Moema Laminites is being used informally to designate a newly identified, possibly Neoproterozoic unit, made up of diamictite and silty-clayey laminites. The strata, so far known at reconnaissance level, crop out extensively in the central-western State of Minas Gerais (Fig. 50.1), below carbonate assigned to the Sete Lagoas Formation (Fm.) of the Bambuı´ Group (Neoproterozoic) (Fig. 50.2). The Moema Laminites crop out discontinuously over a large portion of the southern Sa˜o Francisco Basin, from the latitude of the town of Onc¸a do Pitangui (198430 S, 448W), southward for c. 140 km, at least as far as the town of Cristais (208520 S, 45830 W) (Fig. 50.1). Two exceptionally good exposures of the strata are found at a road cut of BR-354, near the town of Formiga, 1 km north of the intersection with road MG-050, in the southern part of the study area, and at the SAFFRAN quarry (named after the owner company), some 100 km north of the first locality, in the vicinity of Onc¸a do Pitangui, 10 km north of road BR-262. The thickest exposure of the unit is in the vicinity of the town of Moema (Fig. 50.1), but because the name is provisional, no formal type-section has yet been designated.

Structural framework The Moema Laminites are part of the sedimentary cover of the intracratonic Sa˜o Francisco Basin, in the central-western State of Minas Gerais. They occur immediately below the Neoproterozoic Bambui Group along the southwestern margin of the basin. In this area, the horizontal to gently folded platformal cover of the basin fills a depression in the cratonic foreland, located west of the Sete Lagoas High (Fig. 50.1). Deposition of the Moema sediments was contemporaneous with the sag phase of tectonic evolution of the basin (Alkmin 2004). Outcrops of the Moema Laminites are, in general, preserved in topographic lows on the irregular surface of the cratonic basement. Strata are essentially horizontal but have been slightly affected by normal faults. Near Onc¸a do Pitangui they are part of an east –west trending syncline with plunges up to 308E. Rocks of the unit are essentially non-metamorphic.

Stratigraphy The sparcity of outcrops (mostly road cuts), moderate topography and deep weathering make it difficult to log a complete representative stratigraphic section. A regional field reconnaissance was performed along several roads in the area between the towns of Cristais and Bom Despacho, in order to evaluate the distribution and stratigraphic relations of the studied unit (Fig. 50.1). This was complemented by more detailed stratigraphic and sedimentological study of critical outcrops. Laterally, the thickness of the diamictite varies from 30 cm on the Formiga outcrop to, more commonly, c. 5 m, and may be missing in places. In other places, the unit is represented only by remains of the basal diamictite, the upper laminites having been eroded away. The thickness of the overlying laminites reaches at least 60 m in the area of the town of Onc¸a do Pitangui (Fig. 50.1). Strata below and above the lower nonconformity are deeply weathered. Kaolinitized regolith often reaches considerable thickness on top of Archaean basement in the study area and may record a significant and intense phase of pre-Neoproterozoic weathering, superposed by widespread recent alteration. The stratigraphic relations of the Moema Laminites are similar to that of other isolated outcrops of conglomerate/diamictite and associated rocks that are found below carbonate rocks of the Neoproterozoic Sete Lagoas Fm. (Bambuı´ Group). These are known from several places, fringing the western margin of the Sa˜o Francisco Basin (e.g. Member/Facies Carrancas, Costa et al. 1961; Facies Sambura´, Miranda 1953, and so on) and are traditionally correlated with the Jequitaı´ Fm. The latter unit, of well established glacial origin (Isotta et al. 1969), crops out some 400 km north of the study area (Fig. 50.1; see Dardenne et al. (1978) and RochaCampos & Hasui (1981) for a fuller account of these occurrences). At present, there are inconsistencies in the interpretation of the nomenclature, stratigraphic relations and depositional setting of the strata lying below the Sete Lagoas Fm. (Bambuı´ Group), including the Moema Laminites. Pierre Muzzi Magalha˜es, for example, in his Masters dissertation of 1989, recognizes a Conglomeratic Facies (polymitic) and a Pellitic Facies (argillite and siltstone), below carbonate

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 535– 540. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.50

536

A. C. ROCHA-CAMPOS ET AL.

Fig. 50.1. Geological map of the Sa˜o Francisco Basin, Southeastern Brazil, showing location of the best exposures of the Moema Laminites. Squares: 1, Formiga outcrop; 2, SAFFRAN quarry. SFC, Sa˜o Francisco Craton.

rocks of the Sete Lagoas Fm., that are probably equivalent to the Moema Laminites, as a basal unit of the Bambuı` Group (Fig. 50.2). A glacial influence during deposition of these rocks was, however, not recognized. On recent geological maps from the southern part of the Sa˜o Francisco Basin (SEME/COMIG/ CPRM, 2003; CPRM, 2007), rocks considered by us to be part of the Moema Lamites have been called the Serra de Santa Helena Fm. and Sambura´ Fm. The former is a pelitic unit lying above the Sete Lagoas Fm., and the latter corresponds to one of the infra-Bambuı´ Group diamictite occurrences. As pointed out by Dardenne & Castro (2000), the Sambura´ diamictite is a fault-controlled, fan-delta deposit that occurs at different levels of the Bambui Group. Under these circumstances, the use of a provisional name seems more advisable.

Glaciogenic deposits and associated strata The sedimentary characteristics of the Moema Laminites are based on the best exposures in the region: the Formiga area and the SAFFRAN quarry.

Formiga area A cut on the northeastern side of the road BR-364 exposes a lower, discontinuous, thin diamictite of variable thickness (30 –40 cm), resting non-conformably on topographic highs of Archaean orthogneiss or on its regolith, followed with a sharp contact by 15 m of laminites. The smooth surface of the gneiss bears fine, poorly preserved striae (Fig. 50.3). Isolated or semi-detached round boulders of up to 1.5 m lie on top of the orthogneiss and regolith and seem to correspond to corestones (Fig. 50.3). Deep weathering of local rocks has altered their colours, thus enhancing internal features. The diamictite corresponds to a layer of tectonically deformed regolith of the local basement, containing many centimetreto-decimetre clasts, predominantly of the local gneiss, but also of two different lithologies (fine and coarse quartz schist), in a sandy-silty-clayey matrix. Deformation features recognized in the diamictite include boudins, stretched clasts, attenuated and transposed folds and clast alignment. Small thrust faults cut the top of the gneissic basement (Fig. 50.3). The contact between the deformed bed and overlying laminites seems to be sharp. Two subunits are recognized, a lower one of

MOEMA LAMINITES

Serra de Santa Helena Fm. Sete Lagoas Fm.

Bambuí Group

Sandstones

Basement

Mudstones and Siltstones

Stromatolitic structures

537

Thin laminations are present in the lower subunit. Normal graded bedding was observed in centimetre-thick laminae of the upper subunit, as was evidence of local disharmonic, soft sediment folding. Unequivocal evidence of dropstones was not found in the laminites.

Dolomitic calcarenite

Calcarenite with parallel and cross bedding

SAFFRAN quarry

Calcilutite with parallel and cross bedding

At least 60 m of thin (cm) silty-clayey laminites of kaolinitic composition and interbedded, irregular, laterally discontinuous, ferruginous (hematite) bands are exposed on the extraction faces of the quarry (Fig. 50.4). A diamictite bed occurs under the laminites nearby, but its contact is not exposed. The upper boundary with the laminites was not seen in this area. The section is made up of a lower interval where ferruginous beds are more common and an upper thicker, whiteish one, where they are rare (Fig. 50.4). The two intervals differ also in their sedimentary structures. Whereas the lower interval is characterized by intercalation of thin silt-clay lenses (starved ripples?) and normal graded and nongraded laminae, the whiteish clay bands of the upper interval are mostly thinly laminated. A bedding plane at the bottom of the quarry contains two areas bearing sets of parallel, few-metres-long, fine striae and furrows (1 –3 cm wide). The features are rectilinear or curved and their orientation varies from 2898N to 3398N. On the same surface, there are a few examples of striated areas surrounded on both sides by more elevated non-striated ones on the same bedding plane. Rare features, such as masses of sediment laterally covering furrows (Rocha-Campos et al. 1994), suggest that the bed was not consolidated when the striae were produced.

Siltstone

200 m

0

Fig. 50.2. Stratigraphy of the Bambuı´ Group in Southern Sa˜o Francisco Basin. Conglomeratic (not shown) and pellitic facies (siltstone) at the base of the Bambuı´ Group is thought to correspond to the Moema Laminites (Muzzi Magalha˜es, pers. comm. 1989).

dark laminites, 2 m thick, overlain with a sharp contact by a section of light laminites, 12 m thick (Fig. 50.3). The lower interval seems to fill a wide and shallow depression on top of the deformed bed and orthogneiss. Laminae are deformed around corestones of weathered gneiss.

Fig. 50.3. Partial diagrams (a and b) of the Formiga outcrop. Deformation, thought to be glaciotectonic in origin includes stretched clasts, attenuated folds, boudins and comminution (A –C). 1, Orthogneiss (Archaean); 2, regolith; 3, deformed bed (regolith and lower dark laminites); 4, upper light laminites. Note lower dark laminites (unit 3) in (a) are undeformed on the left side of the outcrop.

538

A. C. ROCHA-CAMPOS ET AL.

Top

// 2

Explanation Laminites Ferruginous layer

1

Interbedded laminites and ferruginous layers

0 .m

Fractures/faults

Base

Boundary relations with overlying and underlying non-glacial units Rocks of the Moema Laminites rest non-conformably on an erosive surface formed on top of predominantly granitic terranes of the Archaean basement of the Sa˜o Francisco Basin (Fig. 50.2). The upper contact with the overlying Sete Lagoas Fm. carbonates was not directely observed in the studied area.

Chemostratigraphy There are no known carbonate rocks associated directly with the Moema Laminites and, as such, no chemostratigraphic studies of C, O or Sr have been undertaken to date. Eight samples of laminites from the two areas studied were geochemically analysed to better characterize the two rocks and to

Striae

Fig. 50.4. Stratigraphic log of laminites at the SAFFRAN quarry. showing distribution of ferruginous layer (in black).

contribute to the understanding of their origin. This involved determination of major (FRX) and trace elements (FRX and ICP-MS), REE (ICP-MS), content of Fe2þ and identification of mineralogical phases by X-ray diffraction. REE data were normalized to NASC (North American Shale Composite). Clastic components of the laminites (silt-clay and kaolinite) present a homogeneous composition and chemical signature of more abundant weathering-resistant elements (Cr, Ga, Hf, Rb Sc, Ta, Th and Zr), in comparison with the ferruginous layers (up to 48% hematite). Both clastic and ferruginous layers depict a negative anomaly for Eu and a positive anomaly for Ce.

Other characteristics Laminites from the SAFFRAN Quarry are being exploited as a source of kaolinite for the ceramic industry.

MOEMA LAMINITES

Palaeolatitude and palaeogeography There are no palaeomagnetic data from the Moema Laminites. Palaeomagnetic studies carried out previously on carbonate units of the Sa˜o Francisco Basin revealed that a continental-scale fluid event could have been responsible for the resetting of the palaeo´ Agrella magnetic, as well as the isotopic system, at c. 520 Ma (D Filho et al. 2000; Trindade et al. 2004). Recently, however, Babinski et al. (2007) referred to good-quality palaeomagnetic poles obtained from rocks of the Congo –Sa˜o Francisco craton, penecontemporaneous with the Sete Lagoas Fm., that would place the southern Sa˜o Francisco craton at low-intermediate palaeolatitudes (20– 308).

Geochronological constraints No direct radiometric data are yet available to constrain the depositional age of the Moema Laminites. However, the regional stratigraphic position of other units with similar lithologies below carbonate beds of the Sete Lagoas Fm. along all sections examined, and evidence of a glacial influence in the Moema Laminites, are suggestive of correlation between the Moema Laminites and the Jequitaı´ Fm., which has a cap carbonate dated at 740 + 22 Ma (Babinski et al. 2007). Although the Moema Laminites were deposited on top of the Archaean basement of the Sa˜o Francisco Basin at both localities discussed (Teixeira et al. 2000), Sm –Nd TDM ages yielded by rocks from the Formiga area range from 1.8 to 1.98 Ga. The results indicate that besides the basement, other younger, probably Palaeoproterozoic terranes, may also have contributed as a source of sediments. A similar picture was obtained for samples of laminites and ferruginous bands in the SAFFRAN Quarry. The 147Sm– 144Nd of these rocks are usually very fractionated, with values higher than 0.14, especially for the ferruginous bands, where they can reach values of 0.18. Samples with 147Sm– 144Nd close to 0.12 (normal crustal value) yielded Sm–Nd TDM ages of 1.6–1.9 Ga, again suggesting a much younger source area for the deposits, besides the Archaean basement. On the basis of these data, the age of deposition of the Moema Laminites cannot be older than Palaeoproterozoic.

Discussion Several erosional and depositional features at the Formiga and SAFFRAN quarry outcrops support the existence of glacial influence in this part of the Sa˜o Francisco Basin. At Formiga they include (i) a smooth surface of round topographic highs of orthogneiss, bearing striae and (ii) the presence in the diamictite and gneissic basement of deformation features (Fig. 50.3). These structures are interpreted as having resulted from ductile deformation of the basement rocks (regolith) under a simple shear regimen. Small thrusts affecting the orthogneiss point to brittle behaviour of the rock. Vergence of the features indicates a direction of strain oriented from west to east. Deformation seems to have varied in intensity along the upper surface of the basement, as demonstrated by changes in thickness of the deformed bed, size and shapes of clasts and distribution of deformation features. As a rule, regolith was deformed on topographic highs of the orthogneissic substrate, but not in depressions between them. Such features are commonly found associated with sub-glacial, glacial –tectonic deformation of Recent and Pleistocene glacial deposits (e.g. Bennett & Glasser 1997; Hart 1998), which results in the generation of deformation tills. The same explanation is applicable to the diamictite at the Formiga outcrop. Glacial displacement caused fracturing and small-scale thrusting of the substrate and abrasion, resulting in smooth round surfaces on top

539

of basement highs. Deformation seems to have been associated with lodgement of clasts. The glacier may have moved towards the east over the weathered, deformable bed, in a glacial terrestrial setting that was subsequently inundated following glacier retreat and deglaciation. A second possibility is that a grounded glacier advanced into a preexisting water-filled basin, where laminites had accumulated. The fact that the laminites were not involved in the deformation and that the glacier caused deformation of the regolith favours the first hypothesis (Fig. 50.3). Deposition of the laminites was probably through settling of silt and clay, from suspended plumes fed by the glacier, as well as by turbidity or other types of density currents during the deglaciation phase. Instability of the unconsolidated deposit led to frequent soft-sediment deformation. Lack (or rarity ?) of dropstones in the laminites may indicate that the glacial terminus was far removed from the basin margin or the intervention of mechanisms that either prevent the occurrence of icebergs or result in their fast evacuation, as described from present glacial estuarine environments (Dowdeswell et al. 2000). Sedimentation processes identified in the two localities studied seem to have been grossly similar, as shown by textures and sedimentary structures present. No ice-contact features were found at the SAFFRAN quarry. Interpretation of glacial influence in the area relies only on the presumed origin of the striated laminites by erosion of floating ice masses touching the bottom of the sedimentary basin. Striae on bedding plane of laminites at the SAFFRAN quarry and characteristics such as bending or change of direction are compatible with their origin through scratching of unconsolidated bottom sediments by free-floating ice masses (Woodworth-Lynas & Dowdeswell 1994; Rocha-Campos et al. 1994), either free sea-lake ice or icebergs. The nature of the water body in which the silty-clayey laminites accumulated (either a lake or sea) is difficult to ascertain. The extensive distribution of the pellitic strata in the Sa˜o Francisco Basin seems to favour a marine setting. Geochemical data indicate that Fe from ferruginous layers of the laminites originated from weathering of basic rocks, under more reducing conditions than the present ones and not from the Late Proterozoic and Archaean banded iron formations (BIFs) that crop out in Central Minas Gerais State. Source for kaolinite was probably deeply weathered feldspar-rich Archaean gneissic terrane. Geochemical data also suggest a strong fluviatile contribution in the transportation of sedimentary particles to the depositional site (Graf et al. 1994; Kato et al. 1998). These features denote differences in source areas for the particles and in palaeoenvironmental setting between the Formiga outcrop and SAFFRAN Quarry areas. A more terrestrially influenced (less saline?) environment could then be envisaged for laminites at the quarry. Data from the literature and from this study concur in showing that the Moema Laminites differ in stratigraphic position and lithological composition from the pelitic Serra de Santa Helena Fm., situated above the basal carbonates (Sete Lagoas Fm.) of the Bambuı´ Group, in which they have been included in recent geological maps of the area. Additional fieldwork is needed to clarify the relations of the Moema Laminites with other isolated occurrences of infra-Bambuı´ Group diamictite outcrops, including the Jequitaı´ Fm., known in the Sa˜o Francisco Basin (Dardenne et al. 1978; Rocha-Campos & Hasui 1981), and to establish their mode of origin. Possible correlation among them would significantly enlarge the area known to have been affected by the Early Cryogenian glaciation in Southeastern Brazil. We thank F. M. Canile for her help with the preparation of this text and figures and I. McReath for kindly revising the English language. The original text was much improved by reviews from G. Young and E. Arnaud. This represents a

540

A. C. ROCHA-CAMPOS ET AL.

contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Alkmin, F. F. 2004. O que faz de um cra´ton um cra´ton? O Cra´ton do Sa˜o Francisco e as revelac¸o˜es almedianas ao delimita´-lo. In: Mantesso Neto, V., Bartorelli, A., Carneiro, C., dal Re´, & Brito Neves, B. B. de (organizers) Geologia do Continente Sul-Americano. Beca, 19 –35. Alkmin, F. F. & Marshak, S. 1998. Transamazonian Orogeny in the Southern Sa˜o Francisco Craton Region, Brazil: evidence for Paleoproterozoic collision and collapse in the Quadrila´tero Ferrı´fero. Precambrian Research, 99, 29 –58. Babinski, M., Vieira, L. C. & Trindade, R. I. F. 2007. Direct dating of the Sete Lagoas cap carbonate (Bambuı´ Group, Brazil) and implications for the Neoproterozoic glacial events. Terra Nova, 19, 401– 406. Bennet, M. R. & Glasser, N. F. 1997. Glacial Geology, Ice sheets and Landforms. Wiley. Costa, M. T. da & Branco, J. J. R. 1961 (eds). Introduc¸a˜o, Roteiro para a Excursa˜o Belo Horizonte-Brası´lia. Contribuic¸a˜o ao XIV Congresso Brasileiro de Geologia, Belo Horizonte, Instituto de Pesquisas Radioativas, 15, 9– 25. CPRM – Servic¸o Geolo´gico do Brasil, 2007. Mapa Geolo´gico, Folha Guape´, 1:100.000. D’Agrella Filho, M. S., Babinski, M., Trindade, R. I. F., Van Schmus, W. R. & Ernesto, M. 2000. Simultaneous remagnetization and U– Pb isotope resetting in Neoproterozoic carbonates of the Sa˜o Francisco craton, Brazil. Precambrian Research, 99, 179– 196. Dardenne, M. A. & Castro, P. T. A. 2000. The sedimentology, stratigraphy and tectonic context of the Sa˜o Franscisco Group at the southwest boundary of the Sa˜o Francisco craton, Brazil. Revista Brasileira de Geocieˆncias, 30, 439–441. Dardenne, M. A., Faria, A., Magalha˜es, L. F. & Soares, L. A. 1978. O tilito da base do Grupo Bambuı´ na borda ocidental do cra´ton do Sa˜o Francisco. Sociedade Brasileira de Geologia, Nu´cleo Centro-Oeste, Boletim Informativo, 7/8, 85 –97. Dowdeswell, J. A., Whittington, R. J., Jennings, A. E., Andrews, J. T., Mackensen, A. & Marienfeld, F. 2000. An origin for laminated glacimarine sediments through sea-ice build-up and suppressed iceberg rafting. Sedimentology, 47, 557– 576. Graff, J. L., Jr., O’Connor, E. A. & Van Leewen, P. 1994. Rare earth element evidence of origin and depositional environment of Late Proterozoic ironstone beds and manganese-oxide deposits, SW Brazil

and SE Bolivia. Journal of South American Earth Sciences, 7, 115– 133. Hart, J. 1998. The deforming bed/debris rich basal ice continuum and its implications for the formation of glacial landforms (flutes) and sediments (melt-out till). Quaternary Science Reviews, 17, 737– 754. Isotta, C. A. L., Rocha-Campos, A. C. & Yoshida, R. 1969. Striated pavement of the Upper Pre-Cambrian glaciation in Brazil. Nature, 222, 466– 468. Kato, Y., Ohta, I., Tsunematsu, T., Watanabe, Y., Isozaki, Y., Maruyama, S. & Imai, N. 1998. Rare earth element variation in midArchean banded iron formations: implications for the chemistry of ocean and continent and plate tectonics. Geochimica et Cosmochimica Acta, 62, 3475– 3497. Miranda, J. 1953. Folha Bambuı´-Sec¸a˜o de Geologia. Departamento Nacional da Produc¸a˜o Mineral, Divisa˜o de Geologia e Mineralogia, Relato´rio Anual do Diretor, 1925, 25 – 27. Rocha-Campos, A. C. & Hasui, Y. 1981. Proterozoic diamictites of western Minas Gerais and east Goia´s, central Brazil. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 920–923. Rocha-Campos, A. C., Santos, P. R dos & Canuto, J. R. 1994. Ice scouring structures in Late Paleozoic rhythmites, Parana´ Basin, Brazil. In: Deynoux, M., Miller, J. M. G., Domack, E., Eyles, N., Fairchild, I. J. & Young, G. M. (eds) Earth’s Glacial Record. Cambridge University Press, Cambridge, 234– 240. SEME (Secretaria de Estado de Minas e Energia)/COMIG (Companhia Mineradora de Minas Gerais)/CPRM (Servic¸o Geolo´gico do Brasil), 2003. Projeto Sa˜o Francisco, Geologia e Hidrologia, Text and Maps, 1:250.000 (PDF and CD-ROM). Teixeira, W., Sabate´, P., Barbosa, J., Noce, C. M. & Carneiro, M. A. 2000. Archean and Paleoproterozoic tectonic evolution of the Sa˜o Francisco Craton. In: Cordani, U. G., Milani, E. J., Thomaz Filho, A. & Campos, D. A. (eds) Tectonic Evolution of South America. Sociedade Brasileira de Geologia, Sa˜o Paulo, 101– 137. Trindade, R. I. F., D’Agrella-Filho, M. S., Babinski, M., Font, E. & Brito Neves, B. B. 2004. Paleomagnetism and geochronology of the Bebedouro cap carbonate: evidence for continental-scale Cambrian remagnetization in the Sa˜o Francisco craton, Brazil. Precambrian Research, 128, 83– 103. Woodworth-Lynas, C. M. T. & Dowdeswell, J. A. 1994. Softsediment striated surfaces and massive diamicton facies produced by floating ice. In: Deynoux, M., Miller, J. M. G., Domack, E., Eyles, N., Fairchild, I. J. & Young, G. M. (eds) Earth’s Glacial Record. Cambridge University Press, Cambridge, 241–259.

Chapter 51 The glaciogenic Jequitaı´ Formation, southeastern Brazil A. UHLEIN1*, C. J. S. ALVARENGA2, M. A. DARDENNE2 & R. R. TROMPETTE3 1

Universidade Federal de Minas Gerais – UFMG, Dept. de Geologia, Av. Antoˆnio Carlos, 6627, Campus Pampulha, Belo Horizonte, Minas Gerais, 31270-901, Brazil 2

Universidade de Brası´lia – UnB, Inst. de Geocieˆncias, Campus, Asa Norte, 70910-900, Brası´lia (DF), Brazil 3

35 rue Pascal, 75013, Paris, France

*Corresponding author (e-mail: [email protected]) Abstract: Glaciogenic deposits of the Jequitaı´ Formation (Fm.) are well exposed along the margins of the Serra do Cabral on the Sa˜o Francisco Craton, southeastern Brazil. The Jequitaı´ Formation is thin (0–150 m thick), lenticular and overlies the Espinhac¸o Supergroup on a discrete unconformity. Sandstones show subglacial erosional structures such as grooved and striated pavements oriented ENE– WSW. The Jequitaı´ Fm. consists of massive and stratified diamictites with granules, pebbles and boulders of gneiss, granite, quartzite and carbonate. At the base, the diamictites are massive, whereas the upper part contains many alternating beds of clast-rich and -poor diamictites. They also contain discontinuous, fine-grained sandstones and a few laminated siltstone–mudstone intercalations. This diamictite association indicates glaciomarine sedimentation. The Jequitaı´ Fm. covers the Sa˜o Francisco cratonic domain and its equivalent extends eastward over the Arac¸uaı´ fold belt where it is part of the metasedimentary Macau´bas Group, a thick Neoproterozoic unit with metadiamictites, quartzites and schists. The diamictite–turbidite association of the Macau´bas Group was deposited on the border of the Pan-African– Brasiliano rift as gravity flows.

Three distinct glaciogenic units are present in the Neoproterozoic of Brazil (Alvarenga & Trompette 1992; Nogueira et al. 2003; Alvarenga et al. 2007). The older and more widespread unit is probably of middle Cryogenian age (665 –710? Ma). It is well known on the Sa˜o Francisco Craton where it is represented by the Jequitaı´ Fm. at the base of the Bambuı´ Group in the Sa˜o Francisco basin (Rocha-Campos & Hasui 1981; Karfunkel & Hoppe 1988; Trompette 1994) and as the Bebedoura Fm. in the Una Group, which occurs in a series of segmented sub-basins in the NE of the craton (Guimara˜es et al. 2011). The Jequitaı´ Fm. occurs in the Serra do Cabral region (Fig. 51.1), at the transition between the Sa˜o Francisco craton and the external zone of the Arac¸uaı´ fold belt, in southeastern Brazil (Minas Gerais State). Along the Brası´lia fold belt at the western margin of the Sa˜o Francisco Craton (Goia´s State), these glaciogenic rocks are also exposed in the Cristalina area. The Jequitaı´ Fm. constitutes the base of the Sa˜o Francisco Supergroup, a cover sequence of the Sa˜o Francisco craton. It crops out at the lower part of the Bambuı´ Group along the margins of the Serra do Cabral region, overlying the Palaeo- to Mesoproterozoic Espinhac¸o Supergroup. The Jequitaı´ Fm. was recognized early on as being of glacial origin, first by Branner (1919) and Moraes & Guimara˜es (1930). Subsequently, Isotta et al. (1969) discovered a striated pavement and confirmed the glacial origin.

Structural framework Brasiliano – Pan-African orogenic belts surround the southern Sa˜o Francisco craton: the Brası´lia fold belt to the west and the Arac¸uaı´ fold belt to the east (Brito Neves et al. 1999; Alkmim et al. 2001). The Jequitaı´ Fm. occurs on the Sa˜o Francisco craton, with the type section in the Serra do Cabral region (Fig. 51.1). Meso- to Neoproterozoic rocks form the sedimentary cover of the Sa˜o Francisco Craton, the Sa˜o Francisco basin, a sub-horizontal sequence that was deformed at 560 –550 Ma near the cratonic margins (Uhlein et al. 1998; Alkmim et al. 2006). The Jequitaı´ Fm. covers the cratonic domain and extends eastward into the Arac¸uaı´ fold belt and the metasedimentary Macau´bas Group (Fig. 51.2), a thick Neoproterozoic unit with metadiamictites, quartzites and

schists (Karfunkel & Hoppe 1988; Uhlein et al. 1999; Pedrosa Soares et al. 2001, 2011). The Jequitaı´ Fm. crops out in the Serra do Cabral Anticlinorium, extending eastward under the Bambuı´ Group and reappearing much thicker due to thrust duplication in the Arac¸uaı´ fold belt (Figs 51.1 & 51.2). Tectonism and metamorphism related to the Brasiliano orogeny increases in intensity eastward, reaching epizonal metamorphism. The Arac¸uaı´ fold belt (Macau´bas Group) shows west-vergent thrusts and folds, and greenschist- to amphibolites-facies metamorphism (Fig. 51.2). Over the cratonic domain, rocks of the Mesoproterozoic Espinhac¸o Supergroup are overlain by glaciogenic rocks of the Jequitaı´ Fm. in the Serra do Cabral region (Walde 1978; Uhlein 1991), where they are exposed in large brachy-anticline structures. To the north, near the town of Jequitaı´, ´ gua Fria region displays an anticline structure with the Serra da A sandstones overlain by Jequitaı´ Fm. diamictites and siltstones and carbonates of the Bambuı´ Group (Figs 51.1–51.3).

Stratigraphy The Sa˜o Francisco Basin includes the Espinhac¸o Supergroup and the unconformably overlying Sa˜o Francisco Supergroup, with the Jequitaı´ Fm. at the base overlain by marine siliciclastic and carbonate sediments of the Bambuı´ Group. The Espinhac¸o Supergroup consists of metaquartzites and metasiltstones probably deposited c.1500 Ma (Uhlein et al. 1998; Martins Neto 2000). They are cut by Neoproterozoic tholeiitic mafic dykes that do not cut the overlying Sa˜o Francisco Group. The Jequitaı´ Fm. is thin (0 –150 m thick) and lenticular and overlies the Espinhac¸o Supergroup, on a discrete unconformity, along the margins of the Serra do Cabral. To the north, at the ´ gua Fria, between the towns of Jequitaı´ and Francisco Serra da A Dumond, the basal sandstones show striated pavements (Isotta et al. 1969; Karfunkel & Karfunkel 1976; Rocha-Campos & Hasui 1981; Karfunkel & Hoppe 1988). Grooves and numerous striae are oriented ENE – WSW (1008). The overlying Bambuı´ Group can be subdivided into two major units in the southeastern part of the Sa˜o Francisco basin (Dardenne 1978; Martins 1999). The lower unit is characterized by lenticular

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 541– 546. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.51

542

A. UHLEIN ET AL.

Fig. 51.1. Simplified geological map showing location of the Jequitaı´ Fm. on the Sa˜o Francisco Craton. Modified from Uhlein et al. (1999) and Pedrosa Soares et al. (2001).

carbonates alternating with fine clastics sediments. The limestones and dolomites are commonly oolitic, oncolitic and stromatolitic, and show abundant cross-bedding. The upper unit, the Treˆs Marias Fm., consists of green or brown fine-grained arkoses interlayered with greenish argillaceous siltstones (Uhlein et al. 2004). Three lithostratigraphic units are recognized in the Arac¸uaı´ fold belt: a polycyclic basement, the Palaeo – Mesoproterozoic Espinhac¸o Supergroup, and the Neoproterozoic Macau´bas Group. The latter consists of metadiamictites with pebbles and boulders, some of which are striated and polished (Hettich 1977; Karfunkel & Hoppe 1988; Uhlein et al. 1998, 1999; Pedrosa Soares et al. 2001). The lateral continuity between the Jequitaı´ Fm. into the Macau´bas Group (Fig. 51.2) was originally suggested by Karfunkel & Karfunkel (1976), Karfunkel & Hoppe (1988) and Uhlein et al. (1998, 1999) based on the geographical distribution of facies, the variation of thicknesses, and the rare indication provided by glacial abrasion features on the craton. The Macau´bas Group contains metadiamictites grading vertically and laterally

eastward to rhythmic facies, with metamorphosed sandstone, siltstone and shale intercalations (Fig. 51.2).

Glaciogenic deposits and associated strata The Jequitaı´ Fm., near its type location, is 0 –150 m thick and comprises a diamictite-sandstone-pelite facies association (Figs 51.3 & 51.4). It consists of massive diamictites, as well as very rare and narrow centimetre- to metre-scale alternations of sandstones and pelites (Uhlein et al. 1999; Martins-Neto & Hercos 2002). The diamictites contain granules to large boulders, which are angular to sub-rounded and composed of sandstones, carbonates, pelites, granites and gneisses, vein quartz and volcanic rocks. Carbonate clasts are more common towards the base, whereas crystalline basement clasts dominate towards the top of the Jequitaı´ Fm. (Karfunkel & Hoppe 1988). Coarse material was probably derived from the underlying basement and the Espinhac¸o

Fig. 51.2. Schematic geological cross-section of the (a) Sa˜o Francisco Basin and (b) Arac¸uaı´ Fold Belt. Modified from Uhlein et al. (1999).

THE GLACIOGENIC JEQUITAI´ FORMATION

543

´ gua Fria, Fig. 51.4. Measured vertical sections in Jequitaı´ Fm., Serra da A around Jequitaı´ town, Minas Gerais State. See Figure 51.3 for locations of the sections.

Fig. 51.3. Geological map of the Jequitaı´ region, Minas Gerais State, with locations of the studied sections. Modified from Cukrov et al. ( 2005). See Figure 51.1 for location.

Supergroup, which contains a few limestone and dolostone beds. The diamictites are massive, sheet-like to lenticular bodies, with a high variation in clast/matrix ratio. In the lower portions, the diamictites are massive and show very little variation in clast size or clast/matrix ratio. The upper portions contain many alternating beds of clast-rich and clast-poor diamictite (Fig. 51.4). Clasts float in a texturally and compositionally immature sandstone –siltstone matrix with clay minerals, calcite and sericite. Some diamictite clasts are scratched, polished and faceted (Isotta et al. 1969; Karfunkel & Hoppe 1988). The stratified diamictites also contains discontinous sandstones and a few intercalated laminated pelites. The sandstones are massive (facies Sm in Fig. 51.4), graded or horizontally stratified, fine- to coarse-grained, sheet-like to lenticular bodies, poorly sorted, and overall upward fining. They are interpreted as

subaqueous mass-flow deposits and turbidity currents. Crossstratified sandstones (facies Sc in Fig. 51.4) are fine- to coarsegrained. The diamictites also contain spherical or ovoidal, discontinous, 10–12 m-thick, fine- to medium-grained, massive sandstones intercalations variably interpreted as palaeo-eskers by Karfunkel & Karfunkel (1976) and as large rafts of the Espinhac¸o substratum (Martins-Neto & Hercos 2002). The pelites are composed of millimetre-scale siltstone and claystone alternations (rhythmites) or structureless mudstones, intercalated with the diamictites (Fig. 51.4), and related to low-density turbidity currents or vertical accretion. ´ gua Fria, 10 km SE of Jequitaı´, diamictites of In the Serra da A the Jequitaı´ Fm. rest directly on scratched, grooved and polished pavements developed on sub-horizontal sandstones, interpreted as basement rocks belonging to the Espinhac¸o Supergroup (Isotta et al. 1969; Karfunkel & Hoppe 1988; Uhlein et al. 1999). The erosional features vary from very fine scratches to grooves up 20 cm wide and 5 cm deep (Isotta et al. 1969). They show Vand U-shaped grooves, and crescent fractures have been observed. These features, which have an azimuth of 1008 indicate roughly east – west ice palaeoflow. These glacial abrasion scars are interpreted as land-based and the diamictites are interpreted as tillites resting directly on striated pavements (Isotta et al. 1969; RochaCampos & Hasui 1981; Karfunkel & Hoppe 1988). Striated and faceted clasts, varvites, inter-tillite striated pavements and icethrust features have been identified (Hettich & Karfunkel 1978; Gravenor & Monteiro 1983). These observations indicate that the glaciers that deposited the tillites were continental ice sheets (Isotta et al. 1969; Karfunkel & Karfunkel 1976; Rocha-Campos & Hasui 1981; Karfunkel & Hoppe 1988). Rocha-Campos et al. (1996) suggested a soft-sediment origin for glacial striae and grooves found on top of the sandstones of ´ gua Fria, near Jequitaı´. The main lines of evidence the Serra da A are (i) striae inside grooves laterally covered by slumped plough ridges; (ii) clasts inside furrows partially embedded in the sandstone; (iii) sinuosity of furrows; (iv) crescent-like skip marks and ridges transverse to furrows; (v) occurrence of striated surfaces on two bedding planes that are separated by a 25-cm-thick bed

544

A. UHLEIN ET AL.

of quartzite; (vi) contiguous striated and ripple-marked areas on the same bedding plane; ripple marks a few centimetres below the striated surface are, however, not deformed. This re-interpretation indicates that the striated sandstones may be part of the Jequitaı´ Fm. rather than the Espinhac¸o Supergroup. At the same time, the terrestrial setting of the Jequitaı´ Fm. has been questioned. Uhlein et al. (1999) pointed to the scarcity of outwash deposits around Jequitaı´ to support reinterpretation of the monotonous, massive, clast-poor diamictites as representing glaciomarine sedimentation. Rocha-Campos et al. (1996) reinterpreted the glacial abrasion marks as having been formed in the fluctuating grounding zone of a marine ice sheet or tongue just grazing the sediment surface. Finally, Martins-Neto & Hercos (2002) describe wide grooves with inner steps that yield the palaeo-direction of glacier movement (2858 azimuth) on the ´ gua Fria, suggesting a more western border of the Serra da A complex dispersion pattern. The metadiamictites of the Macau´bas Group in the Arac¸uai fold belt contain clasts from granule to boulder size in a dominantly muddy and/or sandy matrix. The clasts are mainly quartzites, granitoids and gneisses, limestones, quartz and schists. Clasts are commonly tectonically stretched. Clast-rich and clast-poor beds of diamictites alternate and commonly show different contents of sandy and muddy matrix. The diamictite facies are interpreted as subaqueous debris flows deposited along basin margins during a period of tectonic activity. Two main sequences of metadiamictites, 300–2000 m thick, are separated by an interval of metarhythmites. The latter consist of parallel-bedded, centimetre- to metre-scale schists, metasiltstones, calc-silicate rocks and quartzites, which in places show normal grading. Hence, the Macau´bas Group is interpreted as resedimented glacial material deposited by subaqueous debris flow and turbidity currents (Uhlein et al. 1999). From west to east, a depositional model can be proposed for the Jequitaı´ – Macau´bas units: a cratonic diamictite sequence interpreted as glaciomarine (Jequitaı´ Fm.) grading to a diamictite –turbidite sequence deposited on the border of a Neoproterozoic rift and subsequently passive margin basin, the precursor to the Arac¸uaı´ fold belt (Uhlein et al. 1998, 1999; Pedrosa Soares et al. 2001, 2011).

Boundary relations with overlying and underlying non-glacial units In the Serra do Cabral region, the lower contact is a very slight angular unconformity and an erosional disconformity separating the Jequitaı´ Fm. from the underlying Espinhac¸o Supergroup (Walde 1978; Uhlein 1991). Palaeotopographical relief of 300 to 400 m is developed on this contact. The lenticular character of the Jequitaı´ Fm. along the margins of the Serra do Cabral can be related to infilling of palaeotopography. Block tilting, indicated by angular relationships across the Serra do Cabral, reflects Neoproterozoic extensional tectonics related to Rodinia break-up (Uhlein et al. 1999; Martins-Neto & Hercos 2002). The Jequitaı´ Fm. is conformably overlain by metasiltstone of the Bambuı´ Group in the Serra do Cabral region. A hiatus and a discontinuity between the Jequitaı´ Fm. and the Bambuı´ Group can be recognized (Dardenne 1978; Martins 1999).

Chemostratigraphy The lack of biostratigraphic controls and suitable minerals for radiometric age measurements make chemostratigraphy a potential alternative for the establishment of a reliable isotope stratigraphic profile for Neoproterozoic sequences. Meanwhile, in the Serra do Cabral region, the glaciomarine diamictites of the Jequitaı´

Fm. are overlain by siltstones and shales of the Bambuı´ Group, which represent deep-marine platform deposits. Thus it is not possible to recognize a capping carbonate overlying the glacial rocks of ´ gua the Jequitaı´ Fm. in the Serra do Cabral and Serra da A Fria regions. C- and O-isotopic studies have been carried out in Sete Lagoas Fm., basal unit of the Bambuı´ Group in southern Sa˜o Francisco Craton (Santos et al. 2000, 2004; Vieira et al. 2005, 2007), around the cities of Sete Lagoas and Belo Horizonte. The Sete Lagoas Fm. shows sea-floor precipitates with aragonitic pseudomorphs (Peryt et al. 1990) and negative d13C values (c. – 4‰), which permit the characterization of a basal carbonate unit of the Bambuı´ Group as a capping carbonate (Vieira et al. 2007). The Sete Lagoas Fm. displays two upward-shallowing megacycles, with very high d13C (10‰) in the second megacycle (Vieira et al. 2007). Sr-isotopic compositions of least altered carbonates in the Bambguı´ Group range from 0.7074 to 07.7076 (reviewed in Misi et al. 2007).

Mineralization and other characteristics The Macau´bas Group contains economic ferruginous and manganiferous deposits in the Arac¸uaı´ Fold Belt. The Fe deposits form the Riacho dos Poc¸o˜es Member of the Nova Aurora Fm., between Peixe Bravo and Vacaria Rivers (Viveiros et al. 1978). Manganiferous levels originate from the lateritic weathering of rhythmites and schists. Ferruginous diamictites, hematite quartzites and banded iron formations (BIFs), rhythmites, and hematite schists form the ferruginous facies of the Macau´bas Group. They were deposited in the transition zone between a glaciated continental domain and a relatively deep marine environment with turbidity currents. Locally, mafic lavas with pillow structures are intercalated in the Macau´bas Group, suggesting that hydrothermal activity in the basin may have been important.

Palaeolatitude and palaeogeography Palaeomagnetic studies on basic dykes in the eastern Sa˜o Francisco Craton (D’Agrella Filho et al. 1990) place the southern part of the craton between 408 and 608 latitude at c. 1.0 Ga, but there are no available data to constrain palaeolatitudes during deposition of the Jequitaı´ Fm. or Bambuı´ Group. The Jequitaı´ glaciation occurred on the elevated Rodinia supercontinent, where larger areas may have been covered by a continental ice sheet or several ice caps (Trompette 1994).

Geochronological constraints Radiometric age determinations on the Sa˜o Francisco craton provide few concrete age constraints on glaciogenic diamictites of the Jequitaı´ Fm. and equivalent units. Rb –Sr dating of clay minerals yields ages between 932 + 30 Ma and 570 + 7 Ma (Macedo & Bonhomme 1984) and their geological interpretation remains difficult. Badelleyite from mafic rocks intruded into the Espinhac¸o Supergroup yield an age of 906 + 2 Ma (Machado et al. 1989), while the youngest concordant detrital zircon in the Macau´bas Group is 864 + 30 Ma (Pedrosa Soares et al. 2000). Rocks inferred to represent related oceanic crust in the Macau´bas Group have produced a Sm– Nd age of 816 + 72 Ma (Pedrosa Soares et al. 2001) and a 660 + 29 Ma U –Pb age by LAICPMS of a plagiogranite (Pedrosa Soares et al. 2011). The Sete Lagoas Fm., which overlies the Jequitaı´ Fm. on the Sa˜o Francisco craton, has been dated at 740 + 22 Ma by the Pb/Pb carbonate method (Babinski et al. 2007).

THE GLACIOGENIC JEQUITAI´ FORMATION

Discussion Based on the different characteristics of the diamictites and associated sediments, two distinct depositional environments can be recognized: (i) a glaciomarine environment on the Sa˜o Francisco palaeocontinent, with striated bedrock pavements and a 0 –150m-thick diamictite-sandstone-pelite association (i.e. the Jequitaı´ Fm.) and (ii) a deepwater environment dominanted by debris flows (metadiamictites) and turbidites (quartzites, metarhythmites) represented by the Macau´bas Group in the Arac¸uaı´ fold belt, a Neoproterozoic orogen adjacent to the Sa˜o Francisco craton. The Jequitaı´ Fm. was initially interpreted as glaciocontinental and the diamictites regarded as tillites resting directly on striated pavements (Isotta et al. 1969; Karfunkel & Karfunkel 1976; Rocha-Campos & Hasui 1981; Karfunkel & Hoppe 1988). However, the massive clast-poor diamictites and the stratified diamictites of the Jequitaı´ Fm. have been reinterpreted to represent glaciomarine deposits because of its significant thickness (0 –150 m), fine-grained sediment intercalations, and absence of outwash facies. In the Jequitaı´ area, clast-poor diamictites overlying the striated pavement suggests two phases for the record of the glacial event. The first phase, probably on the continent, produced ice erosion (striated pavement); during the second phase, clast-poor diamictite was deposited by gravity flows during ice retreat and sea-level rise (Cukrov et al. 2005) at a time of active extension (Uhlein et al. 1998, 1999; Martins-Neto & Hercos 2002). The lateral transition eastward to the Macau´bas Group is based on the geographical distribution of facies, thickness changes and the presence of glacial abrasion features (Karfunkel & Hoppe 1988; Uhlein et al. 1998, 1999). The Macau´bas Group in the Arac¸uaı´ fold belt is composed of metadiamictites, quartzites, rhythmites, ferruginous diamictites, BIF and schists. It consists of resedimented glacial material deposited by subaqueous debris flows and turbidity currents. The Macau´bas Group represents synglacial, rift-to-drift sedimentation related to the break-up of the Rodinia supercontinent. A rift system opened and was filled by thick gravitational sedimentation, (i.e. debris flows and turbidites; Uhlein et al. 1999; Pedrosa Soares et al. 2001). The elevated rift shoulders with ice caps provided glacial material for subaqueous resedimentation. Tectonic activity on the edge of the rift zone may have triggered slumps that developed into debris flow and turbidity currents. We acknowledge financial support from the CNPq-Brazilian National Research Council (grant 301732/2005-1 PQ) and FAPEMIG-Research Support Foundation of Minas Gerais State (grants CRA-1321/98 and CRA-80772/05), Brazil. F. J. Baars and L. M. Lobato helped improve the English of the original manuscript. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Alkmim, F. F., Marshak, S. & Fonseca, M. A. 2001. Assembling West Gondwana in the Neoproterozoic: clues from the Sa˜o Francisco craton region, Brazil. Geology, 29, 319– 322. Alkmim, F. F., Marshak, S., Pedrosa Soares, A. C., Peres, G. G., Cruz, S. & Whittington, A. 2006. Kinematic evolution of the Arac¸uaı´— ´ frica: Nutcracker tectonics during West Congo orogen in Brazil and A the Neoproterozoic assembly of Gondwana. Precambrian Research, 149, 43– 64. Alvarenga, C. J. S. & Trompette, R. 1992. Glacially influenced sedimentation in the Later Proterozoic of the Paraguay belt (Mato Grosso, Brazil). Palaeogeography, Palaeoclimatology, Palaeoecology, 92, 85– 105. Alvarenga, C. J. S. de, Figueiredo, M. F., Babinski, M. & Pinho, F. E. C. 2007. Glacial diamictites of Serra Azul Formation (Ediacaran, Paraguai belt): evidence of the Gaskiers glacial event in Brazil. Journal of South American Earth Sciences, 23, 236– 241.

545

Babinski, M., Vieira, L. C. & Trindade, R. I. F. 2007. Direct dating of the Sete Lagoas cap carbonate (Bambuı´ Group, Brazil) and implications for the Neoproterozoic glacial events. Terra Nova, 19, 1– 6. Branner, J. C. 1919. Outlines of the geology of Brazil to accompany the Geological Map of Brazil. Geological Society of American Bulletin, 30, 189– 338. Brito Neves, B. B., Campos Neto, M. C. & Fuck, R. A. 1999. From Rodinia to Western Gondwana: an approach to the BrasilianoPanafrican Cycle and orogenic collage. Episodes, 22, 155– 166. Cukrov, N., Alvarenga, C. J. S. de & Uhlein, A. 2005. Litofa´cies da glaciac¸a˜o neoproterozo´ica na porc¸a˜o sul do Cra´ton do Sa˜o Francisco: exemplos de Jequitaı´, MG e Cristalina, GO. Revista Brasileira de Geocieˆncias, 35, 69 –76. D’Agrella Filho, M. S., Pacca, I. G., Teixeira, W., Onstott, T. C. & Renne, P. R. 1990. Paleomagnetic evidence for the evolution of Meso to Neoproterozoic glaciogenic rocks in central-eastern Brazil. Palaeogeography, Palaeoclimatology, Palaeoecology, 80, 255– 265. Dardenne, M. A. 1978. Sı´ntese sobre a estratigrafia do Grupo Bambuı´ no Brasil Central. Anais XXX Congresso Brasileiro de Geologia, SBG, Recife, 2, 597–610. Guimara˜es, J. T., Misi, A., Pedreira, A. J. & Dominguez, J. M. L. 2011. The Bebedouro Formation, Una Group, Bahia (Brazil). In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 503–508. Gravenor, C. P. & Monteiro, R. L. B. P. 1983. Ice-thrust features and a possible intertillite pavement in the Proterozoic Macau´bas Group, Jequitaı´ area, Minas Gerais, Brazil. Journal of Geology, 91, 113– 116. Hettich, M. 1977. A glaciac¸a˜o proterozo´ica no centro-norte de Minas Gerais. Revista Brasileira Geocieˆncias, 7, 87– 101. Hettich, M. & Karfunkel, J. 1978. Um esker, um varvito e seixos estriados no Grupo Macau´bas—norte de Minas Gerais. Revista da Escola de Minas de Ouro Preto, REM, 34, 5– 8. Isotta, C. A. L., Rocha-Campos, A. C. & Yoshida, R. 1969. Striated pavement of the Upper-Precambrian glaciation in Brazil. Nature, 222, 466–468. Karfunkel, B. & Karfunkel, J. 1976. Estudos petro-faciolo´gicos do Grupo Macau´bas na porc¸a˜o mediana da Serra do Espinhac¸o, MG. XIX Congresso Brasileiro de Geologia, 2, 179–188. Karfunkel, J. & Hoppe, A. 1988. Late proterozoic glaciation in centraleastern Brazil: synthesis and model. Palaeogeography, Palaeoclimatology, Palaeoecology, 65, 1 –21. Macedo, M. H. F. & Bonhomme, M. G. 1984. Contribuic¸a˜o a` cronoestratigrafia das Formac¸o˜es Caboclo, Bebedouro e Salitre na Chapada Diamantina (BA) pelos me´todos Rb– Sr e K– Ar. Revista Brasileira de Geocieˆncias, 14, 153– 163. Machado, N., Schrank, A., Abreu, F. R. de, Knauer, L. G. & Abreu, P. A. A. 1989. Resultados preliminares da geocronologia U/Pb na Serra do Espinhac¸o Meridional. An 5 Simp. Geologia Nu´cleo Minas Gerais – 1 Simp. Nu´cleo Brası´lia, Belo Horizonte, 171– 174. Martins, M. 1999. Ana´lise estratigra´fica das sequeˆncias mesoproterozo´icas (borda oeste) e Neoproterozo´icas da bacia do Sa˜o Francisco. MSc thesis, UFRGS. Martins-Neto, M. A. 2000. Tectonics and sedimentation in a PaleoMesoproterozoic rift-sag basin (Espinhac¸o Basin, southeastern Brazil). Precambrian Research, 103, 147–173. Martins-Neto, M. A. & Hercos, C. M. 2002. Sedimentation and tectonic setting of Early Neoproterozoic glacial deposits in south-eastern Brazil. International Association of Sedimentology, Special Publications, 33, 383–403. Misi, A., Kaufman, A. J. et al. 2007. Chemostratigraphic correlation of Neoproterozoic successions in South America. Chemical Geology, 237, 143–167. Moraes, L. J. & Guimara˜es, D. 1930. Geologia da regia˜o norte de Minas Gerais. Anais Academia Brasileira de Cieˆncias, 2, 153–186. Nogueira, A. C. R., Riccomini, C., Sial, A. N., Moura, C. A. V. & Fairchild, T. R. 2003. Soft-sediment deformation at the base of the Neoproterozoic Puga cap carbonate (southwestern Amazon craton, Brazil): confirmation of rapid icehouse to greenhouse transition in snowball Earth. Geology, 31, 613– 616.

546

A. UHLEIN ET AL.

Pedrosa Soares, A. C., Cordani, U. G. & Nutman, A. 2000. Constraining the age of Neoproterozoic glaciation in Eastern Brazil: first U –Pb (SHRIMP) data for detrital zircons. Revista Brasileira de Geocieˆncias, 30, 58 –61. Pedrosa Soares, A. C, Noce, C. M., Vidal, P., Wiedemann, C. M. & Piva-Pinto, C. 2001. The Arac¸uaı´—West Congo Orogen in Brazil: an overview of a confined orogen formed during Gondwanaland assembly. Precambrian Research, 110, 307– 323. Pedrosa-Soares, A. C., Babinski, M., Noce, C., Martins, M., Queiroga, G. & Vilela, F. 2011. The Neoproterozoic Macau´bas Group (Arac¸uaı´ orogen, SE Brazil) with emphasis on the diamictite formations. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 523– 534. Peryt, T. M., Hoppe, A., Bechstadt, T., Koster, J., Pierre, C. J. & Richter, D. K. 1990. Late Proterozoic aragonitic cement crusts, Bambuı´ Group, Minas Gerais, Brazil. Sedimentology, 37, 279– 286. Pimentel, M. M., Alvarenga, C. J. S. de & Armstrong, R. 2002. Provenieˆncia da Formac¸a˜o Jequitaı´, Brasil Central, com base em dados U/Pb Shrimp em zirco˜es detrı´ticos. Anais XLI Congresso Brasileiro de Geologia, SBG, 503, Joa˜o Pessoa (PB). Rocha-Campos, A. C. & Hasui, Y. 1981. Tillites of the Macau´bas Group (Proterozoic) in central Minas Gerais and Southern Bahia, Brazil. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 924– 927. Rocha-Campos, A. C., Young, G. M. & Santos, P. R. 1996. Re-examination of a striated pavement near Jequitaı´, MG: implications for proterozoin stratigraphy and glacial geology. Anais Academia Brasileira de Cieˆncias, 68, 593. Santos, R. V., Alvarenga, C. J. S. de, Dardenne, M. A., Sial, A. N. & Ferreira, V. P. 2000. Carbon and oxigen isotope profiles across Meso-Neoproterozoic limestones from central Brazil: Bambuı´ and Paranoa´ Groups. Precambrian Research, 104, 107– 122. Santos, R. V., Alvarenga, C. J. S. de et al. 2004. Carbon isotopes of Mesoproterozoic – Neoproterozoic sequences from Southern Sa˜o

Francisco craton and Arac¸uaı´ belt, Brazil: paleogeographic implications. Journal of South American Earth Sciences, 18, 27– 39. Trompette, R. 1994. Geology of Western Gondwana (2000– 500 Ma). Pan-African –Brasiliano Aggregation of South America and Africa. Balkema, Rotterdam. Uhlein, A. 1991. Transic¸a˜o cra´ton – faixa dobrada: exemplo do cra´ton do Sa˜o Francisco e da Faixa Arac¸uaı´ (Ciclo Brasiliano) no Estado de Minas Gerais. Aspectos estratigra´ficos e estruturais. PhD thesis, USP, Sa˜o Paulo. Uhlein, A., Trompette, R. & Egydio-Silva, M. 1998. Proterozoic rifting and closure, SE border of the Sa˜o Francisco craton, Brazil. Journal of South American Earth Sciences, 11, 191– 203. Uhlein, A., Trompette, R. & Alvarenga, C. J. S. de 1999. Neoproterozoic glacial and gravitational sedimentation on a continental rifted margin: the Jequitaı´ –Macau´bas sequence (Minas Gerais, Brazil). Journal of South American Earth Sciences, 12, 435– 451. Uhlein, A., Lima, O. N. B., Fantinel, L. M. & Baptista, M. C. 2004. Estratigrafia e evoluc¸a˜o geolo´gica do Grupo Bambuı´, Minas Gerais. Roteiro de Excursa˜o. Excursa˜o 2. Congresso Brasileiro de Geologia. CD-Rom. Vieira, L. C., Trindade, R. I. F. & Nogueira, A. C. R. 2005. Quimioestratigrafia da Formac¸a˜o Sete Lagoas, Grupo Bambuı´, Minas Gerais. III Simposio Cra´ton do Sa˜o Francisco, Salvador (BA), 299–302. Vieira, L. C., Trindade, R. I. F., Nogueira, A. C. R. & Ader, M. 2007. Identification of a Sturtian cap carbonate in the Neoproterozoic Sete Lagoas carbonate platform, Bambuı´ Group, Brazil. Comptes Rendus Geoscience, 339, 240–258. Viveiros, J. F. M., Sa´, E. L., Vilela, O. V., Santos, O. M. & Moreira, J. M. P. 1978. Geologia dos Vales do rios Peixe Bravo e Vacaria, Norte de Minas Gerais. XXX Congresso Brasileiro de Geologia, Recife, 1, 243–254. Walde, D. H. G. 1978. Desenvolvimento faciolo´gico do Precambriano entre a Serra Mineira e a Serra do Cabral (regia˜o sudoeste da Serra do Espinhac¸o, Minas Gerais). Anais XXX Congresso Brasileiro de Geologia, Recife, 2, 711–725.

Chapter 52 The Playa Hermosa Formation, Playa Verde Basin, Uruguay ´ NCHEZ BETTUCCI3 & O. R. TO ´ FALO1 P. J. PAZOS1,2*, A. E. RAPALINI1,2, L. SA 1

Departamento de Ciencias Geolo´gicas, FCEN, Universidad de Buenos Aires, Ciudad Universitaria, Buenos Aires 1428, Argentina 2

Consejo Nacional de Investigaciones Cientı´ficas y Te´cnicas (CONICET)

3

Facultad de Ciencias, Universidad de la Repu´blica, Igua´ 4225, 11400 Montevideo, Uruguay *Corresponding author (e-mail: [email protected])

Abstract: The Playa Hermosa Formation (Fm.) (Playa Verde Basin) is a volcano-sedimentary unit that crops out in the extreme south of the Dom Feliciano Belt in Uruguay. This formation has traditionally been interpreted as non-glacial in origin, but recently it has been suggested that the lower part at least may be glacially influenced. The stratigraphic position of the Playa Hermosa Fm. and its correlation with other Neoproterozoic units of Uruguay, Argentina and southern Brazil remains in dispute. An age of c. 580 Ma is indicated by a hornblende 39Ar/40Ar age on magmatism regarded as coeval with syn-sedimentary volcanism. Preliminary palaeomagnetic data suggest a primary remanence and a mean geomagnetic pole consistent with the proposed apparent polar wander path for the Rı´o de La Plata craton. In combination, these data suggest mid-Ediacaran glacial sedimentation in low to intermediate latitudes on the Rı´o de la Plata craton. However, this conclusion needs to be reinforced by more thorough studies of the age and origin of the Playa Hermosa Fm.

The Ediacaran Playa Hermosa Formation (Masquelin & Sa´nchez Bettucci 1993), regarded as part of the Las Ventanas Formation by Blanco & Gaucher (2005) and Blanco et al. (2009), is commonly omitted in regional correlations and palaeoclimatic interpretations (e.g. Eyles & Janusczack 2004). This formation is exposed in southern Uruguay, close to the city of Piria´polis (between 348490 4100 S, 0558190 0000 W and 348500 0500 S, 0558180 4500 W; see Figs 52.1 & 52.2) and the best outcrops are found along the coast. It was originally interpreted by Masquelin & Sa´nchez Bettucci (1993) as a shallow turbiditic sequence associated with coarse-grained beds deposited in distal fans and canyons sourced from a delta system. Subsequently, Sa´nchez Bettucci & Pazos (1996) ascribed the Playa Hermosa Fm. to the base of the latest Proterozoic –Ordovician Playa Verde basin. The upper part of the lower member of the Playa Hermosa Fm. was deposited contemporaneously with volcanic flows and intrusions of the middle Ediacaran Sierra de Las Animas Complex (Sa´nchez Bettucci 1998; Sa´nchez Bettucci et al. 2009), which provide the most accurate time constraints available for its deposition. A glacial influence in the deposition of the Playa Hermosa was originally proposed by Pazos et al. (1998) and later reinforced by Pazos et al. (2003), although this interpretation was initially not accepted (e.g. Gaucher et al. 2003, 2005). Pazos et al. (2008) published an exhaustive analysis of the available literature on the Playa Hermosa Fm. and concluded that a glaciogenic origin for the lower member of the Playa Hermosa Fm. is unequivocal, although these sediments appear to be dominantly subaqueously reworked glacial debris. Only rainout processes, evidenced by dropstones and coarse-grained rhythmite intervals, are regarded as directly related to ice transport. Correlation between this formation and other glaciogenic and suspected glaciogenic units in the Rı´o de La Plata craton was critically analysed by Pazos et al. (2008).

Structural framework The Uruguayan Precambrian basement is separated into four main tectonic units (Fig. 52.1). From west to east, three units are part of the Rı´o de La Plata craton: (i) the Palaeoproterozoic Piedra Alta

terrane (c. 2.1 Ga) of low- to medium-grade metamorphic orogenic belts, virtually unaffected by younger orogenies, (ii) the Nico Pe´rez terrane represented by Palaeoproterozoic medium- to highgrade metamorphic orogenic belts affected by the Neoproterozoic Brasiliano – Pan-African orogenic cycle, and (iii) the Neoproterozoic to Early Palaeozoic Dom Feliciano belt that developed on the eastern border of the Nico Pe´rez terrane. The fourth and easternmost basement unit in Uruguay is the Punta del Este tectonostratigraphic terrane (also known as Cuchilla Dionisio terrane), which has been interpreted as a fragment of the Kalahari craton accreted to the Rı´o de La Plata craton c. 540 Ma and left behind after the opening of the South Atlantic (Sa´nchez Bettucci et al. 2010). Evidence of the suture is seen in the Sierra Ballena shear zone (Figs 52.1 & 52.2), a continental-scale structure (Fernandes & Koester 1999; Basei et al. 2005). The Playa Hermosa Fm. is located in the southwestern extreme of the Dom Feliciano belt. Sa´nchez Bettucci & Ramos (1999) suggested that the Playa Verde basin originated during postorogenic collapse. In contrast, Gaucher et al. (2003) interpreted this succession as the remnant of a passive margin. The Playa Hermosa Fm. is associated with bimodal (basaltrhyolite) volcanism rich in sub-alkaline subvolcanic (syenite) intrusions. This magmatism has been related to the post-orogenic phase of the Rio Doce Orogeny (Brasiliano Cycle) that developed between c. 590 and 510 Ma (Campos Netto & Figueiredo 1995). The Playa Hermosa Fm. records the opening of the Playa Verde basin in a transtensional tectonic setting (Pazos et al. 2003; Pecoits et al. 2008). The Playa Hermosa Fm. is exposed in a homocline. Its base (348510 S, 0558180 W) is underlain by tonalitic gneisses of the 1735 + 32 Ma Palaeoproterozoic Campanero Unit (Sa´nchez Bettucci et al. 2004, 2009; Mallmann et al. 2007). Its upper contact, however, is covered variably by modern beach sands and Quaternary sediments (Pazos et al. 2003). Tectonic deformation is characterized by extension-related stretching and fracturing. The succession lacks regional metamorphism, despite lying adjacent to the conspicuous tectonic boundary between the Dom Feliciano belt and the Punta del Este terrane (i.e. the Sierra Ballena Shear Zone, Oyhantc¸abal et al. 2007). This major shear zone was apparently active after deposition of the Playa Hermosa

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 547– 553. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.52

548

P. J. PAZOS ET AL.

Fig. 52.1. Location of the studied area and morphotectonic divisions of the eastern part of the Rı´o de la Plata Craton. Modified from Pazos et al. (2008). SYSZ, Sarandı´ del Yı´ shear zone; MAFMSZ, Marı´a Albina-Fraile Muerto shear zone; SBSZ, Sierra Ballena shear zone; CSZ, Cordillera shear zone.

sediments, recording docking of the Punta del Este terrane in the Early Cambrian. Gaucher et al. (2007) proposed that the Nico Pe´rez terrane (Fig. 52.1) was displaced some 500 km northward during the Cambrian, along the Sarandı´ del Yı´ major shear zone that separates this terrane from the Piedra Alta terrane to the west. However, the geotectonic setting and the detailed geological evolution of these rocks is still far from resolved.

Stratigraphy The Playa Verde basin (Sa´nchez Bettucci & Pazos 1996) infill constitutes the Maldonado Group (Pecoits et al. 2008), which is subdivided into the Playa Hermosa, the Las Ventanas and the San Carlos formations (Fig. 52.3). The Las Ventanas and San Carlos formations are discussed in a separate chapter (Pecoits et al. 2011). The glacially influenced Playa Hermosa Fm. comprises two members. The lower member contains conglomerate, finegrained sandstone, pelite and rhythmically laminated fine-grained sediments resembling varves. Coarse-grained lithologies are locally deformed by loading and other syn-sedimentary deformation (Pazos et al. 2003). The upper member is composed of conglomerate, rare sandstone, diamictite beds and volcanic rocks that intruded and brecciated the sediments and produced hyaloclastic features indicative of syn-sedimentary volcanism (Sa´nchez Bettucci et al. 2009). The Las Ventanas Fm. is characterized by conglomerate, sandstone and rhythmites containing out-sized clasts (Pecoits et al. 2008). The San Carlos Fm. is a succession of fluvial metaconglomeratic-sandstone, quartzite, feldspar sandstone and subordinate metapelite (Sanchez Bettucci & Pazos 1996) and rhyolithic flows (Pecoits et al. 2008). This unit is found to the east of the Sierra Ballena shear zone (see Fig. 52.2).

The San Carlos and the Las Ventanas formations were originally assigned to the Ordovician by Sa´nchez Bettucci (1998) based on the occurrence of basalt and syenite clasts in the Las Ventanas Fm. that were thought to be sourced from the Ediacaran to Cam´ nimas Complex (Sa´nchez Bettucci 1998; brian Sierra de Las A Sa´nchez Bettucci & Rapalini 2002). Later, Pecoits et al. (2008) suggested that conglomerate units and rhythmite intervals of the Las Ventanas Fm. are Neoproterozoic with evidence of glacial influence. This succession was later folded during the Early Palaeozoic. Both Pazos et al. (2008) and Pecoits et al. (2008) correlated the lower member of the Playa Hermosa Fm. with the finegrained deposits of the Las Ventanas Fm. based on the similarity of rainout deposits (Fig. 52.3).

Glaciogenic deposits and associated strata The sedimentology of the Playa Hermosa Fm. has been studied by Sa´nchez Bettucci & Pazos (1996), Pazos et al. (1998, 2003) and Fambrini et al. (2003). The Playa Hermosa is divided into two members, with the lower one containing deposits thought to be glaciogenic. In the logged section of the lower member analysed by Pazos et al. (2003), two facies associations (FA I and II) were defined (Fig. 52.4). The lowermost, FA I, is composed of rhythmite, contorted conglomerate and rare normally graded sandstone and pelite beds. The conglomerates have variable matrix content and include sub-rounded and angular pebbles of felsic magmatic rocks as well as intrabasinal rhythmite clasts that are disaggregated and deformed. Massive diamictites contain abundant rip-up rhythmite clasts, as well as some angular feldspar clasts. Evidence for syn-sedimentary deformation is seen in folded stringers of gravels. Conglomerate beds have irregular, sometimes loaded,

THE PLAYA HERMOSA FORMATION

549

Fig. 52.2. Geology of the southern area of the Dom Feliciano Belt. Modified from Pazos et al. (2008).

basal contact, are lenticular at outcrop scale, and have clasts forming aggregates or clusters with clast-supported fabric. The most distinctive lithology of FA I is the fine- to coarsegrained rhythmite. The coarse rhythmite consists of couplets of gravel and sand/mud layers. These rhythmite units contain large, out-sized, bullet-shaped boulders of quartzite. The intervals of

fine-grained rhythmite units rarely occur in FA I and are composed of dark and light green silt and clay couplets. FA II (Fig. 52.4) consists of fine-grained rhythmites, some with brecciated tops, normally graded and rippled sandstones, some containing lenticular, linsen and moderate- to high-angle ripple drift cross-lamination. Diamictite beds are usually sharp based,

550

P. J. PAZOS ET AL.

Fig. 52.3. Comparative stratigraphic chart of the Playa Verde basin, Arroyo del Soldado Group and Tandilia System. Taken and modified from Gaucher et al. (2003), Pecoits et al. (2008) and Pazos & Rapalini (2011). Triangles indicate glaciogenic intervals. The brick symbol corresponds to carbonates. Grey boxes are siliciclastic units.

although some units are interbedded with normal graded sandstone and dark claystone beds. All palaeocurrents measured from ripples indicate palaeoflow towards the NE. Horizons of completely disaggregated mixtures of gravel, sand and mud with relict rhythmite clasts are also present (Bm(i) in Fig. 52.4). Rare out-sized clasts of felsic magmatic origin occur. Close to the top, metre-thick convolute bedding occurs in fine-grained rhythmitic deposits below a second interval of distorted conglomerate. This distorted conglomerate deposit is overlain by extremely thin, tectonically deformed rhythmite beds.

Boundary relations with overlying and underlying non-glacial units The basal contact of the Playa Hermosa Fm. is a disconformity overlying 1.7 Ga tonalitic gneiss (Sa´nchez Bettucci et al. 2009). The contact between the lower glaciogenic member and the upper member and the contact at the top of the Playa Hermosa Fm. are covered by beach sands.

Chemostratigraphy No chemostratigraphic studies have been carried out in the Playa Hermosa Fm. or other units within the Playa Verde Basin.

Palaeolatitude and palaeogeography Preliminary palaeomagnetic data on fine-grained sandstone units of the Playa Hermosa Fm. were reported by Sa´nchez Bettucci & Rapalini (2002). This study included just two sites in the lower part of this succession near Piria´polis (Fig. 52.1). The mean geomagnetic pole suggests a low palaeolatitude (12.7 þ 9.5/ –8.18S) for deposition of this succession, although these data must be confirmed with further analysis before any palaeogeographical interpretation is verified. The entire succession is exposed as a homoclinal sequence tilted some 408 towards WNW, which does not permit the performance of any fold or tilt test to date the remanence. Detailed alternating field (AF) and thermal cleaning and principal component analysis (Kirschvink 1980) assure a reliable isolation and definition of the magnetic components.

Fig. 52.4. Logged section of the lower member of the Playa Hermosa Fm. Modified from Pazos et al. (2003). Cl, claystone; Fs, fine-grained sandstone; Ms, medium-grained sandstone; Cs, coarse-grained sandstone; Fcg, fine-grained conglomerate; Mcg, medium-grained conglomerate.

Isothermal remanent acquisition curves confirm that magnetite is the likely carrier of the remanence in this formation. Lack of resemblance of the characteristic directions (both in situ and after untilting) with post-Ordovician expected directions for South America, coupled with the tectonic stability of the region after the Ordovician suggests a possible primary origin for the characteristic remanence of the Playa Hermosa Fm. However, the very small number of sites (two) and samples (six) make this ‘palaeomagnetic pole’ a simple mean geomagnetic pole with no certainty that non-dipole components of the Earth’s magnetic field have been fully compensated. Under such conditions, ‘true’ palaeolatitudes from palaeomagnetic data are not certain. Despite this uncertainty, the geomagnetic pole for the Playa Hermosa Fm. is consistent with the distribution of other Neoproterozoic palaeomagnetic poles for the Rı´o de La Plata craton

THE PLAYA HERMOSA FORMATION

(Sa´nchez Bettucci & Rapalini 2002; Rapalini 2006; Pazos & Rapalini 2011). Comparison of this pole with the apparent polar wander path suggests an age of c. 590 Ma. This apparent polar wander path also indicates that the Rı´o de La Plata craton was located at intermediate southern latitudes during the Ediacaran and Cambrian (Fig. 52.5). Palaeogeographic reconstruction of the basin is highly uncertain because the Playa Hermosa Fm. only crops out as a narrow belt on the coast, with very little obtainable information on the geometry of the original basin. The systematic palaeocurrent pattern from the SW and quartzite composition of the outsized clasts (dropstones) suggests that the Tandilia System and its eastern prolongation (Rapela et al. 2007; Pazos & Rapalini 2011) may have been the source area. Mylonitized magmatic clasts within the Playa Hermosa Fm. also indicate provenance from an intensely tectonized region.

Geochronological constraints The age of the Playa Hermosa Fm. is not constrained by any direct radiometric dates. However, the interaction of magmatism with unconsolidated sediments in the upper part of the lower member of the Playa Hermosa Fm. (Pazos et al. 2008) strongly suggests that deposition was contemporaneous with this magmatic event. The characteristics of this interaction were described by Sa´nchez Bettucci et al. (2009). The bimodal magmatism corresponds to the first cycle of the Sierra de Las Animas complex and has been loosely constrained between 615 and 550 Ma based on K –Ar and Rb – Sr ages (Sa´nchez Bettucci & Linares 1996; Sa´nchez Betucci & Rapalini 2002) from samples taken from different units of this complex, including flows intercalated in the

551

sedimentary succession. Oyhantc¸abal et al. (2007) obtained an age of 579 + 2 Ma (39Ar/40Ar in hornblende) from a syenitic body (Pan de Azu´car pluton) that is part of this magmatic cycle and is exposed 12 km from the Playa Hermosa outcrops. This age and the geological relationships suggest a roughly similar age for the deposition of the Playa Hermosa sediments. The presence of boulders of conglomerate and sandstone lithologies of the lower member of the Playa Hermosa Fm., enclosed in quartz – syenite breccia of the Sierra de Las Animas complex (Sa´nchez Bettucci et al. 2009), also supports this interpretation.

Discussion The lower member of the Playa Hermosa Fm. is attributable to a glacially influenced environment that records high rates of sedimentation in a tectonically active depositional system. It is separated into two facies associations. FA I is thought to record a pro-glacial depositional setting that grades upward into the more distal FA II (Pazos et al. 2003). Identification of glacial influence is based on a combination of sedimentary facies and features, including ice-rafted debris and rhythmites forming couplets of angular grains (coarse layer) immersed in a fine-grained matrix and dark clay. These couplets are common in Carboniferous deglacial to post-glacial transitions in Argentina (see Pazos et al. 2008 for references) and in the Quaternary. A large out-sized, elongate clast of quartzite that disrupted laminations was interpreted by Pazos et al. (2003) as the most robust evidence for ice-rafting processes. A debris flow origin is ruled out due to the lateral discontinuity of disruption to bedding around the dropstone and the well-laminated nature of the couplets hosting the clast. Very finegrained rhythmite units present in the FA II are similar to varves

Fig. 52.5. Palaeogeographic reconstruction for the Rı´o de la Plata Craton and other blocks during the Ediacaran–Cambrian periods. (a) 600 Ma, (b) 575 Ma, (c) 550 Ma. AM, Amazonia; AR, Arabia; Bal, Baltica; C, Congo; K, Kalahari; Lau, Laurentia; PA, Pampia; RP, Rı´o de la Plata; SF, San Francisco; WA, Western Africa; WN, Western Nile (or Sahara metacraton). Modified from Sa´nchez Bettucci & Rapalini (2002).

552

P. J. PAZOS ET AL.

and suggest a cyclic control on deposition. No tillite deposits have been reported and for this reason, the succession was considered as only glacially influenced by Pazos et al. (2008), consistent with earlier interpretations of a pro-glacial depositional setting (Pazos et al. 2003; Pecoits et al. 2008) for the lower member of the Playa Hermosa Fm. Gravity-driven deposits, such as normally graded sandstone beds with ripple drift cross-lamination and giant convolutions, as well as irregular conglomerate beds, suggest that deposition was abrupt on an unstable slope related to a pro-glacial environment. Top-brecciated rhythmite beds and syn-depositional folding and slumping are interpreted to record syn-sedimentary seismicity. It is difficult to discern between a marine and non-marine depositional setting for the Playa Hermosa Fm. Fambrini et al. (2003) argued for a marine environment based on micro-hummocky structures. However, these structures seem to represent ripple-drift crosslamination rather than storm-wave processes, and hence do not distinguish between marine and lacustrine environments (Pazos et al. 2003). Sa´nchez Bettucci et al. (2009) described orange palagonite glass in the volcanism–sediment interaction that is a common alteration product where basalts extrude into the marine environment. The relationship between the Playa Hermosa Fm. and other units within the Maldonado Group (Playa Verde basin; Fig. 52.3), in particular the purportedly glaciogenic Las Ventanas Fm. (Pecoits et al. 2011), remains subject to debate. Several other Ediacaranaged successions occur on the Rı´o de la Plata craton (Fig. 52.3), including the Cerro do Buggio Fm. (Camaqua˜ basin) and the Arroyo del Soldado Group in the Ribeira belt and the Sierras Bayas Group and equivalent units in the Tandilia System (Pazos & Rapalini 2011). Possible correlations of the Playa Hermosa Fm. with other Ediacaran units cropping out in Brazil have been discussed in detail by Pazos et al. (2003, 2008). In summary, no conclusive correlations can yet be made due to highly fragmentary exposures and insufficient radiometric control. We are grateful to M. A. S. Basei for a constructive review and especially to E. Arnaud for her thorough review and editorial improvements. This is a contribution R-39 to the Instituto Don Pablo Groeber. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Basei, M. A. S., Frimmel, H. E., Nutman, A. P., Preciozzi, F. & Jacob, J. 2005. A connection between the Neoproterozoic Dom Feliciano (Brazil/Uruguay) and Gariep (Namibia/South Africa) orogenic belts – evidence from a reconnaissance provenance study. Precambrian Research, 139, 195–221. Blanco, G. & Gaucher, C. 2005. Estratigrafı´a, paleontologı´a y edad de la Formacio´n Las Ventanas (Neoproterozoico, Uruguay). Latin American Journal of Sedimentology and Basin Analysis, 12, 115–131. Blanco, G., Rajesh, H. M., Gaucher, C., Germs, G. J. B. & Chemale, F., Jr. 2009. Provenance of the Arroyo del Soldado Group (Ediacaran to Cambrian, Uruguay): implications for the paleogeographic evolution of southwestern Gondwana. Precambrian Research, 171, 57 – 73. Campos Neto, M. C. & Figueiredo, M. C. 1995. The Rı´o Doce Orogeny, Southeastern Brazil. Journal of South American Earth Sciences, 8, 143– 162. Eyles, N. & Januszczak, N. 2004. ‘Zipper-rift’: a tectonic model for Neoproterozoic glaciations during the break up of Rodinia after 750 Ma. Earth-Science Reviews, 65, 1– 73. Fambrini, G. L., Paes-de-Almeida, R., Riccomini, C. & FragosoCesar, A. R. S. 2003. Tempestitos com Influeˆncia Glacial da Formac¸a˜o Playa Hermosa (Neoproterozo´ico), Piria´polis, Uruguai. Revista Brasileira de Geocieˆncias, 33, 1 –12. Fernandes, L. A. D. & Koester, E. 1999. An overview of the Neoproterozoic Dorsal de Canguc¸ u strike-slip shear zone and its role in the tectonic evolution of the continental crust in southern Brazil. Journal of African Earth Sciences, 29, 3– 24.

Gaucher, C., Boggiani, P. C., Sprechmann, P., Sial, A. N. & Fairchild, T. 2003. Integrated correlation of the Vendian to Cambrian Arroyo del Soldado and Corumba˜ Groups (Uruguay and Brazil) palaeogeographic, palaeoclimatic and palaeobiologic implications. Precambrian Research, 120, 241– 278. Gaucher, C., Poire´, D., Peral, L. & Chiglino, L. 2005. Litoestratigrafı´a, Bioestratigrafı´a y correlaciones de las sucesiones sedimentarias del Neoproterozoico-Ca´mbrico del Crato´n del Rı´o de La Plata (Uruguay y Argentina). Latin American Journal of Sedimentology and Basin Analysis, 12, 145– 160. Gaucher, C., Poire´, D., Finney, S. C., Vale´ncia, V., Blanco, G., Pamoukaghlian, K. & Go´mez Peral, L. 2007. Zircones detrı´ticos de secuencias neoproterozoicas de Uruguay y Argentina: Inferencias sobre la evolucio´n paleogeogra´fica del craton del Rı´o de La Plata. V Congreso Uruguayo de Geologı´a, Montevideo, Uruguay. CD-Rom (extended paper, 25). Kirschvink, J. L. 1980. The least-squares and plane and the analysis of palaeomagnetic data. Geophysical Journal of the Royal Astronomical Society, 62, 699– 718. Mallmann, G., Chemale, F., Jr., Avila, J. N., Kawashita, K. & Armstrong, R. A. 2007. Isotope geochemistry and geochronology of the Nico Pe´rez Terrane, Rı´o de La Plata Craton, Uruguay. Gondwana Research, 12, 489– 508. Masquelin, H. & Sa´nchez-Bettucci, L. 1993. Propuesta de evolucio´n tectono-sedimentaria para la fosa tardi-Brasliana en la regio´n de Piria´polis, Uruguay. Revista Brasileira de Geociencias, 23, 313– 322. Oyhantc¸abal, P., Siegesmund, S., Wemmer, K., Frei, R. & Layer, P. 2007. Post-collisional transition from calc-alkaline to alkaline magmatism during transcurrent deformation in the southernmost Dom Feliciano Belt (Braziliano –Pan-African, Uruguay). Lithos, 98, 141– 159. Pazos, P. J. & Rapalini, A. 2011. The controversial stratigraphy of the glacial deposits in the Tandilia System, Argentina. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 565–569. Pazos, P. J., Tofalo, R. & Sa´nchez Bettucci, L. 1998. Procesos sedimentarios e indicadores paleoclima´ticos en la seccio´n inferior de la Formacio´n Playa Hermosa, Cuenca Playa Verde, Piria´polis, Uruguay. II Congreso Uruguayo de Geologı´a, Punta del Este, Uruguay, 64 –69. Pazos, P. J., Sa´nchez Bettucci, L. & To´falo, R. O. 2003. The record of the Varanger glaciation at the Rı´o de La Plata craton, VendianCambrian of Uruguay. Gondwana Research, 6, 65 –78. Pazos, P. J., Sa´nchez Bettucci, L. & Loureiro, J. 2008. The Neoproterozoic glacial record in the Rı´o de La Plata Craton: a critical reappraisal. In: Pankhurst, R. J., Trouw, R. A. J., de Brito Neves, B. B. & de Witt, M. J. (eds) West Gondwana Pre-Cenozoic Correlations Across the South Atlantic Region. Geological Society, London, Special Publications, 294, 343–364. Pecoits, E., Gingras, M., Aubet, N. & Konhauser, K. 2008. Ediacaran in Uruguay: palaeoclimatic and palaeobiological implications. Sedimentology, 55, 689– 721. Pecoits, E., Gingras, M. K. & Konhauser, K. O. 2011. Las Ventanas and San Carlos formations, Maldonado Group, Uruguay. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 555–564. Rapalini, A. E. 2006. New Late Proterozoic paleomagnetic pole for the Rı´o de La Plata craton: implications for Gondwana. Precambrian Research, 147, 223– 233. Rapela, C. W., Pankhurst, R. J. et al. 2007. The Rı´o de La Plata craton and the assembly of SW Gondwana. Earth Science Reviews, 83, 49 – 82. Sa´nchez Bettucci, L. 1998. Evolucio´n tecto´nica del Cinturo´n Dom Feliciano en la regio´n Minas – Piria´polis, Uruguay. PhD thesis, Universidad de Buenos Aires, Argentina. Sa´nchez Bettucci, L. & Linares, E. 1996. Primeras Edades en Basaltos ´ nimas, Uruguay. XIII Congreso Geolo´gico del Complejo Sierra de A Argentino and III Congreso de Exploracio´n de Hidrocarburos I, Buenos Aires, Argentina, I, 399–404.

THE PLAYA HERMOSA FORMATION

Sa´nchez Bettucci, L. & Pazos, P. 1996. Ana´lisis Paleoambiental y Marco tecto´nico en la Cuenca Playa Verde, Piria´polis, Uruguay. XIII Congreso Geolo´gico Argentino, Buenos Aires, Argentina, I, 405– 412. Sa´nchez Bettucci, L. & Ramos, V. A. 1999. Aspectos Geolo´gicos de las rocas metavolca´nicas y metasedimentarias del Grupo Lavalleja, Sudeste de Uruguay. Revista Brasilera de Geociencias, 29, 557– 570. Sa´nchez Bettucci, L. & Rapalini, A. E. 2002. Palaeomagnetism of the Sierra de Las Animas Complex, Southern Uruguay: Its Implications in the Assembly of Western Gondwana. Precambrian Research, 118, 243– 265.

553

Sa´nchez Bettucci, L., Oyhantc¸abal, P., Preciozzi, F., Loureiro, J., Ramos, V. A. & Basei, M. A. S. 2004. Mineralizations of the Lavalleja Group (Uruguay), a Neoproterozoic volcano– sedimentary sequence. Gondwana Research, 7, 745–751. Sa´nchez Bettucci, L., Koukharsky, M., Pazos, P. J. & Stareczek, S. 2009. Neoproterozoic subaqueous extrusive –intrusive rocks in the Playa Hermosa Formation in Uruguay: regional and stratigraphic significance. Gondwana Research, 16, 134– 144. Sa´nchez Bettucci, L., Peel, E. & Masquelin, H. 2010. Neoproterozoic tectonic synthesis of Uruguay. International Geology Review, 52, 51– 78.

Chapter 53 Las Ventanas and San Carlos formations, Maldonado Group, Uruguay ERNESTO PECOITS*, MURRAY K. GINGRAS & KURT O. KONHAUSER Department of Earth and Atmospheric Sciences, University of Alberta, 1-26 Earth Sciences Building, Edmonton, AB, T6G 2E3, Canada *Corresponding author (e-mail: [email protected]) Abstract: Together with the Playa Hermosa Formation (Fm.), the Las Ventanas and San Carlos formations constitute the Maldonado Group, which is better developed in the southeastern part of Uruguay and covers an area of c. 200 km2. The total thickness of both units (i.e. Las Ventanas and San Carlos formations) reaches c. 1500 m, and comprises mafic and acidic volcanic rocks, pyroclastic rocks, diamictite, sandstone, conglomerate and pelite. Structurally, the Maldonado Group is extensively deformed, although variably, throughout the region. Strike–slip faults, westward-verging detachment faults, and folds with axis sub-parallel to the strike–slip planes are common features. The presence of pumpellyite, prehnite, chlorite and epidote in mafic rocks indicates very low- to low-grade metamorphic conditions. The Las Ventanas Fm. is characterized by basal conglomerate, diamictite, sandstone and siltstone that pass upwards into finegrained rhythmites (pelite), and is the thickest unit of the Maldonado Group (c. 1250 m). The San Carlos Formation (c. 250 m) comprises fine-grained conglomerate, sandstone and mudstone towards the top. Both units lie on an angular unconformity above Palaeo- and Neoproterozoic basement and are overlain unconformably by late Ediacaran– lowermost Cambrian units. Reliable palaeomagnetic data indicate that the Maldonado Group accumulated at high palaeolatitudes; however, the palaeogeographical evolution of the Rı´o de la Plata Craton during the Neoproterozoic remains conjectural. Radiometric data from intrusive bodies and cross-cutting strike– slip faults place the minimum age of the group at c. 565 Ma, whereas basement volcanic rocks dated at 590 + 2 Ma interbedded with meta-sandstones hosting detrital zircons c. 600 million years old provide the best constraint on the maximum age of deposition. Given the absence of carbonate rocks, no chemostratigraphic studies (e.g. C, O, Sr) are available. The Las Ventanas and San Carlos formations are largely interpreted as units within a thick glacially influenced fan-delta sedimentary system formed during the early Ediacaran in a strike–slip basin. Based on stratigraphic and sedimentological characteristics it has been suggested that this succession, containing glacially influenced diamictite and dropstones, records a glacial period that occurred sometime between c. 570 and 590 million years ago. Ongoing research is focused on establishing the precise age of deposition of the Maldonado Group and on reconstructing the tectonic evolution of the basin. Further palaeomagnetic studies will be especially useful for determining the palaeogeography of the Rı´o de la Plata Craton during the Ediacaran and establishing its relationships with neighbouring strata hosting similar successions.

The Maldonado Group was formally erected by Pecoits et al. (2005) to include the Playa Hermosa (Pazos et al. 2011) and Las Ventanas formations, which are better exposed near the towns of Piria´polis and Pan de Azu´car (Figs 53.1 & 53.2). The San Carlos Fm. was informally included. Subsequent work showed that the succession continues to the SW and NE of Minas and Melo cities (Fig. 53.1). The group reaches a maximum thickness of c. 1500 m and covers an area of c. 200 km2. It comprises acidic and basic volcanic rocks, pyroclastic and sedimentary strata generated in a tectonically active and glacially influenced basin (for a recent review see Pecoits et al. 2008). Midot (1984) originally suggested the Las Ventanas Fm. included conglomerate, sandstone and pelite outcropping at De Las Ventanas Hill and in the surrounding areas (Fig. 53.3). This unit was considered by Midot (1984) and various other authors to be Ordovician in age (e.g. Bossi & Navarro 1991; Masquelin & Sa´nchez 1993; Pazos et al. 2003). This assumption was mainly founded on the inferred development of alluvial fans ´ nimas Complex, sourced from the Cambrian Sierra de las A located westward (Fig. 53.2). However, detailed mapping and stratigraphic analysis of the Las Ventanas Fm. led to its redefinition as a Neoproterozoic volcanic/sedimentary succession (Pecoits 2003a). The sections exposed in the northern and southern parts of the type area were designated as the stratotype and parastratotype of the unit, respectively (Pecoits 2003a). The former is located near Paso del Molino, where 1200 m of Las Ventanas strata are continuously exposed. The parastratotype is situated near the Burguen˜o Quarry and Apolonia Mine, where the unconformable contact between the Las Ventanas Fm. and the basement (Lavalleja Group) is exposed (Fig. 53.3). The San Carlos Fm. was erected by Masquelin (1990), who documented that the unit consists of conglomerate, sandstone and pelite. The depositional environment was likely lacustrine or

fluvio-lacustrine. The stratotype of the formation is located 6 km south of San Carlos town (Fig. 53.2), where 220 m of San Carlos strata are exposed with the base and top of formation not visible (Pecoits et al. 2008). The sedimentary facies and volcanic association of the San Carlos Fm. are similar to those of the middle Las Ventanas Fm. Likewise, palynological macerations carried out in the pelites of both units reveal the occurrence of similar microbiota (Pecoits et al. 2005). These observations led Pecoits et al. (2005) to propose a correlation between the San Carlos and Las Ventanas formations. Whether both units were deposited in the same basin, and subsequently dismantled by the displacement of the Sierra Ballena Shear Zone (Fig. 53.2), or were developed within different depocentres remains uncertain. The first evidence of glacial influence in the Las Ventanas Fm. was recorded by Pecoits (2003a), comprising faceted, outsized clasts in finely laminated rhythmites that were interpreted as dropstones. Recently, Gaucher et al. (2008) reported glacial diamictite with associated dropstones occurring in laminated siltstone to the south of Minas (Fig. 53.2). Additionally, glacial diamictite and fine-grained rhythmites containing striated dropstones are well exposed c. 15 km NW (El Perdido area) of this locality (Pecoits et al. 2008). No glacial evidence has yet been recorded in the San Carlos Fm.

Structural framework In Uruguay, a significant extensional and synkinematic magmatic event corresponding to the final stages of the SW-Gondwana assembly occurred during the Neoproterozoic– lowermost Cambrian (Bossi & Campal 1992; Pecoits 2003b; Oyhantc¸abal 2005). From a structural perspective, the Sierra Ballena Shear Zone (Figs 53.1 & 53.2) constitutes the largest remnant of the

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 555– 564. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.53

556

E. PECOITS ET AL.

50° W WA

3

SOUTH AMERICA

A

SF

C

Florianópolis

ne

Z

Zo

RP

Sh ea r

K (b)

o

Serra Azul Fm

rc in

N

Ge

(a)

ajo r

54° W

Porto Alegre

M

BRAZIL

30° S 2

Melo

ce O ic nt tla

1 Buenos Aires

(c)

58° W

Montevideo

34° S

A

Minas

Sier ra

URUGUAY

an

She a

r Z on

e

Paysandú

Ballena

ARGENTINA

Rivera

Fig. 53.2

54° W

Brasilian –Pan African Orogeny (c. 700– 500 Ma). This highstrain transcurrent structure, which was operative primarily between c. 600 and 580 Ma, contributed significantly to the basinfill architecture of the early Ediacaran units (Oyhantc¸abal 2005). In this regard, the Las Ventanas and San Carlos formations were deposited in a strike –slip basin, as indicated by (i) the diverse depositional facies and their abrupt lateral changes; (ii) apparent migration of the primary depocentre towards the south; (iii) the subparallel trend of the basin with respect to its strike –slip margins; and (iv) synchronous timing with regional shearing (Sierra Ballena Shear Zone). Following the development of the widespread transcurrent system, a gravitational orogenic collapse characterized by high-angle normal faulting and accompanied by marine transgression occurred during the late Ediacaran –Early Cambrian (Pecoits et al. 2008). A similar geotectonic evolution is observed in the associated magmatism, which was initiated with highly fractionated calc-alkaline granite (c. 584 Ma), followed by mildly alkaline granite and shoshonitic volcanics (c. 575 Ma), and concluded with peralkaline intrusions and volcanics (c. 540 –520 Ma) (Oyhantc¸abal et al. 2007). Therefore, the Ediacaran –Early Cambrian in Uruguay is characterized by a transition from a back-arc basin (underlying Lavalleja Group), followed by a strike – slip-related basin (Maldonado Group), to a foreland basin

0

100

200 Km

Fig. 53.1. (a) Distribution of cratonic blocks of west Gondwana: A, Amazonian; C, Congo; K, Kaoko; SF, Sa˜o Francisco; RP, Rı´o de la Plata; WA, West African; Z, Zaire. (b) Regional map showing the location of the units discussed in the text (see (c) for more detail). (c) Distribution of volcano-sedimentary early Ediacaran units of Uruguay and southeastern Brazil: 1, Las Ventanas-San Carlos formations; 2, Camaqua˜ Basin; 3, Itajaı´ Basin.

(overlying Arroyo del Soldado Group), where strike– slip shear zones of crustal scale played a major role in the evolution of the orogen. At the outcrop scale, the Las Ventanas and San Carlos formations show evidence of both brittle and ductile deformation. Small- and large-scale strike –slip faults, westward-verging detachment faults, and folds with axis sub-parallel to the strike – slip planes are common features (Fig. 53.3). Axial plane slaty and sporadic millimetre-spaced fracture cleavages are present in fine-grained facies (pelite). The basic volcanic and pyroclastic rocks show abundant chlorite and epidote as well as pumpellyite and prehnite, demonstrating very low- to low-grade metamorphic conditions (Pecoits 2003a).

Stratigraphy The Las Ventanas and San Carlos formations lie on an angular unconformity above a crystalline basement of undetermined age and the Lavalleja Group (Figs 53.4 & 53.5). Relatively welldated basement granitoids are represented by the Campanero Complex and the Cerro Olivo Complex, with ages of c. 1750 Ma and 1006 + 37 Ma, respectively (Table 53.1; Oyhantc¸abal 2005 and references therein). The lithostratigraphy of the Lavalleja

LAS VENTANAS AND SAN CARLOS FORMATIONS

557

Fig. 53.2. Simplified geological map showing the distribution of the Las Ventanas and San Carlos formations (Maldonado Group) and selected age determinations of southeastern Uruguay (see Table 53.1 and text for explanation). Sources of geological information: Bossi & Navarro (1991), Pecoits et al. (2005), Oyhantc¸abal (2005) and references therein. SYPSZ, Sarandı´ del Yı´-Piria´polis Shear Zone; SBSZ, Sierra Ballena Shear Zone.

Group, although poorly known, is different from that of the Las Ventanas and San Carlos formations. According to Midot (1984), the Lavalleja Group is a volcanosedimentary succession dominated by immature fine-grained siliciclastics, marl, basalt and limestone towards the top. This limestone hosts columnar stromatolites assignable to Conophyton (Poire´ et al. 2005), and although their occurrence extended from the early Proterozoic to the Ediacaran, preliminary radiometric studies suggest an early Ediacaran age for this unit. SHRIMP U – Pb detrital zircon analyses from the Lavalleja Group display ages between 3.4 and 0.6 Ga (Basei et al. 2008). Likewise, interbedded basalt shows a crystallization age of 590 + 2 Ma (U–Pb SHRIMP; Mallmann et al. 2007), indicating that the deposition must have occurred c. 590 million years ago. The whole succession (i.e. the Las Ventanas and San Carlos formations) can be divided (from base to top) into three informal intervals: (i) volcanic and pyroclastic deposits, (ii) conglomeratedominated lithofacies and (iii) pelite-dominated lithofacies. Volcanic and pyroclastic rocks including basalt, mafic hyaloclastic breccias and subaqueous tuff, as well as rhyolite, acidic volcanoclastic and pyroclastic rocks, have been recognized as part of the Maldonado Group (Fig. 53.4a). This bimodal volcanism has long ´ nimas been thought to represent part of the Sierra de las A Complex (e.g. Sa´nchez & Rapalini 2002 and references therein).

The latter, however, shows geochemical signatures, radiometric ages and structural features indicating anorogenic magmatism, which was extruded after the main deformational phase that affected the Maldonado Group (Oyhantc¸abal 2005). In fact, the ´ nimas Complex systematically displays radiometric Sierra de las A ages younger than those of the volcanics assigned to the Maldonado Group (see ‘Geochronological constraints’) and it displays neither ductile deformation nor metamorphism. Field relationships ´ nimas Complex intrudes the Las show that the Sierra de las A Ventanas Fm., providing definite evidence of the older age of the Las Ventanas. Conglomerate-dominated lithofacies (proximal facies association of Pecoits 2003a) dominate the basal part of the Las Ventanas Fm., including clast-supported conglomerate and breccia, diamictite, massive sandstone and conglomerate –sandstone couplets. Upwards, pelites are abundant with occasional conglomerate beds (pelites-dominated lithofacies or distal facies association of Pecoits 2003a). This lithofacies includes laminated siltstone and sandstone-pelite rhythmites, and massive sandstone and conglomerate. Likewise, the San Carlos Fm. consists of basal conglomerate- and upper pelite-dominated lithofacies, but the clast size in the lower conglomerate never reaches that of the Las Ventanas Fm. In this sense, two possible explanations can be drawn. First, the San Carlos Fm. represents a lateral equivalent

558

E. PECOITS ET AL.

Fig. 53.3. Geological map of the type area of the Las Ventanas Fm. The inset shows the location of the stratotype (1) and parastratotype (2) (modified from Pecoits et al. 2008).

LAS VENTANAS AND SAN CARLOS FORMATIONS

559

Fig. 53.4. (a) Stratigraphic column of the Las Ventanas Fm. at its stratotype (point 1, Fig. 53.3). (b) Stratigraphic section for the lower Las Ventanas Fm. at its parastratotype (point 2, Fig. 53.3) (modified from Pecoits et al. 2008). K-Pg, potassium and plagioclase.

to the middle and uppermost part of the Las Ventanas Fm., where coarse-grained conglomerate is rare (see ‘Glaciogenic deposits and associated strata’). Second, both units, although potentially contemporaneous (see below), were deposited in different basins.

The Las Ventanas is thought to be overlain by the Arroyo del Soldado Group, a thick (3000 m) mixed siliciclastic –carbonate succession, mainly represented by intercalating conglomerate, sandstone, siltstone, thick carbonate, Fe-formation, black- and iron-rich shale and chert. It contains a rich fossil assemblage

560

E. PECOITS ET AL.

proposed for the latter, a glacially influenced system in a strike – slip setting was suggested for the Las Ventanas and San Carlos formations (Pecoits et al. 2005, 2008). The presence of abundant organic-walled microfossils and shelly fauna in the Arroyo del Soldado Group points to high biological productivity and an elevated nutrient supply, possibly related to increased weathering during warmer conditions (Gaucher et al. 2004). Based on available geochronology, chemostratigraphy and biostratigraphy, Pecoits et al. (2008) proposed a maximum depositional age of c. 560 Ma for the group, which is younger than earlier suggestions (Gaucher et al. 2004).

Glaciogenic deposits and associated strata

Fig. 53.5. Simplified stratigraphic column of the San Carlos Fm. at its stratotype, 6 km SE of San Carlos town (see Fig. 53.2) (modified from Pecoits et al. 2008).

composed of organic-walled microfossils and small shelly fauna, including the index fossil Cloudina riemkeae (Gaucher et al. 2004). The distinct lithological differences between Las Ventanas and San Carlos formations with the Arroyo del Soldado Group have been explained by different prevailing climatic conditions and tectonic settings. Whereas an evolution towards tropical conditions (Gaucher et al. 2004) and a marine transgression in a foreland setting (Pecoits et al. 2005; Basei et al. 2008) have been

Palaeoenvironmental interpretations indicate that the Las Ventanas –San Carlos system records sheet flood-dominated fan-delta deposits in a glacially influenced setting. Direct evidence of ancient glacial activity comes from the basal and uppermost facies of the Las Ventanas Fm., where some deposits have been described and interpreted as ice-rafted diamictite with striated and faceted clasts and rhythmite-hosting dropstones (Pecoits 2003a; Pecoits et al. 2008). Diamictites are mainly coarse-grained and matrix-supported lithofacies with a massive structure displaying normal and occasionally reversed grading. Clasts are rounded to angular and range from granule to boulder size. Compositionally, the diamictite dominantly contains extrabasinal clasts (rhyolite, basalt, granitoid, gabbro, quartzite). Two types of diamictite can be distinguished: sub-rounded pebbles and cobbles in a massive muddy matrix, and angular to sub-rounded cobbles to boulders in a massive to horizontally laminated silty/clayey matrix. These deposits have a polymodal texture and form relatively thick (2 –10 m) tabular beds with faceted and occasionally striated clasts. Although common in the uppermost part of the succession, finegrained rhythmites are also found interbedded with diamictite at the base of the Las Ventanas Fm. The ,3-m-thick beds are characterized by a millimetre- to centimetre-intercalation of silty and sandy material with clay and also by the presence of out-sized clasts deforming the layering. Based on the diverse composition of these large clasts (granite, basalt, gabbro, etc.), and on the presence of pre-depositional foliation, an interpretation of the clasts as volcanic bombs or other ballistic/pyroclastic material is discarded. These lithofacies (i.e. rhythmites with out-sized clasts and diamictite) were first described at the parastratotype section of the unit (Burguen˜o Quarry and Apolonia Mine; Fig. 53.3), but later discovery in other localities (e.g. NE Minas and Melo; Fig. 53.1) suggesting that they are more extensive than originally thought. The type section of the Las Ventanas Fm. is largely dominated by conglomerate, sandstone and finely laminated siltstone. It begins with a 690-m-thick fining- and thinning-upward cycle (Fig. 53.4a). Conglomerate, sandstone and laminated siltstone dominate the lowermost, medial and uppermost sub-cycles, respectively. The conglomerate is typically granitic and clast-supported with arkosic sandstone present at the top of each sub-cycle. The following changes occur up-section within the lower major cycle: (i) bed thickness progressively decreases, from metre-scale to a few millimetres (laminae); (ii) average grain size decreases from pebbles to silt; (iii) the proportion of granitic clasts becomes smaller; and (iv) planar parallel stratification and lamination become a common feature in the siltstone but are absent in the lower and middle part of the cycle. This finely laminated siltstone shows similar features to those described in the basal part of the unit. Here, the lithofacies is considerably thicker, but the out-sized clasts (dropstones) identified in it are smaller and rarely reach more than 10 cm. The formation passes up-section into a second major cycle that is nearly 560 m thick, and is composed of minor subcycles of

LAS VENTANAS AND SAN CARLOS FORMATIONS

561

Table 53.1. Summary of geochronological data available from southeastern Uruguay (see Fig. 53.2) Stratigraphic unit ´ nimas Complex Sierra de las A A8 del Soldado Group* El Renegado Puntas del Pan de Azu´car Lineament Aigua´ Batholith Solı´s de Mataojo Aguas Blancas Las Ventanas and San Carlos formations* Las Flores Lavalleja Group* Cerro Olivo Complex Zanja del Tigre Fm.* Campanero Complex

Rock type

Method

Porphyres Syenite Granite ‘Mylonite’ Granite Tonalite Mylonite Trachybasalt Quartz-sericite schist Metabasalt Metarhyolite Metarhyolite Orthogneiss Meta-sandstone Orthogneiss Orthogneiss

Rb–Sr Ar–Ar Depositional age: c. 560– 530 Ma Rb–Sr K–ArMusc Rb–Sr U–Pb K–ArMusc Depositional age: c. 590– 570 Ma K–Ar U–Pb(SHRIMP) U–Pb(SHRIMP) U–Pb U–Pb U–Pb U–Pb(SHRIMP) U–Pb U–Pb

Age (Ma) 520 + 5 579 + 1.5 559 + 28 572 + 7 587 + 16 584 + 13 594 + 13 615 + 30 600–3400 590 + 2 624 + 14 667 + 4 1006 + 37 1800–3400 1735 + 2 1754 + 7

Source: Oyhantc¸abal (2005) and references therein, Mallmann et al. (2007), Basei et al. (2008), Pecoits et al. (2008). *Volcanic-sedimentary units.

sandstone and fine-grained conglomerate. The sandstone has a tabular geometry, is massive in appearance, and occasionally has non-erosive basal contacts. The conglomerate is clast-supported, polymictic and has a modal grain size of 3–10 cm. The clast composition is variable, and includes rhyolite (32%), granite (2%), quartz (12%), basic volcanic rocks (11%), alkaline feldspar (10%), plagioclase feldspar (8%) and schist (5%). Clasts, either in conglomerate or sandstone, are fresh and show no signs of chemical weathering.

Chemostratigraphy Owing to the lack of carbonate rocks directly associated with the Las Ventanas or San Carlos formations, no chemostratigraphic studies (e.g. C, O, Sr) have been performed.

Other characteristics No evidence of mineral deposits was found in any of the units described here.

Boundary relations with overlying and underlying non-glacial units Palaeolatitude and palaeogeography The Las Ventanas Fm. rests unconformably on the Lavalleja Group. Palaeoproterozoic orthogneiss (Campanero Complex) and Monzogranite (La Nativa) are of unknown age (Pecoits et al. 2008). A nonconformity separates the San Carlos Fm. from Palaeoproterozoic ortho- and paragneiss (Cerro Olivo Complex; Fig. 53.2). Particularly important is the relationship with the Lavalleja Group. This unit is largely dominated by basalt and immature fine-grained siliciclastic rocks (Midot 1984). An evolution towards warm climate and stable tectono-magmatic conditions is evidenced by thick stromatolitic limestones developed in the uppermost part of the unit (Poire´ et al. 2005), upon which the Las Ventanas Fm. uncomformably lies. However, as discussed below, the hiatus between both units is poorly constrained. Despite this, the transition from the Lavalleja Group to the Las Ventanas Fm. is not only indicated by an angular unconformity but also by a strong change in climatic and tectono-magmatic activity. The relationship and nature of the contact between Las Ventanas and San Carlos formations with the overlying late Ediacaran Arroyo del Soldado Group is not firmly established. The best locality to study such transition is to the north of Minas (Fig. 53.2) where both units are closely exposed. Detailed geological mapping of the area indicates stratal juxtaposition due to tectonic shortening in a zone of major thrusting. In other words, both units would be separated by major structural discontinuities (i.e. thrusts oriented SW– NE), and no conformable contacts have been recorded. The contact between the units – although not well exposed – suggests the presence of an angular unconformity separating the uppermost fine-grained rhythmites of the Las Ventanas Fm. and the sandstone of the basal Arroyo del Soldado Group (Yerbal Fm.).

The location and kinematic history of the blocks involved during the assembly of West-Gondwana in the late Neoproterozoic is poorly known (Rapela et al. 2007). In particular, the palaeogeographical position of the Rı´o de la Plata Craton (Fig. 53.1 inset) during the Ediacaran is highly disputed (e.g. Cordani et al. 2000). In this regard, no palaeomagnetic studies have been performed either in the sedimentary rocks of Las Ventanas Fm. or in the San Carlos Fm. Preliminary mean geomagnetic poles were only obtained from sedimentary rocks of the Playa Hermosa Fm., volcanics from the Las Ventanas Fm., and volcanics and intrusives from ´ nimas Complex (Sa´nchez & Rapalini 2002). the Sierra de las A According to the same authors, the new data support the Apparent Polar Wander Path (APWP) previously suggested for the entire Gondwana since c. 550 Ma, indicating that the Rı´o de la Plata Craton was indeed at that time part of the supercontinent. Furthermore, it was suggested that a mean geomagnetic pole obtained from the Playa Hermosa Fm. (12.7 þ 9.5/–8.18) meant that another example of Neoproterozoic low-latitude glaciation is evident in Uruguay. The most reliable palaeomagnetic pole for early Ediacaran units of the Rı´o de la Plata Craton is derived from the Campo Alegre lavas (Sa´nchez & Rapalini 2002). The Campo Alegre Fm., dated by the U –Pb method at 595 + 5 Ma (Citroni et al. 1999), is located in the Itajaı´ Basin, SE Brazil (Fig. 53.1). Palaeomagnetic reconstructions indicate a moderate palaeolatitude of 33.3 + 9.58S (D’Agrella & Pacca 1988), in marked contrast to the low palaeolatitudes (12.7 þ 9.5/ –8.18) proposed for the Playa Hermosa Fm. (Sa´nchez & Rapalini 2002).

562

E. PECOITS ET AL.

In contrast to the scarce database for the early Ediacaran units, the APWP for Gondwana since 550 Ma is better known. Since 550 Ma, poles for the Rı´o de la Plata Craton and other Gondwanan continents have tended to form a single APWP ranging from c. 308S in the late Ediacaran towards lower palaeolatitudes during the Lower Cambrian (Meert & Van der Voo 1996).

Geochronological constraints Since the definition of both units, the age of the Las Ventanas and San Carlos formations was considered Ordovician (e.g. Midot 1984; Masquelin & Sa´nchez 1993; Pazos et al. 2003). This assumption was challenged by Pecoits et al. (2008) in reporting cross-cutting relationships with several intrusive Ediacaran bodies and major faults that indicated a minimum depositional age of c. 570 Ma. The age of deposition of the Las Ventanas and San Carlos formations is now relatively well constrained to c. 590– 575 Ma by radiometric data based on K – Ar, Rb –Sr and U –Pb dating techniques on basement rocks, interbedded basalt, intrusive syenite, granitic and trachytic dykes, and cross-cutting faults (Table 53.1). This inference is supported by the 590 + 2 Ma age (SHRIMP U –Pb) obtained for a metabasalt (Mallmann et al. 2007), the detrital zircons from the Lavalleja Group in the basement of the Las Ventanas Fm., which show U– Pb (SHRIMP) ages between 3.4 and 0.6 Ga (Basei et al. 2008), and an intrusive syenite that yields Ar/Ar ages of 579 +1.5 Ma (Oyhantc¸abal et al. 2007). Furthermore, the ages are corroborated by (i) basic volcanics interbedded with sedimentary rocks of the Las Ventanas Fm., which display ages between 615 + 30 and 565 + 30 Ma (K –Ar method; Sa´nchez & Linares 1996); (ii) Rb –Sr ages from intrusive granite with an age of 559 + 28 Ma (Preciozzi et al. 1993) and tra´ nimas Complex, which intrude and chyte of the Sierra de Las A overly the Las Ventanas Fm., dated between 520–530 Ma (Bossi et al. 1993; Linares & Sa´nchez 1997); (iii) basic dykes crosscutting the Las Ventanas Fm. to the south of Minas yielding a K –Ar age of 485 +12.5 Ma (Poire´ et al. 2005); (iv) the last reactivation of the Puntas del Pan de Azu´car Lineament, which crosscuts the Las Ventanas Formation (Fig. 53.3), and occurs at 572 + 7 Ma (K –Ar in syn-kinematic muscovites) (Bossi & Campal 1992); and (v) the San Carlos Fm., which is intensively deformed by the Sierra Ballena Shear Zone, in which the third and last deformation phase occurred at c. 550 –500 Ma (Oyhantc¸abal 2005). Finally, the transcurrent tectonics that occurred during the early Ediacaran, which is closely related to the generation of strike –slip basins recorded for example by the Las Ventanas and San Carlos formations, is also associated with a voluminous syn-kinematic magmatism (Pecoits 2003a; Pecoits et al. 2005). Radiometric studies performed in all these bodies yield ages systematically between 570 and 590 Ma (Rb–Sr and U–Pb methods) (Bossi et al. 1993; Hartmann et al. 2002; Oyhantc¸abal 2005). According to Gaucher et al. (2008), the acritarch assemblage recovered from the Las Ventanas Fm. indicates a depositional age between 635 and 582 Ma, supporting previous data (Pecoits 2003a, b). This assemblage, however, comprises and is dominated by individuals with no stratigraphic value, such as Leiosphaeridia and others of doubtful origin (e.g. Soldadophycus) (Butterfield, pers. comm. 2008).

Discussion The facies associations point to the development of sheet flooddominated alluvial fans (Blair & McPherson 1994) intercalated with minor lake deposits in a glacially influenced, transtensional tectonic setting. The proximal facies association comprises massive and horizontally stratified clast-supported conglomerate and rare breccia, massive sandstone, conglomerate – sandstone

couplets and diamictite, while the distal facies includes massive and normally graded sandstone, pebbly sandstone, laminated siltstone, and fine-grained massive and graded conglomerate. The proximal facies association was interpreted by Pecoits (2003a) as a subaerial alluvial fan in which debris-flow deposits (diamictite) and sheet flood deposits (stratified conglomerate and sandstone) constitute the dominant facies. The subaerial alluvial fan succession is characterized by upward-coarsening and upward-thickening trends resulting from fan progradation. The restricted occurrence of debris flow beds and the comparatively high roundness of the clastic fraction indicate that the preserved succession represents middle and outer regions of the alluvial-fan complex. The distal facies association is thought to represent a submarine delta subenvironment with sediment gravity-flow deposits occasionally interbedded with turbidites (massive and graded conglomerate and sandstone) and suspension fallout deposits (laminated siltstone). Although some conglomeratic levels are interpreted to represent shoreline deposits along the distal fan, no evidence of wave reworking has been observed (Pecoits 2003a). The proximal facies association offers evidence of sedimentation under arid climatic conditions, as shown by exceptionally fresh well-rounded clasts (e.g. basalt) in conglomerate and sandstone. Glacial sedimentary evidence comes from the distal facies association. Therein, outsized clasts within finely laminated siltstone have been recorded and interpreted as dropstones. This lithofacies has been described at the base and top of the Las Ventanas Fm. In the first case, the laminated strata are ‘sandwiched’ between massive and bedded diamictite, mostly containing extrabasinal clasts. Faceted, striated and bullet-shaped clasts are consistent with glacial transport and suggest a glacial influence during the deposition of laminated siltstone and diamictite facies. In contrast, the laminated siltstone described at the top of the unit overlies laminated siltstone and fine-grained sandstone, which are interpreted as turbiditic deposits (Pecoits et al. 2008). Here, the glacially influenced laminated siltstone is differentiated from the turbidites by the lack of turbidity current structures, finer grain size and numerous dropstones with impact-induced deformation of underlying laminae. Despite the lack of evidence for glacially influenced sedimentation in the San Carlos Fm., the structural and geochronological framework, stratigraphy and fossil content support the premise that the San Carlos Fm. is correlative with the middle– upper part of the Las Ventanas Fm. (Figs 53.4a & 53.5). This would explain the absence of the basal and uppermost glacially influenced facies described for the Las Ventanas Fm. Only one systematic palaeomagnetic study has been performed in glacial Ediacaran units and associated rocks of Uruguay (Sa´nchez & Rapalini 2002); however, none of the sampled sites corresponds to the Las Ventanas and San Carlos formations. Unfortunately, all the samples from this work were collected near the border of the Dom Feliciano belt, in an area affected by intense Neoproterozoic –Cambrian tectono-magmatic activity, and thus probably affected by widespread remagnetization (e.g. Rapalini & Sa´nchez 2008). The two palaeopoles obtained from ´ nimas volcanic and hypabyssal rocks of the Sierra de las A Complex are, from geochronological and structural points of view, poorly constrained. Much of the radiochronology is based on the K – Ar method, which usually provides a minimum age, and recent dating on the same lithologies using more precise methods has yielded ages as much as 30 Ma older (Oyhantc¸abal et al. 2007). Although the early Ediacaran units of Uruguay are extensively deformed, the interbedded basalt samples for palaeomagnetism were not corrected with respect to the palaeohorizontal. Field relationships have extensively shown that even the youngest ´ nimas Complex (i.e. c. 525 million years rocks of the Sierra de las A old), although not folded, are tilted. Therefore the integrity and usability of these palaeomagnetic data are problematic. Recent palaeogeographical reconstructions locate the Rı´o de la Plata Craton at high palaeolatitudes c. 580 million years ago

LAS VENTANAS AND SAN CARLOS FORMATIONS

(Trinidade & Macouin 2007). Inclination data from deposits slightly older than the Gaskiers equivalent in the Avalon Terrane (Newfoundland) similarly indicate a palaeolatitude of 358S during the early Ediacaran (c. 608 Ma, U –Pb zircon age; Myrow & Kaufman 1999). However, palaeogeographical models between 590 and 560 Ma (i.e. when Gaskiers deposits and their possible Ediacaran correlatives, Squantum, Loch na Cille, and Moelv, were formed) are controversial due to the ambiguous results presented by the Laurentian palaeopoles (Trinidade & Macouin 2007 and references therein). For instance, both low and high latitudes for Laurentia at c. 580 Ma have been proposed. If the latter configuration is confirmed, the glacial strata observed in Laurentia, Baltica, Cadomia, Avalonia and Rı´o de la Plata cratons are compatible with a palaeoclimatic scenario similar to the Phanerozoic glaciations rather than ‘snowball’ conditions. Although early proposals promoted a ‘Marinoan’ age for Las Ventanas and San Carlos formations (Pecoits et al. 2005), an Ediacaran event seems to be a more reasonable alternative (Pecoits et al. 2008) based on the radiometric constraints. Indeed, this would explain the absence of thick ‘cap carbonate’ facies immediately overlying these deposits as is distinctive of Ediacaran glacial deposits. Alternatively, the absence of cap carbonates might be due to deposition in a highly active tectonic setting, characterized by high rates of subsidence and high accumulation rates of siliciclastic sediments, lack of preservation because they were eroded; or, they may simply not have been found yet. However, the typical facies of the Las Ventanas and San Carlos formations indicates an active participation of the hydrological cycle that is incompatible with the ‘Snowball Earth’ model for Cryogenian glaciations (e.g. Hoffman & Schrag 2002). Recently, a similar glacial succession (Tacuarı´ Fm.) was described in NE Uruguay (Veroslavsky et al. 2006). These deposits were long-considered a classic example of the Carboniferous – Permian glaciation in Gondwana (e.g. Bossi & Navarro 1991). According to Veroslavsky et al. (2006) and mainly based on a very similar fossil content to that described for the Las Ventanas Fm., this unit was tentatively assigned to the Neoproterozoic. Ongoing research using radiometric dating on cross-cutting granitic dykes (U–Pb TIMS) and detrital zircons (LA-ICP-MS) has confirmed the Ediacaran age of the succession. The obtained ages constrain the deposition of the unit between 590 and 570 Ma. This is in agreement with the age proposed for the Las Ventanas Fm. (570 –590 Ma) and would suggest regional glacial conditions. Sedimentological, tectonic and magmatic evidence, supported by radiometric ages, suggests some similarities between the Maldonado Group and other successions in Brazil. In this regard, the Las Ventanas and San Carlos formations have been correlated with the Bom Jardim (c. 592–573 Ma) and Cerro do Bugio (573 –559 Ma) allogroups of the Camaqua˜ Basin located in Rio Grande do Sul, southern Brazil (Pecoits 2003b; Fig. 53.1). The Bom Jardim Allogroup is composed of basic to intermediate volcanic rocks, alluvial conglomerate and turbidites (Paim et al. 2000). SHRIMP age dating of the volcanics yielded an age of c. 580 Ma (Paim et al. 2000). The Cerro do Bugio Allogroup consists of acidic and basic rocks, alluvial conglomerate, rhythmites (sandstone-pelite) and pelite. Geochronological studies on acidic rocks yielded a U – Pb age of 573 + 8 Ma (Paim et al. 2000). Both the Bom Jardim and Cerro do Bugio allogroups were deformed by sinistral transcurrent displacement dated to c. 570 Ma and were subsequently intruded by granitic bodies dated at 559 + 7 Ma and 565 + 14 to 561 + 6 Ma (Paim et al. 2000 and references therein). Evidence of a seasonal glacial influence has been suggested by Eerola (2001, 2006) for the Bom Jardim Allogroup. Recently, Alvarenga et al. (2007) reported glacial deposits in the Ediacaran Serra Azul Fm. in the Paraguay belt, Brazil (Alvarenga et al. 2011). The Paraguay belt is located on the south-eastern edge of the Amazon Craton (Fig. 53.1), which in conjunction with the

563

Rı´o de la Plata Craton was probably already amalgamated into a single crustal block by the Ediacaran (e.g. Cordani et al. 2000). Unlike the Las Ventanas –San Carlos formations and their Brazilian correlatives Cerro do Bugio –Bom Jardim allogroups, the Serra Azul Fm. was deposited on a passive margin (Alvarenga et al. 2007), showing no comparable tectono-magmatic activity but probably similar palaeolatitude (see above). The correlation between Ediacaran (c. 590–570 Ma) glacially influenced successions of Brazil and Uruguay strengthen the notion of a post Cryogenian glaciation and suggests that these deposits are distributed more extensively than previously recognized in South America. Future efforts focused on the sedimentological constraints of these and other successions (e.g. Itajaı´ Basin in Brazil) are required to determine the glacial influence on the Brazilian units. The authors wish to thank Natalie Aubet who actively participated in the fieldwork and in the preparation of the figures. Constructive comments were provided by the reviewers Ruben Rieu, Pablo Pazos and editor Emmanuelle Arnaud. The research was partially funded by the Natural Sciences and Engineering Research Council of Canada Discovery Grants to KOK and MKG. This is a contribution to IGCP 512 ‘Neoproterozoic Ice Ages”.

References Alvarenga, C. J. S. de., Figueiredo, M. F., Babinski, M. & Pinho, F. E. C. 2007. Glacial diamictites of Serra Azul formation (Ediacaran, Paraguay belt): evidence of the Gaskier event in Brazil. Journal of South American Earth Sciences, 23, 236– 241. Alvarenga, C. J. S., Boggiani, P. C. et al. 2011. Glacially-influenced sedimentation of the Puga Formation, Cuiaba´ Group and Jacadigo Group, and associated carbonates of the Araras and Corumba´ groups, Paraguay Belt, Brazil. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 487– 497. Basei, M. A. S., Frimmel, H. E., Nutman, A. P. & Preciozzi, F. 2008. West Gondwana amalgamation based on detrital zircon ages from Neoproterozoic Ribeira and Dom Feliciano Belts of South America and comparison with coeval sequences from SW Africa. In: Pankhurst, R. J., Trouw, R. A. J., de Brito Neves, B. B. & de Wit, M. J. (eds) West Gondwana: Pre-Cenozoic Correlations Across the South Atlantic Region. Geological Society, London, Special Publications, 294, 239–256. Blair, T. C. & McPherson, J. G. 1994. Alluvial fans processes and forms. In: Abrahams, A. D. & Parsons, A. J. (eds) Geomorphology of Desert Environments. Chapman and Hall, London, 354–402. Bossi, J. & Navarro, R. 1991. Geologı´a del Uruguay. Departamento de publicaciones de la Universidad de la Repu´blica, Montevideo. Bossi, J. & Campal, N. 1992. Magmatismo y tecto´nica transcurrente durante el Paleozoico Inferior en Uruguay. In: Gutie´rrez-Marco, J. G., Saavedra, J. & Rabano, I. (eds) Paleozoico Inferior de Iberoame´ric. Me´rida, Universidad de Extremadura, 343– 356. Bossi, J., Cingolani, C., Lambias, E., Varela, R. & Campal, N. 1993. Caracterı´sticas del magmatismo post-oroge´nico finibrasiliano en el ´ nimas. Revista Uruguay: formaciones Sierra de Rı´os y Sierra de A Brasileira de Geocieˆncias, 23, 282– 288. Citroni, S. B., Basei, M. A. S., Sato, K. & Siga, O., Jr. 1999. Petrogenesis of the Campo Alegre Basin magmatism, based on geochemical and isotopic data. In: 2nd South-American symposium on isotope geology. Extended abstracts, Co´rdoba, Argentina, 1, 174– 177. Cordani, U. G., Milani, E. J., Thomaz Filho, A. & Campos, D. A. (eds) 2000. Proceedings of the 31st International Geological Congress on Tectonic Evolution of South America. Rio de Janeiro, Brazil. D’Agrella, M. S. F. & Pacca, I. G. 1988. Paleomagnetism of the Itajaı´, Castro and Bom Jardim Groups from Southern Brazil. Geophysical Journal, 93, 365–376. Eerloa, T. 2001. Climate change at the Neoproterozoic – Cambrian transition. In: Zhuravlev, A. Y. & Riding, R. (eds) The Ecology of the Cambrian Radiation. Perspectives in Paleobiology and Earth History. Columbia University Press, 90– 106.

564

E. PECOITS ET AL.

Eerola, T. 2006. Myo¨ha¨isproterotsooiset ilmastonmuutokset Tutkimuksia Etela¨-Brasiliassa. Geologi, 58, 164– 174. Gaucher, C., Sial, A. N., Blanco, G. & Sprechmann, P. 2004. Chemostratigraphy of the lower Arroyo del Soldado Group (Vendian, Uruguay) and palaeoclimatic implications. Gondwana Research, 7, 715– 730. Gaucher, C., Blanco, G., Chiglino, L., Poire, D. & Germs, G. J. B. 2008. Acritarchs of the Las Ventanas Formation (Ediacaran, Uruguay): implications for the timing of coeval rifting and glacial events in western Gondwana. Gondwana Research, 13, 488– 501. Hartmann, L., Santos, O., Bossi, J., Campal, N., Schipilov, A. & McNaughton, N. J. 2002. Zircon and Titanite U– Pb SHRIMP geochronology of Neoproterozoic felsic magmatism on the eastern border of the Rı´o de la Plata Craton, Uruguay. Journal of South American Earth Sciences, 15, 229– 236. Hoffman, P. F. & Schrag, D. P. 2002. The Snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129– 155. Linares, E. & Sa´nchez Bettucci, L. 1997. Edades Rb/Sr y K/Ar del cerro Pan de Azu´car, Piria´polis, Uruguay. South American Symposium on Isotope Geology, San Pablo, 1, 176– 180. ´ vila, J. N., Kawashita, K. & Mallmann, G., Chemale, F., Jr., A Armstrong, R. A. 2007. Isotope geochemistry and geochronology of the Nico Pe´rez Terrane, Rı´o de la Plata Craton, Uruguay. Gondwana Research, 12, 489– 508. Masquelin, H. 1990. Ana´lisis estructural de las zonas de cizalla en las migmatitas de Punta del Este – Uruguay. Acta Geologica Leopoldensia, 30, 139 –158. Masquelin, H. & Sa´nchez, L. 1993. Propuesta de evolucio´n tectonosedimentaria para la fosa tardi-brasiliana en la regio´n de Piria´polis, Uruguay. Revista Brasileira de Geocieˆncias, 23, 313– 322. Meert, J. G. & Van der Voo, R. 1996. Paleomagnetic and 40Ar– 39Ar study of the Sinyai dolerite, Kenya: implications for Gondwana assembly. Journal of Geology, 104, 131– 142. Midot, D. 1984. Etude geologique et diagnostic metalloge´nique pour l’exploration du secteur de Minas (Uruguay). Diplome de Docteur de 3e` Cycle, Universite` Pierre et Marie Curie. Myrow, P. M. & Kaufman, A. J. 1999. A newly discovered cap carbonate above Varanger-age glacial deposits in Newfoundland, Canada. Journal of Sedimentary Research, 6, 784– 793. Oyhantc¸abal, P. 2005. The Sierra Ballena Shear Zone: kinematics, timing and its significance for the geotectonic evolution of southeast Uruguay. PhD thesis, Georg-August-Universita¨t zu Go¨ttingen. Oyhantc¸abal, P., Siegesmund, S., Wemmer, K., Frei, R. & Layer, P. 2007. Post-collisional transition from calc-alkaline to alkaline magmatism during transcurrent deformation in the southernmost Dom Feliciano Belt (Braziliano –Pan-African, Uruguay). Lithos, 98, 141– 159. Paim, P. S. G., Chemale, F., Jr. & Lopes, R. da 2000. A Bacia do Camaqua˜. In: Holz, M. & De Ros, L. F. (eds) A geologı´a do Rio Grande do Sul. CIGO-Universidade Federal Rio Grande do Sul, Rio Grande do Sul, 231– 274.

Pazos, P., To´falo, R. & Sa´nchez Bettucci, L. 2003. The record of the Varanger Glaciation at the Rio de la Plata Craton, Vendian-Cambian of Uruguay. Gondwana Research, 6, 65 – 77. Pazos, P. J., Rapalini, A. E., Sa´nchez-Bettucci, L. & To´falo, O. R. 2011. The Playa Hermosa Formation, Playa Verde Basin, Uruguay. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 547–553. Pecoits, E. 2003a. Sedimentologı´a y consideraciones estratigra´ficas de la Formacio´n Las Ventanas en su a´rea tipo, departamento de Maldonado, Uruguay. Revista de la Sociedad Uruguaya de Geologı´a, Special Issue, 1, 124–140. Pecoits, E. 2003b. Age and preliminary correlation of the Las Ventanas Formation and Bom Jardim-Cerro do Bugio allogroups (Vendian, Uruguay and Brazil). In: Frimmel, H. E. (ed.) 3rd International Colloquium Vendian-Cambrian of W-Gondwana. University of Cape Town, Cape Town, South Africa, 32 – 34. Pecoits, E., Aubet, N., Oyhantc¸abal, P. & Sa´nchez Bettucci, L. 2005. Estratigrafı´a de sucesiones sedimentarias y volcanosedimentarias Neoproterozoicas del Uruguay. Revista de la Sociedad Uruguaya de Geologı´a, 11, 18 –27. Pecoits, E., Gingras, M., Aubet, N. & Konhauser, K. 2008. Ediacaran in Uruguay: palaeoclimatic and palaeobiologic implications. Sedimentology, 55, 689– 719. Poire´, D. G., Gonza´lez, P. D., Canalicchio, J. M., Repetto, F. G. & Canessa, N. D. 2005. Estratigrafı´a del Grupo Mina Verdu´n, Proterozoico de Minas, Uruguay. Latin American Journal of Sedimentology and Basin Analysis, 12, 125– 143. Preciozzi, F., Masquelin, H. & Sa´nchez Bettucci, L. (eds) 1993. Geologı´a de la Porcio´n Sur del Cinturo´n Cuchilla de Dionisio. In: 1st Simp. Int. Neoproterozoico – Ca´mbrico de La Cuenca del Plata. Guı´a de Excursiones, La Paloma, Uruguay. Rapalini, A. E. & Sa´nchez, L. B. 2008. Widespread remagnetization of late Proterozoic sedimentary units of Uruguay and the apparent polar wander path for the Rio de La Plata craton. Geophysical Journal International, 174, 55– 74. Rapela, C. W., Pankhurst, R. J. et al. 2007. The Rı´o de la Plata craton and the assembly of SW Gondwana. Earth-Science Reviews, 83, 49 – 82. Sa´nchez, L. & Linares, E. 1996. Primeras edades en basaltos del ´ nimas, Uruguay. In: 13th Congreso Geolo´Complejo Sierra de las A gico Argentino y 3rd Congreso de Exploracio´n de Hidrocarburos, Actas, La Plata, Argentina, I, 399– 404. Sa´nchez, L. & Rapalini, A. E. 2002. Paleomagnetism of the Sierra de Las Animas Complex, southern Uruguay: its implications in the assembly of western Gondwana. Precambrian Research, 118, 243–265. Trindade, R. I. F. & Macouin, M. 2007. Palaeolatitude of glacial deposits and palaeogeography of Neoproterozoic ice ages. Comptes Rendus Geoscience, 339, 200–211. Veroslavsky, G., de Santa Ana, H. & Daners, G. 2006. Tacuarı´ Formation (Nov. Nom.): lithostratigraphy, facies, environment, age and geological significance (Cerro Largo-Uruguay). Revista de la Sociedad Uruguaya Geologı´a, 13, 23 –35.

Chapter 54 The controversial stratigraphy of the glacial deposits in the Tandilia System, Argentina P. J. PAZOS1,2* & A. RAPALINI1,2 1

Departamento de Ciencias Geolo´gicas, FCEN, Ciudad Universitaria, Buenos Aires 1428, Argentina 2

Consejo Nacional de Investigaciones Cientı´ficas y Te´cnicas (CONICET) *Corresponding author (e-mail: [email protected])

Abstract: In the Ediacaran–Cambrian Tandilia System of central Argentina, the glacial origin of the thin (10 m) Sierra del Volca´n Formation (Fm.) has been recognized for many years (Spalletti & del Valle 1984), being the first undisputed glacial deposits recorded in the Rı´o de la Plata craton. It consists of three units: (i) a basal polymictic diamictite with a kaolin-rich matrix, (ii) a middle pelite with heterolithic levels, undulatory stratification, symmetric ripples, and outsized clasts previously interpreted as dropstones, and (iii) an upper polymictic diamictite with subtle normal grading. Here, this poorly age-constrained formation is reviewed from a stratigraphic and regional context and compared with other putative glaciogenic intervals at the base of the Cerro Largo Fm. and in the Punta Mogotes borehole.

The Tandilia System consists of an Ediacaran – Ordovician package of sedimentary rocks that crops out in a 300-km-long, NW –SE-trending belt south of Buenos Aires, Argentina (Fig. 54.1). The Tandilia System comprises the Sierras Bayas Group and the unconformably overlying Cerro Negro and Balcarce formations. Whereas the stratigraphic and tectonic evolution of the Tandilia System has received some recent attention (e.g. Go´mez Peral et al. 2007), its glaciogenic record remains poorly described in the international literature. The glaciogenic origin of the Sierra del Volca´n Fm., which occurs in the southern part of the Tandilia belt, has long been accepted. However, its age and stratigraphic position, reviewed recently by Pazos et al. (2008), have remained ambiguous, having been variably assigned to the Precambrian, Cambrian and Ordovician. A putative glacial diamictite occurs at the base of the Cerro Largo Fm. in the lower Sierras Bayas Group and crops out near the town of Olavarrı´a in the northwestern part of the Tandilia belt (Fig. 54.1). The relationship between this diamictite and the Sierra del Volca´n Fm., and more generally the stratigraphic framework for the Tandilia System, is poorly understood, but a relatively complete section preserved in the Punta Mogotes borehole in the SE of the belt (Marchese & di Paola 1975) sheds light on the Tandilia stratigraphy.

Structural framework The Tandilia system (Dalla Salda et al. 1988; Cingolani & Dalla Salda 2000) crops out in a 300-km-long chain of low hills extending in a NW –SE direction from the town of Olavarrı´a in the centre of the Buenos Aires province of Mar del Plata on the Atlantic coast (Fig. 54.1). It rests on a Palaeoproterozoic basement assigned to the Buenos Aires Complex (Marchese & di Paola 1975). It is unmetamorphosed (Go´mez Peral et al. 2007) and only lightly folded locally, such as near Olavarrı´a in the NW part of the belt (Massabie & Nestiero 2005). A low-angle unconformity occurs between the Cambrian Cerro Negro Fm. and the overlying Ordovician Balcarce Fm. Glaciogenic deposits of the Sierra del Volca´n Fm. are completely undisturbed, cropping out with subhorizontal bedding and no metamorphic overprint. The Sierras Bayas Group records sedimentation in a shallow marine basin with strong tidal influence and reworking (see Poire´ & Spalletti 2005). The Balcarce Fm. was deposited during a subsequent sedimentary cycle in a more extensive basin. It overlies palaeohighs of basement or the Sierras Bayas Group rocks, with a very low angle of unconformity in places (Pazos et al.

2008). The last confirmed evidence of magmatism in the Tandilia System is a series of diabase sills (Los Barrientos) dated by Rapela et al. (1974) as lower Ordovician and probably associated with the Ordovician Famatinian magmatic cycle in the western Proto-Andean continental margin of Gondwana (Pankhurst & Rapela 1998). Teruggi & Kilmurray (1975) proposed that the present landscape of the Tandilia system is due to Miocene Andean reactivation of basement faults. However, from a detailed geomorphological study, Demoulin et al. (2005) recognized two regional planation surfaces, which suggests that most of the present landscape is significantly older (Palaeogene or even Cretaceous), and that the region has experienced slow denudation rates.

Stratigraphy The main geological units within the Tandilia belt are a Palaeoproterozoic igneous and metamorphic basement (The Buenos Aires Complex) and a Neoproterozoic to Early Palaeozoic cover comprised of Ediacaran (?) calcareous and clastic sediments of the Sierras Bayas Group, the clastic (Cambrian) Cerro Negro Fm., and quartzite of the Ordovician Balcarce Fm., which gently overlaps the underlying units and marks an extensive transgression across the basin. The Buenos Aires Complex includes various igneous and metamorphic suites. Sm–Nd model ages averaging 2620 + 80 Ma on several magmatic units suggest a Late Archaean history (Pankhurst et al. 2003). Hartmann et al. (2002) recognized orogenic peaks around 2.16 Ga and 2.08 Ga in Uruguay (eastern part of the craton). The Sierras Bayas Group corresponds approximately to what was originally described as the La Tinta Fm. (In˜iguez Rodriguez et al. 1989; Marchese & di Paola 1975) and the Sierras Bayas Fm. (Borrelo 1962). This group, according to Poire´ & Spalletti (2005), presents a simple stratigraphy and covers the basement rocks of the Buenos Aires Complex (Fig. 54.2). The basal Villa Mo´nica Fm. (52–70 m thick) is composed of quartz-arenite, pelite, conglomerate and cherty dolomite beds with stromatolites (Poire´ & Spalletti 2005). It is overlain by the Cerro Largo Fm., which is c. 40 m thick and includes a diamictite containing mainly chert clasts and slumped beds at the base. Poire´ & Spalletti (2005) interpret this diamictite as glaciogenic in origin. The diamictite passes upward into multicolour pelite, glauconitic sandstone units and whitish quartzite beds containing hummocky, planar and trough cross-stratification. The succession ends with a shallowing-upward

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 565– 569. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.54

566

P. J. PAZOS & A. RAPALINI

Fig. 54.1. The Tandilia System and outcrops of the Sierra del Volca´n Fm. at del Volca´n Hill. The Punta Mogotes Borehole is located just south of Mar del Plata.

claystone cycle (Go´mez Peral et al. 2007). The Olavarrı´a Fm., which follows upwards, is composed of red, fine-grained rippled sandstone, pelite and heterolithic sediments deposited in a shallow, marginal marine environment judging from the occurrence of bimodal palaeocurrents (Poire´ et al. 2005). The youngest unit of the Sierras Bayas Group is the Loma Negra Fm., which is c. 40 m thick and composed almost exclusively of reddish and black limestone deposited in an open marine ramp to lagoonal environment. It contains Cloudina, which indicates a late Ediacaran age and tropical depositional environment (Poire´ & Spalletti 2005; Gaucher et al. 2003, 2005; Go´mez Peral et al. 2007, Pecoits et al. 2008). The overlying Cerro Negro Fm. is disconformably separated from the upper Sierras Bayas Group by a palaeokarst surface (Barrio et al. 1991). This .150-m-thick unit is composed of reddish fine-grained sandstone to claystone deposited in a shallow marine to tidal environment. Poire´ & Spalletti (2005) correlated the Cerro Negro Fm. with the Sierra del Volca´n and Punta Mogotes formations, which occur in the southeastern part of the belt, and suggested a Cambrian age for all three units. Go´mez Peral et al. (2007, fig. 2) instead argued that the Cerro Negro

ERAS PERIODS ORDOVICIAN

LITOSTRATIGRAPHIC UNITS SE REGION NW REGION Balcarce Formation

CAMBRIAN

LATE EDIACARAN

EDIACARAN

PALAEO-MESO PROTEROZOIC

Sierras Bayas Group

Cerro Negro Formation Loma Negra Formation Olavarría Formation

Borehole

CENTRAL REGION

Balcarce Formation

Balcarce Formation

Cerro Negro Formation Loma Negra Formation

Cerro Largo Formation

Las Águilas and Cerro Largo Formation

Villa Mónica Formation

Villa Mónica Formation

CerroNegro Formation (units 2-3)

URUGUAY upper Las Ventanas Formation

Arroyo del Soldado Group

Palaeohigh ? Del Volcán Formation

BUENOS AIRES COMPLEX

UNIT 4

Playa Hermosa Formation

Sierras Bayas Group (units 5-7)

Punta Mogotes Formation (unit 8)

ARCHAEAN COMPLEX

Fig. 54.2. Integrated stratigraphy (after Poire´ & Spalletti 2005), including the Punta Mogotes borehole stratigraphy suggested in this chapter, and the suggested stratigraphic position for the Sierra del Volca´n Fm. Diamictite intervals marked with triangles.

´ guilas, the Punta Mogotes Fm. is laterally equivalent to the Las A and the Sierra del Volca´n formations. We propose a distinct stratigraphic framework (Fig. 54.2) based on a description by Marchese & di Paola (1975) of the Tandilia stratigraphy in the Punta Magotes borehole in the southestern part of the belt. These authors recognized eight units within the well. The lowermost, unit 8 (89 m thick), is composed of metapelite beds that they redefined as the Punta Mogotes Fm. and correlated with the regional basement. Units 7– 5 comprise quartz arentite and heterolithic sediments. Unit 4 is a diamictite (6 m thick), which we suspect to be glaciogenic (see below). It is followed by more siliciclastic units 2–3, which may correlate with either the Cerro Negro or Balcarce formations. Unit 1 is assigned to the Balcarce Fm.

Glaciogenic deposits and associated strata The glaciogenic deposits of the Sierra del Volca´n Fm. are extremely thin (10 m thick), and the outcrops are almost entirely covered by vegetation and debris from the overlying Balcarce Fm. Exposures occur on private property with restricted access (Estancia Cerro Del Volca´n). The sedimentological description is based on recent fieldwork and descriptions given in Spalletti & del Valle (1984). The unit was divided into three intervals: (i) a lower diamictite, (ii) a middle heterolithic and pelitic unit, and (iii) an upper diamictite. The lower diamictite is polymictic with clasts of the Buenos Aires complex, but no carbonate clasts. The matrix contains abundant kaolinite, reminiscent of the saprolitized basement and the matrix of the Balcarce Fm. The middle interval is rich in outsized clasts disrupting the stratification, which exhibits undulatory and faintly developed ripples. The overlying diamictite is normally graded (Spalletti & del Valle 1984). Evidence of a glacial influence in deposition of the Sierra del Volca´n Fm. includes glacial pavements, rainout deposits, faceted clasts and matrix-supported diamictite beds with immature composition (Pazos et al. 2008). Another diamictite occurs at the contact between the Villa Monica and Cerro Largo formations (Go´mez Peral et al. 2007, Fig. 54.3) in the northwestern sector of the belt. The slumped diamictite contains mainly chert clasts and was interpreted by Poire´ & Spalletti (2005) as glacial in origin, although no direct evidence of glacial influence has been reported. Another diamictite occurs at a depth of 264 –270 m (unit 4) in the Punta Mogotes borehole (Marchese & di Paola 1975). This 6-m-thick diamictite contains fragments of feldspar and of metamorphosed pelite believed to be derived from the Punta Mogotes Fm. and is compositionally distinct from bracketing sediments. No unequivocal evidence for glaciation has been observed, but based on the stratigraphic arguments described below, we tentatively correlate unit 4 with the Sierra del Volca´n Fm.

THE TANDILIA SYSTEM, ARGENTINA

567

low-angle unconformity (Spalletti & del Valle 1984). The nature of the contacts of the diamictite comprising unit 4 in the Punta Mogotes borehole is unkown.

Chemostratigraphy Go´mez Peral et al. (2007) published reconnaissance chemostratigraphic data on the Villa Mo´nica and Loma Negra formations of the Sierras Bayas Group. To summarize, the Villa Mo´nica carbonates are heavily diagenetically altered, with very low Sr concentrations and enrichments in Fe and Mn. d13C values define a narrow range from –0.5 to 2.2‰. The Loma Negra Fm. is moderately altered, with Sr contents of 300 –400 ppm. d13C carbonate values cluster between 3 and 4‰, whereas organic matter values lie between – 27 and –28‰.

Palaeolatitude and palaeogeography The available Neoproterozoic palaeomagnetic data from the Rı´o de la Plata (RP) craton were reviewed in Tohver et al. (2006) and Rapalini (2005, 2006). A preliminary apparent polar wander track for the interval 600– 500 Ma has been proposed recently for this craton (Rapalini 2006; Fig. 54.4). The available palaeomagnetic database to constrain the palaeogeographical evolution of the RP craton consists of four palaeomagnetic poles and one mean geomagnetic pole. The latter means that there is no guarantee of full compensation of non-dipole components of the Earth magnetic field. The oldest available pole was obtained from the Campo Alegre ´ Agrella Filho & Pacca 1988) in Brazil. These rocks lavas (CA, D were accurately dated as 595 + 5 Ma (U – Pb, Citroni et al. 1999). The similar position of the mean geomagnetic pole for the Playa Hermosa sediments (PH) to CA (Fig. 54.4) suggests a broadly similar age for both (Pazos et al. 2011). The other poles in the APWP are nearly identical to those of the older units of the Sierra de Animas Complex (SA2, Sanchez Bettucci & Rapalini 2002) and Los Barrientos claystone (LB, Rapalini 2006) in the Tandilia System (Table 54.1). Originally, an age of c. 550 Ma was assigned to SA2 based on old K –Ar and Rb –Sr ages. Oyhantc¸abal et al. (2007) published an Ar/Ar age of 579 + 2 Ma on hornblende for the Pan de Azucar syenite, interpreted as a cooling age of that body. LB is a high-quality pole with no firm age constraints, although its position being similar to SA2 suggests an age between 580 and 570 Ma. Taking into consideration the apparent polar wander path in Figure 54.4 and Ediacaran to Cambrian poles from other

Fig. 54.3. Detailed logged section of the Sierra del Volca´n Fm. taken from Pazos et al. (2008). (a) Location of the interval studied. (b) Logged section originally from Spalletti & del Valle (1984).

Boundary relations with overlying and underlying non-glacial units The Sierras Bayas Group is bounded by stratigraphic unconformities (e.g. Poire´ & Spalletti 2005; Go´mez Peral et al. 2007). The non-conformity that separates the basement (Buenos Aires Complex) from the Sierras Bayas Group represents hundreds of millions of years during which the basement was intensely chemically weathered. An unconformity also separates the Villa Mo´nica Fm. from the overlying Cerro Largo Fm., but the nature and magnitude of the gap is subject to debate (Pazos et al. 2008). The Sierra del Volca´n Fm. sits non-conformably on the basement and is separated from the overlying Balcarce Fm. by a

Table 54.1. Available Neoproterozoic palaeomagnetic poles for the Rio de la Plata Geographical location

Formation/lithology

Uruguay

Sierra de las Animas Complex (SA1) Sierra de las Animas Complex (SA2) Sierra de los Barrientos claystones (LB) Playa Hermosa Fm. (PH) Campo Alegre lavas (CA)

Uruguay Tandilla System

Uruguay Brazil

Palaeomagnetic results

Palaeopoles age

Q

Ref.

5.98 338.18 (19.68/26.78)

AF/Th demag, N ¼ 7 (n ¼ 33), IRM

3

1

–16.98 250.98 (15.98/21.58)

AF/Th demag, N ¼ 6 (n ¼ 27), IRM, both polarities AF/Th demag, N ¼ 12 (n ¼ 56), both polarities, reversal of EMF recorded. Correction for remanence anisotropy AF/Th demag, N ¼ 2 (n ¼ 6), IRM, small number of samples AF/Th demag, N ¼ 6 (n ¼ 46), both polarities.

Cambrian (520– 500 Ma) Ediacaran (579 Ma) Ediacaran?

4

1

6

2

Early Ediacaran?

3

1

Early Ediacaran (595 Ma)

5

3

Palaeomagnetic pole (PP) (lat.) (long.) A95 (dp/dm)

–15.18 252.68 (10.98/14.28)

–43.08 198.48 (8.68/16.08) –57.08 223.08 9.08

´ Agrella Filho & Q, quality factor of the pole (maximum, 7) according to criteria set by Van der Voo (1990). 1, Sa´nchez Bettucci & Rapalini (2002); 2, Rapalini (2006); 3, D Pacca (1988). EMF, Earth magnetic field; IRM, isothermal remanent magnetization; AF, alternating field; Th, thermal.

568

P. J. PAZOS & A. RAPALINI

observations and interpretations of the Sierra del Volca´n Fm. can be summarized as follows: 500 SA1

525

510 535

550 LB SA2 (579) CA (595 ) PH Fig. 54.4. Palaeomagnetic poles and their respective circles (or ovals) of confidence for the Rio de la Plata (RP) craton (black squares with white ovals), and other Gondwana blocks (white circles with grey ovals). Poles from remagnetized units (Rapalini & Sanchez Bettucci 2008) have been omitted. The hypothetical apparent polar wander track for this craton is shown in dashed lines. Numbers indicate approximate ages of poles in Ma. Data for RP poles are presented in Table 54.1. Gondwana poles selected from Trindade et al. (2006). SA1, Sierra de Animas 1; SA2, Sierra de Animas 2; PH, Playa Hermosa; LB, Los Barrientos; CA, Campo Alegre poles.

Gondwanan cratonic blocks (Trindade et al. 2006), the RP craton appears to have lain between 308S and 608S through the Ediacaran and Early Cambrian.

Geochronology In the Villa Mo´nica Fm., Rb –Sr data on diagenetic clay minerals yielded an age of 795 + 28 Ma on a green shale. A Rb – Sr age of 769 + 12 Ma was obtained on shale deposits of the overlying Cerro Largo Fm. (Bonhomme & Cingolani 1980), while Cingolani & Bonhomme (1988) and Kawashita et al. (1999) suggested a depositional age of c. 730 Ma for the same formation based on Rb –Sr whole-rock data. However, these ages are all regarded as too old and, based on the palaeomagnetic considerations described above, an Ediacaran age for the Sierra Bayas Group is more likely. Detrital zircon geochronology (Rapela et al. 2007; Gaucher et al. 2007) has provided no meaningful age constraints on the Sierras Bayas Group, but has established a maximum Early Ordovician age for the Balcarce Fm. (Rapela et al. 2007). The body fossil Cloudina riemkeae occurs in the Loma Negra Fm. (Gaucher et al. 2005; Pecoits et al. 2008) and confirms an Ediacaran age, at least for the upper Sierras Bayas Group. Chauria circularis has been reported from Villa Mo´nica and Cerro Negro formations and leiosphaeridia-dominated acritarch assemblages are found from the upper Villa Mo´nica Fm. to the middle Cerro Negro Fm. (Gaucher et al. 2005).

Discussion The diamictite beds of the Sierra del Volca´n Fm. are identified as glacial in origin based on a range of features and interpretations described in detail in Pazos et al. (2008). The important

† The diamictite at the base and top of the unit (Fig. 54.3) appears to have been deposited by debris flows, as is particularly evident in the normal grading of the upper diamictite. † The clasts in the diamictite beds are derived mainly from the metamorphic rocks of the Buenos Aires Complex, as well as unmetamorphosed quartzite clasts and blocks, probably belonging to the quartzite beds of the Sierras Bayas Group present in the Villa Mo´nica (lower quartzite) and Cerro Largo (upper quartzite) formations. † The micromorphology of quartz grains is consistent with a glacial origin (van Staden et al. 2005). † The middle unit contains dropstones that clearly disrupt bedding (Spalletti & del Valle 1984) and symmetric ripples indicating a combination of open water and floating ice. The diamictite documented at the base of the Cerro Largo Fm. is a cherty breccia that lacks evidence of glacially influenced sedimentation. The diamictite interval documented in the Punta Mogotes borehole (unit 4) described in detail by Marchese & di Paola (1975) is very thin (6 m thick) and differs compositionally with both underlying and overlying sediments. It contains fresh feldspar and angular basement clasts floating in a clayish matrix. These features indicate a low degree of weathering of clast material, but not sufficient to assign an unambiguous glacial origin (e.g. Eyles & Januszczak 2004). The absence of carbonate clasts, despite the low weathering index, imply that the unit 4 diamictite is older than the Loma Negra Fm., and thus indisputably Neoproterozoic in age. The Sierra del Volca´n Fm. similarly lacks carbonate clasts and we suggest that it correlates with the unit 4 diamictite (Fig. 54.2). This correlation requires an explanation for the absence of the Loma Negra carbonates in the Punta Mogotes borehole. We interpret this to be the result of a topographic high in the Punta Mogotes region, which was likely also a source area for clasts in the Sierra del Volca´n Fm. We thank Dr R. Somoza and Professor V. Ramos for providing bibliography, discussions and stimulation, although we stress that we accept full responsibility for the ideas presented here. I want to express my gratitude (P.P.) to Cheryl Turkington for her unconditional support. This is the contribution R-38 of the Instituto de Estudios Andinos Don Pablo Groeber. This represents a contribution of the IUGS and UNESCO-funded IGCP (International Geoscience Programme) project #512.

References Barrio, C. A., Poire´, D. G. & In˜iguez Rodrı´guez, A. M. 1991. El contacto entre la Formacio´n Loma Negra (Grupo Sierras Bayas) y la Formacio´n Cerro Negro, un ejemplo de paleokarst, Olavarrı´a, provincia de Buenos Aires. Revista de la Asociacio´n Geolo´gica Argentina, 46, 69–76. Bonhomme, M. G. & Cingolani, C. 1980. Mineralogı´a y geocronologı´a Rb –Sr y K–Ar de fracciones finas de la Formacion ‘La Tinta’, provincia de Buenos Aires. Revista de la Asociacio´n Geolo´gica Argentina, 35, 519–538. Borello, A. 1962. Formacio´n Punta Mogotes (Eopaleozoico Provincia de Buenos Aires). Comisio´n Nacional de Investigaciones Cientı´ficas. Notas 1, 1. La Plata. Cingolani, C. & Bonhomme, M. G. 1988. Resultados geocronolo´gicos en niveles pelı´ticos intercalados en las dolomı´as de las Sierras Bayas (Grupo La Tinta), provincia de Buenos Aires. Actas Segundas Jornadas Geolo´gicas, Actas I, 2283– 2389. Cingolani, C. & Dalla Salda, L. 2000. Buenos Aires Cratonic region. In: Cordani, U., Milani, E., Thomaz Filho, A. & Campos, D. (eds) Tectonic Evolution of South America. 31st International Geological Congress, Brazil, Rio Janeiro, 139–147. Citroni, S. B., Basei, M. A. S., Sato, K. & Siga, O. Jr. 1999. Petrogenesis of the Campo Alegre basin magmatism, based on geochemical and isotopic data. II South American Symposium on Isotope Geology, Cordoba, Argentina, Actas, 174– 177.

THE TANDILIA SYSTEM, ARGENTINA

D’Agrella Filho, M. S. & Pacca, I. I. G. 1988. Paleomagnetism of the Itajaı´, Castro and Bom Jardim Groups from Southern Brazil. Geophysical Journal International, 93, 365–376. Dalla Salda, L., Bossi, J. & Cingolani, C. 1988. The Rio de la Plata cratonic regio´n of southwestern Gondwana. Episodes, 11, 263–269. Demoulin, A., Za´rate, M. & Rabassa, J. 2005. Long-term landscape development: a perspective from the southern Buenos Aires ranges of east central Argentina. Journal of South American Earth Sciences, 19, 193– 204. Eyles, N. & Januszczak, N. 2004. ‘Zipper-rift’: a tectonic model for Neoproterozoic glaciations during the break up of Rodinia after 750 Ma. Earth-Science Reviews, 65, 1 – 73. Gaucher, C., Boggiani, P. C., Sprechmann, P., Sial, A. N. & Fiarchild, T. 2003. Integrated correlation of the Vendian to Cambrian Arroyo del Soldado and Corumba˜ Groups (Uruguay and Brazil) – palaeogeographic, palaeoclimatic and palaeobiologic implications. Precambrian Research, 120, 241– 278. Gaucher, C., Poire´, D. G., Peral, L. & Chiglino, L. 2005. Litoestratigrafı´a, Bioestratigrafı´a y correlaciones de las sucesiones sedimentarias del Neoproterozoico –Cambrico del Crato´n del Rı´o de la Plata (Uruguay y Argentina). Latin American Journal of Sedimentology and Basin Analysis, 12, 145–160. Gaucher, C., Poire´, D., Finney, S. C., Valencia, V., Blanco, G., Pamoukaghlian, K. & Go´mez Peral, L. 2007. Zircones detrı´ticos de secuencias neoproterozoicas de Uruguay y Argentina: inferencias sobre la evolucio´n paleogeogra´fica del carto´n del Rı´o de la Plata. V Congreso Uruguayo de Geologı´a, Montevideo, Uruguay. (CD-ROM, extended paper). Go´mez Peral, L., Poire´, D. G., Strauss, H. & Zimmermann, U. 2007. Chemostratigraphy and diagenetic constraints on Neoproterozoic carbonate successions from the Sierras Bayas Group, Tandilia System, Argentina. Chemical Geology, 237, 109–128. Hartmann, L. A., Santos, J. O. S., Cingolani, C. A. & MacNaugthon, N. J. 2002. Two Paleoproterozoic orogenies in the evolution of the Tandilia Belt, Buenos Aires, as evidenced by zircon U –Pb SHRIMP geochronology. Int. Geology Rev., 44, 528– 543. In˜iguez Rodrı´guez, A. M., Del Valle, A., Poire´, D., Spalletti, L. A. & Zalba, P. 1989. Cuenca Preca´mbrica/Palaeozoica inferior de Tandilia, Provincia de Buenos Aires. In: Chebli, W. & Spalletti, L. A. (eds) Cuencas Sedimentarias Argentinas. Serie Correlacio´n Geolo´gica, San Miguel del Tucuma´n, 6, 245– 263. Kawashita, K., Varela, R. et al. 1999. Geocronology and chemostratigraphy of ‘La Tinta’ Neoproterozoic sedimentary rocks, Buenos Aires Province, Argentina. Actas II. South American Symposium on Isotope, 403–407. Marchese, H. G. & di Paola, E. C. 1975. Miogeosinclinal Tandil. Revista de la Asociacio´n Geolo´gica Argentina, 30, 161–179. Massabie, A. C. & Nestiero, O. E. 2005. La estructura del Grupo Sierras Bayas en el sector norte de las Sierras Septentrionales de Buenos Aires. Revista de la Asociacio´n Geolo´gica Argentina, 60, 135–146 Oyhantc¸abal, P., Siegesmund, A., Wemmer, K., Frei, R. & Layeretal, P. 2007. Post-collisional transition from calc-alkaline to alkaline magmatism during transcurrent deformation in the southernmost Dom Feliciano Belt (Braziliano –Pan-African, Uruguay). Lithos, 98, 141– 159. Pankhurst, R. J. & Rapela, C. W. (eds) 1998. The Proto-Andean Margin of Gondwana. Geological Society of London, Special Publication, 142. Pankhurst, R. J., Ramos, A. & Linares, E. 2003. Antiquity of the Rı´o de la Plata craton in Tandilia, southern Buenos Aires Province, Argentina. Journal of South American Earth Sciences, 16, 5 –13. Pazos, P. J., Sa´nchez Bettucci, L. & Loureiro, J. 2008. The Neoproterozoic glacial record in the Rı´o de la Plata Craton: a critical reappraisal. In: Pankhurst, R. J., Trouw, R. A. J., de Brito Neves,

569

B. B. & de Witt, M. J. (eds) West Gondwana Pre-Cenozoic Correlations Across the South Atlantic Region. Geological Society, London, Special Publications, 294, 343–364. Pazos, P. J., Rapalini, A., Sa´nchez Bettucci, L. & To´falo, R. O. 2011. The Playa Hermosa Formation, Playa Verde Basin, Uruguay. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 547– 553. Pecoits, E., Gingras, M., Aubet, N. & Konhauser, K. 2008. Ediacaran in Uruguay: palaeoclimatic and palaeobiological implications. Sedimentology, 55, 689– 721. Poire´, D. G. & Spalletti, L. A. 2005. La cubierta sedimentaria preca´mbrica/paleozoica inferior del Sistema de Tandilia. In: Barrio, R. E., Etcheverry, R. O., Caballe´, M. F. & Llambı´as, E. J. (eds) Geologı´a y Recursos Minerales de la provincia de Buenos Aires. Relatorio del XVI Congreso Geolo´gico Argentino, 51 –68. Poire´, D. G., Canalicchio, J. & Alonso, G. 2005. Las calizas del Sistema de Tandilia y su utilizacio´n en la industria cementera. In: Barrio, R. E., Etcheverry, R. O., Caballe´, M. F. & Llambı´as, E. J. (eds) Geologı´a y Recursos Minerales de la provincia de Buenos Aires. Relatorio del XVI Congreso Geolo´gico Argentino, 387– 396. Rapalini, A. E. 2005. The accretionary history of Southern South America from the latest Proterozoic to the Late Paleozoic: some paleomagnetic constraints. In: Vaughan, A., Leat, P. & Pankhurst, R. J. (eds) Terrane Processes at the Margins of Gondwana. Geological Society, London, Special Publication, 246, 305– 328. Rapalini, A. E. 2006. New Late Proterozoic paleomagnetic pole for the Rio de la Plata craton: implications for Gondwana. Precambrian Research, 147, 223–233. Rapalini, A. E. & Sanchez Bettucci, L. 2008. Widespread remagnetization of Late Proterozoic sedimentary units of Uruguay and the apparent polar wander path for the Rio de la Plata craton. Geophysical Journal International, 174, 55 – 74. Rapela, C., Dalla Salda, L. & Cingolani, C. 1974. Un intrusivo ba´sico Ordovı´cico en la Formacio´n La Tinta (Sierra de los Barrientos), Provincia de Buenos Aires. Revista de la Asociacio´n Geolo´gica Argentina, 29, 319–331. Rapela, C. W., Pankhurst, R. J. et al. 2007. The Rio de la Plata craton and the assembly of SW Gondwana. Earth Science Reviews, 83, 49– 82. Sa´nchez Bettucci, L. & Rapalini, A. E. 2002. Paleomagnetism of the Sierra de Las Animas Complex, Southern Uruguay: Its Implications in the Assembly of Western Gondwana. Precambrian Research, 118, 243–265. Spalletti, L. & Del Valle, A. 1984. Las diamictitas del sector oriental de Tandilia: caracteres sedimentolo´gicos y origen. Revista de la Asociacio´n Geolo´gica Argentina, 39, 188– 206. Teruggi, M. E. & Kilmurray, J. 1975. Tandilia. In: Geologı´a de la Provincia de Buenos Aires. Relatorio del VI Congreso Geolo´gico Argentino, Buenos Aires, 55 –77. Tohver, E., D’Agrella Filho, M. S. & Trindade, R. I. F. 2006. Paleomagnetic record of Africa and South America for the 1200–500 Ma interval, and evaluation of Rodinia and Gondwana assemblies. Precambrian Research, 147, 193– 222. Trindade, R. I. F., D’Agrella-Filho, M. S., Epof, I. & Brito Neves, B. B. 2006. Paleomagnetism of Early Cambrian Itabaiana mafic dikes (NE Brazil) and the final assembly of Gondwana. Earth Planetary Science Letters, 244, 361– 377. Van der Voo, R. 1990. The reliability of paleomagnetic data. Tectonophysics, 184, 1– 9. van Staden, A., Zimmermann, U. & Germs, G. J. 2005. Provenance and depositional study on tillites from the Volcan Hill, Tandilia System in east Argentina: preliminary results. XVI Congreso Geolo´gico Argentino, Actas, I, 239–246.

Chapter 55 Glacial sediments and associated strata of the Polarisbreen Group, northeastern Svalbard GALEN P. HALVERSON1,2 1

School of Earth and Environmental Sciences, The University of Adelaide North Terrace, Adelaide, SA 5005, Australia

2

Present address: Department of Earth and Planetary Sciences, McGill University, 3450 University Street, Montreal, QC, H3A 2A7, Canada (e-mail: [email protected]) Abstract: Northeastern Svalbard hosts exceptionally well-preserved Neoproterozoic sediments. The glaciogenic Petrovbreen Member and Wilsonbreen Formation (Fm.) of the Polarisbreen Group crop out in a narrow, Caledonian-aged fold-and-thrust belt spanning from Olav V Land on Spitsbergen in the south to western Nordaustlandet in the north. The older Petrovbreen Member is thin (0– 52 m) and patchily preserved, comprising mainly poorly stratified, dolomite-matrix diamictite likely deposited in a marine setting. The basal contact of the Petrovbreen Member erosionally truncates the upper Russøya Member, which preserves a large (12‰) negative C-isotope anomaly. The Petrovbreen Member is overlain by 200 m of dark, monotonous shales of the MacDonaldryggen Member, followed by cherty dolomite grainstone and microbiolamintes of the Slangen Member. The upper Slangen Member is an exposure surface in the southern part of the belt, but in northern Spitsbergen and on Nordaustlandet is transitional into sands of the northward-thickening Bra˚vika Member. The Wilsonbreen Fm. is typically 100– 150 m thick and consists mostly of massively to poorly stratified diamictite, with subordinate sand beds, conglomerate lenses, and carbonates deposited in a terrestrial environment. It is overlain by the colourful Dracoisen Fm., which records, at its base, a typical post-glacial negative d13C anomaly. There are no direct radiometric age constraints or reliable palaeomagnetic data from the Polarisbreen Group, but it is widely accepted that northeastern Svalbard was contiguous with East Greenland during the Neoproterozoic.

Late Neoproterozoic glaciogenic rocks occur in both western and northeastern Svalbard (Harland et al. 1993). The much better studied and preserved glaciogenic sediments and bracketing strata of the upper Hecla Hoek succession in northeastern Svalbard are the subject of this chapter. These rocks occur in Olav V Land and Ny Friesland in northeastern Spitsbergen and in Gustav V Land in northwestern Nordaustlandet (Fig. 55.1). These rocks are best exposed on nunataks in Olav V Land and southern Ny Friesland, but occur also over a large area in Nordaustlandet along the eastern side of Hinlopenstretet, and to a lesser extent, in coastal exposures in northern Ny Friesland. In Spitsbergen, two distinct glacial units, named the Petrovbreen Member (older) and Wilsonbreen Fm. (younger) occur within the middle of the mixed clastic-carbonate Polarisbreen Group (Fig. 55.2). Both the Petrovbreen Member and the Wilsonbreen Fm., together with bracketing strata, are readily identified in Nordaustlandet (Halverson et al. 2004), where a separate nomenclature has historically been used (Kulling 1934). Because correlation across Hinlopenstretet (Fig. 55.1) is unambiguous (Kulling 1934; Harland et al. 1993; Fairchild & Hambrey 1995), only the better known nomenclature from Spitsbergen (Harland 1997) is used here (Fig. 55.2). The history of geological investigation of Precambrian glacial deposits was recently summarized in Harland (2007). In short, the Neoproterozoic sedimentary succession in northeastern Svalbard was first studied by Nordenskio¨ld (1863). A glacial origin for the Polarisbreen diamictites (specifically the Wilsonbreen Fm.) was originally proposed by Kulling (1934), who mapped and named these rocks in Nordaustlandet and northern Ny Friesland during the Swedish– Norwegian Arctic Expedition of 1931. Fleming & Edmonds (1941) further investigated coastal exposures of the Hecla Hoek, including the Wilsonbreen Fm., but it was not until W. B. Harland and C. B. Wilson worked systematically on the upper Hecla Hoek succession on both coastal and inland sections on Spitsbergen (Harland & Wilson 1956; Wilson 1961; Wilson & Harland 1964) that specific attention was focused on the Neoproterozoic glacial rocks of Svalbard and their significance in Earth history (Harland 2007). Since this time, and with their global importance established (Harland 1964), several detailed sedimentological studies on the Polarisbreen glaciogenic

sediments have been carried out (e.g. Chumakov 1968; Edwards 1976; Hambrey 1982; Fairchild 1983; Dowdeswell et al. 1985), with the results culminating in a monograph by Harland et al. (1993). Carbonates within and bounding the Polarisbreen diamictites have also been the focus of coupled sedimentological – geochemical investigations (Fairchild & Hambrey 1984; Fairchild & Spiro 1987; Halverson et al. 2004). More broadly, Neoproterozoic sediments in Svalbard and East Greenland were the subject of the seminal chemostratigraphic study by Knoll et al. (1986) that first firmly established the intimate association between negative C-isotope anomalies and Precambrian glaciation. The upper Hecla Hoek succession also hosts an unusually rich microfossil record, which has been studied in detail by A. H. Knoll, N. J. Butterfield and colleagues (e.g. Knoll 1982, 1984; Butterfield et al. 1988, 1994). The excellent preservation of these sediments and their role in the ongoing debate over the cause and severity of Neoproterozoic glaciations overcompensate for the logistical difficulty involved in studying them. This contribution is intended to update the review by Hambrey et al. (1981) in the context of more recent research on the Polarisbreen Group and significant advances in our understanding of the Neoproterozoic Earth system since that time.

Structural framework The Svalbard archipelago, situated on the northwestern corner of the Barents Shelf, comprises three tectonic terranes that were amalgamated in the Silurian – Devonian Ny Friesland (Caledonian) orogeny (Harland & Gayer 1972; Harland et al. 1992; Gee & Page 1994; Lyberis & Manby 1999). The thick Hecla Hoek succession occurs in the northern part of the Eastern Terrane, which is the only exposed part of the Barentsia microcontinent (Breivik et al. 2002). The Polarisbreen Group is part of the relatively undeformed upper Hecla Hoek succession, which includes the Lomfjorden and Hinlopenstretet supergroups (Fig. 55.2). In Ny Friesland, the basal Lomfjorden Supergroup is juxtaposed against high-grade metasediments of the Stubendorfbreen Supergroup along the Eolussletta Shear Zone (Fig. 55.1; Lyberis & Manby 1999). The nature of this contact and the relationship between the lower and

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 571– 579. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.55

572

G. P. HALVERSON

Fig. 55.1. Geological sketch map of the region in northeastern Svalbard where the Polarisbreen Group and equivalent rocks crop out. The mapping is from Dallmann et al. (2002), with modifications to the Neoproterozoic strata based on mapping by the author. Letters a to e refer to principal regions in which the Polarisbreen Group has been studied: a, Akademikerbreen area; b, Ditlovtoppen; c, Dracoisen/Oslobreen; d, Sveanor (south coast of Murchisonfjord), e, Aldousbreen (north coast of Wahlenbergfjord). ESZ, Eolusletta Shear Zone (cf. Lyberis & Manby 1999); LFZ, Lomfjorden Fault Zone.

upper Hecla Hoek has been the subject of a long-standing debate, with one side arguing that the two packages form a single, conformable stratigraphic sequence (e.g. Harland 1959; Harland et al. 1992). Most recent workers, however, agree that this contact represents a significant hiatus associated with Grenvillian orogenesis (e.g. Gee et al. 1995; Manby & Lyberis 1995;

Witt-Nilsson et al. 1998), an argument that is supported by mapping and geochronology in Nordaustlandet where the Lomfjord Supergroup rests unconformably on early Neoproterozoic volcanics and subvolcanic intrusives (Fig. 55.2), which themselves intrude and overlie high-grade, late Mesoproterozoic metasediments (Gee et al. 1995; Johansson et al. 2005). The broad,

THE POLARISBREEN GROUP, NORTHEASTERN SVALBARD

Fig. 55.2. Stratigraphic nomenclature and correlation of the Neoproterozoic successions in northeastern Spitsbergen and northwestern Nordaustlandet. Note that only the nomenclature from Spitsbergen is used in this chapter. Hinlopen. Spgp., Hinlopenstretet Supergroup; Lom. S., Lomfjorden Supergroup; St. S., Stubendorfbreen Supergroup. The ‘upper Hecla Hoek succession’, referred to in the text and many publications, is the Lomfjorden plus Hinlopenstretet supergroups.

conformable and aggradational nature of the upper Hecla Hoek and its continuity with the equivalent units in East Greenland (Fairchild & Hambrey 1995) imply that the glacial sediments of the Polarisbreen Group were deposited in a thermally subsiding basin (Maloof et al. 2006). The upper Hecla Hoek crops out in a 120-km-long, north – south belt (Fig. 55.1) that was deformed during the Caledonian orogeny. The Polarisbreen Group sits in the core of the north-plunging Hinlopenstrettet Synclinorium, which spans the length of the belt (Fig. 55.1). The eastern limb of the synclinorium, exposed in Nordaustlandet, is a fold-and-thrust belt, dissected by a network of conjugate NW –SE and SW –NE strike –slip faults (Sokolov et al. 1968). The western limb of the synclinorium is less deformed, but is similarly cut by Caledonian-aged reverse and strike –slip faults. (Lyberis & Manby 1999). In the central part of the belt, the upper Hecla Hoek is intruded by Silurian granitoids (Teben’kov et al. 1996). The upper Hecla Hoek was eroded and draped by Carboniferous sediments during post-orogenic, west– east extension (Harland 1959), but normal faulting was largely concentrated along the prominent Lomfjorden Fault Zone (LFZ; Fig. 55.1). Dolerites related to the Late Jurassic – Early Cretaceous High Arctic Large Igneous Province (Maher 2001) intrude the Hecla Hoek within the northern part of the belt in the Hinlopenstretet region (Fig. 55.1). The Late Cretaceous –Early Tertiary West Spitsbergen Orogen, which heavily deformed the Western Terrane (Harland 1959), only lightly affected the Eastern Terrane, but did result in reactivation of major lineaments, such as the LFZ (Larsen 1987; Lyberis & Manby 1993).

573

is sharp (Halverson et al. 2004). The top of the Polarisbreen Group is bound by the Cambrian Tokammane Fm. (base of the Oslobreen Group; Fig. 55.2). Although no erosional break at this contact is evident from mapping or outcrop exposures, a large depositional hiatus is implied by the biostratigraphy (Knoll & Swett 1987). The Polarisbreen Group (Figs 55.2 & 55.3) is subdivided into three formations (Wilson & Harland 1964): Elbobreen, Wilsonbreen and Dracoisen. The Elbobreen Fm. is further subdivided into four regionally mappable members, each of which could be accorded formation status in its own right: (from bottom to top) the E1 or Russøya Member, the E2 or Petrovbreen Member, the E3 or MacDonaldryggen Member, and the E4 or Slangen Member (Fig. 55.2). A fifth unit that only occurs in the north of the outcrop belt, the Bra˚vika member, was tentatively assigned to the Wilsonbreen Fm. (Halverson et al. 2004), but is clearly transitional below with Slangen Member, and thus may more appropriately belong to the Elbobreen Fm. The Petrovbreen Member and Wilsonbreen Fm. comprise the two glaciogenic units, both of which are distinct from one another sedimentologically and within the context of their bounding strata (Figs 55.3 & 55.4).

Glaciogenic deposits and associated strata Russøya Member (E1) The Russøya Member varies from 75 to 170 m in thickness, thinning from south to north (Halverson et al. 2004). It is a lithologically variable unit, being predominantly carbonate in Nordaustlandet and containing an increasing fraction of sandstone and siltstone to the south. Despite its heterogeneity, the stratigraphic geometry of the Russøya Member is consistent, comprising a relatively thick (40– 160 m), basal, shoaling-upward sequence, overlain by up to four complete parasequences, 5–20 m thick (Fig. 55.3). The lowest parasequence consists of a basal dark shale that grades upward into increasingly carbonate-rich sediments, including limestone ribbonites and grainstones. Molar tooth structures are prominent in this sequence, but occur exclusively within the ribbonite facies (Fig. 55.3). The upper parasequences typically consist of shale at the base and sandstone or dolomitic grainstone, stromatolites or microbial laminites at the top. The uppermost parasequence is capped by a 2– 6-m-thick biostrome of a spiralling, columnar stromatolite (Kussiella) and brecciated and variably truncated on an outcrop scale (Halverson et al. 2004). Although erosional truncation below the Petrovbreen Members is ,2 m on the outcrop scale, regional correlations, assisted by chemostratigraphy (see below), suggest as much as 50 m of erosion beneath the Petrovbreen Member (Halverson et al. 2004).

Stratigraphy

Petrovbreen Member (E2)

The Polarisbreen Group (formerly Series) was originally defined by Wilson & Harland (1964) in the Dracoisen area in southern Ny Friesland, near the confluence of the Polarisbreen and Chydeniusbreen glaciers (Fig. 55.2). It conformably overlies the thick stack (2 km) of Akademikerbreen Group carbonates and broadly marks a transition from predominantly carbonate to mixed carbonate-clastic sedimentation. The contact between these two groups is easily identified throughout the outcrop belt, with the lithologically distinctive, pale yellow ‘Dartboard Dolomite’ (Wilson 1961; Wilson & Harland 1964) of the uppermost Akademikerbreen Group contrasting with the laminated black shale and bluish-grey to black limestone of the lower Polarisbreen Group (Halverson et al. 2004). The nature of this contact is variable. In some places, it is unambiguously transitional. Elsewhere, the upper Dartboard Dolomite is brecciated and silicified, and contains tepee-like structures, and the contact with the overlying limestone

The Petrovbreen Member is on average only about 10 m thick (Fig. 55.4), but is highly variable (,52 m), its thickness broadly reciprocating the depth of erosion on the sub-glacial sequence boundary (Halverson et al. 2004). In Nordaustlandet, the Petrovbreen Member is typically very thin and had been recognized only in northernmost sections (Flood et al. 1969; Hambrey 1982; Harland et al. 1993), but recent mapping has shown that it can be identified in most sections (Halverson et al. 2004). The yellow- to orange-weathering Petrovbreen Member is predominantly composed of poorly stratified and massive diamictite and wackestone with a yellow- to orange-weathering dolomite matrix. Other common facies include dolomitic rhythmites, tabular clast conglomerates and sandy dolostone (Fairchild & Hambrey 1984). Climbing ripples occur near the top of the member in some sections. The clasts in the diamictite range up to 90 cm in diameter, are subangular to angular, and consist mainly of light grey dolomite

574

G. P. HALVERSON

Fig. 55.3. Composite stratigraphic column through the Polarisbreen Group (after Halverson et al. 2004) and associated chemostratigraphic and biostratigraphic data. Sr-isotope data are from Halverson et al. (2007), C-isotope data (v. VPDB) on organic matter (d13Corg) from Knoll et al. (1986) and Kaufman et al. (1997) (as compiled in the latter), and C-isotope data on carbonates (d13Ccarb) from Fairchild & Spiro (1987), Halverson et al. (2004) and Kaufman et al. (1997). The two principal acritarch assemblages in the Polarisbreen Group (right side of figure) are summarized from Knoll (1982) and Knoll & Swett (1987).

(grainstone and stromatolite) and black chert. Minor sandstone, siltstone and rare volcanic clasts are also found (Hambrey 1982; Harland et al. 1993). Evidence for direct glacial influence in the Petrovbreen Member comes from rare striated clasts and the abundance of dropstones in finely laminated sediments (Hambrey 1982; Harland et al. 1993), as well as evidence of glacial rock flour in the form of submicrometre-sized dolomite in the matrix of the diamictite (Fairchild 1983). Fe and 18O-enrichment of this dolomitic matrix relative to clasts suggests early marine alteration (Fairchild et al. 1989). The absence of features indicative of a terrestrial environment, coupled with rapid facies changes and the occurrence of dropstones, suggest glaciomarine (or glaciolacustrine) deposition near an ice-grounding line (Hambrey 1982; Fairchild & Hambrey 1984; Harland et al. 1993).

MacDonaldryggen Member (E3) The contact between the upper Petrovbreen Member and the base of the MacDonaldryggen Member is everywhere sharp but conformable. In Olav V Land, where the MacDonaldryggen Member was named (Harland et al. 1993), the base consists of a thin (10–40 cm), finely laminated carbonate (variably limestone and dolomite) with wispy organic-rich laminae (Halverson et al. 2004). Elsewhere, the upper Petrovbreen Member is directly overlain by the same finely laminated, olive green to dark grey mudstone that comprises the bulk of the c. 200-m-thick MacDonaldryggen Member. The fraction of silt increases up-section, with thin, fine sandstone beds present in some sections. In Nordaustlandet, authigenic calcite nodules up to 2 cm in diameter, and in places densely clustered within distinct layers, speckle the upper MacDonaldryggen Member some 20–40 m beneath the contact with the overlying Slangen Member (Fig. 55.3). These nodules are commonly

rosette-shaped and have canted and square prismatic habits (Halverson et al. 2004). The upper MacDonaldryggen shales and siltstones are everywhere transitional into flaggy dolomites that mark the bottom of the Slangen Member. In southern sections, this transition includes an increase in carbonate, including limestone ribbons. Mud-cracks have been reported from this level (Fairchild & Hambrey 1984), but appear to be rare.

Slangen Member (E4) The Slangen Member is typically a 20 –30 m regressive parasequence of cherty dolostone. The dominant facies in the Slangen Member is a vuggy, trough and tabular cross-bedded grainstone; however, fenaestral microbial laminates also occur as thin interbeds within the grainstone and as distinct beds up to 5 m thick in the upper Slangen Member. Ribbons occur mainly in the transition with the underlying MacDonaldryggen Member, but are also important in the uppermost Slangen Member in Nordauslandet. The occurrence of length-slow chalcedony, high Na concentrations, and rare anhydrite in the Slangen Member suggest deposition in a restricted environment such as a coastal sabkha (Fairchild & Hambrey 1984). The upper contact of the Slangen Member is variable. In most Spitsbergen sections, it is sharp, silicified and brecciated, with clasts showing no evidence for reworking (Fairchild & Hambrey 1984). At Ditlovtoppen (Fig. 55.1) the grainstone passes transitionally upward into 2.5 m of mud-cracked siltstone beneath the basal diamictites of the Wilsonbreen Fm. In Nordaustlandet, the grainstones and microbial laminites are commonly overlain by silty dolomite ribbons, which grade upwards through ripple crosslaminated dolomitic siltstone and fine sandstone and into dolomitic sandstone of the basal Bra˚vika Member (Fig. 55.4).

THE POLARISBREEN GROUP, NORTHEASTERN SVALBARD

575

Fig. 55.4. Characteristic stratigraphic sections of the Wilsonbreen Fm. and Petrovbreen Member glacials from each of principal regions in which the glacial intervals have been studied (locations a to e are linked to Fig. 55.1; the Petrovbreen Member does not occur at location e). Stratigraphic depth is given relative to a 0 m datum at the top of each glacial unit. Note the difference in scale between the Wilsonbreen Fm. and Petrovbreen Member. The measured sections are compiled from published columns in Edwards (1976), Harland et al. (1993), Halverson et al. (2004) and unpublished data of the author. The breakdown and naming of members here follows that of Harland et al. (1993), but is slightly different from that used in previous papers, where the names of the Ormen and Gropbreen Members are inverted and a fourth Wilsonbreen member is included (e.g. Hambrey 1982).

The Wilsonbreen Fm. In most sections, the Wilsonbreen Fm. consists of c. 100–150 m of green- to red-weathering, massively to poorly bedded diamictite; however, the diamictite thins dramatically northwards in Nordaustlandet, being reciprocated by a northward-thickening wedge of quartz sands, informally named the Bra˚vika member (Halverson et al. 2004; Fig. 55.4). In the northernmost exposure of the Wilsonbreen Fm. (at Langrunesset), the Bra˚vika member is 250 m thick and diamictite is absent. In Spitsbergen, the Wilsonbreen Fm. is subdivided into the Ormen (W1), Middle Carbonate (W2) and Gropbreen (W3) members (Harland et al. 1993; Fig. 55.4). The Ormen Member varies from 20 to 75 m in thickness and consists mainly of poorly stratified, shaley diamictite, but also includes lenses and interbeds of sandstone, conglomerate and breccia (Fig. 55.4; Fairchild & Hambrey 1984). Clast lithology is variable, but dominated by carbonates derived from the underlying Elbobreen Fm. and Akademikerbreen Group. Chert, silt, sandstone and basement clasts, including exotic fragments of gneiss and undeformed pink granite, are also found. Striated clasts occur throughout the unit and dropstones, although not common in most sections, are

abundant at Aldousbreen (Fig. 55.4e; Edwards 1976). Sandstone wedges occur near or at the top of the Ormen Member in the Akademikerbreen area (Fig. 55.4a) and at Ditlovtoppen (Fig. 55.4b) (Hambrey 1982; Fairchild & Hambrey 1984). The Middle Carbonate Member is thin (,25 m) and comprises mainly dolomitic, medium-grained quartz sandstone. It is, however, recognized and distinguished by the presence of thin (,1.5 m) beds, lenses and fragments of primary carbonate. The nature of the carbonate is variable, being corrugated silty dolomitic microbial laminites at Ormen (in the Akademikerbreen area), mixed limestone and dolomite rhythmites and ribbons at Ditlovtoppen (Fig. 55.4b), and isolated limestone stromatolites within sandstone at Dracoisen (Fig. 55.4c). Based on their geochemistry, these carbonates are interpreted to have formed under the influence of evaporation in periglacial lakes (Fairchild et al. 1989). The Gropbreen Member is lithologically similar to the Ormen Member, with dominantly weakly stratified to massive maroonweathering diamictite, and subordinate sandstone lenses and beds. Clast composition is also similar to the Ormen Member, but with a higher proportion of basement clasts. In many sections, a medium brown sandstone up to a few metres thick occurs at the very top of the Gropbreen Member and is continuous with well

576

G. P. HALVERSON

developed sandstone wedges that penetrate up to 2 m into the underlying diamictite. Whether the sand is present or not, the contact between the Wilsonbreen Fm. and the overlying Dracoisen Fm. is always sharp. In Nordaustlandet, the Wilsonbreen is subdivided into two members. The Bra˚vika member overall is a coarsening-upwards package of quartz sandstone, with irregular dolomite lenses and intraclasts, and in the northernmost sections, minor limestone rhythmites and ribbons. The northward-thickening sandstone contains both planar and trough cross-bedding throughout, indicating a palaeoflow direction to the north (Halverson et al. 2004). The prevalence of low-angle cross-sets, coupled with pinstripe laminations and a bimodal grain population, suggests an aeolian origin, although the trough cross-bedding and dolomite precipitation indicate that the sand was in places reworked subaqueously. In the southern exposures in Nordaustlandet, the Bra˚vika sandstones are transitional with overlying diamictites, but in the Sveanor (Fig. 55.4d) are also transitional below with the Slangen Member (Halverson et al. 2004). The diamictite-dominated unit overlying the Bra˚vika member in Nordaustlandet resembles the Ormen and Gropbreen members on Spitsbergen, only containing a higher percentage of undeformed igneous clasts. Rhythmites with dropstones and fine interbedded turbidites are well preserved in the southernmost Nordaustlandet section at Aldousbreen (Fig. 55.4e), in Wahlenbergfjorden (Edwards 1976). Unstratified siltstones, interpreted as loessites (Edwards 1976), also occur in the same section (Fig. 55.4).

The Dracoisen Fm. The colourful lower part of the Draocoisen Fm. is a superb marker interval (Wilson & Harland 1964). A yellow-weathering dolomite bed everywhere overlies the Wilsonbreen Fm., with no evidence for reworking or depositional hiatus. This dolostone varies in thickness from as much as 18 m at Ditlovtoppen to less than 3 m in parts of Nordaustlandet (Halverson et al. 2004) and always comprises the transgressive component of a thick (c. 170 m) sequence (Fig. 55.3). The dolostone is mostly finely laminated dololutite, with minor low-angle erosional truncations and cross-laminations. Inversely graded laminae and pockets of peloids are also common. Large-scale oscillation ripples with amplitudes of up to 40 cm and wavelengths up to 6 m (Allen & Hoffman 2005) occur in the upper part of the dolostone in most northern sections. The top of the dolostone is transitional into marly red siltstone, which then passes upwards into green then black shales. The change to black shale, at c. 30–40 m above the base of the Dracoisen Fm., marks the maximum flooding surface of the sequence. Black shales with common carbonate concretions persist for over 100 m and are transitional into mud-cracked, variegated, finely bedded siltstone and shale that constitutes much of the upper part of the formation. The interval of mud-cracked siltstone is interrupted by 9 m of microbial-laminated dolomite, internally deformed by cauliflower chert.

Chemostratigraphy In one of the earliest systematic chemostratigraphic studies of a Neoproterozoic succession, Knoll et al. (1986) produced a coupled d13Ccarb – d13Corg record through the Veteranen, Akademikerbreen and Polarisbreen groups in both Spitsbergen and Nordaustlandet (as well as East Greenland). Although by modern standards the resolution of this record is low, with only a handful of data points from the Polarisbreen Group, this paper established the concurrence of glaciations and negative C-isotope anomalies, as well as the prevalence of high d13C values during the much of the Neoproterozoic. Subsequent C-isotope stratigraphy on the Polarisbreen Group was published by Fairchild & Spiro (1987), Kaufman et al. (1997) and Halverson

et al. (2004, 2005). Separate studies have focused on the isotopic composition of dolomites and limestones within the Wilsonbreen Fm. as a means of reconstructing the palaeoenvironment of syn-glacial carbonate precipitation (Fairchid & Spiro 1987, 1990; Fairchild et al. 1989). A compilation of C-isotope data from these various publications is plotted alongside a composite stratigraphic column in Figure 55.3. The Polarisbreen Group d13Ccarb and d13Corg records broadly mirror one another (Fig. 55.3), with an average isotopic difference of c. 30‰, which is typical for the Neoproterozoic (Hayes et al. 1999). The most striking feature of the C-isotope record is a 12‰ negative anomaly just below the Petrovbreen Member. This anomaly is reproduced in multiple sections and essentially spans the upper two parasequences of the Russøya Member (Halverson et al. 2004). d13C values are still negative in carbonates within the lowermost MacDonaldyrggen Fm., but quickly rise to positive values, where they remain into the Wilsonbreen Fm. Intraglacial carbonates show a range in d18O and d13C values between –11 and þ11‰ and þ1 and 5‰, respectively (Fairchild & Spiro 1987; Halverson et al. 2004). Fairchild & Spiro (1987, 1990) argued that the Wilsonbreen carbonates were precipitated by evaporation in a periglacial lake based, in part, on these unusually high d18O compositions. More recently, Bao et al. (2009) reported D17O (triple oxygen isotopes) compositions in carbonate-associated sulphate (CAS) within these carbonates of as low as –1.6‰, which are the lowest values ever reported from natural specimens. Bao et al. (2009) interpreted these anomalous values to record either extraordinarily high pCO2 or unusually sluggish O2 cycling during the glaciation. A second negative C-isotope anomaly is recorded in the basal Dracoisen dolostone and reproduced in multiple sections, where d13C values gradually decline from –3 to – 5‰ (Halverson et al. 2004). d13C values then spike to .10‰ in evaporitic dolomites in the upper Dracoisen Fm. (Fig. 55.3). Four Sr-isotope data from the Polarisbreen Group were first published by Derry et al. (1989), and a few new analyses from the original sample set were added in Jacobsen & Kaufman (1999). Additional 87Sr/86Sr data from the Polarisbreen were published in Halverson et al. (2005, 2007). The highest-quality data from all these studies consistently show 87Sr/86Sr ¼ 0.7067– 0.7068 in the Russøya Member (Fig. 55.3). The absence of unaltered marine limestone above the Russøya Member precludes application of Sr-isotope chemostratigraphy to the remainder of the Polarisbreen Group.

Palaeolatitude and palaeogeography Barentsia is widely assumed to have been contiguous with East Greenland during the Neoproterozoic based on the remarkable stratigraphic similarity between the two regions (Harland & Gayer 1972; Knoll et al. 1986; Fairchild & Hambrey 1995). However, the precise location of Eastern Svalbard relative to the East Greenland Caledonides (Johansson et al. 2005) and the geometry and origin of the East Greenland –eastern Svalbard platform within the broader framework of the connection and eventual rifting between Laurentia, Baltica and Amazonia are controversial. In the conventional configuration, western Baltica is the conjugate margin to eastern Greenland (and Barentsia), virtually identical to the Caledonian fit (e.g. Weil et al. 1998). In a different reconstruction, Hartz & Torsvik (2002) placed the East Greenland –eastern Svalbard platform adjacent to the southern peri-Urals of Baltica and proposed that it developed as a sinistral pull-apart basin between Baltica and Amazonia. This model has been challenged by more recent palaeomagnetic data from Baltica (Cawood & Pisarevsky 2006), and in a modification of earlier west-Baltica – eastern Greenland fits, Pisarevsky et al. (2008) proposed that the west Norwegian margin was conjugate to southeastern Greenland. This model implies that the eastern Svalbard and the

THE POLARISBREEN GROUP, NORTHEASTERN SVALBARD

Neoproteroozic successions in northern Norway were part of a long passive margin isolated from the other coeval sedimentary basins to the south. Direct palaeomagnetic constraints for Svalbard’s Neoproterozoic palaeography are limited to a new suite of palaeomagnetic data from the Akademikerbreen Group that show a stable, preCaledonian remanent magnetization indicating that the EGES platform resided in tropical latitudes (I  158) during the midNeoproterozoic (Maloof et al. 2006). No reliable palaeomagnetic results have been obtained from the Polarisbreen Group, and Svalbard’s complicated and uncertain tectonic history since this time precludes any extrapolation of palaeolatitude for the glacial deposits from elsewhere on the Laurentia craton (Evans 2000).

Geochronological constraints No direct radiometric ages have been obtained on the upper Hecla Hoek. U –Pb ages on detrital zircons in the Planetfjella Group in Spitsbergen and presumed primary zircons from volcanics and subvolcanic intrusives beneath the Veteranen Group in Nordaustlandet indicate a maximum age for the base of the Lomfjorden Supergroup of c. 946 Ma (Johansson et al. 2000, 2005; Fig. 55.2). Similar ages have been obtained from igneous clasts in the Wilsonbreen Fm. in Nordaustlandet (Johansson et al. 2000). The top of the Polarisbreen Group, which is in disconformable contact with the overlying Cambrian Tokamanne Fm. (Knoll & Swett 1987), is inferred to be ,542 Ma (the age of the Precambrian –Cambrian boundary; Amthor et al. 2003). If the Dracoisen Fm. is correlative with the Maieberg Fm. in northern Namibia and the Doushantuo Fm. in South China (Hoffmann et al. 2004; Condon et al. 2005), as implied by the chemostratigraphy and biostratigraphy of the Polarisbreen Group, then the Wilsonbreen glaciation ended at 635 Ma (Halverson et al. 2005).

Biostratigraphy In contrast to the rich and diverse fossil assemblage from the underlying Akademikerbreen Group (Knoll 1982; Butterfield et al. 1994, and references therein), the biostratigraphic record of the Polarisbreen Group is most notable for its depauperate microfossil assemblage, despite the generally mild thermal history of the rocks (Knoll & Swett 1987). In the MacDonaldryggen Member and in shales within the Wilsonbreen Fm., the acritarch Bavlinella faveolata dominates the microfossil assemblage (Knoll 1982; Knoll & Swett 1987). B. faveolata also occurs in the Dracoisen Fm. and is locally abundant, but overall is subordinate to a low diversity assemblage of small, smooth-walled leiosphere acritarchs (Fig. 55.3). The diverse assemblage of the large, acanthomorphic acritarchs that characterize middle– upper Ediacaran successions in Australia (Grey et al. 2003) and elsewhere (Knoll 2000) is absent in the Polarisbreen Group (Knoll & Swett 1987). Coupled with the lack of Ediacaran fauna, this lacuna strongly suggests that the contact between the Polarisbreen Group and the overlying Oslobreen Group represents a significant depositional hiatus (Knoll & Swett 1987) and that at least the last 35 myr of the Ediacaran Period are missing in Svalbard.

Discussion The Neoproterozoic succession in Svalbard has played a prominent role in questions and controversies regarding Precambrian glaciations and environmental change since Harland (1964) first proposed a pan-global infra-Cambrian glaciation. Although a few researchers maintained scepticism over the glacial origin of diamictites in the Polarisbreen Group (Krasil’shchov 1973; Schermerhorn 1974), the detailed sedimentological work of Edwards, Hambrey, Fairchild and others (e.g. Edwards 1976; Hambrey

577

1982; Fairchild & Hambrey 1984; Dowdeswell et al. 1985) effectively put to rest any doubt. Evidence for a glacial environment, specifically in the Wilsonbreen Fm., is found in the abundance of striated and faceted clasts, dropstones, the extrabasinal origin of some stones, and sandstone wedges. The Petrovbreen Member diamictite and associated lithofacies are typically interpreted to have been deposited in relatively deep water, most likely in a marine environment (Fairchild & Hambrey 1984; Harland et al. 1993). This interpretation is consistent with the lack of evidence for any obvious change in relative sea level at the contact between the Petrovbreen and MacDonaldryggen Members (Halverson et al. 2004). In contrast, the heterogeneous Wilsonbreen Fm., which includes diamictites interpreted as lodgement tills (Dowdeswell et al. 1985), direct evidence for subaerial exposure (sandstone wedges), and lacustrine carbonates is inferred to have been deposited in a predominantly terrestrial setting (Fairchild & Hambrey 1984; Dowdeswell et al. 1985; Halverson et al. 2004). Dowdeswell et al. (1985) argued that the Wilsonbreen diamictites were deposited beneath ice sheets or ice caps (rather than valley glaciers) based on the preponderance of exotic clasts and a lack of angular supraglacial debris. A great deal of effort has been expended on attempting to correlate the Polarisbreen diamictites with other Neoproterozoic glacial units in the North Atlantic (e.g. Harland & Gayer 1972; Hambrey 1983; Nystuen 1985; Fairchild & Hambrey 1995) and globally (e.g. Kaufman et al. 1997; Kennedy et al. 1998; Halverson et al. 2005). Based on biostratigraphic and chemostratigraphic data, it has generally been argued that the Polarisbreen Group is late ‘Vendian’ in age, meaning essentially that it post-dates the earliest Neoproterozoic glaciations. Unfortunately, the initial optimism that the chronostratigraphy and correlations would be firmed up by geochronological and palaeomagnetic data (Hambrey 1983) has not been realized, and no undisputed correlation scheme exists for the North Atlantic region, let alone the whole of the globe. In a twist on the conventional Vendian interpretation of the age of the Polarisbreen Group diamictites, Halverson et al. (2004) argued that both the Petrovbreen Member and Wilsonbreen Fm. belonged to a single glacial episode, ending at 635 Ma (the approximate age of the end of the Ghaub glaciation in Namibia and Nantuo glaciation in South China; Hoffmann et al. 2004; Condon et al. 2005). This model was based on (i) the similar stratigraphic context and magnitude of the upper Russøya negative d13C anomaly (Fig. 55.3) and the so-called Trezona anomaly, which precedes the Ghaub glaciation in Namibia (Halverson et al. 2002), coupled with lack of evidence for an analogous anomaly preceding any of the older glaciations; (ii) the occurrence of glendonites in the upper MacDonaldryggen Member; and (iii) the virtually identical geochemistry, sedimentology and stratigraphy of the Dracoisen and post-Ghaub Maieberg (and other) cap-carbonate sequences. In light of new Sr-isotope data from the level of the negative anomaly in the Russøya Member (87Sr/86Sr ¼ 0.7068 v. 0.7072 in the Trezona anomaly; Halverson et al. 2007) and evidence that the lowermost of three distinct glacial units in the Dalradian Supergroup is preceded by a negative C-isotope anomaly of similar magnitude to that in the upper Russøya Member (McCay et al. 2006), it now seems more likely that the Petrovbreen Member diamictite must represent a separate, older glaciation. Although it remains likely that the Wilsonbreen Fm. is c. 635 Ma, the ongoing ambiguity in Cryogenian–Ediacaran age constraints precludes any unequivocal assignment of age to either of the Polarisbreen glaciations. The more conventional two-glaciation interpretation also raises the question of the climatic significance of the glendonite nodules in the upper MacDonaldryggen Member. Glendonites have also been reported in possibly equivalent-aged strata in the Windermere Supergroup in the Mackenzie Mountains (James et al. 2005). Insofar as the MacDonaldryggen pseudomorphs originated as ikaite, which forms at near-freezing temperatures in alkalinecharged waters (Buchart et al. 1997), then their occurrence in

578

G. P. HALVERSON

continental shelf sediments with no other evidence of glaciation, and just below sabkha deposits, poses yet another Neoproterozoic climatic conundrum. Thus, it seems certain that the Polarisbreen Group will continue to play a central role in the ongoing effort to understand Neoproterozoic climate. This chapter stems in part from the author’s PhD and subsequent research in Svalbard, funded by NSF and NASA. Original mapping, measured sections, and other contributions presented here are the results of collaborative fieldwork with A. C. Maloof, P. F. Hoffman, E. W. Domack, M. T. Hurtgen, J. Eigenbrode, and many field assistants. M. Hambrey and I. Fairchild provided constructive reviews of the manuscript. The contribution forms TRaX record #169. This represents a contribution of the IUGS and UNESCO-funded IGCP (International Geoscience Programme) project #512.

References Allen, P. A. & Hoffman, P. F. 2005. Extreme winds and waves in the aftermath of a Neoproterozoic glaciation. Nature, 433, 123– 127. Amthor, J. E., Grotzinger, J. P., Schro¨der, S., Bowring, S. A., Ramezani, J., Martin, M. W. & Matter, A. 2003. Extinction of Cloudina and Namacalathus at the Precambrian – Cambrian boundary in Oman. Geology, 31, 431–434. Bao, H., Fairchild, I. J., Wynn, P. M. & Spo¨tl, C. 2009. Stretching the envelope of past surface environments: Neoproterzoic glacial lakes from Svalbard. Science, 323, 119–122. Breivik, A. J., Mjelde, R., Grogan, P., Shimamura, H., Murai, Y., Nishimura, Y. & Kuwana, A. 2002. A possible Caledonide arm through the Barents Sea imaged by OBS data. Tectonophysics, 355, 67 –97. Buchart, B., Seaman, P. et al. 1997. Submarine columns of ikaite tufa. Nature, 390, 129–130. Butterfield, N. J., Knoll, A. H. & Swett, K. 1988. Exceptional preservation of fossils in an Upper Proterozoic shale. Nature, 334, 424– 427. Butterfield, N. J., Knoll, A. H. & Swett, K. 1994. Paleobiology of the Neoproterozoic Svanbergfjellet Formation, Spitsbergen. Fossils and Strata, 34, 1– 81. Cawood, P. A. & Pisarevsky, S. A. 2006. Was Baltica right-way-up or upside-down in the Neoproterozoic? Journal of the Geological Society, London, 163, 753–759. Chumakov, N. 1968. On the character of the Late Precambrian glaciation of Spitsbergen (translated title). Doklady An SSSR, Seriya Geologicheskaya, 180, 1446– 1449. Condon, D., Zhu, M., Bowring, S., Jin, Y., Wang, W. & Yang, A. 2005. From the Marinoan glaciation to the oldest bilaterians: U– Pb ages from the Doushantou Formation, China. Science, 308, 95– 98. Derry, L. A., Keto, L. S., Jacobsen, S. B., Knoll, A. H. & Swett, K. 1989. Strontium isotopic variations in Upper Proterozoic carbonates from Svalbard and East Greenland. Geochimica et Cosmochimica Acta, 53, 2331– 2339. Dallmann, W. K., Ohta, Y., Elvevold, S. & Blomeier, D. 2002. Bedrock map of Svalbard and Jan Mayen (1:750,000 scale geological map). Norsk Polarinstitutt Temekart, 33. Dowdeswell, J. A., Hambrey, M. J. & Wu, R. 1985. A Comparison of clast fabric and shape in Late Precambrian and Modern glaciogenic sediments. Journal of Sedimentary Petrology, 55, 691– 704. Edwards, M. B. 1976. Sedimentology of Late Precambrian Sveanor and Kapp Sparre Formations at Aldousbreen, Wahlenbergfjorden, ˚ rbok, 1974, 51 –61. Nordaustlandet. Norsk Polarinstitutt A Edwards, M. B. 1978. Late Precambrian glacial loessites from north Norway and Svalbard. Journal of Sedimentary Petrology, 49, 85 – 91. Evans, D. A. D. 2000. Stratigraphic, geochronological and paleomagnetic constraints upon the Neoproterozoic climatic paradoxes. American Journal of Science, 300, 347– 443. Fairchild, I. J. 1983. Effects of glacial transport and neomorphism on Precambrian dolomite crystal sizes. Nature, 304, 714– 716. Fairchild, I. J. & Hambrey, M. J. 1984. The Vendian succession of northeastern Spitsbergen: petrogenesis of a dolomite– tillite association. Precambrian Research, 26, 111–167.

Fairchild, I. J. & Hambrey, M. J. 1995. Vendian basin evolution in East Greenland and NE Svalbard. Precambrian Research, 73, 217– 233. Fairchild, I. J. & Spiro, B. 1987. Petrological and isotopic implications of some contrasting Late Precambrian carbonates, NE Spitsbergen. Sedimentology, 34, 973– 989. Fairchild, I. J. & Spiro, B. 1990. Carbonate minerals in glacial sediments: geochemical clues to palaeoenvironment. In: Dowdeswell, J. A. & Scourse, J. D. (eds) Glacimarine Environments: Processes and Sediments. Geological Society, London, Special Publications, 53, 201– 216. Fairchild, I. J., Hambrey, M. J., Spiro, B. & Jefferson, T. H. 1989. Late Proterozoic glacial carbonates in northeast Spitsbergen: new insights into the carbonate –tillite association. Geological Magazine, 126, 469– 490. Fleming, W. L. S. & Edmonds, J. M. 1941. Hecla Hoek rocks of New Friesland (Spitsbergen). Geological Magazine, 78, 405– 428. Flood, B., Gee, D. G., Hjelle, A., Siggerud, T. & Winsnes, T. 1969. The geology of Nordaustlandet, northern and central parts. Norsk Polarinstitutt Skrifter, 146, 1– 139 (þ1:250,000 map). Gee, D. G. & Page, L. M. 1994. Caledonian terrane assembly on Svalbard: new evidence from 40Ar/39Ar dating in Ny Friesland. American Journal of Science, 294, 1166–1186. ˚ . et al. 1995. Grenvillian basement and a major Gee, D. G., Johansson, A unconformity within the Caledonides of Nordaustlandet, Svalbard. Precambrian Research, 70, 215– 234. Grey, K., Walter, M. R. & Calver, C. R. 2003. Neoproterozoic biotic diversification: Snowball Earth or aftermath of the Acraman impact. Geology, 31, 459– 462. Halverson, G. P., Hoffman, P. F., Schrag, D. P. & Kaufman, A. J. 2002. A major perturbation of the carbon cycle before the Ghaub glaciation (Neoproterozoic) in Namibia: prelude to snowball Earth? Geochemistry, Geophysics, Geosystems, 3, doi: 10.1029/ 2001GC000244. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. 2005. Towards a Neoproterozoic composite carbon-isotope record. Geological Society of America Bulletin, 117, 1181–1207. Halverson, G. P., Maloof, A. C. & Hoffman, P. F. 2004. The Marinoan glaciation (Neoproterozoic) in Svalbard. Basin Research, 16, 297– 324. Halverson, G. P., Dudas, F. O., Maloof, A. C. & Bowring, S. A. 2007. Evolution of the 87Sr/86Sr composition of Neoproterozoic seawater. Palaeogeography, Palaeoclimatology, Palaeoecology, 256, 103– 129. Hambrey, M. J. 1982. Late Precambrian diamictites of northeastern Svalbard. Geological Magazine, 119, 527– 551. Hambrey, M. J. 1983. Correlation of late Proterozoic tillites in the North Atlantic region and Europe. Geological Magazine, 120, 290– 320. Hambrey, M. J., Harland, W. B. & Waddams, P. 1981. Late Precambrian tillites of Svalbard. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, 592–600. Harland, W. B. 1959. The Caledonian sequence in Ny Friesland, Spitsbergen. Quarterly Journal of the Geological Society, 114, 307– 342. Harland, W. B. 1964. Critical evidence for a great infra-Cambrian glaciation. Geologische Rundschau, 54, 45 –61. Harland, W. B. 1997. The Geology of Svalbard. Geological Society of London Memoir, 17, London, 521. Harland, W. B. 2007. The Ny Friesland Orogen, Spitsbergen. Geological Magazine, 144, 633– 642. Harland, W. B. & Gayer, R. A. 1972. The Arctic Caledonides and earlier oceans. Geological Magazine, 109, 289–314. Harland, W. B. & Wilson, C. B. 1956. The Hecla Hoek succession in Ny Friesland, Spitsbergen. Geological Magazine, 93, 265– 286. Harland, W. B., Scott, R. A., Auckland, K. A. & Snape, I. 1992. The Ny Friesland Orogen, Spitsbergen. Geological Magazine, 129, 679– 708. Harland, W. B., Hambrey, M. J. & Waddams, P. 1993. Vendian Geology of Svalbard. Norsk Polarinstitutt Skrifter (Oslo), 193, 150. Hartz, E. & Torsvik, T. 2002. Baltica upside down: a new plate tectonic model for Rodinia and the Iapetus Ocean. Geology, 30, 225– 258. Hayes, J. M., Strauss, H. & Kaufman, A. J. 1999. The abundance of C in marine organic carbon and isotopic fractionation in the global

THE POLARISBREEN GROUP, NORTHEASTERN SVALBARD

biogeochemical cycle of carbon during the past 800 Ma. Chemical Geology, 161, 103– 125. Hoffmann, K. H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. A U– Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817–820. Jacobsen, S. B. & Kaufman, A. J. 1999. The Sr, C and O isotopic evolution of Neoproterozoic seawater. Chemical Geology, 161, 37 –57. James, N. P., Narbonne, G. M., Dalrymple, R. W. & Kyser, C. 2005. Glendonites in Neoproterozoic low-latitude, interglacial sedimentary rocks, northwest Canada: insights on the Cryogenian ocean and Precambrian cold-water carbonates. Geology, 33, 9 –12. ˚ ., Larianov, A. N., Tebenkov, A. M., Gee, D. G., WhiteJohansson, A house, M. J. & Vestin, J. 2000. Grenvillian magmatism of western and central Nordaustlandet, northeastern Svalbard. Transactions of the Royal Society of Edinburgh, 90, 221– 234. Johansson, A., Gee, D. G., Larionov, A. N., Ohta, Y. & Tebenkov, A. M. 2005. Grenvillian and Caledonian evolution of eastern Svalbard – a tale of two orogenies. Terra Nova, 17, 317– 325. Kaufman, A. J., Knoll, A. H. & Narbonne, G. M. 1997. Isotopes, ice ages and terminal Proterozoic earth history. Proceedings of the National Academy of Science (USA), 94, 6600–6605. Kennedy, M. J., Runnegar, B., Prave, A. R., Hoffmann, K.-H. & Arthur, M. 1998. Two or four Neoproterozoic glaciations? Geology, 26, 1059 – 1063. Knoll, A. H. 1982. Microfossil based biostratigraphy of the Precambrian Hecla Hoek sequence of Nordaustlandet, Svalbard. Geological Magazine, 199, 269– 279. Knoll, A. H. 1984. Microbiotas of the late Precambrian Hunnberg Formation Nordaustlandet, Svalbard. Journal of Palaeontology, 58, 131– 162. Knoll, A. H. 2000. Learning to tell Neoproterozoic time. Precambrian Research, 100, 3– 20. Knoll, A. H. & Swett, K. 1987. Micropaleontology across the Precambrian – Cambrian boundary in Spitsbergen. Journal of Palaeontology, 61, 898– 926. Knoll, A. H., Hayes, J. M., Kaufman, A. J., Swett, K. & Lambert, I. B. 1986. Secular variation in carbon isotope ratios from Upper Proterozoic successions of Svalbard and east Greenland. Nature, 321, 832– 837. Krasil’shchikov, A. A. 1973. The stratigraphy and paleotectonics of the Precambrian – Early Paleozoic of Spitsbergen (translated title). Trudy Arkticheskogo Nauchno-Issledovatel’skogo Instituta, 172, 1 –120. Kulling, O. 1934. The Hecla Hoek Formation round Hinlopenstredet. Geografiska Annaler, 14, 161–253. Larsen, V. B. 1987. A synthesis of tectonically related stratigraphy in the North Atlantic – Arctic region from Aalenian to Cenomanian time. Norsk Geologisk Tidsskrift, 67, 323– 338. Lyberis, N. & Manby, G. 1993. The origin of the West Spitsbergen Fold Belt from geological constraints and plate kinematics: implications for the Arctic. Tectonophysics, 224, 371–391.

579

Lyberis, N. & Manby, G. 1999. Continental collision and lateral escape deformation in the lower and upper crust: an example from Caledonide Svalbard. Tectonics, 18, 40 – 63. Maher, H. D. Jr. 2001. Manifestations of the Cretaceous High Arctic Large Igneous Province in Svalbard. Journal of Geology, 109, 91– 104. Maloof, A. C., Halverson, G. P., Kirschvink, J. L., Schrag, D. P., Weiss, B. P. & Hoffman, P. F. 2006. Combined paleomagnetic, isotopic and stratigraphic evidence for true polar wander from the Neoproterozoic Akademikerbreen Group, Svalbard, Norway. Geological Society of America Bulletin, 118, 1099– 2014. Manby, G. & Lyberis, N. 1995. Discussion on the Ny Friesland Orogen, Spitsbergen. Geological Magazine, 132, 351– 356. McCay, G. A., Prave, A. R., Alsop, G. I. & Fallick, A. E. 2006. Glacial trinity: Neoproterozoic Earth history within the British –Irish Caledonides. Geology, 34, 909– 912. Nordenskio¨ld, A. E. 1863. Geografisk och geognostisk beskrifning over noro¨stra delarna af Spetbergen och Hinlopen Strait [Geographic and geognostic descriptions of the northeast part of Spitsbergen and Hinlopen Straight]. Kungliga Svenska Vetenskapsakademiens Handlingar, 4. Stockholm. Nystuen, J. P. 1985. Facies and preservation of glaciogenic sequences from the Varanger ice age in Scandinavia and other parts of the North Atlantic region. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 209–229. Pisarevsky, S. A., Murphy, J. B., Cawood, P. A. & Collins, A. S. 2008. Late Neoproterozoic and Early Cambrian palaeogeography: models and problems. In: Pankhurst, R. J., Trouw, R. A. J., de Brito Neves, B. B. & de Wit, M. (eds) West Gondwana: Pre-Cenozoic Correlations Across the South Atlantic Region. Geological Society, London, Special Publications, 294, 9 –31. Schermerhorn, L. J. G. 1974. Late Precambrian mixtites: glacial and/or non-glacial. American Journal of Science, 274, 673–824. Sokolov, V. N., Krasil’chov, A. A. & Livshits, YU. YA. 1968. The main features of the tectonic structure of Spitsbergen. Geological Magazine, 105, 95– 115. Teben’kov, A. M., Ohta, Y., Balashov, J. A. & Sirotkin, A. N. 1996. Newtontoppen granitoid rocks; their geology, chemistry and Rb –Sr age. Polar Research, 15, 67– 80. Weil, A. B., Van der Voo, R., Mac Niocaill, C. & Meert, J. G. 1998. The Proterozoic supercontinent Rodinia: paleomagnetically-derived reconstructions for 1100– 800 Ma. Earth and Planetary Science Letters, 154, 13– 24. Wilson, C. B. 1961. The upper Middle Hecla Hoek rocks of Ny Friesland, Spitsbergen. Geological Magazine, 98, 89 – 116. Wilson, C. B. & Harland, W. B. 1964. The Polarisbreen Series and other evidences of late Pre-Cambrian ice ages in Spitsbergen. Geological Magazine, 101, 198– 219. Witt-Nilsson, P., Gee, D. G. & Hellman, F. J. 1998. Tectonostratigraphy of the Caledonian Atomfjella Antiform of the northern Ny Friesland, Svalbard. Norsk Geologiske Tidsskrift, 78, 67 –80.

Chapter 56 Neoproterozoic (Cryogenian –Ediacaran) deposits in East and North-East Greenland SVEND STOUGE1*, JØRGEN LØYE CHRISTIANSEN2, DAVID A. T. HARPER1, MICHAEL HOUMARK-NIELSEN3, ˚ RD6 KASPER KRISTIANSEN4, CONALL MACNIOCAILL5 & BJØRN BUCHARDT-WESTERGA 1

Geological Museum, University of Copenhagen, DK-1350 Copenhagen K, Denmark 2

University College of Zealand, DK-4300 Holbaek, Denmark

3

GeoGenetics Centre, Natural History Museum, University of Copenhagen, DK-1350 Copenhagen K, Denmark, 4

Maersk Oil and Gas, Denmark

5

Department of Earth Sciences, Oxford University, UK

6

Department of Geography and Geology, University of Copenhagen, DK-1350 Copenhagen K, Denmark *Corresponding author (e-mail: [email protected])

Abstract: The Neoproterozoic succession of East and North-East Greenland (over 14 000 m thick) includes the Eleonore Bay Supergroup (?Tonian– Cryogenian) and the Tillite Group (Cryogenian–Ediacaran). The upper units of the Eleonore Bay Supergroup consist of shallow to deeper-water carbonates, succeeded by siliciclastic fine-grained sediments (Bedgroup 19) that characterize the top unit of the supergroup. The Tillite Group includes two diamictite-bearing units (Ulvesø and Storeelv formations) of glaciogenic origin and two upper, upwards-shallowing strata (Canyon and Spiral Creek formations) that were deposited during semiarid conditions and concluded the Neoproterozoic depositional cycle. Diamictite is preserved on the craton and compares with the Storeelv Formation (Fm.) of the Tillite Group. Detailed investigations of the diamictite-bearing units (i.e. Ulvesø and Storeelv formations) demonstrate that the lower of the two formations is mainly of marine origin, whereas the upper one has both marine and terrestrial origins. Chemostratigraphic data include analyses on total carbon (TC), total organic carbon (TOC), total sulphur (TS) and d13C. The data set for d13C shows a substantial and abrupt shift towards negative values of 10%, from below Bedgroup 19. Low-diversity acritarch assemblages (Cryogenian) are recorded from the Andre´e Land and Tillite groups; a thin cherty dolostone unit present above the Storeelv Fm. suggests that the diamictite units are of late Cryogenian age and the upper part of the Tillite Group is Ediacaran. Bedgroup 19 disconformably overlies older carbonates and the unit is a prelude to the succeeding (upper Cryogenian– lower Ediacaran) diamictite sediments of the Tillite Group. A disconformity separates the Tillite Group from the overlying Lower Palaeozoic sediments. Both disconformities are, according to palaeomagnetic data, related to rift–drift episodes that occurred during the late Neoproterozoic. Alternatively, the isotope data suggest that the diamictites were deposited during the late Cryogenian glaciation and the older disconformity may be interpreted as a significant gap developed by the lowering of sea level during an early Cryogenian glaciation.

The Neoproterozoic sediments in East and North-East Greenland are well preserved and, despite their remote location, the area is well known and has served as key area for the study of Neoproterozoic successions in the Arctic (Figs 56.1 & 56.2). It should be noted that this region was previously referred to as East Greenland in the literature but is now referred to as East and North-East Greenland according to new administrative boundaries. Two diamictite bearing units (Ulvesø and Storeelv formations, formerly referred to as lower and upper tillite respectively) are thought to record glacial conditions in the region during this time period. Other diamictite units occur in para-autochthonous windows and are broadly correlated with the Storeelv Fm. The Ulvesø and Storeelv formations are part of the Tillite Group (Haller 1971; Henriksen & Higgins 1976; Hambrey & Spencer 1987), which Kulling (1929) initially referred to as the Tillite Series. Even though the name does not follow modern guidelines for stratigraphic terminology, it has been used continuously in the literature since 1929, and it is retained here following the recommendation of Hambrey & Spencer (1987, p. 6). The Tillite Group is exposed in a north – south-trending belt extending through the central fjord region from Canning Land in the south (728250 N) to Payer Land in the north (748200 N). The exposure is excellent in many places, especially along the coastlines of the fjords. The Ulvesø Fm. was formally defined by Hambrey & Spencer (1987) with its type locality in Kløftelv, c. 200 m north of

Ulvesø on Ella Ø (728520 N, 258050 W) and with a coastal section beginning in Tømmerbugt and extending into Bastion Bugt, serving as the reference section (Figs 56.3 & 56.4). It formed part of the Cape Oswald Fm. of Poulsen (1930), and Schaub (1950) placed the unit in his Tømmerbugt Group. The Storeelv Fm. takes its name from Storeelv on Ella Ø (728510 N, 258070 W) (Fig. 56.1). Systematic research of the upper Neoproterozoic sediments began in the mid-1920s and continued in the 1950s, when several expeditions visited the area under the leadership of Lauge Koch (Koch 1929, 1930; Koch & Haller 1971; for a summary see Haller 1971 and Henriksen & Higgins 2008). At that time, ship and aircraft support had become widely available, which made it possible to visit these remote parts of East and North-East Greenland. Stratigraphical, sedimentological and facies studies from the mid-1970s onwards were undertaken by the Geological Survey of Greenland as part of its general mapping programme (Henriksen & Higgins 1976; Benga˚rd 1991; Henriksen 1999, 2003; Henriksen & Higgins 2008). Geologists from Harvard and Cambridge University participated in expeditions to the region in the 1970s and 1980s, which resulted in several new stratigraphical and sedimentological findings (Hambrey 1983, 1989; Hambrey & Spencer 1987; Moncrieff & Hambrey 1988, 1990; Hambrey et al. 1989; Herrington & Fairchild 1989; Manby & Hambrey 1989; Swett & Knoll 1989) and descriptions of microfossils assemblages (Knoll et al. 1986; Green et al. 1987, 1988, 1989). The present

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 581– 592. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.56

582

S. STOUGE ET AL.

Fig. 56.1. Location map showing positions of exposures, and the structure and geology of East and North-East Greenland. (Modified from Frederiksen 2000 and Higgins et al. 2004). Used with permission from the National Geological Survey of Denmark and Greenland (GEUS) #.

lithostratigraphic scheme for the upper Neoproterozoic deposits in North-East Greenland (Hambrey & Spencer 1987) appeared at around the same time, when Knoll et al. (1986) published their geochemistry-driven data for the broadly correlative successions on Spitsbergen. More recent research on stratigraphy, sedimentology, micropalaeontology, palaeomagnetism and geochemistry started in the late 1990s (Frederiksen et al. 1999; Frederiksen 2000) and continues now into the 21st century (Stouge et al. 2001; Sønderholm et al. 2008).

Structural framework Most of the Neoproterozoic deposits of East and North-East Greenland accumulated in the Eleonore Bay basin that existed for c. 350 Ma. This sedimentary basin was nearly 500 km long, NE –SW trending and 200– 400 km wide, and a pile of c. 14– 15 km of sediments accumulated (Sønderholm et al. 2008). Eleonore Bay Basin sediments, preserved today in the Franz Joseph Allochton of the Caledonian orogen were deposited

NEOPROTEROZOIC NE GREENLAND

Era

Period

Supergroup

Group

Palaeozoic Cambrian (not named) Kong Oscar Fjord

Formation Kløftelv

583

Events

Sequence Ma

Iapetus passive margin

Sauk

542 Ediacarian

Hiatus Spiral Creek

Eleonore Bay

Cryogenian

Canyon

Neoproterozoic

Uplift 580

Tillite Group

III 635

Storeelv

Maximal separation SQB

Arena

II Ulvesø SQB

Drift Final carbonate deposition

Bedgroup 20

I Rifting

Bedgroup 19 Andrée Land

SQB Bedgroup 18

c. 200– 400 km away, farther to the east (Higgins & Leslie 2000, 2008) than the present day. The Neoproterozoic succession in the basin is thought to record a rift –drift transition associated with the fragmentation of Rodinia (Herrington & Fairchild 1989; Stouge et al. 2001; Eyles & Januszczak 2004; Nystuen et al. 2008; Sønderholm et al. 2008), although volcanic deposits typically associated with rifted margins have not been reported from the succession (Soper 1994; Higgins & Leslie 2008; Sønderholm et al. 2008). This Neoproterozoic sedimentary pile is incorporated in the Caledonian mountain belts of eastern Greenland, extending from 708 to 828N and flanking the eastern edge of the Greenland ice sheet (Henriksen & Higgins 1976, 2008; Henriksen 1999; Higgins et al. 2004; Gee et al. 2008; Higgins & Leslie 2008). The outcrop of Caledonian bedrock in eastern Greenland (Fig. 56.1) is c. 300 km wide in the south, at 708N, and narrows northwards to c. 100 km, striking obliquely offshore into the northern Greenland continental shelf. It forms the western part of the Caledonian Orogen dominated by thrust sheets, which were emplaced west-northwestwards onto the Laurentian Craton with its Lower Palaeozoic cover. The allochthons are composed of two major thrust sheets, the Niggli Spids thrust sheet and the Hagar Bjerg thrust sheet (Higgins et al. 2001, 2004; Higgins & Leslie 2008) (Fig. 56.1). The lower Niggli Spids thrust sheet comprises Archaean and Palaeoproterozoic basement gneisses overlain by upper Mesoproterozoic to lower Neoproterozoic metasediments or the Krummedal supracrustal sequence, which is several kilometres thick. The Hagar Bjerg thrust sheet overlies the Niggle Spids thrust sheet; it includes the Krummedal supracrustal sequence intruded by c. 950–920 Ma granites (Kalsbeek et al. 2001, 2008). These are overlain by the c. 18.5-km-thick Neoproterozoic siliciclastic and carbonate successions (Eleonore Bay Supergroup), siliciclastic sediments with diamictite (Tillite Group), and Cambrian–Middle Ordovician siliciclastic and platform carbonate formations (Kong Oscar Fjord Group). This sedimentary package comprises the upper part of the Hagar Bjerg thrust sheet and is recognized as the Franz Joseph allochthon (Higgins et al. 2004; Higgins & Leslie 2008). All the transported rocks in eastern Greenland were derived from the Laurentian margin, including both Archaean and Palaeoproterozoic crystalline basement and the younger, shallow-marine sedimentary cover. WNW-directed thrust transport of the allochthons has been estimated to be at least 200 km (Higgins & Leslie 2000, 2008), along with sinistral strike displacements; hence, prior to the Caledonian Orogeny, the platform margin of Laurentia was nearly twice as wide as it is today.

Foundering of carbonate platform Carbonate platform

Fig. 56.2. Stratigraphical nomenclature used in this chapter and summary of the series of events that occurred in the Eleonore Bay basin during the Cryogenian and Ediacaran. The succession comprises three stratigraphic sequences, and the hiatus between the Tillite Group and the Lower Palaeozoic Kløftelv Fm. is extensive. No radiometric ages are available in this succession; ages are inferred based on interpretation of the stratigraphic and biostratigraphic record. See text for discussion.

The upper part of the Andre´e Land Group is of special interest in this context, and together with the Tillite Group crop out over an area of c. 200 km from north to south and c. 50 km in an east – west direction (Fig. 56.1). The outcrop area is bounded to the west by the older sediments of the Eleonore Bay Supergroup and to the east by the Lower Palaeozoic outcrops or by major north – south-trending faults. The sediments are gently folded into synforms and antiforms within the East and North-East Greenland Caledonides and are only mildly affected by Caledonian metamorphism; the succession is locally jointed and faulted by younger late Palaeozoic (Carboniferous) orogenic events. Additional occurrences of upper Neoproterozoic sediments are preserved in the Charcot Land, Ga˚seland and Ma˚lebjerg windows in the lower allochthon and are para-autochthonous deposits (Fig. 56.1; Wenk 1961; Phillips et al. 1973; Montcrieff 1989; Smith & Robertson 1999; Higgins et al. 2001, 2004; Sønderholm et al. 2008). The deposits are overlain by the Hagar Bjerg and the Niggli Spids thrust sheets (Higgins et al. 2004, 2008; Fig. 56.1). These sediments accumulated at least 200 km to the west of the Eleonore Bay basin but are thought to be broadly correlative with the Tillite Group (Higgins & Leslie 2000, 2008).

Stratigraphy The allochthonous upper Neoproterozoic sediments in East and North-East Greenland are collectively referred to as the Eleonore Bay Supergroup (Andre´e Land Group) and the overlying Tillite Group (Haller 1971; Sønderholm & Tirsgaard 1993; Sønderholm et al. 2008) (Fig. 56.2). Associated para-autochtonous deposits are broadly correlated with the Storeelv Fm. of the Tillite Group in the Eleonore Bay Basin (Montcrieff 1989).

Andre´e Land Group The top unit of the Eleonore Bay succession is the Andre´e Land Group (Teichert 1933; Tirsgaard 1993, 1996; Sønderholm & Tirsgaard 1993; Tirsgaard & Sønderholm 1997) (Fig. 56.2), which consists primarily of shallow to relatively deeper-water marine carbonates succeeded by marine deep-water siltstone-shale, cherty rhythmites and carbonate. It totals c. 1500 m in thickness and was recognized previously as the ‘Die Kalk-Dolomit Serie’ (i.e. ‘Limestone-Dolomite Series’; Teichert 1933), which was subdivided into a number of ‘bed groups’ (Teichert 1933; Schaub

584

S. STOUGE ET AL.

Fig. 56.3. Stratigraphy of the uppermost Andre´e Land Group (Bedgroups 19 and 20) and Tillite Group. The sedimentary logs from Andre´e Land and Ella Ø represent the northern and southern development of the succession respectively. AS, sandstone (aeolian) unit forming at the top of Bedgroup 20; CD, cap dolomite following above the diamictite of the Storeelv Fm.

1950; Katz 1952; Eha 1953; Fra¨nkl 1953; Haller 1953, 1971; Cowie & Adams 1957; Sommer 1957). These informal bed groups have since been widely used as standard reference nomenclature in the literature for the Eleonore Bay Supergroup (Haller 1971; Hambrey & Spencer 1987; Herrington & Fairchild 1989; Swett & Knoll 1989; Fairchild & Hambrey 1995). The Andre´e Land Group was formally established by Sønderholm & Tirsgaard (1993), who also introduced informal litho-units (AL1 –7), which should serve as replacements for the bedgroups. However, the older stratigraphic terminology summarized by Haller (1971) is still applicable and is used here as stratigraphical framework (Fig. 56.2). The upper Andre´e Land Group, named Bedgroup 18 (¼ AL5 of Sønderholm & Tirsgaard 1993), is composed of marine,

shallow-water carbonates that accumulated on a ramp situated along a former low-latitude passive continental margin (Frederiksen et al. 1999; Frederiksen 2000; Stouge et al. 2001; Mac Niocaill et al. 2004, 2008; Sønderholm et al. 2008). The uppermost strata of the Andre´e Land Group are referred to as Bedgroup 19 and Bedgroup 20 (¼ AL6 and AL7 of Sønderholm & Tirsgaard 1993). These are up to c. 400 m thick and consist of fine-grained clastic sediments and carbonate.

Tillite Group The upper part of the Neoproterozoic succession comprises the siliciclastic deposits and minor carbonates of the Tillite Group

NEOPROTEROZOIC NE GREENLAND

585

Fig. 56.4. Detailed stratigraphic logs of the Ulvesø and Storeelv formations. The Ulvesø Formation is from the type section at Ulvesø in Kløftelv on Ella Ø and the reference section from Tømmerbugt til Bastion Bugt, Ella Ø. Stratigraphic log of the Storeelv Formation is from the coast section in Bastion Bugt, Ella Ø.

586

S. STOUGE ET AL.

(c. 900 m), representing terrestrial and marine depositional settings in five different formations (Figs 56.2 & 56.3). The marine deposits accumulated at some distance from the continental margin in a deep-water, oceanic setting. The diamictite facies of the Ulvesø and Storeelv formations are separated by mudstone and sandstone facies of the Arena Fm. and succeeded, first by deep-water sediments and then by shallowing-upwards successions, where the uppermost unit concludes with evaporitic environments (Canyon and Spiral Creek formations; Fig. 56.2). A regional synthesis of the depositional environments and stratigraphical nomenclature was reviewed by Haller (1971) and revisions were published by Cowie & Adams (1957), Hambrey & Spencer (1987), Fairchild & Hambrey (1995), Sønderholm & Tirsgaard (1993) and Sønderholm et al. (2008). The switch from deposition of carbonate to deep-water, mostly fine-clastic, turbidites on the slope in the upper part of the Andre´e Land Group heralded the deposition of the Cryogenian –Ediacaran diamictite units in the region. The Eleonore Bay Supergroup disconformably overlies the Krummedal supracrustal succession (Higgins et al. 2004, 2008; Nystuen et al. 2008). Granites, which are between 950 and 920 Ma old (Kalsbeek et al. 2001, 2008), intrude the Krummedal supracrustal sequence, and the c. 900 Ma date for the granites may represent a maximum age for the initial start of deposition in the Eleonore Bay Basin. The late Neoproterozoic succession is unconformably overlain by the quartzitic sandstone of the Kløftelv Fm. (Lower Cambrian) of the Kong Oscar Fjord Group in the central fjord zone (Fig. 56.2).

Associated para-autochthonous strata Diamictite beds have been recorded from Charcot Land (Montcrieff 1989), Ga˚seland (Wenk 1961; Phillips et al. 1973) and Ma˚lebjerg windows (Smith & Robertson 1999) (Fig. 56.1). The diamictite deposits are preserved in pockets on the surface of the basement and below the higher thrusts. The Støvfanget Fm. (Montcrieff 1989) is named from the succession in the Ga˚seland window. The strata from the Charcot Land and Ma˚lebjerg windows were named the Tillite Nunatak Fm. (Montcrieff 1989; Smith & Robertson 1999). The sediments are considered to be lateral equivalents, so the name Støvfanget Fm. is used here for these deposits.

In detail, Unit A begins with quartzitic, fine-grained sandstone overlying the black limestone of Bedgroup 18. The quartzitic sandstone is succeeded by bedded chert and rhythmically bedded grey chert. The chert beds are 1–5 cm thick and have millimetre-scale partings of grey-green shale. This unit is succeeded by rhythmically bedded chert with partings consisting of 0.1 –0.5 mm thin green-grey shale as well as interbeds of grey-red-brown shale overlain by finely laminated chert and grey shale. The chert is characterized by light yellow to brown weathering surfaces. Besides occasional ripples, sedimentary structures are rare to absent in the unit. The basal quartzitic unit of Unit A is sparsely exposed, so it has not been reported previously. It is well exposed in the region from Andre´e Land and east of Eleonore Bay and towards the north (Figs 56.1 & 56.3). On Ella Ø, Schaub (1950) and Hambrey & Spencer (1987) described Unit A in detail and divided it into several beds, but Unit A has also an extensive lateral distribution. It is recorded across the whole basin and with a nearly uniform development. Many of the individual beds can be traced over long distances; one prominent chert/carbonate horizon with shale partings, 7 to 10 m thick, can easily be traced laterally from Lyell Land to Ella Ø to Ole Rømer Land and farther to the north (i.e. for more than 200 km); it can thus be used as a marker bed across the whole basin. Unit B is composed mainly of grey and black relatively organic-rich shale that is interbedded with subordinate parted to ribbon limestone and limestone breccia. The ribbon and parted limestone forms the base of the unit and is very fine-grained, dark grey to black, finely laminated and with limestone beds varying in thickness from 4 to 40 cm. Several horizons of the ribbon and parted limestone are slumped and contorted. Above the interval with parted to ribbon limestone and breccia beds follows a unit up to 100 m thick, dark grey to black shale. Unit B is recorded only in the southern extension of the basin (Sønderholm et al. 2008). The limestone breccias of Unit B have a chaotic fabric with a matrix composed of grey to black, dirty green to dull brown, purple-to-yellow weathering calcareous mud and lime mudstone. Purple-to-yellow weathering carbonate and elongated plates of low lithological variety are the most common clasts. They are composed of grey, blue-grey, to dark grey limestone and dolomite plates. The lime breccias are mostly thin but well exposed. On Ella Ø, one breccia horizon up to 10 m thick serves as a local lithological marker band in this region (Hambrey & Spencer 1987, fig. 6; Herrington & Fairchild 1989).

Glaciogenic deposits and associated strata Bedgroup 19, Andre´e Land Group

Bedgroup 20, Andre´e Land Group

Bedgroup 19 is characterized by deep-water black, green, grey and red shales, interbedded with chert, dolostone, minor limestone, interbedded limestone and shale and carbonate breccias and conglomerates (Fig. 56.2). Bedgroup 19 is overlain either by Bedgroup 20 or the Ulvesø Fm. of the Tillite Group (Sønderholm et al. 2008). The lateral distribution of facies is heterogeneous and the succession shows a northern and a southern development (Fig. 56.3). Bed group 19 is here subdivided into two informal units: Unit A and Unit B. Unit A is well developed throughout the basin, whereas Unit B is only found in the southern portion of the basin. Unit A is the stratigraphically oldest strata of Bedgroup 19 of the upper Andre´e Land Group (Sønderholm et al. 2008). It is 200 m or more thick, and is characterized by mixed chert and fine-grained to very fine-grained siliclastic sediments associated with minor carbonate. Lithologies include argillite (shale); siltstone; bedded light grey, dense thin-bedded chert, developed as rhythmites; and siliceous argillite interbedded with variable amounts of thinbedded and very fine-grained grey, green to violet chert. Yellow to yellow-brown weathering is characteristic of the finely laminated grey to green to violet chert.

Bedgroup 20 (AL7 as per Sønderholm & Tirsga˚rd 1993) is mainly a carbonate unit (Katz 1952; Eha 1953; Fra¨nkl 1953; Haller 1953; Hambrey & Spencer 1987; Herrington & Fairchild 1989; Sønderstro¨m et al. 2008), which conformably overlies Unit A of Bedgroup 19. Bedgroup 20 is typically composed of (i) silty, planar bedded ribbon limestone at the base, (ii) lime mudstone and bedded limestone in the main part of the unit (Fig. 56.3) and (iii) planar lamellites associated with small domal stromatolites and minor nodular cherts. Oolitic-pisolitic limestone, up to 40 m thick, prevails in the higher part of the unit on Ole Rømers Land (Swett & Knoll 1989; Frederiksen 2000). Bedgroup 20 varies in thickness from 240 m in Ole Rømer Land and decreases towards the south where 2 m are recorded on the southern part of Ymer Ø. The unit extends northwards as far as Hudson Land. Farther to the south and towards the southern extension of the Eleonore Bay Basin, the carbonates of Bedgroup 20 are absent. The uppermost part of Bedgroup 20 is composed of yellow, medium-grained sandstone with well-developed and large crossbeds (Fig. 56.3). This sandstone subunit (c. 40 m) is shown in

NEOPROTEROZOIC NE GREENLAND

Fig. 56.3 as ‘aeolian sandstone’ (AS). This subunit displays the same geographical distribution as Bedgroup 20 (Sønderholm et al. 2008).

Ulvesø Fm., Tillite Group At Ella Ø, the Ulvesø Fm. comprises a generally coarsening upwards succession composed of diamictite, associated with minor laminite and shale, massive sandstone and conglomerate (Figs 56.3 & 56.4). The crudely bedded diamictite is composed predominantly of carbonate clasts derived mainly from the older Andre´e Land Group and sediments of Bedgroup 19 in the southern part of the basin, and from carbonate units of the older Andre´e Land Group and Bedgroup 20 to the north of Ymer Ø. Beds are normally 30–50 cm thick, and clasts are occasionally up to 1 m in size. The matrix is grey-green and composed of shale, silt and sand, which weather characteristically as dull yellow to brown. The formation varies laterally in thickness from nearly 10 m in the north to a maximum thickness of c. 320 m on northern Scoresby Land. The erosive-based diamictite is rich in striated outsized clasts up to a metre in size and slump folds and mega breccias occur frequently throughout the succession. Towards the top, diamictite interfingers with erosive-based sandstone and clastsupported conglomerate, beds of shale and brecciated limestone derived from the breakdown of the underlying carbonate-shale platform. Together with the lowermost part of the sandstones of the overlying Arena Fm., the diamictite of the Ulvesø Fm. appears strongly deformed, occurring as metre-sized ball and pillow structures.

Arena Fm., Tillite Group The Arena Fm. is composed of grey, dark-grey to green shale, siltstone and sandstone (Figs 56.2 & 56.3). The formation was formally defined by Hambrey & Spencer (1987), with its type locality at Arenaen on Gunnar Andersson Land, Ymer Ø (738200 N, 248500 W). The formation was previously included in the Tillite Series (Kulling 1929; Schaub 1955; Katz 1954, 1961) and the Cape Oswald Fm. of Poulsen (1930). Later it was named the Inter Tillite Member (Schaub 1950) or Inter-Tillite (Haller 1971; Henriksen & Higgins 1976; Higgins 1981). The formation varies from c. 100 m in thickness in Strindberg Land and Ole Rømer Land, through 220– 360 m in thickness in Albert Heim Bjerge (Hudson Land). The formation begins with shale and silt beds followed by thick-bedded green to grey sandstone. The shale and siltstone are the dominant lithologies, with a minor increase in the amount of sandstone occurring towards the top of the Arena Fm. The recessive Arena Fm. separates the diamictite units and is not easy to access in inland outcrops, and the beds of the unit are commonly covered. On Ella Ø, the lowest part of the unit is composed of yellow sandstone, which is similar to the sandstone on Scoresby Land. The latter includes thin diamictite horizons (Hambrey & Spencer 1987). Sedimentary structures in the Arena Fm. comprise cross-lamination, wave ripples, graded bedding, load-casts and slump structures.

587

alternate with breccias of limestone and sandstone. More commonly, they interfinger with cross-bedded sandstone and laminated shale with outsized clasts (Figs 56.3 & 56.4). Conglomerate-filled mega-channels truncate the diamictite, mainly in the middle part of the formation. Towards the top, diamictite units once more dominate the succession. However, here it interfingers with minor beds of laminated shale with dropstones and bedded sandstone.

Canyon Fm., Tillite Group The Canyon Fm. was first named the Canyon Series by Schaub (1955) and Fra¨nkl (1953). Later, Haller (1971) modified it to the Canyon Fm. The name is derived from Tillite Canyon Fm. (Poulsen 1930), but the name Tillite was dropped by subsequent authors. In the northern part of the basin, this formation includes a cherty dolostone horizon at the base. The cherty carbonate unit, up to 10 m thick, is situated at the base of the formation and extends laterally from Andre´e Land (Tillite Kløft and west of Kap Weber) and northwards to Albert Heim Bjerge (Cowie & Adams 1957). The unit is composed of light grey, dense and finely laminated chert or siliceous dolostone. Above follows a rhythmically bedded shale and chert. The shale is commonly very red, and the chert beds are pale grey to nearly white, so the shale/chert unit is easy to recognize from a distance. This cherty carbonate unit is not present on Ella Ø, and red shale follows above the Storeelv diamictite. The characteristic, rhythmically bedded, red shale and grey finely laminated chert appear c. 10 m above the Storeelv Fm. The subsequent and main part of the succession of the Canyon Fm. is composed of red, grey green and black shale. Towards the top and on Ella Ø, the Canyon Fm. becomes increasingly sandy and the sandstone and shale display well-developed, upwards coarsening sequences. On Ella Ø, the top of the formation is developed as c. 50 m yellow weathering dolostone with small stromatolitic reefs and algal laminations. This facies of the Canyon Fm. is only seen on Ella Ø. Elsewhere, black shale with prominent yellow-weathering surfaces marks the top of the formation.

Spiral Creek Fm., Tillite Group Spiral Creek Fm. is the topmost unit of the Neoproterozoic succession in East and North-East Greenland (Fig. 56.2). The formation is named after Spiral Creek, a locality on the northern shore of Andre´e Land (Poulsen 1930, p. 307), where the unit is well displayed. It is also known from Lyell Land, Ella Ø and Ole Rømer Land. Spiral Creek Fm. is up to 45 m thick and is composed of siltstone, sandstone and stromatolitic dolostone. The colours of the sediments vary from yellow, grey to green and maroon. The formation is divided into several beds or horizons (Poulsen 1930; Schaub 1950; Cowie & Adams 1957). Halite pseudomorphs, mudcracks, ripples and intraformational breccia are frequent in the beds. The sediments probably deposited in a playa-like setting and under semiarid conditions (Hambrey & Spencer 1987; Fairchild & Hambrey 1995).

Storeelv Fm., Tillite Group

Støvfanget Fm., para-autochthonous deposits

The Storeelv Fm. ranges in thickness from 220 m on Ella Ø (Fig. 56.3) and Arenaen to 60 m in Strindberg Land in the north and Canning Land in the south. On Ella Ø, the formation conformably rests on a striated glacial boulder pavement of outsized far travelled erratics. Crudely bedded reddish diamictite with clasts of local and extra-basinal origin, including igneous and metamorphic rocks of the Archaean basement, occasionally

The sediments of the Støvfanget Fm. include diamictite, sandstone and laminated mudstone with dropstone (Smith et al. 2004; Sønderholm et al. 2008). The composition of the clasts is heterogeneous and derived from local sources; granitic blocks are common. The blocks vary in size and are up to 2 m, rarely 6 m, in size. The succession varies from a couple of tens of metres to a maximum of c. 200 m in thickness.

588

S. STOUGE ET AL.

Boundary relations with overlying and underlying non-glacial units Boundary relations within the Neoproterozoic succession The lower boundary of Bedgroup 19 is not well exposed in the southern part of its extent of outcrop. The boundary relationships, however, are well displayed in Andre´e Land and Strindberg Land, where the boundary is a disconformity. The carbonate top surface of Bedgroup 18 is dissolved and has karst-like features (Frederiksen 2000; Sønderholm et al. 2008). This carbonate surface is overlain by a thin, quartzitic and medium-grained sandstone. The quartzitic bedded sandstone is succeeded by bedded finegrained to grey chert deposits with thin shale partings. The lower boundary of Bedgroup 20 is conformable, with a gradual transition from dark grey to black shale, through interbedded silty laminated limestone to bedded limestone. The base of the overlying Tillite Group is usually sharp in the deep part of the basin, where black shale (Unit B) of Bedgroup 19 is overlain directly by the Ulvesø Formation diamictite (Hambrey & Spencer 1987). The boundary towards the north is different and the top carbonate strata of Bedgroup 20 are separated from the diamictite of the Ulvesø Fm. by a yellow (aeolian), quartzitic sandstone. This sandstone is conformably overlain by the Ulvesø Fm. of the Tillite Group. On Ymer Ø, Bedgroup 20 is in faulted contact with the overlying Ulvesø Fm. and boundary relationships are obscure (Hambrey & Spencer 1987; Moncrieff & Hambrey 1988). The upper boundary of the Ulvesø Fm. is in most places gradual and is marked by a transition from dark grey diamictite, conglomerate or sandstone to the lower sandstone of the Arena Fm. On Ella Ø the base of the overlying Arena Fm. is sharp and strongly deformed with the underlying Ulvesø into metre-sized load-casts. The upper Neoproterozoic para-authochthonous sediments rest directly on basement rocks and are tectonically constrained upwards by thrust sheets; however, the diamictite in the Ma˚lebjerg window is overlain disconformably by Lower Cambrian quartzitic sandstone referred to as the Slottet Fm. (Smith et al. 2004).

Boundary relations with the Lower Palaeozoic succession The Lower Cambrian Kløftelv Fm. of the Kong Oscar Fjord Group (Lower Cambrian to Middle Ordovician) unconformably overlies the Tillite Group across the whole Eleonore Bay basin and the boundary is a low-angle angular unconformity (Henriksen & Higgins 1976; Hambrey 1989; Hambrey et al. 1989; Stouge et al. 2001, 2002; Sønderholm et al. 2008). Thus the boundary relationships from the north to south are variably developed. In Andre´e Land and on Ella Ø, the upper Neoproterozoic evaporitic sediments of Spiral Creek Fm. are overlain unconfomably by the Lower Cambrian Kløftelv Fm., and the Lower Palaeozoic succession begins with a basal conglomerate (Stouge et al. 2001, 2002). Away from Andre´e Land and Ella Ø, the Spiral Creek Fm. and the upper stromatolitic limestone facies of the Canyon Fm. are absent and black shales of Canyon Fm. are directly overlain by a basal conglomerate followed by the arenitic sandstone of the Lower Cambrian Klo¨ftelv Fm. The hiatus separating the Neoproterozoic from the Palaeozoic strata is proposed to have lasted for about 35 Ma (see Sønderholm et al. 2008 for further discussion).

Chemostratigraphy Early studies of carbonate and organic carbon geochemistry were carried out by Knoll et al. (1986). Samples from Bedgroup 7 through 20 (Eleonore Bay Supergroup) as well as from the

Canyon and Spiral Creek formations (Tillite Group) revealed negative d13Corg values between –20 and – 35‰PDB. The d13Ccarb values were predominantly positive throughout the Eleonore Bay Supergroup with a sharp decrease to negative values ( –4 to –9‰PBD) in the uppermost Bedgroup 19/20 and negative values (c. 0 to –5‰PDB) in the Canyon and Spiral Creek formations. Both geochemical data sets followed very similar profiles suggesting a primary signature and an unusual secular variation in the Neoproterozoic carbon cycle (Knoll et al. 1986). Total carbon (TC), total organic carbon (TOC) and total sulphur (TS) have been recorded through an interval ranging from the upper part of the Andre´e Land Group and up to and including the top of Tillite Group; the results will be presented elsewhere. Stable C-isotopes from the upper part of the Andre´e Land Group have also been analysed and will be presented elsewhere (see Halverson et al. 2005; Sønderholm et al. 2008 for further details).

Other characteristics Organic constructed organisms (acritarchs) in the succession represent the plankton of the Andre´e Land –Tillite Group basin. The microflora assemblage is of low diversity and consists of both small and large smooth leosperoid types associated with only few acanthomorphs (Vidal 1976, 1979, 1985). The biostratigraphic value of the acritarch assemblage is at present limited, but in general the microflora from Andre´e Land Group is assigned to the Cryogenian (Vidal 1976, 1979; Green et al. 1987, 1988, 1989). Other microfossils reported from the succession are vaseshaped protists (Green et al. 1988), similar to those reported in other Neoproterozoic successions (Porter & Knoll 2000; Porter et al. 2003).

Palaeolatitude and palaeogeography Palaeogeographical reconstructions of East and North-East Greenland have long been tied to the palaeogeographical evolution of the Neoproterozoic succession in northeastern Svalbard based on detailed lithostratigraphic similarity between these two basins (Harland & Gayer 1972; Fairchild & Hambrey 1995; Halverson et al. 2004; Sønderholm et al. 2008; Halverson 2011). The Hekla Hoek succession in northeastern Spitsbergen is similar to the Eleonore Bay Supergroup of northeastern Greenland and shows development from siliciclastic to carbonate deposition followed by glaciogenic deposits. This similarity is also strengthened by comparable and substantial thicknesses of the lithological units in the two regions (Harland & Geyer 1972; Hambrey 1989; Fairchild & Hambrey 1995; Harland 1997). The chemostratigraphy shows similarities (Knoll et al. 1986) suggesting close affinities between the two areas. Halverson et al. (2005) identified the distinctive pre-Marinoan negative d13C anomaly or the ‘Trezona anomaly’ beneath the lowest diamictite in northeastern Svalbard. The same anomaly has been recognized from within the upper unit of the Eleonore Bay Supergroup in East and North-East Greenland (Sønderholm 2008; Kristiansen unpublished data), which also provides a strong argument for the similarity of the two regions. Palynology studies in both East and North-East Greenland and Svalbard (Vidal 1985; Green et al. 1989; Swett & Knoll 1989) also indicate close relationships between the two areas and as a whole the two regions are considered to be deposited within the same basin or series of ensialic basins, with East and North-East Greenland located near eastern Svalbard (Sønderholm et al. 2008). Recent and ongoing palaeomagnetic studies have shown that East and North-East Greenland was situated at equatorial latitudes at about 38S during the Cryogenian, with palaeomagnetic data from 23 sites (105 specimens) towards the top of the Eleonore Bay Supergroup yielding a mean palaeomagnetic inclination

NEOPROTEROZOIC NE GREENLAND

of 58 (k ¼ 12; a95 ¼ 98) (Mac Niocaill et al. 2004, 2008; Kilner 2005). These data are constrained by a positive fold test (Caledonian folding) and a positive conglomerate test on clasts in the overlying lower Tillite. At that time the carbonate sedimentary succession of the Andre´e Land Group was deposited in a platform setting on a ramp. The region formed the eastern part of the northeastern margin of the Rodinia palaeocontinent. A change in the latitudinal position began at the base of Bedgroup 19 as the continental margin region started to move southwards. During deposition of the Storeelv Fm., the area was situated at c. 668 southern latitude, with sparse palaeomagnetic data from six sites (nine specimens) in the Tillite Group yielding a mean palaeomagnetic inclination of 788 (k ¼ 10; a95 ¼ 228) (Mac Niocaill et al. 2004, 2008; Kilner 2005), although this awaits confirmation with further analyses in progress. The movement towards higher, southern latitudes is associated with the shift from stable carbonate accumulation of Andre´e Land Group to the siliciclastic deposition of Bedgroup 19 and the start of foundering of the platform. The presence of Bedgroup 20 signifies that carbonate accumulation – for a short while – returned to the region, but finally ceased in the late Cryogenian.

Geochronological constraints Radiometric ages have not been obtained from the Andre´e Land and Tillite groups. A Cryogenian age is confirmed based on the biostratigraphic record (see previous section).

Discussion Depositional setting The stratigraphy of the upper Andre´e Land Group has been considered to conform to a normal superposition succession, where Bedgroup 20 overlies Bedgroup 19. However, Eha (1953) proposed that the carbonate sediments of Bedgroup 19 (¼ base of Unit B of this chapter) represented the lateral equivalents of Bedgroup 20. In this review, the upper part of Bedgroup 19, that is, the sediments of Unit B and Bedgroup 20 (¼ AL7), are considered coeval (Fig. 56.3), as this interpretation is supported by the sedimentology, intragroup stratigraphy and the overall stratigraphical succession developed during recent fieldwork. Unit B is considered the fine-grained and deep-water equivalent of Bedgroup 20, dominated by parted to ribbon limestone, debris-flow deposits, lime-breccias containing intrabasinal clasts and distal turbiditic shale at the basin margin and organic-rich shale in the basin. Bedgroup 19 (Eha 1953; Fra¨nkl 1953; Haller 1971) of the uppermost Andre´e Land Group significantly provides the first record of fine-grained to very fine-grained siliciclastic sediment and chert rhythmites (i.e. Unit A), which followed disconformably above the shallow-water stromatolite-bearing limestone and slightly deeper-water carbonate rocks. Limestone (Bedgroup 20) and shale and ‘lime-breccias’ (i.e. Bedgroup 19, Unit B) underlie the first diamictite unit (i.e. Ulvesø Fm.) of the Tillite Group. This succession is interpreted as recording deepening within the basin and foundering of a carbonate platform that had developed on a stable craton and had been in existence for a long period of time (i.e. during deposition of the underlying Ymer Ø Group and most of the Andre´e Land Group). In addition, the rapid facies change from carbonate accumulation to siliciclastic deposition shows that rift-related subsidence was initiated in the Cryogenian, creating north –south elongated basins with isolated and carbonatecapped platforms and deep-water troughs. Bedgroup 20 carbonates probably developed a low-relief bank profile with mobile pisolite shoals at the margin, and shallow sub-tidal carbonate and peritidal deposits in the interior. The final carbonate deposition in the region (i.e. before the occurrence of diamictite and represented by

589

Bedgroup 20) was due to a marked regression and perhaps also due to a change in palaeo-climate. The diamictites of the Tillite Group have long been considered glaciogenic deposits (Kulling 1929; Poulsen 1930 and subsequent authors). Hambrey & Spencer (1987) and Moncrieff & Hambrey (1988, 1990) provided detailed models for glacial depositional environments. Halverson et al. (2004, 2005) placed the Tillite Group deposits within the scope of ‘Snowball Earth’. The Ulvesø and Storeelv formations are indeed dominated by diamictite, with megabreccias and sculptured clasts alternating with minor quantities of shale, sandstone and conglomerate. A terrestrial origin can most probably be applied to the lowermost and middle parts of the Storeelv Fm.; the facies successions here indicate a highly diverse depositional system. Glaciers grounded on the shelf or in fjords delivered till and glacio-fluvial deposits interbedded with flow diamictites and lacustrine and aeolian facies. Other parts of the Storeelv and most of the Ulvesø Fm. were most probably deposited below the grounding line in sub-aqueous environments. Consequently, the Tillite Group contains a limited amount of tillite, and although most components in the succession originated from glacial debris, most diamictites finally settled in marine settings, especially those of the Ulvesø Fm. A marine origin is supported by the occurrence of algal life forms throughout the Tillite Group, except where sediments show signs of oxidation (Stouge & Piasecki unpublished data). The diamictite units were principally deposited from suspensions of glacial debris released beneath floating ice shelves and by ice-rafting in a glaciomarine environment. These deposits suffered slumping and reworking by sediment gravity flows in a near shelf break environment. Only parts of the Storeelv Fm. indicate that glaciers occasionally reached grounding line or more rarely that terrestrial ice sheets deposited tillite and glaciofluvial deposits. It is not clear how much conglomerate and sandstone that originated in on-shore fluvial and aeolian environments was actually reworked to eventually settle in submarine canyons and in current- and wavedominated shallow water.

Sequence stratigraphy The upper part of the Andre´e Land Group and the Tillite Group represent three depositional sequences that mostly accumulated as deeper- to deep-water sediments in oceanic settings (Fig. 56.2; Sønderholm et al. 2008). The oldest sequence is represented by Bedgroups 19 and 20. The lower sequence boundary is an erosional or a dissolution surface. Siliciclastic silt, shale and chert were deposited in the transgressive system tract during the progradation of the coastline outside the study area. The maximum flooding surface can be traced across the whole basin, and the following highstand deposition created the regressive platform of Bedgroup 20 and the distal deposits of Unit B. The regressive top of Bedgroup 20 is composed of shallow-water silt and sandstone deposited in tidal flat environment and aeolian sandstones. The second sequence comprises the Ulvesø and Arena formations (Fairchild & Hambrey 1995: Sønderholm et al. 2008), which are developed exclusively in clastic facies. The third sequence comprises the transgressive Storeelv Fm. – including the thin carbonate unit – and the Canyon Fm., which becomes regressive upwards, permitting shallow-water carbonates to accumulate. The overlying Spiral Creek Fm. represents the ultimate part of the highstand, and the regressive top of the Tillite Group accumulated in shoreface and semi-arid, evaporitic lagoonal environments (Fairchild & Herrington 1989).

Timing of glaciation Possibly two widespread or even global glaciations took place during the late Neoproterozoic (Hambrey & Harland 1995; Knoll

590

S. STOUGE ET AL.

2000; Hoffman & Schrag 2002). It is, however, uncertain if the two diamictite units in East and North-East Greenland represent one or two separate glaciations and in what respect they correlate with other Neoproterozoic phases of glaciations (e.g. Kennedy et al. 1988; Kaufman et al. 1997; Brasier & Shields 2000; Hoffmann & Schrag 2002; Robb et al. 2004). Halverson et al. (2004, 2005) suggested that the two diamictite units on Svalbard represent one glaciation. This idea is based on the d13C-isotope curve (Knoll et al. 1986; Halverson et al. 2004, 2005) and the presence of a distinctive d13C anomaly (¼‘Trezona anomaly’) found beneath the lowest tillite unit in Svalbard. In addition, the presence of the thin carbonate unit, which should be equal to a ‘cap dolomite’ (Fairchild & Hambrey 1995), is typical of the late Cryogenian glaciation. The diamictite units in Svalbard were therefore related genetically to the late Cryogenian glaciation (Marinoan) by Halverson et al. (2004, 2005). Based on their stratigraphic similarity (mentioned previously), the two diamictite units in East and North-East Greenland may also contain tillite deposits of the Marinoan (late Cryogenian) glaciation. The trend of the d13C-isotope curve (Knoll et al. 1986; Kristiansen et al. unpublished data) from the upper part of the Andre´e land Group carbonates, combined with the presence of a possible cap dolomite above the Storeelv Formation in the northern part of the basin, suggests that the two diamictite units in East and North-East Greenland could be related to one glaciation. A problem with this interpretation arises, however, because any trace of an older Cryogenian glaciation often found in other Neoproterozoic successions is apparently missing in the Greenland succession. Although no diamictite-bearing units occur below the Ulvesø Fm., the extensive disconformity developed at the boundary between Bedgroup 18 and Bedgroup 19 may be the response to a global or eustatic lowering of sea level causing subaerial exposure and subsequent dissolution of the shallow marine sea bed in the region. The subsequent deepening (see ‘Sequence stratigraphy’ section) could thus be a result of the rapid melting of the ice sheet producing fine- to very fine-grained sedimentation.

Palaeogeographical setting The transition from the upper Andre´e Land Group to Bedgroup 19 and further to the Tillite Group has also been interpreted as deposition in an early rift basin related to the splitting of Rodinia and before the opening phase of the Iapetus Ocean (Fig. 56.2; Herrington & Fairchild 1989; Stouge et al. 2001; Eyles & Januszczak 2004; Nystuen et al. 2008; Sønderholm et al. 2008). According to the interpretation of Eyles & Januszczak (2004), the upper part of the Andre´e Land Group and the main part of the Tillite Group accumulated during active rifting, and deep-water sediments accumulated in oceanic settings bordered by glaciated continents. Upwards, the uppermost part of the Andre´e Land Group and the top of the Tillite Group became shallower and the regressive top of the Andre´e Land Group is marked by a prominent dissolution surface, which is overlain by marine sandstone. The palaeomagnetic signal supports rifting and the development of an extension-related basin and the rapid drift of the region from low to higher southern latitudes (Mac Niocaill et al. 2004, 2008). A Cryogenian age for the onset of rifting as potentially recorded in the East and North-East Greenland succession is earlier than that initially proposed for the region by previous authors (i.e. during the Ediacaran; e.g. Soper 1994; Higgins & Leslie 2000). An additional and more significant problem with the rift-basin model is that no evidence or structures such as rift-related dolerite dykes or volcanic or pyroclastic deposits have been observed in the region, which might suggest a rifting stage (Soper 1994; Higgins & Leslie 2008; Sønderholm et al. 2008). One explanation could be that the eastern Laurentian margin in East and North-East Greenland was, in fact, located in a cratonward position relative to the rift zone.

The authors thank the reviewers M. J. Hambrey and A. H. Knoll for their valuable comments and suggestions, from which the manuscript greatly benefited. Fieldwork in Greenland was funded both by the Carlsberg Foundation and the Danish Research Council. The Geological Survey of Denmark and Greenland (GEUS) gave permission to reproduce elements in Figure 56.1. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Bengaard, H.-J. 1991. Upper Proterozoic (Eleonore Bay Supergroup) to Devonian central fjord zone, East Greenland, geological map, 1: 250000. Copenhagen, Geological Survey of Greenland. Brasier, M. D. & Shields, G. 2000. Neoproterozoic chemostratigraphy and correlation of the Port Askaig glaciation, Dalradian Supergroup of Scotland. Journal of the Geological Society, London, 157, 909– 914. Cowie, J. W. & Adams, P. J. 1957. The geology of the CambroOrdovician rocks of central Greenland. Part 1: Stratigraphy and structure. Meddelelser om Grønland, 153, 193. Eha, S. 1953. The pre-Devonian sediments on Ymers Ø, Suess Land, and Ella Ø (East Greenland) and their tectonics. Meddelelser om Grønland, 111, 105. Eyles, N. & Januszczak, N. 2004. ‘Zipper-rift’: a tectonic model for Neoproterozoic glaciations. Earth Science Reviews, 65, 1 –73. Fairchild, I. J. & Herrington, P. M. 1989. A tempestite-stromatoliteevaporite association (Late Vendian, East Greenland): a shorefacelagoon model. Precambrian Research, 43, 101– 127. Fairchild, I. J. & Hambrey, M. J. 1995. Vendian basin evolution in East Greenland and NE Svalbard. Precambrian Research, 73, 217– 223. Fra¨nkl, E. 1953. Geologische Untersuchungen in Ost-Andrees Land (NE-Grønland). Meddelelser om Grønland, 113, 160. Frederiksen, K. S. 2000. Evolution of a Late Proterozoic carbonate ramp Ymer Ø and Andre´e Land Groups, Eleonore Bay Supergroup, East Greenland: response to relative sea-level rise. Polarforschung, 68, 125– 130. Frederiksen, K. S., Craig, L. E. & Skipper, C. B 1999. New observations of the stratigraphy and sedimentology of the Upper Proterozoic Andre´e Land Group, East Greenland: supporting evidence for a drowned carbonate ramp. Danmarks og Grønlands Geologiske Undersøgelse Rapport, 1999/19, 145– 158. Gee, D. G., Fossen, H., Henriksen, N. & Higgins, A. K. 2008. From the Early Paleozoic platforms of Baltica and Laurentia to the Caldonide Orogen of Scandinavia and Greenland. Episodes, 31, 1 – 8. Green, J. W., Knoll, A. H., Golubic, S. & Swett, K. 1987. Paleobiology of distinctive benthic microfossils from the Upper Proterozoic limestone-dolomite ‘Series’, central East Greenland. American Journal of Botany, 74, 928– 940. Green, J. W., Knoll, A. H. & Swett, K. 1988. Microfossils from oolites and pisolites of the Upper Proterozoic Eleonore Bay Group, central East Greenland. Journal of Paleontology, 62, 835– 852. Green, J. W., Knoll, A. H. & Swett, K. 1989. Microfossils from silicified stromatolitic carbonates of the Upper Proterozoic LimestoneDolomite ‘Series’, central East Greenland. Geological Magazine, 126, 567– 585. Haller, J. 1953. Geologie und Petrographie von West-Andre´es Land und Ost-Frænkels Land (NE-Gro¨nland). Meddelelser om Grønland, 113, 196. Haller, J. 1971. Geology of the East Greenland Caledonides. Interscience Publishers, New York. Halverson, G. P. 2011. Glacial sediments and associated strata of the Polarisbreen Group, northeastern Svalbard. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 571–580. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Toward a Neoproterozoic composite carbonisotope record. Geological Society of America, Bulletin, 117, 1181– 1207.

NEOPROTEROZOIC NE GREENLAND

Halverson, G. P., Maloof, A. C. & Hoffman, P. F. 2004. The Marinoan glaciation (Neoproterozoic) in northeast Svalbard. Basin Research, 16, 297– 324. Hambrey, M. J. 1983. Correlation of late Proterozoic tillites in the North Atlantic region and Europe. Geological Magazine, 120, 290– 320. Hambrey, M. J. 1989. The Late Proterozoic sedimentary record of East Greenland: its place in understanding the evolution of the Caledonide Orogen. In: Gayer, R. A. (ed.) The Caledonide Geology of Scandinavia. Graham & Trotman, London, 352, 257– 262. Hambrey, M. J. & Spencer, A. M. 1987. Late Precambrian glaciation of central East Greenland: Meddelelser om Grønland, Geoscience, 19, 50. Hambrey, M. J. & Harland, W. B. 1995. The Late Proterozoic glacial era. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 255– 272. Hambrey, M. J., Peel, J. S. & Smith, M. P. 1989. Upper Proterozoic and Lower Palaeozoic in northern East Greenland. Rapport Grønlands Geologiske Undersøgelse, 145, 103– 108. Harland, W. B. 1997. The Geology of Svalbard. Geological Society of London, Memoirs, 17, 521. Harland, W. B. & Gayer, R. A. 1972. The Arctic Caledonides and earlier oceans. Geological Magazine, 109, 289–314. Henriksen, N. 1999. Conclusion of the 1 : 500000 mapping project in the Caledonian fold belt in North-East Greenland. Geology of Greenland Survey Bulletin 183, 10– 22. Henriksen, N. 2003. Caledonian orogen East Greenland 708– 828N, lithotectonic map 1: 1 000000. Geological Survey of Denmark and Greenland, Copenhagen. Henriksen, N. & Higgins, A. K. 1976. East Greenland Caledonian fold belt. In: Escher, A. & Watt, W. S. (eds) Geology of Greenland. Geological Survey of Greenland, Copenhagen, 182–246. Henriksen, N. & Higgins, A. K. 2008. Geological research and mapping in the Caledonian orogen in East Greenland, 708N–828N. 1 – 28. In: Higgins, A. K., Gilotti, J. & Smith, P. M. (eds) The Greenland Caledonides: Evolution of the Northeast Margin of Laurentia. Memoir Geological Society of America, 202, 368. Herrington, P. M. & Fairchild, I. J. 1989. Carbonate shelf and slope facies evolution prior to Vendian glaciation, central East Greenland. In: Gayer, R. A. (ed.) The Caledonide Geology of Scandinavia. Graham & Trotman, London, 263–273. Higgins, A. K. 1981. The Late Precambrian Tillite Group of the Kong Oscars Fjord and Kejser Franz Josefs Fjord region of East Greenland. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 778– 781. Higgins, A. K. & Leslie, A. G. 2000. Restoring thrusting in the East Greenland Caledonides. Geology, 28, 1019– 1022. Higgins, A. K. & Leslie, A. G. 2008. Architecture and evolution of the East Greenland Caledonides – an introduction. In: Higgins, A. K., Gilotti, J. A. & Smith, M. P. (eds) The Greenland Caledonides: Evolution of the Northeast Margin of Laurentia. Geological Society of America, Memoir, 202, 369, 29– 53. Higgins, A. K, Leslie, A. G. & Smith, M. P. 2001. Neoproterozoic – Lower Palaeozoic stratigraphical relationships in the marginal thin-skinned thrust belt of the East Greenland Caledonides: comparisons with the foreland of Scotland. Geological Magazine, 138, 143– 160. Higgins, A. K., Elvevold, S. et al. 2004. The foreland-propagating thrust sheet architecture of the East Greenland Caledonides 728 – 758N. Journal of the Geological Society (London), 161, 1009– 1026. Higgins, A. K., Gilotti, J. A. & Smith, M. P. (eds) 2008. The Greenland Caledonides: Evolution of the Northeast Margin of Laurentia. Geological Society of America, Memoir, 202, 369. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129– 155. Kalsbeek, F., Thrane, K., Nutman, A. P. & Jepsen, H. F. 2000. Late Mesoproterozoic to early Neoproterozoic history of the East Greenland Caledonides: evidence for Grenvillian orogenesis? Journal of the Geological Society (London), 157, 1215– 1225. Kalsbeek, F., Jepsen, H. F. & Jones, K. A. 2001. Geochemistry and petrogenesis of S-type granites in the East Greenland Caledonides. Lithos, 57, 91 – 109, doi: 10.1016/S0024–4937(01)00038-X.

591

Kalsbeek, F., Higgins, A. K. & Jepsen, H. F. 2008. Granites and granites in the East Greenland Caledonides. 227– 249. In: Higgins, A. K., Gilotti, J. A. & Smith, M. P. (eds) The Greenland Caledonides: Evolution of the Northeast Margin of Laurentia. Memoir, Geological Society of America, 202, 369. Katz, H. R. 1952. Zur Geologie von Strindbergs Land (NE-Gro¨nland). Meddelelser om Grønland, 111, 150. Katz, H. R. 1954. Einige Benmerkungen zur Lithologie und Stratigraphie der Tillitprofile in Gebiet des Kejser Franz Josephs Fjord, Ostgro¨nland. Meddelelser om Grønland, 72, 63. Katz, H. R. 1961. Late Precambrian to Cambrian stratigraphy in East Greenland. In: Raasch, G. O. (ed.) Geology of the Arctic. Vol. 1. University of Toronto Press, Toronto, 299– 328. Kaufman, A. J., Knoll, A. H. & Narbonne, G. M. 1997. Isotopes, ice ages, and terminal Proterozoic earth history. Proceedings of the National Academy of Science, 94, 6600– 6605. Kennedy, M. J., Runnegar, B., Prave, A. R., Hoffman, K. H. & Arthur, M. 1998. Two or four Neoproterozoic glaciations? Geology, 26, 1059– 1063. Kilner, B. R. 2005. Low latitude Neoproterozoic glaciation: palaeomagnetic investigations of the Precambrian sequences of Oman and East Greenland. Unpublished D.Phil thesis, University of Oxford. Knoll, A. H. 2000. Learning to tell Neoproterozoic time. Precambrian Research, 100, 3– 20. Knoll, A. H., Hayes, J. M., Kaufman, A. J., Swett, K. & Lambert, I. B. 1986. Secular variation in carbon isotope ratios from Upper Proterozoic successions of Svalbard and East Greenland. Nature, 321, 832–838. Koch, L. 1929. The geology of East Greenland. Meddelelser om Grønland, 73, II(I), 204. Koch, L. 1930. Die technische Entwicklung Gro¨nlands. Geologische Rundschau, 21, 345– 347. Koch, J. & Haller, J. 1971. Geological map of East Greenland 72 –76 N. Lat. (1 : 250.000). Meddelelser om Grønland, 183, 26, 13 maps. Kulling, O. 1929. Stratigraphic studies of the geology of North-east Greenland. Meddelelser om Grønland, 74, 317–346. Mac NioCaill, C., Stouge, S., Harper, D. A. T., Christiansen, J., Kilner, B., Johnson, A. & Watts, C. 2004. Preliminary paleomagnetic results from the late Neoproterozoic of eastern Greenland: A low-latitude Sturtian glaciation? EOS Transactions of the American Geophysical Union, 85, 165. Mac Niocaill, C., Kilner, B., Stouge, S., Knudsen, M. F., Harper, D. A. T. & Christiansen, J. L. 2008. The Neoproterozoic drift history of Laurentia: a critical evaluation and new palaeomagnetic data from Northern and Eastern Greenland. EOS Transactions of the American Geophysical Union, 89, S514– 05. Manby, G. M. & Hambrey, M. J. 1989. The structural setting of the Late Proterozoic tillites of East Greenland. In: Gayer, R. A. (ed.) The Caledonide Geology of Scandinavia. Graham & Trotman, London, 352, 299–312. Moncrieff, A. C. M. 1989. The Tillite Group and related rocks of East Greenland: implications for Late Proterozoic palaeogeography. In: Gayer, R. A. (ed.) The Caledonian Geology of Scandinavia. Graham and Trotman, London, 352, 285–297. Moncrieff, A. C. M. & Hambrey, M. J. 1988. Late Precambrian glacially-related grooved and striated surfaces in the Tillite Group of Central East Greenland. Palaeogeography, Palaeoclimatology, Palaeoecology, 65, 183–200. Moncrieff, A. C. M. & Hambrey, M. J. 1990. Marginal-marine glacial sedimentation in the late Precambrian succession of East Greenland. In: Dowdeswell, J. A. & Scourse, J. D. (eds) Glacimarine Environments: Processes and Sediments. Geological Society, London, Special Publication, 53, 397– 410. Nystuen, J. P., Andresen, A., Kumpulainen, R. A. & Siedlecka, A. 2008. Neoproterozoic basin evolution in Fennoscandia, East Greenland, and Svalbard. Episodes, 31, 35 – 43. Phillips, W. E. A, Stillman, C. J., Friderichsen, J. D. & Jemelin, L. 1973. Preliminary results of mapping in the western gneiss and schist zone around Vestfjord and inner Ga˚sefjord, south-west Scoresby Sund. Rapport Grønlands Geologiske Undersøgelse, 58, 17– 32. Porter, S. M. & Knoll, A. H. 2000. Testate amoebae in the Neoproterozoic era: evidence from vase-shaped microfossils in the Chuar Group, Grand Canyon. Paleobiology, 26, 360–385.

592

S. STOUGE ET AL.

Porter, S. M., Meisterfeld, R. & Knoll, A. H. 2003. Vase-shaped microfossils from the Neoproterozoic Chuar Group, Grand Canyon: a classification guided by modern testate amoebae. Journal of Paleontology, 77, 409–429. Poulsen, C. 1930. Contributions to the stratigraphy of the Cambro – Ordovician of East Greenland. Meddelelser om Grønland, 74, 297–316. Robb, L. J., Knoll, A. H., Plumb, K. A., Shields, G. A., Strauss, H. & Veizer, J. 2004. The Precambrian: the Archaean and Proterozoic eons. In: Gradstein, F. M., Ogg, J. G. & Smith, A. G. (eds) A Geologic Time Scale 2004. Cambridge University Press, Cambridge, 129– 140. Schaub, H. P. 1950. On the pre-Cambrian sedimentation in NEGreenland. Meddelelser om Grønland, 114, 50. Schaub, H. P. 1955. Tectonics and morphology of Kap Oswald (NE-Greenland). Meddelelser om Grønland, 103, 1– 33. Smith, M. P. & Robertson, S. 1999. Vendian – Lower Palaeozoic stratigraphy of the parautochthon in the Ma˚lebjerg and Eleonore Sø windows, East Greenland Caledonides. Danmarks og Grønlands Geologiske Undersøgelse Rapport, 1999/19, 169– 182. Smith, M. P., Rasmussen, J. A., Robertson, S., Higgins, A. K & Leslie, A. G. 2004. Lower Palaeozoic stratigraphy of the East Greenland Caledonides. Geological Survey of Denmark and Greenland Bulletin, 6, 5 –28. Sommer, M. 1957. Geologie von Lyells Land (NE-Gro¨nland). Meddelelser om Grønland, 155, 157. Soper, N. J. 1994. Neoproterozoic sedimentation on the NE margin of Laurentia and the opening of Iapetus. Geological Magazine, 131, 291– 299. Stouge, S., Boyce, W. D., Christiansen, J. L., Harper, D. A. T. & Knight, I. 2001. Vendian – Lower Ordovician stratigraphy of Ella Ø, North-East Greenland: new investigations. Geology of Greenland Survey Bulletin, 189, 107–114. Stouge, S., Boyce, W. D., Christiansen, J. L., Harper, D. A. T. & Knight, I. 2002. Lower –Middle Ordovician stratigraphy of NorthEast Greenland. Geology of Greenland Survey Bulletin, 191, 117– 125. Sønderholm, M. & Tirsgaard, H. 1993. Lithostratigraphic framework of the Upper Proterozoic Eleonore Bay Supergroup of East and

North-East Greenland. Bulletin Grønlands Geologiske Undersøgelse, 167, 38 Sønderholm, M., Frederiksen, K. S., Smith, M. P. & Tirsgaard, H. 2008. Neoproterozoic sedimentary basins with glacigenic deposits of the East Greenland Caledonides. In: Higgins, A. K., Gilotti, J. A. & Smith, M. P. (eds) The Greenland Caledonides: Evolution of the Northeast Margin of Laurentia. Memoir, Geological Society of America, 202, 99– 136. Swett, K. & Knoll, A. H. 1989. Marine pisolites from Upper Proterozoic carbonates of East Greenland and Spitsbergen. Sedimentology, 36, 75 – 93. Teichert, C. 1933. Untersuchungen zum Bau des kaledonischen Gebirges in Ostgro¨nland. Meddelelser om Grønland, 95, 121. Tirsgaard, H. 1993. The architecture of Precambrian high energy tidal channels: an example from the Lyell Land Group (Eleonore Bay Supergroup), East Greenland. Sedimentary Geology, 88, 137–152. Tirsgaard, H. 1996. Cyclic sedimentation of carbonate and siliciclastic deposits on a late Precambrian ramp: the Elisabeth Bjerg Formation (Eleonore Bay Supergroup), East Greenland. Journal of Sedimentary Research, 66B, 699– 712. Tirsgaard, H. & Sønderholm, M. 1997. Lithostratigraphy, sedimentary evolution and sequence stratigraphy of the Upper Proterozoic Lyell Land Group (Eleonore Bay Supergroup) of East and North-East Greenland. Geology of Greenland Survey Bulletin, 178, 60. Vidal, G. 1976. Late Precambrian acritarchs from the Eleonore Bay Group and Tillite Group in East Greenland. Grønlands Geologiske Undersøgelse Rapport, 78, 19. Vidal, G. 1979. Acritarchs from the Upper Proterozoic and Lower Cambrian of East Greenland. Grønlands Geologiske Undersøgelse Rapport, 134, 40. Vidal, G. 1985. Biostratigraphic correlation of the Upper Proterozoic and Lower Cambrian of the Fennoscandian Shield and the Caledonides of East Greenland and Svalbard. In: Gee, D. G. & Sturt, B. A. (eds) The Caledonide Orogen – Scandinavia and Related Areas. John Wiley and Sons Ltd., London, 331– 338. Wenk, E. 1961. On the crystalline basement and the basal part of the preCambrian Eleonore Bay Group in the southwestern part of Scoresby Sund. Meddelelser om Grønland, 168, 54.

Chapter 57 Glaciogenic rocks of the Neoproterozoic Smalfjord and Mortensnes formations, Vestertana Group, E. Finnmark, Norway A. HUGH N. RICE1 *, MARC B. EDWARDS2, TOR A. HANSEN3, EMMANUELLE ARNAUD4 & GALEN P. HALVERSON5,6 1

Vienna University, Structural Processes Group, Department of Geodynamics and Sedimentology, Althanstrasse 14, 1090 Vienna, Austria, Europe 2

5430 Dumfries Drive, Houston, Texas, USA

3

Talisman Energy Norge AS, Verven 4, PO Box 649, Stavanger, Norway

4

Department of Land Resource Science, University of Guelph, Guelph, Ontario, N1G 2W1, Canada

5

Department of Geology & Geophysics, School of Earth & Environmental Sciences, University of Adelaide, North Terrace, Adelaide, SA 5005, Australia

6

Present address: Department of Earth and Planetary Sciences, McGill University, 3450 University Street, Montreal, Quebec, H3A 2A7, Canada *Corresponding author (e-mail: [email protected]) Abstract: The Vestertana Group in East Finnmark, North Norway, contains two Neoproterozoic glaciogenic sequences, the Smalfjord and Mortensnes formations, preserved on the northern edge of Baltica. The former comprises up to 420 m of aeolian, fluvioglacial and glaciomarine sediments and terrestrial diamictite. The latter consists of up to 50 m of predominantly diamictite. The Smalfjord Formation (Fm.) is underlain by dolostones (Grasdalen Fm., Tanafjorden Group), only locally preserved due to the sub-Smalfjord Fm. unconformity, which cuts down-section through a c. 2.5 km dominantly clastic sequence to rest on Baltic Shield gneisses. The two glaciogenic successions are separated by c. 350 m of mostly clastic sediments (Nyborg Fm.), with thin dolostones at the base and towards the top. The latter are generally absent due to the sub-Mortensnes Fm. unconformity, which also cuts down southwards through the Nyborg and Smalfjord formations to the Baltic Shield. No robust isotopic age constraints are available for the succession. d13C data, together with cap dolostone characteristics, offer paradigmic correlations with other areas (Smalfjord ; Marinoan; Mortensnes ; Gaskiers). A limited Ediacaran fauna, including Aspidella, give only broad age constraints. Palaeomagnetic data are ambiguous; some suggest Baltica lay at equatorial (158S) to mid-latitudes (508S) for the period 750– 550 Ma, respectively, while other interpretations place it at either 308N or S at c. 550 Ma.

This chapter is concerned with the glaciogenic sediments of the Smalfjord and Mortensnes formations (Lower and Upper Tillites, respectively, in older literature), lying at the base of the Vestertana Group, within a c. 4.7-km-thick, predominantly clastic sequence of Tonian/Cryogenian to Tremadocian age. The former glaciogenic unit only occurs in three areas, all in E Finnmark; Laksefjordvidda, Tanafjord and Varangerfjord (Fig. 57.1). In contrast, the latter unit has tentatively been correlated with many other diamictite-bearing units in the Scandinavian Caledonides (Kumpulainen 2011; Kumpulianen & Greiling 2011; Nystuen & Lamminen 2011; Stodt et al. 2011). Owing to significant facies variations, no single outcrop/ area can be taken as a type locality/region. The rocks have been studied by many authors (see references in Føyn 1937; Bjørlykke 1967; Føyn & Siedlecki 1980; Laajoki 2002), but detailed sedimentological studies remain few (Reading & Walker 1966; Edwards 1972, 1975, 1979, 1984; Hansen 1992; Arnaud & Eyles 2002; Baarli et al. 2006; Levine et al. 2006; Arnaud 2008). An overview of d13C data from associated dolostones was given by Halverson et al. (2005). The Smalfjord Fm. is significant in having provided some of the first evidence for a glacial event specifically attributed to a prePleistocene age, at Oaibacˇcˇanjar’ga (formerly Bigganjar’ga; Reusch 1891). Interpretation of this outcrop has remained a bone of contention since its discovery; recent results indicate that intense high-temperature brittle deformation occurred during rapid glacial movement (Bestmann et al. 2006). The two glaciogenic sequences became an informal ‘type sequence’ for late Precambrian glaciations in the North Atlantic region (e.g. Hambrey 1983), known as the Varang(er)ian glaciation (Varangeristiden; Kulling 1951), later developed into

the Varanger Epoch (cf Harland et al. 1989). Use of the term Varang(er)ian is here discouraged for reasons given below.

Structural framework During the Neoproterozoic in northern Scandinavia, the NNE – SSW-trending western Baltica continental shelf, now preserved in the Scandinavian Caledonides, was joined to the WNW –ESEtrending Timan Basin in NE Baltica (Siedlecka 1985; Gayer & Rice 1989; Siedlecka et al. 2004). Both of these margins are thought to have experienced extensional rift tectonics from c. 1000 to c. 630 Ma, associated with the break-up of Rodinia, followed by relatively quiescent passive margin conditions during the deposition of the two glaciogenic units (Røe 2003; Siedlecka et al. 2004). The Timan Basin has been divided into northerly basinal and southerly shelf regions (Siedlecka et al. 1995), with the WNW – ESE-trending dextral strike – slip Trollfjorden – Komagelva Fault (cf Rice et al. 1989) forming the boundary between these two regions (Fig. 57.1). The Smalfjord Fm. was deposited (or has only been preserved in; but see Kumpulainen 2011) the eastern part of the Gaissa Basin, which palaeogeographically formed a linking area between the Timanian shelf region and the western Baltica continental shelf. In contrast, the Mortensnes Fm. and its presumed correlatives were deposited and preserved over much of western Baltica. The rocks were shortened predominantly during SilurianDevonian Caledonian orogenesis (Scandian event). Deformation was essentially in-sequence, with c. 50% shortening in the external imbricate zone (Gaissa Thrust Belt; imbricated Gaissa Basin

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 593– 602. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.57

594

A. H. N. RICE ET AL.

Fig. 57.1. Map of NE Scandinavian Caledonides showing the main localities and distribution of glaciogenic lithologies. A, Alduskai’di; An, Andabakoai’vi; ˇ ikkoja˚kka; D, Digermul; G, Gaessenjar’ga; Gr, Grasdal; Gu, Gulgofjord; H, Handelsneset; L, Lappaluokoai’vi; Lp, Leirpollen; Lv, Laksefjordvidda; M, Mortensnes; C, C Mi, Miel’keja˚kka; Mn, Manjunnas; N, Nyborg; Ne, Nesseby; Nj, Njukcˇagai’sa; O, Oaibacˇcˇannjar’ga; R, Ruos’soai’vi; S, Selesnjar’ga; Sk, Skja˚holm; Sm, Smalfjord; St, Stappugied’di; U, Uccaskai’di; V, Vestertana; Vb, Vaððasbak’te; Vr, Vieranjar’ga.

sediments) west of Laksefjordvidda (Townsend et al. 1986). East of Laksefjordvidda, shortening decreased to c. 15% (Chapman et al. 1985) and dies out very gradually across Varangerhalvøya. In the Digermul –Tanafjord region, large-scale folds developed, sometimes with a back-thrusting vergence (Reading 1965; Chapman et al. 1985; Siedlecka 1987). These likely formed by buttressing, as the sole thrust attempted to footwall-shortcut through NNE –SSW-trending extensional structures. Although the strain decreases to the east, most pelitic rocks have a cleavage, locally with crenulations of sedimentary laminae in flexural-slip fold hinges, while diamictites may carry a spaced (approximately centimetre scale) anastamosing cleavage. Pressuresolution was common in carbonate rocks, including dolomite-rich diamictites. However, in sandier lithologies, finite strains are generally low, with sedimentary structures well preserved.

Stratigraphy The glaciogenic sequences in E. Finnmark lie at the base of the Vestertana Group, within the c. 4.7-km-thick, predominantly shallow marine and siliciclastic, Tonian (inferred) to Tremadocian East Finnmark Supergroup (Fig. 57.2; Føyn 1937; Johnson et al. 1978; Siedlecka et al. 2004). The sub-Smalfjord Fm. unconformity at the base of the Vestertana Group cuts down-section to the south until it lies on the Baltic Shield around Varangerfjord (Fig. 57.1). Between the Tana River west of Varangerfjord and Andabakoaivi (Fig. 57.1), the Vestertana Group forms the most southerly Caledonian Neoproterozoic rocks, although the basement –cover contact is not exposed. The Smalfjord Fm. has been preserved in two major palaeovalleys (Varanger and Krokvatn palaeovalleys; Bjørlykke 1967; Føyn

& Siedlecki 1980) and in the intervening Tanafjord area. The maximum preserved thickness occurs in the Krokvatn palaeovalley (c. 420 m), but facies variations preclude formation-wide member names. The Mortensnes Fm. (and equivalents; see Kumpulainen 2011; Kumpulainen & Greiling 2011; Nystuen & Lamminen 2011; Stodt et al. 2011), which is more widespread across the orogen, often lying unconformably on autochthonous or allochthonous basement, is ,50 m thick. The two glaciogenic units are separated by c. 350 m of clastic and minor carbonate sediments of the Nyborg Fm. (Edwards 1984). The uppermost sediments of the Tanafjorden Group that underlie the Smalfjord Fm. are the mixed carbonate and clastic sediments of the Grasdalen and Fadnuvag’gi formations. The Mortensnes Fm. is overlain by mudstones and sandstones of the Stappogiedde Fm. (Lillevannet Member). The Precambrian –Cambrian boundary lies near the base of the overlying Breidvika Fm. (Føyn & Glaessner 1979; Farmer et al. 1992; Crimes & McIlroy 1999).

Glaciogenic deposits and associated strata Grasdalen Fm., Tanafjord The Grasdalen Fm. (230 m), which only crops out in the GrasdalGulgofjord area, very close to the top of the Tanafjorden Group, was divided into three major parts by Siedlecka & Siedlecki (1971) and Johnson et al. (1978): a lower mixed sand/silt-marldolostone unit (111 m), a massive dolostone (58 m) and an interbedded dolostone-clastic sequence (61 m). In detail, however, the stratigraphy is considerably more complex (Rice, unpublished data; Fig. 57.2).

NEOPROTEROZOIC SMALFJORD AND MORTENSNES FORMATIONS

595

Fig. 57.2. Regional stratigraphic profiles showing position of glaciogenic units (modified from Johnson et al. 1978; Føyn & Siedlecki 1980; Edwards 1984; Rice & Townsend 1996; Halverson et al. 2005). The Lille Molvika Fm. and Ekkerøya group (formerly known as the Ekkerøya Fm., within the Vadsø Group) were informally defined based on regional correlations by Rice & Townsend (1996). UE, LE, position of uppermost and lowermost Ediacaran fauna; Pl, position of Platysolenites antiquissimus (Eichw.). Note different vertical scales for columns (a) and (c) compared to (b).

Fadnuvag’gi Fm., Tanafjord Although the Grasdalen Fm. is usually taken to continue up to the base of the Smalfjord Fm. (Siedlecka & Siedlecki 1971), the interbedded dolostone-clastic sequence is overlain by c. 20 m of a distinctly more fissile, darker, pyritiferous and finely interbanded shale silt/sandstone without carbonate interlayers, informally termed the Fadnuvaggi Fm. (Halverson et al. 2005).

Smalfjord Fm., Krokvatn Palaeovalley, Laksefjordvidda Føyn & Siedlecki (1980) documented three diamictite units with intervening sandstones (Lower, Middle and Upper Krokvatn Diamictites, Lower and Upper Krokvatn Sandstones; Fig. 57.2)

deposited within an essentially north –south-oriented palaeovalley cut into the Tanafjorden Group on Laksefjordvidda (Fig. 57.1). Figure 57.3a shows a semi-schematic profile across the northern part of the palaeovalley (Uccaskai’di-Vaððasbak’te); further south (Cˇikkojokka; Fig 57.1), the sequence comprises only the middle diamictite and upper sandstone (Føyn & Siedlecki 1980). The Tanafjorden Group dips to the NW compared to the 420-m-thick palaeovalley infill, with 600 m of tilted stratigraphy eroded. The Lower and Middle Krovatn Diamictites are both ,25 m thick, with a grey to reddish colour and a sandy matrix. The former includes a 2-m-thick sandstone, whereas the latter has interlayered sandstones in the basal part and has faint banding and rare dropstones (Føyn & Siedlecki 1980). Clasts, sometimes faceted, are predominantly from the Tanafjorden Group, with rare

596

A. H. N. RICE ET AL.

Fig. 57.3. (a) Schematic profile from Uccaskai’di to Vaððasbak’te through the glacial successions in the Krokvatn Palaeovalley (modified from Føyn & Siedlecki 1980). LKD, MKD, UKD, Lower, Middle and Upper Krokvatn Diamictites; (b) Schematic profile from Varangerfjord to Gulgofjord (modified from Edwards & Føyn 1981). Nyborg facies: N2, fan channel; N3, N4, fine and coarse submarine fans, respectively; N5, tidal distributary; N6, bay/lagoon; N7, offshore barrier. lsm, usm, lower and upper sub-members of the Lillevannet Member.

kilometre-sized dolostone blocks at the base (Edwards et al. 1973) and some basement material. The upper diamictite (,100 m thick) is more variable, with dolomitic and sandy matrices and a buff or grey colour. These are interbedded with sandstones and siltstones with a few lonestones. On Uccaskai’di, the Upper Krokvatn Diamictite has abundant dolostone clasts, possibly reflecting the nearby presence of Tanafjorden Group dolostones (Fig. 57.1; Føyn & Siedlecki 1980). In contrast, in the northeastern part of the valley, the Nyborg Fm. rests unconformably on the Lower Krokvatn Diamictite and Lower Sandstone, the higher units having been removed either during deposition of the third diamictite elsewhere, or afterwards. The massive, homogeneous very fine-grained sandstones and very coarse-grained siltstones are essentially devoid of sedimentary structures throughout the succession; ripples occur at only one locality. Beds are from 1 m to several metres thick, without sorting of the angular grains, although larger grains (.0.2 mm) are rounded. Hand-sized basement lonestones occur at two localities. Directly above the Middle Krokvatn Diamictite, the sediments are red (Føyn & Siedlecki 1980).

Smalfjord Fm., Varanger Palaeovalley, Varangerfjord The Smalfjord Fm. in the Varangerfjord area (Bjørlykke 1967; Edwards 1975, 1984; Arnaud & Eyles 2002; Arnaud 2008; Laajoki 2001, 2002; Baarli et al. 2006) has been correlated with the lower to middle part of the Krokvatn succession (Føyn & Siedlecki 1980). The following descriptions are based on

outcrops at Vieranjar’ga, Nesseby, Handelsneset, Skja˚holm, Oaibacˇcˇanjar’ga and Selesnajar’ga (Edwards 1975, 1984; Arnaud & Eyles 2002; Baarli et al. 2006; Arnaud 2008). Two diamictite and four sandstone and conglomerate (S1– S4) facies have been identified in this area. The first type of diamictite is a massive, matrix-supported, pink diamictite with a mediumto coarse-grained matrix with only a few granitic and sandstone clasts. These are up to 20 cm across, sub-rounded to sub-angular, and without striae or faceting. This unit, in beds up to 1 m thick, can be traced over several hundred metres. The second diamictite, which is also mostly matrix-supported and massive, with a distinctly grey colour in outcrop, consists of a muddy, medium grained sandstone matrix and abundant granite and sandstone clasts, ,50 cm across, some with faceting. Bjørlykke (1967) also noted striated clasts in grey diamictite. The S1 facies comprises well-stratified pebble conglomerates, occasionally imbricated, sandstones showing cross-bedding and parallel lamination, with fining-upwards trends between these lithologies and rare mud-drapes. S2 is typified by large-scale inclined bedding, 5–10 m thick and continuous for up to several kilometres. These include steeply dipping poorly sorted pebble conglomerates to well-sorted gently dipping parallel-laminated sandstones. S3 comprises medium-bedded parallel-sided to slightly lenticular sandstones, usually internally massive, but at times parallel-laminated or rippled. Mudstone partings are thin or absent but soft-sediment deformation, with convolute bedding, recumbent folds and faulting occur. Steep-sided isolated channels filled with poorly sorted conglomerate are also present. S4 comprises interbeddded sandstones and mudstones, the former in thin to

NEOPROTEROZOIC SMALFJORD AND MORTENSNES FORMATIONS

597

medium, laterally continuous beds. These display many turbiditic features (erosive surfaces, sole marks, grading, convolute bedding). A 3-m-thick, crudely stratified breccia within S4 contains angular sandstone clasts ,40 cm across, sometimes graded, in a coarsegrained sandstone matrix. The stratigraphic and lateral distribution of these facies vary from place to place. Laterally, lithofacies can often be traced for several tens of metres, although abrupt facies changes are common (cf Edwards 1984). Complex soft-sediment deformation in sandstone and conglomerate facies is particularly notable at the base of the Smalfjord Fm. at Handelsneset (Arnaud 2008). Overall, the diamictite, breccia and S4 facies tend to occur near the base of the sequence, whereas the S1 facies occurs near or at the top.

Edwards (1984), but ordered them differently, without any depositional cyclicity. In the Grasdal area, Edwards (1979) described a 3 –6-m-thick dark grey siltstone with an unusually high silt and low clay content; this may also occur on the southern side of Leirpollen (Fig. 57.1). Clasts consist of angular quartz and feldspar grains of very fine sand to coarse silt size, forming c. 77% of the rock, with a statistically non-random preferred north –south to NE– SW orientation. This unit has sharp contacts with the over- and underlying units (the latter is a very dark diamictite with abundant dolomitic clasts) and an internal stratification defined by isolated, typically dolomitic, clasts (,7.5 mm sized) in discrete layers c. 30 cm apart, together with very faint grey colour variations.

Smalfjord Fm., Tanafjord area

Nyborg Fm., E. Finnmark (0– 350 m)

The Smalfjord Fm. in this region (Reading & Walker 1966; Edwards 1984; Hansen 1992; Arnaud & Eyles 2002) has been correlated with the Upper Krokvatn Diamictite (Laksefjordvidda region; Føyn & Siedlecki 1980; Edwards 1984). The best outcrops in this area are those exposed at Gaessenjar’ga and Luovtat (Fig. 57.1). Based on outcrops in the Vestertana to Smalfjord area, with additional important outcrops between Njukcˇagai’sa and east of the Tana River, Edwards (1984) recognized a fivefold (A –E) repetitive cycle of erosion surface, diamictite and laminated mudstone, occasionally with intervening sandstones. Some units can be traced over hundreds of square kilometres. Diamictite units are from 2–40 m thick, structureless or stratified, and have erosive bases, at both outcrop and regional scales. Diamictites are all matrix-supported, with clasts up to 1 m in diameter (most are ,35 cm) in a siltstone to sandstone matrix. Material from underlying beds was often incorporated into the overlying diamictite, forming inclusions of diamictite, conglomerate, sandstone or mudstone, folds and faults, load-casts, boudinage, shear bands and flame structures. Some diamictite units have gradational contacts (Hansen 1992; Arnaud & Eyles 2002). Stratified sandstone/conglomerate bodies are present within the diamictite; these deposits resemble the facies in the Varangerfjord area (see above). In some cases, the basal diamictite represents mixing of far-travelled and locally derived material. The dominant clast and matrix source areas determine the diamictite lithology; buff and brown colours indicate a dolostone source (cycles B, C and E). Cycle A is purple, with a hematitic matrix, and was derived from ferruginous sandstones of the Dakkovarre Fm. (Fig. 57.2). Diamictites are also often distinguished by variable clast abundance. Some diamictite units exhibit normal grading with a sandier matrix at the base and muddier matrix at the top (Hansen 1992; Arnaud & Eyles 2002). The Smalfjord Fm. in the Tanafjord area has a much higher proportion of diamictite and is laterally much more consistent compared to outcrops and lithofacies distributions in the Varangerfjord area; in the Gaessenjar’ ga area, over 85% of the sections consist of diamictite units and these tend to be traceable over several kilometres (Edwards 1984; Arnaud & Eyles 2002). One exception is that of a 1–5-m-thick purple diamictite that grades into a laminated mudstone with outsized clasts within several hundred metres of lateral exposure at Gaessenjar’ga (Arnaud & Eyles 2002). Sandstone beds are massive, graded (normal or reverse), rippled and cross-bedded (trough and planar). Laminated mudstones, 0.3– 10 m thick, vary in colour, lamination prominence, whether this is random or rhythmic, and size, abundance and composition of lonestones. The latter are predominantly dolostones, ,1–30 cm in size and are either dispersed throughout the lithology or lie in discrete layers. Some laminated mudstone units are interbedded with sandstone or diamictite. In the Gaessenjar’ga area, Hansen (1992) and Arnaud & Eyles (2002) recognized essentially the same lithologies and units as

The Smalfjord and Mortensnes formations are separated by the Nyborg Fm. The lowest part of this unit (Member A) comprises lateral and vertical gradations between massive buff-weathering dolomicrites with sheet cracks and pseudo-tepees, through massive to thinly bedded dolomicrites to interbeddded red sandstones/shales and dolomicrites, locally reworked as edgewise breccias, to dolomite-cemented orange-red sandstones, as well as white-weathering green-grey sandstones and red and dark grey shales (Edwards 1984; Reading & Walker 1966; Rice & Hofmann 2001). Such sequences are usually only few tens of metres thick (Reading & Walker 1966; Edwards 1984) although the succession is c. 160 m thick at Alduskaidi (Fig. 57.1; Føyn & Siedlecki 1980). Generally, successions become more clastic upwards. On Ruos’soai’vi, west of Varangerfjord, 10-mm sulphate crystal fans (now pseudomorphed by quartz) grew on the basement, preserved within a 12-mm-thick biotite extraclast recrystallized dolostone overlain by white-weathering sandstones. Similar white-weathering sandstones, likely cannibalized from the Smalfjord Fm., also directly overlie very irregular basement surfaces on Ruos’soai’vi. White-weathering green-grey sandstones (18.5 cm thick) also unconformably overlie c. 14 m of massive sheetcracked dolomicrite in the Miel’keja˚kka area, SW of Tanafjord (Fig. 57.1). Members B –D of the Nyborg Fm. comprise, respectively, interbedded reddish-purple shales/sandstones (c. 200 m), interbedded grey-green shales/sandstones (c. 150–200 m) and purple sandstone/grey-green shales (c. 70 m; Edwards 1984). Member E (c. 25 m thick), exposed only in the Gulgofjord-Grasdal area (Fig. 57.1), consists of white to grey sandstones with two thin (c. 1 m thick) buff-weathering siliclastic-extraclast dolomicrites at the base (Edwards 1984).

Mortensnes Fm., E. Finnmark (,50 m) Edwards (1984) recognized three members in this formation, the lowest (,30 m) being a northwards-thinning wedge of predominantly grey-green to purple massive diamictite (depending on the substrate colour) of highly variable clast/matrix composition, with basement clasts and minor intrabasinal clasts from the underlying substrate. This dies out around the southern end of Vestertana. Brecciation of the substrate has been found below some diamictites, with substrate blocks up to 20 m long and 1 m thick incorporated into the overlying diamictite. Lenses and bands of extra-basinal dark diamictite occur within more mixed lithologies. Basic clasts may show facets and striations (Banks et al. 1971), while clast sizes decrease overall to the north (Edwards 1984). The middle member has a gradational contact to the underlying rocks and is distinguished from it by its buff-brown weathering, dolomitic composition of matrix and clasts with subordinate chert. This member rapidly increases in thickness from ,4 m in the south to .10 m, approximately coincident with the southern end of Vestertana, giving two sub-members. The thin sub-member

598

A. H. N. RICE ET AL.

comprises 2– 4 m of stratified diamictite with primary and softsediment deformation structures. The thick sub-member comprises five lithofacies; a blanket of massive purple to grey-green deformation diamictite, a zone of large tabular blocks (of the Nyborg Fm., diamictite and white sandstone), a relatively rare stratified dolomitic diamictite, a prominent buff-brown diamictite with a sandy matrix, which thins from 20 to 8 m from south to north, and, at the top, a bedded buff brown diamictite. The unconformably overlying upper member (,40 m) consists of dark grey massive diamictite overlain north of Stappugied’di and Leirpollen by a 40-cm-thick dolomitic matrix diamictite with small dolostone and occasional large basement clasts. A 20–30-cm-thick polymict conglomerate overlies the formation over a large area (Edwards 1984).

Lillevannet Member, Stappogiedde Fm. (40 – 110 m) The Mortensnes Fm. is overlain by the Lillevannet Member of the Stappogiedde Fm. (Reading & Walker 1966), which thins from south to north, approximately coincident with the southern end of Vestertana (Edwards 1984). The lower sub-member (3 –55 m) comprises grey, parallel laminated mudstones, silty to sandy in the north, grading upwards into a siltstone with some ripple crosslamination and with fine- to medium-grained lenticular sandstones (Edwards 1984). The upper sub-member is a complex assemblage of sandstone and shale facies, including coarse arkosic sandstones, poorly cross-bedded and granule conglomerates, medium-grained sub-arkosic sandstone, relatively well sorted and rounded sandstones, thin to medium bedded fine to very fine lenticular, erosivebased sandstones, dark grey, brown weathering rippled and finely laminated micaceous silty-sandy mudstone, sometimes in coarsening-upwards cycles and finely parallel laminated grey mudstones (Edwards 1984).

Boundary relations with overlying and underlying non-glacial units

Chemostratigraphy d13C data from E. Finnmark were briefly reviewed by Halverson et al. (2005); although more data are now available the results are essentially the same as those previously reported. All data are standardized to VPDB; see Halverson et al. (2005) for analytical methods. The Grasdalen Fm. has variable values, passing from –2 to –3‰ at the base, gradually rising to þ6‰ and then falling very rapidly to – 3‰; the two stratigraphically highest samples record values of þ3 to þ4‰. Ten profiles through the dolostone (or interlayered dolostones– shales/siltstones/sandstones) sequences lying directly above the Smalfjord Fm. have been analysed over a wide geographical area and include the c. 160-m-thick dolostone at Alduskaidi (Fig. 57.1; Føyn & Siedlecki 1980). Although few profiles expose both the bottom and top of the dolomite unaffected by either tectonic or penecontemporaneous erosional processes, all samples show d13C values between –1.0 and – 6.0‰ (VPDB). Most have essentially constant d13C values or show either a slight increase or decrease in concentration. Only the 14-m-thick profile at Miel’keja˚kka (Fig. 57.1) shows a major variation, passing from – 2.55‰ at the base to –5.92‰ at the erosive upper contact. These data show a poor concave upwards distribution, comparable to the uppermost part of the slope-bank d13C profile documented by Hoffman et al. (2007). Thin dolostones in Member E of the Nyborg Fm., lying c. 20 m below the sub-Mortensnes unconformity in the Gulgofjord area (Edwards 1984; Figs 57.1 & 57.2) gave d13C values between –7.6 and – 9.9‰ (VPDB). Significantly, the thin carbonate-matrix diamictite forming the uppermost part of the Mortensnes Fm. on the NW side of Tanafjord (Edwards 1984) also gave extreme negative values (down to – 10.44‰ VPDB), taken to reflect erosion of the underlying dolostones of Member E in the Nyborg Fm. (Rice & Halverson, unpublished data).

Palaeolatitude and palaeogeography Southwards, the Smalfjord Fm. cuts down through up to 2.5 km of the Tanafjorden, Ekkerøya and Vadsø Groups, to lie on the basement around Varangerfjord. Minor outliers of Neoproterozoic sediments overlie the basement, and minor inliers of basement occur within the Smalfjord Fm., together preserving an irregular unconformity, locally palaeo-frost-shattered (Bjørlykke 1967; Siedlecka 1990; Rice & Hofmann 2000; Rice et al. 2001; Laajokki 2001, 2002; Edwards 1984, 1975). The base of the Smalfjord Fm. is characterized by rare east –west to WNW –ESE oriented subglacial striations; the best developed occur at Oaibacˇcˇanjar’ga (Bigganjar’ga) in Varangerfjord (Reusch 1891; Strahan 1897), although others have been reported (Bjørlykke 1967; Rice & Hoffman 2000; Laajoki 2002). Rice & Hofmann (2000) and Bestmann et al. (2006) documented a thin glacially formed breccia within and as ridges beside the striations at Oaibacˇcˇanjar’ga. Regionally, the upper contact of the Smalfjord Fm. with the overlying Nyborg Fm. is also an unconformity, because the latter rests on different parts of the former in different areas; specifically, the upper part of the Smalfjord Fm. in the Vaððasbak’te area of the Krokvatn Palaeovalley and in the Varanger palaeovalley are absent (Fig. 57.1; Føyn & Siedlecki 1980; Edwards 1984). In the Ruos’soai’vi-Lap’paluokoai’vi area (Fig. 57.1), the Nyborg Fm. rests directly on the basement (Siedlecka 1990; Rice unpublished data). The base of the Mortensnes Fm. is an invariably planar surface at outcrop-scale that cuts down-section towards the south at a very low angle. The contact often exhibits brecciation or homogenization of the underlying sediments (deformation diamictite; Edwards 1984). At its upper contact, the Mortensnes Fm. is overlain with a sharp or rapid transition by the Lillevannet Member (Fig. 57.2; Reading & Walker 1966; Edwards 1984).

Palaeomagnetic analyses on the Nyborg Fm. have yielded palaeolatitudes for this area of 338S to 418S (Torsvik et al. 1995), although no robust geochronological constraint is associated with this palaeolatitude. Samples were subjected to thermal demagnetization and some also to alternating field demagnetization. High-temperature components from two field areas (55 samples) were identified with a positive fold test at 95% significance level (k ¼ 4.9 in situ and 14.6 100% unfolded; a95 ¼ 33.8 in situ and 18.1 100% unfolded, Torsvik et al. 1995). The palaeogeography of Baltica as a whole has been constrained from other sites in Baltica, although well-dated palaeopoles remain few (see reviews in Bingen et al. 2005, Cocks & Torsvik 2005). At c. 750 Ma, a southwards facing Finnmark-Kola region at c. 158S formed the southern margin of Baltica (Hartz & Torsvik 2002). This region lay adjacent to a roughly east –west-trending rift (Timanian margin) that to the west linked with the north –south trending rift/spreading axis developing between Baltica and Laurentia (Greenland). At 616 Ma, Baltica is thought to have been at polar latitudes (758S) based on the well-dated Egersund dolerite dykes in southern Norway (Bingen et al. 2005). By c. 550 Ma, E. Finnmark was lying at c. 508S (Cocks & Torsvik 2005). In contrast, Cawood & Pisarevsky (2006) and Pisarevsky et al. (2008) place Baltica in a more equatorial position (c. 308) although whether it lay in the northern or southern hemisphere is unclear.

Other characteristics No economic deposits have been reported in these rocks. Acritarchs have been documented from the Vadsø, Ekkerøya,

NEOPROTEROZOIC SMALFJORD AND MORTENSNES FORMATIONS

Tanafjorden and Vestertana Groups, although only reworked acritarchs and vase-shaped fossils have so far been found in the Smalfjord and Mortensnes formations (Vidal 1981; Vidal & Siedlecka 1983). A limited Ediacaran fauna (Farmer et al. 1992; Crimes & McIlroy 1999), dominated by discoidal forms (cf Gehling et al. 2000) and including Aspidella (Narbonne, pers. comm. 2008), has been documented, the oldest forms occurring in the Innerelv Member (Fig 57.2).

Geochronological constraints No robust and high-resolution geochronological constraints are available. Rb –Sr dating of shales associated with the glaciogenic strata provides broad constraints for the deposition of the Smalfjord and Mortensnes formations (630 – 560 Ma; Pringle 1973; Gorohkov et al. 2001).

Discussion Føyn & Siedlecki (1980) interpreted the Lower Krokvatn Diamictite as a massive indurated ground moraine and the Middle and Upper Krokvatn diamictites as partly ground moraine and partly deposited under water. Aeolian or fluvial processes deposited the intervening sandstones into quiet lakes or marine basins in the palaeovalley, while the red sediments above the middle diamictite are likely loessites. Essentially, Føyn & Siedlecki (1980) interpreted the sequence as alternating glacial stadial and interstadial events. However, Eyles (1993) argued that this model was presumptive, with the interpretation that the rocks are glacial in origin being based on their stratigraphic position rather than any unequivocal diagnostic glacial criteria, and thus the alternations were simply presumed to represent cyclical glacial advances and retreats, without substantive proof. Bjørlykke (1967) and Edwards (1984) interpreted the Varangerfjord succession as an infill of a glacially scoured valley, the ice moving towards the NW (Laajoki 2003; Baarli et al. 2006). Irregular glacier retreat left ice-cored moraines and basement highs that were submerged by rising water levels (Levine et al. 2006). Bathymetric lows were filled first by sediment gravity flows and overlain subsequently by rapidly prograding deltas and sandur plains. Later diamictites may represent glacial advances and/or sediment slumping. At Oaibacˇcˇanjar’ga, Edwards (1975) suggested that the diamictite overlying the striations formed as a meltout diamictite from sediment-laden dead-ice, with the margins slumping and being eroded as the ice-entrained sediment lost cohesion. In contrast, Schermerhorn (1974), Jensen & Wulff-Pedersen (1996), Crowell (1999) and Arnaud & Eyles (2002) argued that the diamictite formed during slumping, possibly of glaciogenic deposits, with some suggestions that the striations also formed during slumping as the overriding diamictite was deposited; some of these authors did not realize that the pavement is part of a major, deep regional angular unconformity. The very thin breccia preserved in a number of striations and as ridges beside striations on the pavement has been ascribed to glacio-tectonic brittle deformation processes, preserved by the instantaneous recrystallization of the highly strained comminuted material, associated with flash-heating of the substrate, possibly up to c. 1700 8C, during rapid glacial movement. This was likely associated with glacial earthquakes (Rice & Hofmann 2000; Bestmann et al. 2006). The comminuted material represents proto-rock-flour. In the Tanafjord area, Edwards (1984) interpreted the Smalfjord Fm. (equivalent to the Upper Krokvatn Diamictite of Føyn & Siedlecki 1980; Fig. 57.2) as a cyclic deposit of glacial advances and retreats, with the former yielding a basal massive lodgement tillite, locally with sandstones deposited in situ by sub-glacial meltwater at the ice margin. At the base of cycles, ice movement

599

resulted in the formation of deformation tillites, comprised solely of deformed substrate sediments, while mixing of local and fartravelled material formed banded tillites. Erosional bases of the cycles are recorded at the outcrop-scale by deformation structures and mixing with subjacent material and on a regional scale by the absence of stratigraphy below diamictites. During retreat, poorly sorted sand and silt with pebbles were deposited at the ice margin by tractional underflows and gravity flows. In interglacial periods, finer-grained sediments accumulated, occasionally with dropstones, attesting to a glaciomarine environment. The dark silt-dominated deposit at Grasdal was interpreted to be a loessite (Edwards 1979). Hansen (1992), while agreeing that most of the diamictites were of glacial origin, argued for a shelf depositional environment dominated by undermelt- and flow-tillites, some heavily deformed by an advancing glacier. Rapid vertical and lateral thickness variations and terminations of both tillites and associated facies were interpreted to have been mostly associated with changes in glacial movement directions and local bathymetrical conditions rather than glacial erosion and lodgement. Glaciogenic mass-movement tillites (flow-tillite) laterally grading into sandstones, and at other stratigraphic levels grading into rhythmites, indicates transport directions and the relative location of the basin/continent. In contrast, Arnaud & Eyles (2002) suggested that the diamictites in both the Gaessenjar’ga and the southern and western shores of Varangerfjord accumulated from subaqueous gravity flows from the basin margin, with the latter forming part of a debris apron (see also Crowell 1999; Schermerhorn 1974; Jensen & Wulff-Pedersen 1996). Thus, although the slumped material may have been of glacial origin, seen in rare faceted and striated clasts and lonestones, the diamictites were not directly deposited by ice and so cannot be considered to be tillites. Instead, Arnaud & Eyles (2002) suggested that they record deposition of unstable sediments on the edge of an active extensional basin (but see Edwards 2004; Arnaud & Eyles 2004), in which icebergs or sea-ice contributed dropstones. Ice proximal settings were inferred only for the eastern shore of Skja˚holmen, where sediment gravity flow diamictite is closely associated with well stratified glaciofluvial sandstones and conglomerates and at Handelsneset, where complex deformation observed in conglomerates and sandstones suggests active deformation by overriding ice (Arnaud 2008). The proposed source areas, and thus sediment transport (ice or otherwise) directions for the Smalfjord Fm., also vary. Føyn & Siedlecki (1980) argue that the Middle and probably also the Lower Krokvatn Diamictite were derived from the south. Arnaud & Eyles (2002) also proposed a southerly source area for the probable equivalent rocks at Vieranjarg’a in the Varangerfjord area. Edwards (1984) inferred a westward ice flow along the Varanger palaeovalley, with two dominant striation directions documented: NW –SE and east – west (Bjørlykke 1967; Edwards 1975; Jensen & Pedersen 1996; Rice & Hofmann 2000; Laajoki 2002). Edwards (1984) also proposed that of the five advance/retreat cycles identified in the Tanafjord area, the first was southerly derived, the subsequent three northerly derived and the last of unknown origin. Hansen (1992), in contrast, suggested that the lower part of the succession was NE-derived and the latter part was SE-derived. These differences are not simply due to the different order of the units proposed by the two authors. Member A of the Nyborg Fm. represents post-glacial transgression. Although Edwards (1984) suggested that the dolostones formed around topographically high areas, correlation of the carbonates with Marinoan cap dolostones (see below) suggests that deposition was not directly related to palaeodepth (cf Hoffman et al. 2007). Variations in the clastic input certainly controlled which lithology formed. Subsequently, Members B–D reflect a rapid deepening of the basin (Fig. 57.3), followed by a more gradual shallowing, with Member E representing a barrier lagoon facies, with intermittent carbonate deposition (Edwards 1984). The Mortensnes Fm. represents two advances and retreats of ice (Edwards 1984); the first was initially derived both from the south

600

A. H. N. RICE ET AL.

(lower member) and later from the north (middle member). The source area of the second advance (upper member), which had a major change in provenance, is not known, although the extreme negative d13C values of the dolomitic diamictite at the top of the formation are thought to indicate derivation from the underlying Member E of the Nyborg Fm. For both cycles, lodgement tillite was followed by floating ice, giving finer-grained sedimentation and dropstones. The polymict conglomerate draping the formation has been interpreted as a lag-conglomerate, formed during local isostatic uplift, after glacial retreat (Edwards 1984). The succession passed essentially conformably upwards into the post-glacial Lillevannet Member, representing fluviodeltaic and subsequent marine deposition. Halverson et al. (2005) gave probable correlations of the glaciogenic units in E. Finnmark with those documented elsewhere, based on chemostratigraphy and carbonate lithologies. Dolostones at the base of the Nyborg Fm. have negative d13C values (–1 to –5.9‰), and, where massive, have sheet cracks, pseudo-tepees and, at one locality, sulphate crystal fans. This, in combination with their buff-weathering colour strongly suggests that this is a Marinoan-type cap dolostone (cf Kennedy et al. 1998). d13C data from the 230-m-thick Grasdalen Fm. in the Gulgofjord area (Fig. 57.1) show positive values (6‰), falling rapidly to negative values and then recovering to positive values below the c. 20-m-thick Fadnuvag’gi Fm., which lies directly under the Smalfjord Fm. in the Grasdal area (Fig. 57.1). Halverson et al. (2005) equated this d13C pattern with the Trezona anomaly, but since a broadly similar anomaly occurs under the Port Askaig Fm., which is no longer considered to be a Marinoan equivalent (McCay et al. 2006), this is not a robust constraint. Nevertheless, the characteristics of the cap dolostone secure the Smalfjord Fm. as an equivalent of the 635 Ma (Hoffmann et al. 2004) Marinoan glaciogenic succession. Thin dolostones in the upper part of the Nyborg Fm. (Member E), with d13C values of –7.6 to –9.9‰, are likely correlatives of the Wonoka anomaly, which elsewhere has d13C values as low as –12.81‰ (VPDB: Le Guerroue´ 2006); this implies that the almost immediately overlying Mortensnes Fm. is broadly a correlative of the 580 Ma Gaskiers diamictite (cf Halverson et al. 2005). Kulling (1951) introduced the term Varangeristiden in a discussion of diamictites in northern Sweden. Although those diamictites are now correlated solely with the Mortensnes Fm. (cf. Stodt et al. 2011), the term originally included all the Neoproterozoic glaciogenic rocks in E. Finnmark and implied one glacial period. However, as the Smalfjord and Mortensnes formations are now correlated with two distinct glacial events (Halverson et al. 2005), one of which was worldwide and the other only localized, the concept of a Varangeristiden (Varangian/Varangerian ice age) is no longer valid. Further, taking a 12 Ma duration for the Marinoan glaciation (Bodiselitsch et al. 2005) and 1 Ma for the Gaskiers (Bowring et al. 2003) implies a total time span of 647–579 Ma for the two glacial events, of which only 13 Ma (18%) was actually spent under ice; scarcely enough to justify the term ‘ice age’. For these reasons, use of the term Varang(er)ian ice age is here discouraged. No evidence of an older, Sturtian equivalent, glaciation has been recorded in this area. The Lille Molvika Fm. (Ekkerøya Group; Fig. 57.2) is bounded by two regional unconformities (Rice 1994; Siedlecka 1995), either of which may ‘hide’ a pre-Marinoan glacial event. These unconformities are of post-800 Ma age, based on correlations in the Manjunnas area (Fig. 57.1), where the Ekkerøya Group overlies the Ba˚tsfjord Fm. of the Barents Sea Group (Rice 1994). d13C data from the 1400–1600-m-thick Ba˚tsfjord Fm. in the North Varanger Region (Fig. 57.1) are consistently negative (Rice & Halverson, unpublished data), suggesting a correlation with the c. 800 Ma old Bitter Springs anomaly (cf Halverson et al. 2005). In the North Varanger Region (Fig. 57.1), the Barents Sea (max. 9 km thick) and overlying Løkvikfjell (c. 5.7 km) Groups are

separated by a major angular unconformity (Johnson et al. 1978; Siedlecki & Levell 1978); no glacial deposits occur in these sequences. Biostratigraphic data (Vidal & Siedlecka 1983) are too imprecise to estimate the age of the youngest rocks in the Barents Sea Group (Moczydlowska-Vidal, pers. comm. 2007), although it is younger than 800 Ma (see above). Mafic dykes cutting the unconformity and basal part of the Løkvikfjell Group were intruded at c. 577 + 14 Ma (Rice et al. 2004). Thus evidence for glacial conditions recorded by the Smalfjord and Mortensnes formations may not have been preserved in this basin. A. H. N. R. thanks A. and J. Pettersen, G. Stensvold, P. Sørflaten, Ø. Hauge and the extended Larsen family (Vestertana) for hospitality and boating assistance during fieldwork in E. Finnmark, and C. and R. Hofmann and M. Ebner for assistance in the field. M. B. E. thanks A. and J. Pettersen for their hospitality; H. Reading, S. Føyn, S. Siedlecki, A. Siedlecka, S.-L. Røe, T. Spencer, K. Bjørlykke and G. Boulton for scientific assistance in the field and/or discussions; the New York State Higher Education Assistance Program and family for financial and moral support. Other acknowledgements are given in Edwards (1984). E.A. thanks the Natural Sciences and Engineering Research Council of Canada for funding, and C. Trotter and S. Aspden for assistance in the field. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Arnaud, E. 2008. Deformation in the Neoproterozoic Smalfjord Formation, Northern Norway: an indicator of glacial depositional conditions? Sedimentology, 55, 335– 356. Arnaud, E. & Eyles, C. H. 2002. Glacial influence on Neoproterozoic sedimentation: the Smalfjord Formation, northern Norway. Sedimentology, 49, 765– 788. Arnaud, E. & Eyles, C. H. 2004. Glacial influence on Neoproterozoic sedimentation: the Smalfjord Formation, northern Norway — reply. Sedimentology, 51, 1423– 1430. Baarli, B. G., Levine, R. & Johnson, M. E. 2006. The Late Neoproterozoic Smalfjord Formation of the Varanger peninsula in northern Norway: a shallow fjord deposit. Norwegian Journal of Geology, 86, 133– 150. Banks, N. L., Edwards, M. B., Geddes, W. P., Hobday, D. K. & Reading, H. G. 1971. Late Precambrian and Cambro-Ordovician sedimentation in East Finnmark. Norges geologiske Undersøkelse, 269, 197– 236. Bestmann, M., Rice, A. H. N., Langenhorst, F., Grasemann, B. & Heidelbach, F. 2006. Subglacial bedrock welding associated with glacial earthquakes. Journal of the Geological Society, 163, 417– 420. Bingen, B., Griffin, W. L., Torsvik, T. H. & Saeed, A. 2005. Timing of Late Neoproterozoic glaciation on Baltica constrained by detrital zircon geochronology in the Hedmark Group, south-west Norway. Terra Nova, 17, 250–258. Bjørlykke, K. O. 1967. The Eocambrian ‘Reusch Moraine’ at Bigganjargga and the geology around Varangerfjord, Northern Norway. Norges geologiske Undersøkelse, 251, 18 –44. Bodiselitsch, B., Koeberl, C., Master, S. & Reimold, W. U. 2005. Estimating duration and intensity of Neoproterozoic snowball glaciations from Ir anomalies. Science, 308, 239–242. Bowring, S., Myrow, P., Landing, E., Ramezani, J. & Grotzinger, J. 2003. Geochronological constraints on terminal Neoproterozoic events and the rise of metazoans. Geophysical Research Abstracts, 5, 13219. Cawood, P. A. & Pisarevsk, S. A. 2006. Was Baltica right-way-up or upside-down in the Neoprotereozoic? Journal of the Geological Society, London, 163, 753–759. Chapman, T. J., Gayer, R. A. & Williams, G. D. 1985. Structural crosssections through the Finnmark Caledonides and timing of the Finnmarkian event. In: Gee, D. G. & Sturt, B. A. (eds) The Caledonide Orogen — Scandinavia and Related Areas. J. Wiley & Sons, Chichester, 593–610. Cocks, L. R. M. & Torsvik, T. H. 2005. Baltica from the late Precambrian to mid-Palaeozoic times: the gain and loss of a terranes identity. Earth Science Reviews, 72, 39 –66.

NEOPROTEROZOIC SMALFJORD AND MORTENSNES FORMATIONS

Crimes, T. P. & McIlroy, D. 1999. A biota of Ediacaran aspect from the Lower Cambrian strata on the Digermul Peninsula, Arctic Norway. Geological Magazine, 136, 633– 642. Crowell, J. C. 1999. Pre-Mesozoic Ice-ages. Their bearing on Understanding the Climate System. Geological Society of America Memoir, 192, 106. Edwards, M. B. 1972. Glacial, interglacial and postglacial sedimentation in a late Precambrian shelf environment. DPhil, Oxford University. Available at http://www.marcedwards.com/publications_nnorway. htm. Edwards, M. B. 1975. Glacial retreat sedimentation in the Smalfjord Formation, Late Precambrian, North Norway. Sedimentology, 22, 75 – 94. Edwards, M. B. 1979. Late Precambrian glacial loessites from North Norway and Svalbard. Journal of Sedimentary Petrology, 49, 85– 92. Edwards, M. B. 1984. Sedimentology of the Upper Proterozoic glacial record, Vestertana Group, Finnmark, North Norway. Norges geologiske Undersøkelse Bulletin, 394, 1– 76. Edwards, M. B. 2004. Glacial influence on Neoproterozoic sedimentation: the Smalfjord Formation, northern Norway — discusssion. Sedimentology, 51, 1409– 1417. Edwards, M. B. & Føyn, S. 1981. Late Precambrian tillites in Finnmark, North Norway. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 606– 610. Edwards, M. B., Baylis, P., Gibling, M., Goffe, W., Potter, M. & Suthren, R. J. 1973. Stratigraphy of the ‘Older Sandstone Series’ (Tanafjord Group) and Vestertana Group North of Stallogaissa, Laksefjord District, Finnmark. Norges geologiske Undersøkelse, 294, 25 – 41. Eyles, N. 1993. Earth’s glacial record and its tectonic setting. Earth Science Reviews, 35, 1 –248. Farmer, J., Vidal, G., Moczydłowska, A., Strauss, H., Ahlberg, P. & Siedlecka, A. 1992. Ediacaran fossils from the Innerelv Member (late Proterozoic) of the Tanafjorden area, northeastern Finnmark. Geological Magazine, 129, 181– 195. Føyn, S. 1937. The Eo-Cambrian series of the Tana District, northern Norway. Norsk Geologisk Tidsskrift, 17, 65– 164. Føyn, S. & Glaessner, M. F. 1979. Platysolenites, other animal fossils, and the Precambrian – Cambrian transition in Norway. Norsk Geologisk Tiddskrift, 59, 25 – 46. Føyn, S. & Siedlecki, S. 1980. Glacial Stadials and Interstadials in the Late Precambrian Smalfjord Tillite on Laksefjordvidda, Finnmark, North Norway. Norges geologiske Undersøkelse, 358, 31 – 45. Gayer, R. A. & Rice, A. H. N. 1989. Palaeogeographic reconstruction of the pre- to syn-Iapetus rifting sediments in the Caledonides of Finnmark, N. Norway. In: Gayer, R. A. (ed.) The Caledonide Geology of Scandinavia. Graham & Trotman, London, 127– 139. Gehling, J. G., Narbonne, G. M. & Anderson, M. M. 2000. The first named Ediacaran body fossil, Aspidella terranovic. Palaeontology, 43, 427– 456. Gorokhov, I. M., Siedlecka, A., Roberts, D., Melnikov, N. N. & Turchenko, T. L. 2001. Rb –Sr dating of diagenetic illite in Neoproterozoic shales, Varanger Peninsula, northern Norway. Geological Magazine, 138, 541– 562. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Towards a Neoproterozoic composite carbon isotope record. Geological Society of America Bulletin, 117, 1181– 1207. Hambrey, M. J. 1983. Correlation of Late Proterozoic tillites in the North Atlantic region and Europe. Geological Magazine, 120, 209–232. Hansen, T. A. 1992. Sedimentologiske og Stratigrafiske Undersøkelser av den Sen-Prekambriske Smalfjord formasjonen, Øst-Finnmark. Cand. Real thesis, Universitet i Troms. Harland, W. B., Armstrong, R. L., Cox, A. V., Craig, L. E., Smith, A. G. & Smith, D. G. 1989. A Geological Time Scale 1989. Cambridge University Press. Hartz, E. H. & Torsvik, T. H. 2002. Baltica upside down: a new plate tectonic model for Rodinia and the Iapetus Ocean. Geology, 30, 255– 258.

601

Hoffman, P. F., Halverson, G. P., Domack, E. W., Husson, J. M., Higgins, J. A. & Schrag, D. P. 2007. Are basal Ediacaran (635 Ma) post-glacial ‘cap dolostones’ diachronous. Earth and Planetary Science Letters, 258, 114–131. Hoffmann, K.-H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. U –Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan Glaciation. Geology, 32, 817– 820. Jensen, P. A. & Wulff-Pedersen, E. 1996. Glacial or non-glacial origin for the Bigganjarga tillite, Finnmark, northern Norway. Geological Magazine, 133, 137– 145. Johnson, H. D., Levell, B. K. & Siedlecki, S. 1978. Late Precambrian sedimentary rocks in East Finnmark, North Norway and their relationship to the Trollfjord-Komagelva Fault. Journal of the Geological Society, London, 135, 517– 533. Kennedy, M. J., Runnegar, B., Prave, A. R., Hoffman, K. H. & Arthur, M. A. 1998. Two or four Neoproterozoic glaciations? Geology, 26, 1059– 1063. Kulling, O. 1951. Spa˚r av Varangeristiden i Norbotten. Sveriges geologiska Underso¨kning A˚rsbok 43 (1949), C 503, 1 –45. Kumpulainen, R. A. 2011. The Neoproterozoic Lillfja¨llet Formation, southern Swedish Caledonides. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 629– 634. Kumpulainen, R. A. & Greiling, R. O. 2011. Evidence for late Neoproterozoic glaciation in the central Scandinavian Caledonides. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 623–628. Laajoki, K. 2001. Additional observations on the late Proterozoic Varangerfjorden unconformity, Finnmark, northern Norway. Bulletin of the Geological Survey of Finland, 73, 17 –34. Laajoki, K. 2002. New evidence of glacial abrasion of the Late Proterozoic unconformity around Varangerfjorden, northern Norway. Special Publication of the International Association of Sedimentologists, 33, 405– 436. Laajoki, K. 2003. The Larajæg’gi outcrop — a large combined Neoproterozoic/Pleistocene roche moutonne´e at Karlebotn, Finnmark, North Norway. Norwegian Journal of Geology, 84, 107– 115. Le Guerroue´, E., Allen, P. A. & Cozzi, A. 2006. Chemostratigraphic and sedimentological framework of the largest negative carbon isotopic excursion in Earth history: the Neoproterozoic Shuram Formation (Nafun Group, Oman). Precambrian Research, 146, 68 –92. Levine, R., Baarli, G. & Johnson, M. E. 2006. Glacial and rocky-shore dynamics of the northern Karlebotn monadnocks: late Neoproterozoic of northern Norway. Canadian Journal of Earth Sciences, 43, 1215–1228. McCay, G. A., Prave, A. R., Alsop, G. I. & Fallick, A. E. 2006. Glacial trinity: Neoproterozoic Earth history within the British –Irish Caledonides. Geology, 34, 909– 912. Nystuen, J. P. & Lamminen, J. T. 2011 Neoproterozoic glaciation of South Norway: from continental interior to rift and pericratonic basins in western Baltica. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 613–622. Pisarevsky, S. A., Murphy, J. B., Cawood, P. A. & Collins, A. S. 2008. Late Neoproterozoic and Early Cambrian palaeogeography: models and problems. In: Pankhurst, R. J., Trouw, R. A. J., de Brito Neves, B. B. & de Wit, M. J. (eds) West Gondwana: Pre-Cenozoic Correlations Across the South Atlantic Region. Geological Society, London, Special Publication, 294, 9–31. Pringle, I. R. 1973. Rb –Sr age determinations on shales associated with the Varanger ice age. Geological Magazine, 109, 465– 472. Reading, H. G. 1965. Eocambrian and Lower Palaeozoic geology of the Digermul Peninsula, Tanafjord, Finnmark. Norges geologiske Undersøkelse, 234, 167– 191. Reading, H. G. & Walker, R. G. 1966. Sedimentation of Eocambrian tillites and associated sediments in Finnmark, North Norway. Palaeogeography, Palaeoclimatology, Palaeoecology, 2, 177– 212. Reusch, H. 1891. Skuringsmærker og morængrus eftervist i Finnmarken fra en periode meget ældre end “istiden”. Norges geologiske Undersøkelse, 1, 78 –85 (97–100 English summary).

602

A. H. N. RICE ET AL.

Rice, A. H. N. 1994. Stratigraphic overlap of the late Proterozoic Vadsø and Barents Sea Groups and correlation across the Trollfjorden– Komagelva Fault, Finnmark, North Norway. Norsk Geologisk Tidsskrift, 74, 48 –57. Rice, A. H. N. & Townsend, C. 1996. Correlation of the late Precambrian Ekkerøya Formation (Vadsø Group; E. Finnmark) and the Brennelvfjord Interbedded member (Porsangerfjord Group; W. Finnmark), N. Norwegian Caledonides. Norsk Geologisk Tidsskrift, 76, 55– 61. Rice, A. H. N. & Hofmann, Ch.-Ch. 2000. Evidence for a glacial origin of the Neoproterozoic III striations at Oaibacˇcˇanjar’ga, Finnmark, northern Norway. Geological Magazine, 137, 355– 366. Rice, A. H. N. & Hofmann, Ch.-Ch. 2001. The transition from Neoproterozoic glacial to interglacial sedimentation near Hammarnes, East Finnmark, North Norway. Norwegian Journal of Geology, 81, 257– 262. Rice, A. H. N., Gayer, R. A., Robinson, D. & Bevins, R. E. 1989. Strikeslip restoration of the Barents Sea Caledonides Terrane, Finnmark, North Norway. Tectonics, 8, 247– 264. Rice, A. H. N., Hofmann, Ch.-Ch. & Pettersen, A. 2001. A new sedimentary succession from the southern margin of the Neoproterozoic Gaissa Basin, south Varangerfjord, North Norway; the Lattanjar´ga unit. Norsk Geologisk Tidsskrift, 81, 41 –48. Rice, A. H. N., Ntaflos, T., Gayer, R. A. & Beckinsale, R. D. 2004. Metadolerite geochronology and dolerite geochemistry from East Finnmark, Northern Scandinavian Caledonides. Geological Magazine, 141, 301– 318. Røe, S.-L. 2003. Neoproterozoic peripheral-basin deposits in eastern Finnmark, N. Norway: stratigraphic revision and palaeotectonic implications. Norwegian Journal of Geology, 83, 259– 274. Schermerhorn, L. J. G. 1974. Late Precambrian mixtites: glacial or non-glacial. American Journal of Science, 274, 673–824. Siedlecka, A. 1985. Development of the Upper Proterozoic sedimentary basins of the Varanger Peninsula, East Finnmark. Geological Survey of Finland Bulletin, 331, 175–185. Siedlecka, A. 1987. Trollfjorden berggrunnskart 2336 3, 1:50,000, foreløpig utgave. Norges geologiske Undersøkelse. Siedlecka, A. 1990. Varangerbotn berggrunnskart 2335 3, 1:50,000, foreløpig utgave. Norges geologisk Undersøkelse. Siedlecka, A. 1995. Neoproterozoic sedimentation on the Rybachi and Sredni Peninsulas and Kildin Island, NW Kola, Russia. Norges geologiske Undersøkelse Bulletin, 427, 52– 55.

Siedlecka, A. & Siedlecki, S. 1971. Late Precambrian sedimentary rocks of the Tanafjord-Varangerfjord region of Varanger Peninsula, Northern Norway. Norges geologiske Undersøkelse, 269, 246–294. Siedlecka, A., Lyubtsov, V. V. & Negrutsa, V. Z. 1995. Correlations between Upper Proterozoic successions in the TanafjordenVarangerfjorden Region of Varanger Peninsula, northern Norway, and on Sredni Peninsula and Kildin Island in the northern coastal area of Kola Peninsula in Russia. Norges geologiske Undersøkelse Special Publication, 7, 217– 232. Siedlecka, A., Roberts, D., Nystuen, J. P. & Olovyanishnikov, V. G. 2004. Northeastern and northwestern margins of Baltica in Neoproterozoic time: evidence from the Timanian and Caledonian Orogens. In: Gee, D. G. & Pease, V. (eds) The Neoproterozoic Timanide Orogen of Eastern Baltic. Geological Society, London, Memoirs, 30, 169–190. Siedlecki, S. & Levell, B. K. 1978. Lithostratigraphy of the Late Precambrian Løkvikfjell Group on Varanger Peninsula, East Finnmark, North Norway. Norges geologiske Undersøkelse, 343, 73 – 85. Stodt, F., Rice, A. H. N., Bjo¨rklund, L., Bax, G. & Pharaoh, T. C. 2011. Evidence of late Neoproterozoic glaciation in the Caledonides of NW Scandinavia. In: Arnaud, E., Halverson, G. P. & ShieldsZhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 603–611. Strahan, A. 1897. On Glacial Phenomena of Palaeozoic age in the Varanger Fjord. Quarterly Journal of the Geological Society, 53, 137– 146. Torsvik, T., Lohmann, K. C. & Sturt, B. A. 1995. Vendian glaciations and their relation to the dispersal of Rodinia: palaeomagnetic constraints. Geology, 23, 727– 730. Townsend, C., Roberts, D., Rice, A. H. N. & Gayer, R. A. 1986. The Gaissa Nappe, Finnmark, North Norway: an example of a deeply eroded external imbricate zone within the Scandinavian Caledonides. Journal of Structural Geology, 8, 431–440. Vidal, G. 1981. Micropalaeontology and biostratigraphy of the Upper Proterozoic and Lower Cambrian sequences in East Finnmark, northern Norway. Norges geologiske Undersøkelse, 362, 1 –53. Vidal, G. & Siedlecka, A. 1983. Planktonic, acid-resistant microfossils from the Upper Proterozoic strata of the Barents Sea region of Varanger Peninsula, East Finnmark, Northern Norway. Norges geologiske Undersøkelse, 382, 45– 79.

Chapter 58 Evidence of late Neoproterozoic glaciation in the Caledonides of NW Scandinavia ¨ RKLUND3, G. BAX4, G. P. HALVERSON5,6 & T. C. PHARAOH7 F. STODT1, A. H. N. RICE2*, L. BJO 1

Savignystrasse 9, 35037 Marburg, Germany

2

Vienna University, Structural Processes Group, Department of Geodynamics and Sedimentology, Althanstrasse 14, 1090 Vienna, Austria, Europe 3

Go¨teborg University, Earth Sciences Centre, Box 460, SE-405 30 Go¨teborg, Sweden

4

Department of Earth Sciences & Centre for Image Analysis, Uppsala University, SE-752 36, Uppsala, Sweden

5

Department of Geology & Geophysics, School of Earth & Environmental Sciences, University of Adelaide, North Terrace, Adelaide, SA 5005, Australia 6

Present address: Department of Earth and Planetary Sciences, McGill University, 3450 University Street, Montreal, Quebec, H3A 2A7, Canada 7

Geophysics & Marine Geosciences, British Geological Survey, Keyworth, Nottinghamshire, NG12 5GG, UK *Corresponding author (e-mail: [email protected]) Abstract: The northwestern part of the Scandinavian Caledonides, formed by SE- to ESE-directed thrusting through the Neoproterozoic W. Baltica continental shelf, contains numerous small and often isolated outcrops of diamictite and associated strata. No precise biostratigraphic or isotopic data are available to constrain the age of these sediments, but, on the basis of their stratigraphic position, most are correlated with the Mortensnes Formation (Fm.) in E. Finnmark and also presumed to be of glaciogenic origin. The Mortensnes Fm. has been correlated with the 580 Ma Gaskiers glacial event on the basis of d13C isotope studies. Structurally, the deposits occur in the Autochthon (below the Tornetra¨sk Fm.), within an external imbricate zone (Lower Allochthon), within cover successions lying unconformably on allochthonous basement (Window Allochthon) palaeogeographically derived from below or outboard of the Lower Allochthon and, more rarely, within the Middle Allochthon, derived from outboard of the Window Allochthon. Evidence for a glaciogenic origin is typically poor or lacking. Only in the Komagfjord Antiformal Stack (Window Allochthon), where an up to 40-m-thick succession of three fining upwards cycles has been mapped, are the deposits comparable in thickness and complexity to the Mortensnes Fm. Other sequences are sometimes ,1 m thick and unconformably overlain by post-‘glacial’ deposits. The Vakkejokk Breccia, a submarine slump in the Tornetra¨sk area of the Autochthon closely underlies the correlative Precambrian –Cambrian lithostratigraphic boundary in E. Finnmark but overlies the first appearance of the boundary marker fossil Treptichnus pedum. Although sometimes interpreted as periglacial, this seems unlikely in view of the 30–508 palaeolatitude during deposition. Calcite nodules (,1 cm size) in the Vakkejokk Breccia have previously been interpreted as glendonite, but the microstructure and palaeolatitude makes this unlikely; they are likely a replacement of gypsum. Diamictites of uncertain origin have also been found in the Ediacaran Lower Siltstone Member of the Tornetra¨sk Fm. and unconformably under the ?Lower Cambrian Lomvatn Fm. in the Komagfjord Antiformal Stack.

This chapter primarily covers (meta-)sediments correlated with the Mortensnes Fm., the younger of the two glaciogenic units of E. Finnmark, N. Norway (cf Rice et al. 2011). Such deposits typically occur as small and isolated outcrops within both the Autochthon and nappes of the Scandinavian Caledonides, often resting unconformably on basement rocks (Kumpulainen 2011; Kumpulainen & Greiling 2011; Nystuen & Lamminen 2011) After restoration of the Caledonian nappes, these diamictite outcrops indicate that glacial deposits covered an area of some 140 000 km2 in N. Scandinavia (including E. Finnmark). Studies in Norrbotten, Sweden, in the southernmost part of the area covered here (Fig. 58.1), led Kulling (1951) to introduce the term Varangeristiden (Varang(er)ian glaciation) to encompass both the glacial events described here and the earlier Smalfjord Fm. (Marinoan). However, for reasons summarized below, this term should no longer be used.

Structural framework Restoration of the Scandinavian Caledonides, using balanced cross-sections, demonstrates that the NNE –SSW-trending (present orientation) continental shelf of W. Baltica comprised an outboard basin lying adjacent to the continental edge and an intermittently developed, somewhat shallower inboard basin

(Fig. 58.2; Gayer et al. 1987; Gayer & Greiling 1989; Rice 1998, 2001, 2006). In NE Scandinavia, this continental shelf linked with the WNW –ESE-trending Timanian Basin (Siedlecka 1985; Gayer & Rice 1989; Siedlecka et al. 2004; cf Rice et al. 2011) that formed after the accretion of several minor terranes to Baltica in the middle Neoproterozoic (cf Cocks & Torsvik 2005). The Timanian Basin has been divided into northerly deep water and southerly shelf parts (Siedlecka et al. 1995, 2004), separated by the Trollfjorden – Komagelva Fault (cf Rice et al. 1989). In Scandinavia, the shelf part comprises the Gaissa Basin, which, in the west, was also an inboard basin of the NNE –SSW-trending W. Baltica shelf. Predominantly in-sequence ESE- to SE-directed Caledonian shortening occurred mostly in Silurian – Devonian times. This passed through the outboard basin (Middle Allochthon), then a palaeogeographic basement-high (Window Allochthon, Rice 2001) and then the inboard basin, forming an external imbricate zone (external Lower Allochthon), finally emplacing the nappe pile onto the Autochthon. In some areas, the internal part of the Lower Allochthon restores to above the Window Allochthon and was imbricated earlier (Fig. 58.2). Both basins have essentially the same lithostratigraphy; in most areas, the basement-high was finally drowned in late Ediacaran to mid-Cambrian –Tremadocian times, essentially contemporary with the mainland to the SE. In the Tysfjord-Akkajaure area, both the Middle and Lower Allochthons

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 603– 611. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.58

604

F. STODT ET AL.

Fig. 58.1. Map of north Scandinavian Caledonides showing the distribution of glaciogenic lithologies and non-glaciogenic diamictites outside the E. Finnmark (Laksefjordvidda-Tanafjord-Mortensnes) area. A, Altenes Window; An, Andabakoaivi; Ak, Akkajaure; AK, Alta-Kvænangen Window; Au, Autajaure Window; B, Bulljovagge; G, Gæv’dnjajav’ri; H, Halkavarre; J, Jerta; K, Komagfjord Window; Ka, Karnjelaja˚kka; Ko, Kuokkel Window; Lv, Laksefjordvidda; Mn, Mauken Window; P, Porsavatn; Po, Porsangerfjord; Pa, Paittasja¨rvi; Ri, Ritsemjaure Window; RS, Rombak-Sjangeli Window; Ru, Ruoddojokka; Si, Sitojaure; Sj, Stora Sjo¨fallet; Sk, Skoganvarre; Sv, Sarvapakte; T, Tornetra¨sk; Ty, Tysfjord; Vk, Vakkeja˚kka (now called Sarva´johka); Vu, Vuojtasrika.

are dominated by basement rocks, with only a thin cover succession (Bjo¨rklund 1985, 1989), indicating that very restricted basin development occurred in this region. In the area covered here, much of the Middle Allochthon comprises pre-Caledonian basement (Bjo¨rklund 1985, 1989; Kirkland et al. 2006). Window Allochthon rocks are exposed in the cores of large domal/periclinal structures, with the major units being the Komagfjord Antiformal Stack (Komagfjord, Altenes & AltaKvaenangen Windows combined), the Rombak-Sjangeli, and

adjacent Kuokkel Windows, and to the south, the relatively small Autajaure and Ritsemjaure Windows (Fig. 58.1). Although a basal thrust is typically not exposed, their allochthonous status has been inferred from structural and metamorphic criteria (Gayer et al. 1987; Anderson 1989; Bax 1989; Bjo¨rklund 1989; Rice 2001). The Gaissa Thrust Belt (derived from the inboard, Gaissa Basin), with 50% shortening in the west (Townsend et al. 1986, 1989), is the main development of the Lower Allochthon. On

Fig. 58.2. Schematic (V = H) WNW– ESE-oriented palinspastic cross-section showing the development of inboard and outboard basins in pre-diamictite times, separated by a basement high. The basins and basement high were then buried under glacial and post-glacial deposits. The main tectonic units are numbered in the order of their imbrication, with schematic main and minor thrusts. Localities along the top are given in the caption to Figure 58.1. Note that in many areas only one basin formed and in some areas neither basin formed.

THE CALEDONIDES OF NW SCANDINAVIA

Laksefjordvidda and to the east (Fig. 58.1), shortening decreases to c. 15% (Chapman et al. 1985). SW of Ruoddojokka, most of the Lower Allochthon has been eroded away (cf. Hossack & Cooper 1986; Anderson 1989), although several relicts have been preserved (Fig. 58.1; Jerta Nappe, Rautas Complex, Lower Thrust Complex – see below).

605

Stratigraphy The deposits documented here are associated with condensed sequences correlated directly or indirectly with the late Precambrian sediments exposed in E. Finnmark (predominantly the Vestertana Group; cf. Rice et al. 2011; Fig. 58.3).

Fig. 58.3. Regional stratigraphic profiles. (a) Alta-Kvaenangen, S. Komagfjord and N. Komagfjord refer to where each part of the log-section was measured. In the S. Komagfjord zone, Borras Group, Rafsbotn Fm. and Slettfjell Fm. are the three lithostratigraphic equivalents from the three tectonic windows (from south to north) in the Komagfjord Antiformal Stack. The section shown is of the Slettfjell Fm. (Pharaoh 1985). (b) Section from Tornetra¨sk (Thelander 1982; modified after Bax 1984 and Stodt 1987). (c) E. Finnmark and W. Finnmark refer to the successions in the eastern (Johnson et al. 1978) and western (Williams 1976; Rice & Townsend 1996) parts of the Gaissa Basin. Pl, Platysolenties antiquissimus; T, Treptichnus pedum; UE, LE, upper and lower limit of reported Ediacaran fauna; C, calcite nodules.

606

F. STODT ET AL.

West of Andabakoaivi (Fig. 58.1), the Caledonian Autochthon comprises the Dividal Group, correlated with the post-Mortensnes Fm. part of the Vestertana Group and overlying Digermul Group in E. Finnmark (Fig. 58.3; Kulling 1964; Føyn 1967; Vogt 1967; Thelander 1982). The constituent Tornetra¨sk (c. 100– 260 m) and overlying Alum Shale (c. 80 m) formations are locally separated from the underlying pre-Caledonian basement (often a regolith; Bax 1984; Stodt 1987) by small patches of diamictite (Fig. 58.1); typically these are unnamed. The locally developed, but important Vakkejokk Breccia (cf. Kulling 1964; Thelander 1982; Stodt 1987) cuts across the lower part of the Tornetra¨sk Fm. (Fig. 58.3). Although Kulling (1964) defined a Middle Sandstone Fm. (; unit C of Vogt 1967), Thelander (1982), who downgraded the rank of all the units from formations to members, merged this into the Lower Siltstone Member (Fig. 58.3). (Note that the Vakkejokk Breccia type locality, Vakkeja˚kka, or Orddajohka on some maps, is now called Sarva´johka; but this is not the same as the Sarvaja˚kka, now Sarva´gorsa, described by Thelander, 1982). In the western part of the Gaissa Thrust Belt (external Lower Allochthon; imbricated Gaissa Basin; Fig. 58.2; Townsend et al. 1986, 1989), the Tanafjord Group is thinner than in the type area, although correlatives of the Grasdal and Lille Molvika formations (Stabbursdal plus Porsanger formations and Brennelvfjord Fm., respectively; Fig. 58.3) are thicker (cf. Rice & Townsend 1996). The poorly studied Airoaivi Group is thinner and lithologically different to the Vadsø Group, its possible chronostratigraphic equivalent (Townsend et al. 1989; Rice & Townsend 1996). No diamictite has been preserved from this part of the Gaissa Basin. In the Jerta Fm. (Jerta Nappe; Lower Allochthon), Skjerlie & Tan (1961; Fig. 58.1) recorded a predominantly clastic sequence with a basal diamictite. This whole sequence, estimated at c. 500 m thick, is likely a correlative of the Vestertana Group, although the presence of pyritiferous black shales suggest it may be partially correlated with the Digermul Group/Alum Shale Fm. (Reading 1965; Thelander 1982). The successions in both the Rautas Complex and the Lower Thrust Complex (near Tornetra¨sk and Stora Sjo¨fallet, respectively; Fig. 58.1) have been reliably correlated with the tectonically underlying autochthonous Dividal Group (Bjo¨rklund 1985; Bax 1989). In the Komagfjord Antiformal Stack (Window Allochthon), the cover sequence youngs northwards, with ,185 m of Bossekop Group (; Tanafjord Group; Fig. 58.3; Føyn 1964, 1985) lying unconformably on the basement in the Alta-Kvaenangen Window (Fig. 58.1). The Bossekop Group is unconformably overlain by the Borras Group, correlated with the post-Nyborg Fm. part of the Vestertana Group, including c. 10 m of basal diamictite (Fig. 58.3; Føyn 1964, 1985). In the Altenes and southern part of the Komagfjord Window, patchy outcrops of diamictite up to 40 m thick rest directly on the basement and are overlain by a ,150-m-thick sequence again directly comparable to the lower part of the Tornetra¨sk Fm. (Rafsbotn & Slettfjell formations; Roberts & Fareth 1974; Pharaoh 1985) whereas in the northern part of the Komagfjord Window a small patch of diamictite (here termed the Porsavatn Diamictite Bed) is overlain by ,160 m of the Lomvatn Fm. (Pharoah 1985), broadly correlated in this chapter with the upper parts of the Tornetra¨sk Fm. (Fig. 58.3). Other successions within the area, from the Lower, Window and Middle Allochthons, that overlie diamictites have been directly correlated with the Dividal Group and have not been given local stratigraphic names.

Glaciogenic and associated deposits Autochthon Near Skoganvarre, Siedlecka (1987; Fig. 58.1) recorded three outcrops of polymict conglomerate within a 1km distance, comprising

1–50 cm clasts of amphibolite, gneiss and quartzite, lying below the distinctive basal conglomerate of the Tornetra¨sk Fm. To the SW of Ruoddojokka, Holmsen (1956) noted a very variable stratigraphy under the Tornetra¨sk Fm.; in some places only occasional lonestones are present at the base of the sequence, essentially lying on the basement, whereas elsewhere 2–4 m of grey diamictite, with 1 m clasts, is overlain by 0.5– 1m of carbonate-bearing sandstone. At Bulljovagge, c. 2 m of red brown and grey diamictite overlie quartzitic sandstone, but the contact with the basement is not exposed (Holmsen 1957; Mathiesen in Skjerlie & Tan 1960). In these outcrops, the diamictite is often dominated by clasts derived from the immediately underlying rocks. In the Paittasja¨rvi – Sitojaure area and further south (Fig. 58.1), Kautsky (1949) described undeformed chloritic diamictites (Sito diamictite; Stromberg 1981) up to 4 m thick, with a dark matrix. The unsorted, angular to poorly rounded blocks are up to cubic metres in size and were derived from the underlying basement, which shows no signs of pre-diamictite weathering. The matrix is siliciclastic, comprising small quartz and felspar fragments; no carbonate is present. Near Sitojaure, the diamictite is overlain by .30 m of green mudstones with frequent blue quartzites and these are probably overlain by the Dividal Group, which directly overlies diamictites in nearby areas (Kulling 1951). Kulling (1951) also recorded banded silts and clays in the Sitojaure area. Other thin diamictites in this region were documented by Kulling (1951; Fig. 58.1), some with faceted clasts, although it is often unclear whether the outcrops documented belong to the Autochthon or overlying Lower Allochthon (cf Stro¨mberg 1981). The Vakkejokk Breccia lies in the upper part of the Lower Siltstone Member of the Tornetra¨sk Fm. on the NE side of Tornetra¨sk (Kulling 1964; Thelander 1982; Stodt 1987; Figs 58.1 & 58.3), thinning from 3– 10 m (authors disagree on the thickness) in the Vakkeja˚kka (Sarva´gorsa) area to c. 0.5 m some 7.5 km to the ESE. This locally cuts down through the Lower Sandstone Member to the basement (Stodt 1987). Where the breccia is thin, it may be overlain by sandstones of the uppermost part of the Lower Siltstone Member (Thelander 1982). Stodt (1987) documented four breccia types. † Type A (recorded only at Gaev’dnjajav’ri; Fig. 58.1): conglomerate with ,4-cm-sized, well-rounded dark-grey clay-silt clasts and some quartz fragments, forming the base of a 12-cm-thick BoumaABCD sequence lacking a structureless sandstone and dewatering structures. This lies with an erosional contact on the underlying thin-bedded sandstones of the Lower Siltsone Member, c. 1 m below the base of the Middle Siltstone Member. † Type B: angular to rounded granitic, vein-quartz, silt and shale matrix-supported clasts up to 15 cm in size in sandstone beds up to 30 cm thick. Platy clasts are bedding parallel. In some cases, the beds show normal grading. Matrix sandstone grains are rounded to well-rounded and predominantly quartz. The carbonate cement may have been derived from a carbonate-mud matrix. † Type C: normally .90% granitic clasts with minor elongate, randomly oriented sedimentary clasts, often folded. The former are predominantly clast-supported, equidimensional to slightly elongate, mostly angular but sometimes well-rounded. Intraclasts are platy and show abundant fold structures. The matrix comprises grey, green and red silt/sandstone, often internally brecciated. Two such flows have been recognized, except at Vakkeja˚kka, where five flows occur. South of Sarvapakte (Sarvabakti; Fig. 58.1) one flow contains up to 1-cm-sized calcite nodules interpreted as glendonite by Stodt (1987; but see ‘Discussion’). The flows usually lie conformably or slightly erosively on the Lower Siltstone Member. † Type D: granitic blocks 2– 100 m in size, lying either within and above siltstones that are often folded and disturbed or on type C breccias, along sharp boundaries, exposed for c. 4 km east of Vakkeja˚kka. SE of Vakkeja˚kka, 2– 3 m clasts lie directly

THE CALEDONIDES OF NW SCANDINAVIA

on the basement. Some breccia clasts, derived from earlier flows, are also present (breccia-in-breccia structure). Granitic blocks and clasts are basement-derived.

Lower Allochthon No outcrops of diamictites have been recorded within the Gaissa Thrust Belt in the Porsangerfjord area; exposures from areas further east are reviewed in Rice et al. (2011). In the Jerta Fm., Skerlie & Tan (1960; Fig. 58.1) described a diamictite with 5–30 cm angular to rounded, unsorted clasts of light grey, yellowweathering dolomite, as well as greenstones and quartzites. The matrix is dark to brownish-grey, with angular dolomite and smaller poorly rounded quartz grains. Some finer, stratified and possibly graded beds are also present. No thicknesses were given, but photographs indicate a c. 2 m or greater thickness. To the north of Sitojaure, in the Stora Sjo¨fa¨llet area, Kulling (1948, 1951, 1982) reported a close association of thinly bedded calcareous shales and diamictite or diamictite-like deposits with lithologies similar to the diamictites in the Ritsemjaure Window (see below).

Window Allochthon In the Alta-Kvaenangen window, the most southerly part of the Komagfjord Antiformal Stack, 185 m of quartzites and shales of inferred shallow marine origin (Bossekop Group) are unconformably overlain by c. 10 m of red-brown diamictite at the base of the Borras Group, passing up gradually to conglomerates correlated with the base of the Tornetra¨sk Fm. and thence to sandstones and shales (Føyn 1964, 1985; Fig. 58.3). Further north, in the Altenes Window, Roberts & Fareth (1974) very locally found c. 2 m of reddish-brown diamictite with clasts (,30 cm size; 30% of the rock) of mixed cover and basement lithologies, many derived from the immediate substrate, lying on the basement. Nearby, the Rafsbotn Fm., comprising a basal well-rounded pebble-sized quartz conglomerate, c. 1 m thick, and then by green and red mudstones, slates and siltstones, directly overlies the basement. The Rafsbotn Fm. is equivalent to the basal part of the Tornetra¨sk Fm. Along the southern rim of the Komagfjord Window (northern part of the Komagfjord Antiformal Stack), outcrops of the essentially undeformed Nyvoll Tillite Member (Slettfjell Fm.) are preserved as lenses up to 1km long and 44 m thick (although this particularly thick outcrop thins to nothing in 300 m along strike; Pharaoh 1980, 1985). Pharaoh (1980) described four facies, with up to three fining-upwards cycles: † Facies A: poorly sorted, unstratified polymict diamictites with .20% clasts in a reddish-brown matrix, locally clast supported at the base, where the clast content may rise to 40%. Most clasts are ,5 cm across, but range up to 60 cm in size and vary from angular to sub-rounded, and are sometimes facetted. A large percentage was locally derived, although gneiss, granite and quartzite clasts might have been externally derived. This facies is up to 5 m thick and always forms the base of the sequence, grading up into Facies B, although it reappears higher in the sequence. † Facies B: poorly sorted polymict diamictite similar to facies A, but with significantly fewer (,20%) and smaller clasts (c. 10 cm max.), locally with a carbonate matrix/cement towards the top. Thickness varies from 1 to over 20 m. † Facies C: poorly sorted, poorly stratified diamictite, gradational from facies B by a continued decline in clast size (,1cm) and content (,5%). ‘Bedding’ is defined by thin laminae of quart grains in a reddish brown muddy matrix. This facies is generally thinly developed, although a thickness of 23 m occurs in one section.

607

† Facies D: intercalated sandstones and mudstone in bands 1– 5 cm thick, usually reddish brown, although the sandstones may be paler and greyish-green. Sandstones are frequently graded, with microconglomeratic erosive bases, ripples, load casts and intraformational mud-flake breccias. Impact pits are present, as is evidence of soft-sediment folding and both thrust and normal soft-sediment faulting. Bedding in the diamictite is parallel to that in the overlying basal conglomerate of the Vargelv Member (Fig. 58.3), indicating that diamictite thickness variations reflect a syn-sedimentary uneven basement surface; a palaeorelief, up to 3 m high, has been preserved at outcrops and an overall pre-diamictite topography of 40 m has been mapped. Pharaoh (1980) suggested that the Nyvoll Member was deposited in a NW – SE-trending palaeovalley c. 1km wide and up to 40 m deep, accounting for the sequence having a thickness atypical outside the E. Finnmark area (cf Rice et al. 2011). The northern margin of the valley is poorly defined due to scarce exposure. Clasts in the diamictites comprise feldspathic metasandstones, jasper, vein quartz, metagabbro, serpentinite and trondjemite, all locally derived; sometimes they can be linked to nearby palaeo-topographic highs. These sediments are overlain by a quartz conglomerate, with little topographic relief at the contact, and then by red and green siltstones/shales, comparable with the Borras Group/Rafsbotn Fm. to the south (Fig. 58.3). In the northern part of the Komagfjord Window, a ,0.5-m-thick, very poorly sorted matrix-supported diamictite (Pharaoh 1985), here termed the Porsavatn Diamictite Bed, comprising 0.5 –3 cm angular clasts of vein quartz and quartzite in a grey muddy matrix, underlies the 180-m-thick Lomvatn Fm. at one area (Figs 58.1 & 58.3). Elsewhere, the 0.6–2-m-thick basal conglomerate of this formation, consisting of well-rounded vein quartz and quartzite pebbles, rests unconformably on the basement and is overlain by sandstones and shales (Pharaoh 1985), probably correlatives of the upper part of the Tornetra¨sk Fm. (Fig. 58.3). The Gearbelja´vri Fm., a thin sedimentary veneer that crops out around the rim of the Rombak-Sjangeli and Kuokkel Windows and on minor tectonic klippen within them, has also been correlated with the basal part of the Tornetra¨sk Fm. (Fig. 58.1, Brown & Wells 1966; Tull et al. 1985; Bax 1989, 2001). In places, a basal conglomerate lies at the basement –cover contact, but elsewhere the contact grades from unweathered basement, through regolith into quartzites. At Vuojtasrika, Brown & Wells (1966) documented 0.5-m-deep, diamictite-filled fissures in the basement, but Bax (unpublished data) reinterpreted these as fault breccias. Along the northern margin of the Rombak-Sjangeli Window, Tull et al. (1985) described grey to white arkosic cross-bedded sandstones, interlayered with conglomerates and pelitic schists, locally overlain by diamictites with clasts of granitic, ?dioritic and pelitic lithologies up to 25 cm in diameter. Coarsely graded sequences are common. Current directions indicate flow to the south to SSW, although the basement –cover contact is a palaeopeneplain. At one outcrop in the SE part of the Kuokkel Window, a 1-m-thick diamictite is underlain by quartzites correlated with Lower Sandstone Member (Tornetra¨sk Fm.). This diamictite, which lies c. 2 m above the basement –cover unconformity, although the contact is not exposed directly here (Bax 1984), comprises angular, unsorted, carbonate and sandstone clasts (no basement clasts) up to 20 cm in size, surrounded by a fine-grained matrix. No clast sorting or preferred orientation has been observed. Upwards in the section, above the overlying cross-bedded sandstones, ripple marks are common in sandy layers interbedded with brown shales (Bax 1984). In the Ritsemjaure Window, within the Middle Allochthon (Akkajaure Nappe Complex), vertical fissure fillings trending 0108 occur in mesoperthitic granite. The fissures have highly

608

F. STODT ET AL.

irregular margins and widths of up to 5 m, but are of unknown depths (Fig. 58.1; Bjo¨rklund 1989). The fissures are filled with an unsorted sedimentary breccia, consisting of a dark grey, silty matrix, chaotically mixed with all sizes (,1m) of angular clasts from the host granite and very subordinate shale and dolomite clasts. The same type of breccia overlies the basement granite, but with smaller (,10 cm) clasts. Microscopic clasts reflect the mesoperthitic feldspars of the subjacent granite. The breccias grade upwards into dark conglomerates with up to pebble-sized clasts and a weak clast size stratification and thence into dark shales with ,5 cm lonestones. Local shearing within the diamictite sequence makes thickness estimates uncertain, but the diamictite lying on the granite is 5 –8 m thick and the overlying conglomeratic to shaly part is c. 5 m thick in the least disturbed areas. Overlying this, along a sharp contact, is a ,3-m-thick, typically blue-grey orthoquartzite with a basal quartz conglomerate including porphyry clasts, correlated with the base of the Tornetra¨sk Fm. (Bjo¨rklund 1989). Further west, on the NE side of the Tysfjord Culmination, several local pockets of strongly sheared ,0.5-m-thick schistose unsorted diamictites with granite clasts in a dark grey matrix have been mapped with irregular contacts on the basement (Fig. 58.1; Bjo¨rklund 1989). These are overlain by c. 1-m-thick white to grey quartzites and thence by overthrust rocks of the Lower Allochthon.

Middle Allochthon In the eastern part of the lowest imbricate of the Akkajaure Nappe Complex (in Karnjelaja˚kka; Fig. 58.1), a 1–2-m-thick diamictite with centimetre-sized angular mesoperthitic granite clasts in a carbonate-bearing matrix rests on an uneven granitic basement surface. The diamictite is sharply overlain by a c. 5-m-thick conglomerate with sub-rounded to sub-angular quartz and less common mesoperthite clasts fining upwards to gravelly quartzite, followed by grey quartz phyllite, all correlated with the base of the Tornetra¨sk Fm. Westwards within this imbricate, a transition from mesoscopic clast-rich diamictites to strongly weathered calcitedolomite mica schists occurs, overlain by quartz mylonites. Clast size diminishes westwards. The schists have microscopic subangular clasts of mesoperthite, quartz and albite similar to those in the diamictite; chemically, the calcite-dolomite mica schists and the diamictite matrix are very similar (Bjo¨rklund 1989). On the higher, more westerly derived thrust sheets, micaceous marbles are interlayered with these schists.

Boundary relations with overlying and underlying non-glacial units The sediments (mostly diamictites) generally lie unconformably on autochthonous or allochthonous Baltic Shield-derived basement rocks and are unconformably overlain by the distinctive quartz-rich basal conglomerate of the Tornetra¨sk Fm. or (deformed) correlatives. In some cases, a few metres of finergrained sediments lie between the diamictites and the basal conglomerate. The succession in the Komagfjord Antiformal Stack differs in that in the south, around the Alta-Kvaengen Window, the diamictites lie on a condensed sequence correlated with the mid- to upper part of the Tanafjord Group (Fig. 58.3; Føyn 1964, 1985), whilst further north they lie on allochthonous basement (Roberts & Fareth 1974; Pharoah 1985). The upper contact in the Komagfjord Antiformal Stack is also atypical, in that it is gradational with the Borras Group and Slettfjell formations, correlatives of the Lillevatn Fm. (¼ base Tornetra¨sk Fm.) in E. Finnmark (Føyn 1964; Pharaoh 1980, 1985). The upper contact of the Porsavatn Diamictite Bed is erosively overlain by the Lomvatn Fm. (Fig. 58.3).

In contrast, the diamictite in the Kuokkel Window lies within a sandstone/quartzite succession correlated with the Lower Sandstone Member of the Tornetra¨sk Fm., above the basal conglomerate. Rather similarly, the Vakkejokk Breccia rests with a marked erosive unconformity on the underlying sediments of the Tornetra¨sk Fm., cutting down-section to, very locally, the basement, and is conformably overlain by younger sediments (Stodt 1987).

Chemostratigraphy No stable isotope data are available due to the general lack of carbonates in the sequence. d13C values from calcite nodules in the Vakkejokk Breccia range between –4.03 and 0.79‰ (VPDB; mean, – 1.57‰, +1.38, N ¼ 11; crystals were drilled out and analysed using the technique of Halverson et al. 2005).

Palaeolatitude and palaeogeography Well-dated palaeopoles that can be used to constrain the Ediacaran palaeogeogeography of Baltica are scarce (cf Torsvik et al. 1996; Bingen et al. 2005; Cocks & Torsvik 2005). The southern rim of an inverted Baltica lay at c. 158S at c. 750 Ma (Hartz & Torsvik 2002). At 616 Ma, data from mafic dykes in southern Scandinvia indicate that Baltica lay at 758S although by c. 550 Ma, the area covered here lay at c. 45–508S (Cocks & Torsvik 2005). In contrast, Cawood & Pisarevsky (2006) and Pisarevsky et al. (2008) place Baltica in a more equatorial position (c. 308), although whether it lay in the northern or southern hemisphere is unclear.

Geochronological constraints No robust isotopic age constraints are available from this area. Correlation of the sub-Dividal Group sediments with the Mortensnes Fm. gives a broad age constraint of c. 580 Ma, by correlation with the Gaskiers glaciation (Bowring et al. 2003). Although it is likely that the Gaskiers event was diachronous, the close association of the Mortensnes Fm. with extremely negative d13C values in Finnmark, correlated with the Wonoka anomaly (Halverson et al. 2005; Rice et al. 2011), probably makes the effect of diachroneity on the timing of glaciation relatively small. The Vakkejokk Breccia is closely underlain (within a few metres) by the fossil-rich Kullingia Beds, within which Treptichnus pedum has been recorded (Stodt 1987). Although Treptichnus pedum was established as the index trace fossil for the base of the Cambrian (Brasier et al. 1994), it has subsequently been found in the youngest Ediacaran successions (Gehling et al. 2001). Thus the Vakkejokk Breccia is constrained to very latest Precambrian to early Cambrian times. This boundary lies somewhat below the lithostratigraphic correlation with the proposed Precambrian – Cambrian boundary from E. Finnmark (Fig. 58.3).

Discussion In East Finnmark, the glaciogenic Smalfjord Fm. has a typical Marinoan-type cap dolostone, forming the base of the Nyborg Fm. (cf Halverson et al. 2005; Rice et al. 2011). Thin dolostones in the upper part of the Nyborg Fm. (Member E; Edwards 1984; Fig. 58.3) have d13C values of –7.6 and –9.9‰ (VPDB); these have been correlated with the extreme negative d13C values recorded in the Wonoka anomaly in other parts of the world (down to – 12‰; cf. Halverson et al. 2005, Le Guerroue´ et al. 2006) to suggest that the almost immediately overlying Mortnesnes Fm. is broadly a correlative of the 580 Ma Gaskiers diamictite (cf Halverson et al. 2005). Correlation of the sub-Torneta¨sk

THE CALEDONIDES OF NW SCANDINAVIA

Fm. diamictites and other ‘glaciogenic’ sediments in the present area with the Mortensnes Fm. is based on the robust lithostratigraphic correlation of the Stappogiedde Fm. in the Vestertana area with the Dividal Group in the Andabakaoaivi-Halkavarre area (Føyn 1967) and from there to regions to the SW (Føyn 1964; Vogt 1967; Thelander 1982, and references cited above). The limited palaeontological data (Platysolenites antiquissimus Eichwald, Treptichnus pedum) support these correlations (Kulling 1964; Hamar 1967; Føyn & Glaessner 1979; Stodt 1987; Crimes & McIlroy 1999; Fig. 58.3). The outcrops documented here are often small and rather uninspiring. Some have little or no direct (reported) evidence of glacial activity and it is their stratigraphic position, underlying the distinctive quartz-conglomerate forming the base of the Dividal Group, that has been used to infer a glacial (direct or very indirect) origin. Exposures such as the polymict conglomerates near Skoganvarre (Siedlecka 1987) are clearly of this type. Similarly, the finer-grained, stratified deposits associated with diamictites in the Jerta Fm. (Skerlie & Tan 1960) were inferred to be fluvioglacial solely based on the presence of diamictites; however, these are also an unreliable indicator of glaciogenic facies. Criteria used for suggesting a glacial origin include faceted clasts, impact pits (reflecting dropstones), soft-sediment folding and faulting, taken to reflect glacial shear stresses, and the diamictitic composition and texture of the rocks. In the Komagfjord Window, Pharoah (1980) recognized small roches moutone´es forming part of the 3-m-high palaeorelief within the 40-m-deep palaeovalley. The presence of regular ‘varved’ sequences associated with diamictite has also been cited as evidence of glacial activity (Kulling 1951). Although most of these criteria are no longer regarded as being unequivocally diagnostic of glacial activity, their co-occurrence, together with their stratigraphic position, forms a reasonably compelling argument for a glacial (sensu lato) origin; whether the diamictites are tillites or slightly reworked material is unknown. Within the lowest imbricate of the Akkajaure Nappe Complex there is some evidence for a transition from coarse diamictites in the ESE to finer-grained diamictites (now schists) in the WNW. Both lithologies have comparable angular mesoperthitic clasts, with clast sizes decreasing to the west, and similar geochemistries. This has been inferred to reflect a transition from a proximal glacial to a more distal, possibly glaciomarine, depositional environment (Bjo¨rklund 1989). Kulling (1951) suggested that the Vakkejokk Breccia is a tillite, but later studies indicated a mass-flow origin (cf Stodt 1987). A peri-glacial environment has been proposed (Stro¨mberg 1981; Thelander 1982) and this seemed to be confirmed by the discovery of glendonite in the breccia (Stodt 1987). Similarly, the breccia clasts found in the Type D breccia (breccia-in-breccia structure) were interpreted to indicate that the material was frozen during transportation. Stodt (1987) also reported up to 20-cm-diameter lonestones in the Lower Siltstone Member, under the Vakkejokk Breccia, although no dropstones were seen. However, detailed examination of the ‘glendonite’ does not show any evidence for the 30% volume loss associated with the breakdown of ikaite. Instead, each radial ‘arm’ comprises a single, slightly distorted calcite crystal, with curved (strained) twin planes and radially varying (undulose) extinction orientations. Essentially, the internal microstructure seems too regular to be a replacement of ikaite and is instead thought to represent replacement of an initial gypsum nodule (Peckmann, pers. comm. 2011). The nodules are too irregular, partly due to bedding parallel pressure solution, to make a diagnostic determination of the crystal form. This non-glacial reinterpretation is consistent with the palaeomagnetic data and the lack of widespread glacial deposits in places nearer the poles at that time. The type A conglomerate in the Vakkejokk Breccia is likely a rapidly accumulated mid-fan deposit. In the overlying breccias, the parallelism of platy clasts indicates laminar flow and the

609

internal brecciation is indicative of rigid plugs, both typical of debris flows (Stodt 1987). Stodt (1987) proposed that the breccia formed due to the uplift and collapse of a palaeotopographic high; based on the asymmentry of soft-sediment folds within the breccia, this lay to the west, with east-directed slumping. Similar, but slightly younger normal faulting has been documented at the Caledonian front SW of Akkajaure (Hansen 1989). The diamictites at the northern margin of the Rombak-Sjangeli Window and in the SE corner of the Kuokkel Window also lie within, rather than under, sediments correlated with the lower part of the Tornetra¨sk Fm. (Bax 1984; Tull et al. 1985). Although it cannot be wholly discounted, it seems unlikely that these are tillites. A more likely origin is that they are debris-flows, possibly reworked nearby sub-Tornetra¨sk Fm. tills or the commonly reported regolith in the area (Stodt 1987; Bax 1989). The most northerly part of the Komagfjord Antiformal Stack, a palaeo-basement high, lay well above the Mortensnes post-glacial sea level, with drowning occurring some time after it occurred in more southerly parts of the high; glacial sediments deposited in this area would thus have been strongly affected by erosion, with a low preservation potential. The Porsavatn Diamictite Bed, lying under a sharp erosional contact with the basal conglomerates of the Lomvatn Fm. is the only potential relict of such deposits found (Fig. 58.3) and, if of glaciogenic origin (sensu lato) is most likely reworked (debris-flow), rather than primary. Kulling (1951) introduced the term Varangeristiden (Varang(er)ian ice age) in a discussion of diamictites in northern Sweden. Although these deposits are now correlated solely with the Mortensnes Fm., the term originally encompassed all the Neoproterozoic glaciogenic rocks of the Smalfjord and Mortensnes formations in E. Finnmark (and elsewhere in Scandinavia). The Smalfjord and Mortensnes formations have now been correlated with the Marinoan and Gaskiers glacial events based on, respectively, the development of a typical Marinoan cap dolostone and a close association with extreme negative d13C values taken to reflect the Wonoka anomaly (Halverson et al. 2005; Rice et al. 2011). One glacial event was worldwide in scope, while the other was much more localized. Taking a 12 Ma duration of the Marinoan glaciation (Bodiselitsch et al. 2005) and 1 Ma for the Gaskiers (Bowring et al. 2003) implies a total time span of 647– 579 Ma for the two glacial events, of which only 13 Ma (18%) were actually spent under ice, scarcely enough to justify the term ‘ice age’. Further, Harland et al. (1989) used the term for a Precambrian epoch. In view of the very confused implications given by the term, we discourage the use of the term Varang(er)ian in any sense. In summary, the many isolated, but regionally persistent subDividal Group diamictites in northern Scandinavia, occurring in the Autochthon and at all tectonic levels of the Caledonian nappes derived from Baltica, testify to a mid-Ediacaran glaciation broadly equated with the 580 Ma Gaskiers event (which was probably diachronous). Lithological, structural and mineralogical characteristics suggest terrestrial as well as glaciomarine depositional environments. These deposits covered an area of 140 000 km2 in the region considered. Applying the same structural model as used here to the whole orogen (cf Gayer & Greiling 1989; Rice 2006) indicates that patches of Mortensnes Fm. equivalents (Kumpulainen 2011; Kumpulainen & Greiling 2011; Nystuen & Lamminen 2011) cover a restored area of c. 780 000 km2. A. H. N. R. thanks A. and J. Pettersen and U. and P.-Ø. Gjøvik for many years of hospitality during fieldwork in Finnmark, and C. and R. Hofmann for assistance in the field. T.C.P.’s contribution appears with the permission of the Director, British Geological Survey (Natural Environment Research Council). D. Roberts and R. Kumpulainen are thanked for providing information. The latter and G. Shields are thanked for their reviews. G. Narbonne is thanked for confirming the identification of Treptichnus pedum in the Tornetra¨sk Fm. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

610

F. STODT ET AL.

References Anderson, M. W. 1989. Basement-cover evolution during Caledonian Orogenesis, Troms, N. Norway. In: Gayer, R. A. (ed.) The Caledonide Geology of Scandinavia. Graham & Trotman, London, 101– 110. Bax, G. 1984. Geologie des Tornehamngebietes am Westufer des Tornetra¨sk, Schwedisch-Lappland. Diploma thesis, University of Marburg. Bax, G. 1989. Caledonian structural evolution and tectonostratigraphy in the Rombak-Sjangeli Window and its covering sequences, northern Scandinavian Caledonides. Norges geologiske Undersøkelse Bulletin, 415, 87 –104. Bax, G. 2001. Application of remote sensing techniques for the geological mapping of tectonic klippen in the northern Scandinavian Caledonides. In: Buchroithner, M. F. (ed.) A Decade of Trans-European Remote Sensing Cooperation. Balkema, Rotterdam, 245– 249. Bingen, B., Griffin, W. L., Torsvik, T. H. & Saeed, A. 2005. Timing of Late Neoproterozoic glaciation on Baltica constrained by detrital zircon geochronology in the Hedmark Group, SE Norway. Terra Nova, 17, 250– 258. Bjo¨rklund, L. 1985. The Middle and Lower Allochthons in the Akkajaure-Tysfjord area, northern Scandinavian Caledonides. In: Gee, D. G. & Sturt, B. A. (eds) The Caledonide Orogen — Scandinavia and Related Areas. J. Wiley and Sons, Chichester, 515– 528. Bjo¨rklund, L. 1989. Geology of the Akkajaure-Tysfjord-Lofoten traverse, N. Scandinavian Caledonides. PhD thesis, Go¨teborg University Publ. A 59, ISSN 0348-2367. Bodiselitsch, B., Koeberl, C., Master, S. & Reimold, W. U. 2005. Estimating duration and intensity of Neoproterozoic snowball glaciations from Ir anomalies. Science, 308, 239– 242. Bowring, S., Myrow, P., Landing, E., Ramezani, J. & Grotzinger, J. 2003. Geochronological constraints on terminal Neoproterozoic events and the rise of metazoans. Geophysical Research Abstracts, 5, 13219. Brasier, M., Cowie, J. & Taylor, M. 1994. Decision on the Precambrian –Cambrian boundary stratotype. Episodes, 17, 95 –100. Brown, B. R. & Wells, M. L. 1966. A contribution to the geology of the Vassijaure-Sjangel area of Swedish Lapland. Geologiska Fo¨reningens i Stockholm Fo¨rhandlingar, 87, 527– 547. Cawood, P. A. & Pisarevsky, S. A. 2006. Was Baltica right-way-up or upside-down in the Neoprotereozoic? Journal of the Geological Society, London, 163, 753–759. Chapman, T. J., Gayer, R. A. & Williams, G. D. 1985. Structural crosssections through the Finnmark Caledonides and timing of the Finnmarkian event. In: Gee, D. G. & Sturt, B. A. (eds) The Caledonide Orogen — Scandinavia and Related Areas. J. Wiley and Sons, Chichester, 593–610. Cocks, L. R. M. & Torsvik, T. H. 2005. Baltica from the late Precambrian to mid-Palaeozoic times: the gain and loss of a terranes identity. Earth Science Reviews, 72, 39 –66. Crimes, T. P. & McIlroy, D. 1999. A biota of Ediacaran aspect from the Lower Cambrian strata on the Digermul Peninsula, Arctic Norway. Geological Magazine, 136, 633–642. Edwards, M. B. 1984. Sedimentology of the Upper Proterozoic glacial record, Vestertana Group, Finnmark, North Norway. Norges geologiske Undersøkelse Bulletin, 394, 1– 76. Føyn, S. 1964. Den tillitførende formasjonsgruppe i Alta — en jevnføring med Øst-Finnmark og med indre Finnmark. Norges geologiske Undersøkelse, 228, 139– 150. Føyn, S. 1967. Dividal-gruppen (HYOLITHUS-sonen) i Finnmarke og dens forhold til de eokambrisk-kambriske formasjoner. Norges geologiske Undersøkelse, 249, 1– 84. Føyn, S. 1985. The Late Precambrian in northern Scandinavia. In: Gee, D. G. & Sturt, B. A. (eds) The Caledonide Orogen — Scandinavia and related areas. J. Wiley and Sons, Chichester, 233– 245. Føyn, S. & Glaessner, M. F. 1979. Platysolenites, other animal fossils, and the Precambrian –Cambrian transition in Norway. Norsk Geologisk Tidsskrift, 59, 25– 46. Gayer, R. A. & Greiling, R. O. 1989. Caledonian nappe geometry in north-central Sweden and basin evolution on the Baltoscandian margin. Geological Magazine, 126, 499– 513.

Gayer, R. A. & Rice, A. H. N. 1989. Palaeogeographic reconstruction of the pre- to syn-Iapetus rifting sediments in the Caledonides of Finnmark, N. Norway. In: Gayer, R. A. (ed.) The Caledonide Geology of Scandinavia. Graham & Trotman, London, 127–139. Gayer, R. A., Rice, A. H. N., Roberts, D., Townsend, C. & Welbon, A. 1987. Restoration of the Caledonian Baltoscandian margin from balanced cross-sections: the problem of excess continental crust. Transactions of the Royal Society of Edinburgh: Earth Sciences, 78, 197– 217. Gehling, J. G., Jensen, S., Droser, M. L., Myrow, P. M. & Narbonne, G. M. 2001. Burrrowing below the basal Cambrian GSSP, Fortune Head, Newfoundland. Geological Magazine, 138, 213– 218. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Towards a Neoproterozoic composite carbon isotope record. Geological Society of America Bulletin, 117, 1181–1207. Hamar, G. 1967. Platysolenites antiquissimus Eich. (Vermes) from the Lower Cambrian of northern Norway. Norges geologiske Undersøkelse, 249, 87 – 95. Hansen, L. 1989. Age relationships between normal and thrust faults near the Caledonian front at the Vietas hydropower station, northern Sweden. In: Gayer, R. A. (ed.) The Caledonide Geology of Scandinavia. Graham & Trotman, London, 91 – 100. Harland, W. B., Armstrong, R. L., Cox, A. V., Craig, L. E., Smith, A. G. & Smith, D. G. 1989. A Geological Time Scale 1989. Cambridge University Press, Cambridge. Hartz, E. H. & Torsvik, T. H. 2002. Baltica upside down: a new plate tectonic model for Rodinia and the Iapetus Ocean. Geology, 30, 255– 258. Holmsen, P. 1956. Hyolithus-sonens basale lag i Vest-Finnmark. Norges geologiske Undersøkelse, 195, 65 –72. Holmsen, P. 1957. De eokambriske lag under hyolithussonen mellen Cˇarajavvre og Cˇaskias, Vestfinnmark. Norges geologiske Undersøkelse, 200, 47 – 50. Hossack, J. R. & Cooper, M. A. 1986. Collision Tectonics in the Scandinavian Caledonides. Geological Society of London, Special Publications, 19, 287–304. Johnson, H. D., Levell, B. K. & Siedlecki, S. 1978. Late Precambrian sedimentary rocks in East Finnmark, North Norway and their relationship to the Trollfjord–Komagelva Fault. Journal of the Geological Society, London, 135, 517–533. Kautsky, G. 1949. Eokambrische Tillitvorkommen in Norbotten, Schweden. Geologiska Fo¨reningens i Stockholm Fo¨rhandlingar, 71, 595– 603. Kirkland, C. L., Daly, S. J. & Whitehouse, M. J. 2006. Provenance and Terrane Evolution of the Kalak Nappe Complex, Norwegian Caledonides: Implications for Neoproterozoic Palaeogeography and Tectonics. Journal of Geology, 115, 21– 41. Kulling, O. 1948. Om berggrunden i Sareks randomra˚den. Geologiska Fo¨reningens i Stockholm Fo¨rhandlingar, 70, 661–672. Kulling, O. 1951. Spa˚r av Varangeristiden i Norbotten. Sveriges geologiska Underso¨kning A˚rsbok 43 (1949), C 503, 1 –45. ¨ versikt over norra Norrbottensfja¨llens kaledonbergKulling, O. 1964. O grund (with English summary). Sveriges geologiska Underso¨kning, Ba 19, 166. ¨ versikt o¨ver so¨dra Norrbottenfja¨llens KaledonbergKulling, O. 1982. O grund. Sveriges Geologiska Underso¨kning, Ba 26, 295. Kumpulainen, R. A. 2011 The Neoproterozoic Lillfja¨llet Formation, southern Swedish Caledonides. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 629– 634. Kumpulainen, R. A. & Greiling, R. O. 2011. Evidence for late Neoproterozoic glaciation in the central Scandinavian Caledonides. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 623–628. Le Guerroue´, E., Allen, P. A. & Cozzi, A. 2006. Chemostratigraphic and sedimentological framework of the largest negative carbon isotopic excursion in Earth history: the Neoproterozoic Shuram Formation (Nafun Group, Oman). Precambrian Research, 146, 68– 92. Nystuen, J. P. & Lamminen, J. T. 2011. Neoproterozoic glaciation of South Norway: from continental interior to rift and pericratonic

THE CALEDONIDES OF NW SCANDINAVIA

basins in western Baltica. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 613– 622. Pharaoh, T. C. 1980. The geological history of the Komagfjord Tectonic Window, Finnmark, Northern Norway. PhD thesis, University of Dundee. Pharaoh, T. C. 1985. The stratigraphy and sedimentology of autochthonous metasediments in the Repparfjord-Komagfjord Tectonic Window, west Finnmark. In: Gee, D. G. & Sturt, B. A. (eds) The Caledonide Orogen — Scandinavia and Related Areas. J. Wiley and Sons, Chichester, 347– 357. Pisarevsky, S. A., Murphy, J. B., Cawood, P. A. & Collins, A. S. 2008. Late Neoproterozoic and Early Cambrian palaeogeography: models and problems. In: Pankhurst, R. J., Trouw, R. A. J., De Brito Neves, B. B. & De Wit, M. J. (eds) West Canadian: Pro-Cenozoic Correlations Across the South Atlantic Region. Geological Society, London, Special Publication, 294, 9 – 31. Reading, H. G. 1965. Eocambrian and Lower Palaeozoic geology of the Digermul Peninsula, Tanafjord, Finnmark. Norges geologiske Undersøkelsei, 234, 167– 191. Rice, A. H. N. 1998. Stretching lineations and structural evolution of the Kalak Nappe Complex (Middle Allochthon) in the RepparfjordFægfjord area, Finnmark, N. Norway. Norsk Geologisk Tidsskrift, 78, 277– 289. Rice, A. H. N. 2001. Field evidence for thrusting of the basement rocks coring tectonic windows in the Scandinavian Caledonides; an insight from the Kunes Nappe, Finnmark, Norway. Norsk Geologisk Tidsskrift, 81, 321– 328. Rice, A. H. N. 2006. Quantifying the exhumation of UHP-rocks in the Western Gneiss Region, S. W. Norway: a branch-line — balanced cross-section model. Austrian Journal of Earth Sciences, 98, 2 –21. Rice, A. H. N. & Townsend, C. 1996. Correlation of the late Precambrian Ekkerøya Formation (Vadsø Group; E. Finnmark) and the Brennelvfjord Interbedded member (Porsangerfjord Group; W. Finnmark, N. Norwegian Caledonides. Norsk Geologisk Tidsskrift, 76, 55 –61. Rice, A. H. N., Edwards, M. B., Hansen, T. A., Arnaud, E. & Halverson, G. P. 2011. Glaciogenic rocks of the Smalfjord and Mortensnes Formations, Vestertana Group, E. Finnmark, Norway. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 593–602. Rice, A. H. N., Gayer, R. A., Robinson, D. & Bevins, R. E. 1989. Strikeslip restoration of the Barents Sea Caledonides Terrane, Finnmark, North Norway. Tectonics, 8, 247– 264. Roberts, D. & Fareth, E. 1974. Correlations of autochthonous stratigraphical sequences in the Alta-Reppparfjord Region, West Finnmark. Norsk Geologisk Tidsskrift, 54, 123– 129. Siedlecka, A. 1985. Development of the Upper Proterozoic sedimentary basins of the Varanger Peninsula, East Finnmark. Geological Survey of Finland Bulletin, 331, 175–185.

611

Siedlecka, A. 1987. Skoganvarre berggrunnskart 2034 4, 1:50,000. Foreløpig utgave. Norges geologiske Undersøkelse. Siedlecka, A., Lyubtsov, V. V. & Negrutsa, V. Z. 1995. Correlations between Upper Proterozoic successions in the TanafjordenVarangerfjorden Region of Varanger Peninsula, northern Norway, and on Sredni Peninsula and Kildin Island in the northern coastal area of Kola Peninsula in Russia. Norges geologiske Undersøkelse, Special Publicationi, 7, 217– 232. Siedlecka, A., Roberts, D., Nystuen, J. P. & Olovyanishnikov, V. G. 2004. Northeastern and northwestern margins of Baltica in Neoproterozoic time: evidence from the Timanian and Caledonian Orogens. In: Gee, D. G. & Pease, V. (eds) The Neoproterozoic Timanide Orogen of Eastern Baltic. Geological Society, London, Memoirs, 30, 169– 190. Skjerlie, F. J. & Tan, T. H. 1960. The geology of the Caledonides of the Reisa Valley area, Troms-Finnmark, Northern Norway. Norges geologiske Undersøkelse, 213, 175–196. Stodt, F. 1987. Sedimentologie, Spurenfossilien und Weichko¨rperMetazoan der Dividal Gruppe (Wendium-Unterkambrium) im Tornetra¨skgebiet/Nordschweden. PhD thesis, Phillips-University Marburg-Lahn. Stromberg, A. G. B. 1981. The Late Precambrian Sito tillite and the Vakkejokk breccia in the northern Swedish Caledonides. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 611– 614. Thelander, T. 1982. The Tornetra¨sk Formation of the Dividal Group, northern Swedish Caledonides. Sveriges geologiska Underso¨kning, C 789, 1– 41. Torsvik, T. H., Smethurst, M. A. et al. 1996. Continental break-up and collision in the Neoproterozoic and Palaeozoic — a tale of Baltica and Laurentia. Earth Science Reviews, 40, 229– 258. Townsend, C., Roberts, D., Rice, A. H. N. & Gayer, R. A. 1986. The Gaissa Nappe, Finnmark, North Norway: an example of a deeply eroded external imbricate zone within the Scandinavian Caledonides. Journal of Structural Geology, 8, 431– 440. Townsend, C., Rice, A. H. N. & Mackay, A. 1989. The structure and stratigraphy of the southwestern portion of the Gaissa Thrust Belt and adjacent Kalak Nappe Complex, Finnmark, N Norway. In: Gayer, R. A. (ed.) The Caledonide Geology of Scandinavia. Graham & Trotman, London, 111–126. Tull, J. F., Bartley, J. M., Hodges, K. V., Andresen, A., Steltenpohl, M. G. & White, J. M. 1985. The Caledonides in the Ofoten region (68 –698N), north Norway: key aspects of tectonic evolution. In: Gee, D. G. & Sturt, B. A. (eds) The Caledonide Orogen — Scandinavia and Related Areas. J. Wiley and Sons, Chichester, 553– 568. Vogt, T. 1967. Fjellkjedestudier in den ostlige del av Troms. Norge geologiske Undersøkelse, 248, 1 –59. Williams, D. M. 1976. A revised stratigraphy of the Gaissa Nappe, Finnmark. Norges geologiske Undersøkelse, 324, 63 –78.

Chapter 59 Neoproterozoic glaciation of South Norway: from continental interior to rift and pericratonic basins in western Baltica JOHAN P. NYSTUEN* & JARKKO T. LAMMINEN Department of Geosciences, University of Oslo, P.O. Box 1047 Blindern, NO-0316 Oslo, Norway *Corresponding author (e-mail: [email protected]) Abstract: Neoproterozoic glacial deposits of South Norway comprise the Moelv and Koppang formations. The former occurs on Baltican crystalline basement in autochthonous position at the Caledonian erosional nappe front, on basement windows and basement thrust sheets in the Caledonian nappe region, and in thick sedimentary successions in the allochthonous Hedmark and Valdres rift basins. The Koppang Formation (Fm.) occurs on top of platform carbonates in the allochthonous pericratonic Engerdalen Basin. The glacial deposits are dominated by diamictite interpreted as basal till from warm-based grounded ice, whereas stratified successions of diamictite beds, sandstone and laminated siltstone with outsized stones represent local ice-margin deposits and/or subglacially infilled water bodies, and the final glaciomarine stage. Palinspastic reconstruction of Caledonian nappe complexes carrying the glacial formations indicates that the glacial deposits were deposited over a wide area by a large western Baltoscandian ice sheet, probably during the Gaskiers (c. 580 Ma) glacial event (or events), but the age of the glaciation in South Norway needs to be better constrained.

Neoproterozoic glacial deposits in South Norway have been recognized since Holtedahl (1922) interpreted the ‘Moelv conglomerate’ by Lake Mjøsa as an ‘Eo-Cambrian’ tillite. The ‘Moelv tillite’, formally termed the Moelv Fm. (Bjørlykke et al. 1967), has since been recorded in a series of outcrops in autochthonous position and in the Caledonian Osen-Røa Nappe Complex and the Valdres Nappe Complex in the Sparagmite region of central and eastern South Norway (the term sparagmite, from Greek ‘sparagma’ meaning fragment, was introduced by Esmark (1829) for the arkosic Neoproterozoic sandstones that dominate the region). The Koppang Fm., a diamictite unit that has been correlated with the Moelv Fm., occurs in the Kvitvola Nappe Complex of the northern part of the Sparagmite region (Bjørlykke & Nystuen 1981) (Figs 59.1 & 59.2). The outcrops of both formations are scattered and generally small, from a few to several hundreds of square metres. The best exposures are present in road and railway sections and along very steep cliff sides. The largest outcrop of the Moelv Fm. measures c. 3 km in length along the western side of the mountain Fonna˚sfjellet (Fig. 59.2). In its type area, the Moelv Fm. has recently been exposed in a series of new road sections and bedrock surfaces in a residential area within the town of Moelv (Nystuen 2008). The Koppang Fm. (To¨rnebohm 1896, p. 31), being poorly exposed in the Koppang type area, is best exposed in its hypostratotype in the Engerdalen area further to the east (Nystuen 1980). The Moelv and Koppang formations were recorded during regional mapping in the 1960s–1980s (e.g. Loesche & Nickelsen 1968; Nickelsen 1974; Siedlecka et al. 1987). Sedimentological and stratigraphic studies of the Moelv Fm. were also performed during this period (Bjørlykke 1966, 1969, 1974; Englund 1966, 1973a; Løberg 1970; Bjørlykke et al. 1976; Nystuen 1976a, b, 1985; Nystuen & Sæther 1979; Nystuen & Ilebekk 1981; Sæther & Nystuen 1981; Siedlecka & Ilebekk 1982). The glaciogenic formations were shown to have been deposited in four major principal depositional settings: directly upon crystalline basement and in three palaeobasins derived from the western marginal zone of Neoproterozoic Baltica, represented by the Moelv Fm. in the Hedmark and Valdres basins and the Koppang Fm. in the Engerdalen Basin (Nystuen 1985; Siedlecka et al. 2004). These palaeobasins, together with similar allochthonous basins with Neoproterozoic glacial formations in Sweden (Kumpulainen 2011; Kumpulainen & Greiling 2011) are referred to the western Baltoscandian basins (Kumpulainen & Nystuen 1985).

In the present review of the Neoproterozoic glacial deposits in South Norway, main emphasis will be given to facies, stratigraphic and structural position and implication for depositional extent, palaeogeography, character of glaciation, correlation and age. Geochronological and geochemical data given without any references correspond to the authors’ own yet unpublished results from provenance studies.

Structural framework Structural setting of the Moelv Fm. The Moelv Fm. rests with depositional contact on crystalline basement rocks in three types of structural settings: (i) in autochthonous position at the erosional nappe front at the eastern side of lake Storsjøen, (ii) in thrust sheets within the Osen-Røa Nappe Complex east of lake Storsjøen and along the river Mistra at the northern end of lake Storsjøen and in the Valdres Nappe Complex at Ormtjernskampen and (iii) in the Snødøla, Atnsjøen, Øversjødalen and Tufsingdalen basement windows in the northern part of the Sparagmite region (Siedlecka et al. 2004) (Figs 59.2 & 59.3). Basement underlying the diamictite in the Valdres Nappe Complex consists of gabbroic rocks, whereas in the other settings the crystalline basement rocks are granitic with minor bodies of gabbro and dolerite (Siedlecka et al. 1987). In the window structures, Precambrian basement rocks and a succession of late Neoproterozoic –Ordovician rocks up to c. 130 m in thickness occur beneath the Osen-Røa and Kvitvola nappe complexes; the Moelv Fm. is the oldest unit in this cover succession. The granitic rocks of the structural windows yield U –Pb zircon ages of c. 1650 Ma, whereas similar rocks at the nappe front have been dated to 1680 Ma. The granitic rocks belong to the younger intrusives of the Transscandinavian Igneous Belt (TIB) in central southern Fennoscandia (e.g. Heim et al. 1996; Ho¨gdahl et al. 2004; Andersen et al. 2009). The window structures in the northern part of the Sparagmite region reveal the same lithology and stratigraphy as the autochthon at the nappe front. For this reason the window structures have been interpreted as domes of the underlying Baltica basement and its sedimentary cover (Nystuen 1981). Morley (1986) suggested the window structures to be allochthonous. This is supported by regional studies of similar basement windows in the Scandinavian

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 613– 622. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.59

614

J. P. NYSTUEN & J. T. LAMMINEN

Fig. 59.1. Location of outcrop area (framed area) for the Neoprotereozoic glacial Moelv and Koppang formations in South Norway, with outcrops shown in the map of Figure 59.2. Major geological crustal domains within the pre-glacial Fennoscandian basement are shown (modified from Bingen et al. 2008). The Caledonian crustal segment Western Gneiss Region (WGR) comprises the northwesternmost coastal part of South Norway. The Cambrian– Devonian Caledonian domain and small areas with Phanerozoic cover rocks in the Fennoscandian shield are not shown.

Caledonides (e.g. Gayer et al. 1987; Gayer & Greiling 1989; Gee et al. 2010). In accordance with Morley (1986) and Rice (2005), the Snødøla, Atnsjøen, Spekedalen, Øversjødalen, Tufsingdalen and other window structures are here considered allochthonous, in a tectonostratigraphic position beneath the Osen-Røa Nappe

Complex. The windows may thus represent structural elements (horsts) of large duplex structures in nappe complexes of the Lower and Middle Allochthon, with the Caledonian basal thrust as the lower bounding surface, a thrust not being exposed in the northern part of the Sparagmite region.

Fig. 59.2. Simplified geological map of the Sparagmite region in South Norway, showing outcrops of the Neoproterozoic glacial Moelv and Koppang formations. The Western Gneiss Region (WGR) mentioned in the text is located NW of the area covered by the map (Figs 59.1 & 59.4). Map modified from Nystuen (1987).

NEOPROTEROZOIC GLACIATION OF SOUTH NORWAY

615

Fig. 59.3. Representative examples of facies and stratigraphic context of the glacial Moelv and Koppang formations. The inset map covers the same area as the map of Figure 59.2, in which localities are also shown. The Moelv Fm. in the Hedmark Group is in an autochthonous position in the Osen-Røa Nappe Complex and tectonic ˚ stdalen (A ˚ s); 8, Kjølsjøberget (Kj); 9, Øvre Rendal (ØR); windows: 1, Slemdalen (Sl); 2, Osdalen (Os); 3, Sjølisand (Sj); 4, Andra˚ (An); 5, Rena; 6, Moelv; 7, A 10, Atnsjøen window (At); 11, Famphøgdene (Fa); 12, Prestkampen (Pr). Koppang Fm. in the Engerdalen Group of the Kvitvola Nappe Complex: 13, Hyllera˚sen (Hy); 14, Sjoa (Sjo). Equivalent to the Moelv Fm. in the Valdres Group of the Valdres Nappe Complex: 15, Ormtjernskampen East (OrE); 16, Ormtjernskampen West (OrW); 17, Mellene (Mel).

The second major structural setting of the Moelv Fm. is within the thick rift basin successions of the Hedmark and Valdres basins of the Osen-Røa and Valdres nappe complexes, respectively (Fig. 59.2). The rift character of these allochthonous palaeobasins is reflected by at least 5000 –6000-m-thick successions with large lateral facies variations, from marginal alluvial or deep-marine conglomerate fans to central basinal facies of either deep-marine sandstone and shale or alluvial plain sandstone, and in the Hedmark Basin also by tholeiitic basalt flows. The Moelv Fm. occurs both in marginal and distal positions of the basin successions (Siedlecka et al. 2004). In most settings where the Moelv Fm. rests on crystalline basement, the primary features are well preserved with little distortion of clasts and matrix of the diamictite, except close to overlying thrust boundaries. In the Hedmark and Valdres basin successions, deformation and metamorphism generally increase from south to north. In high-strain zones along thrusts and shear zones, clasts in the diamictite can be flattened and elongated parallel to foliation in the matrix. The metamorphism is of lower greenschist facies. Non-deformed diamictite is, however, very common in most parts of the outcrop area of the Moelv Fm.

basin, the Engerdalen Basin (Siedlecka et al. 2004). In this basin, sandstone, carbonate and diamictite formations have a wide lateral extent. The Kvitvola Nappe Complex is located above the Osen-Røa Nappe Complex and beneath the Upper Allochthonous Trondheim Nappe Complex (Fig. 59.2). Compared to the Moelv Fm., the Koppang Fm. is generally more deformed. The formation is located very close to the thrust boundary of the Kvitvola Nappe Complex. Together with the underlying carbonate-shale formation, the diamictite of the Koppang Fm. has acted as a detachment horizon during the Caledonian thrust movements. In some areas, stretching and flattening of clasts makes the identification of clast lithologies very difficult; in some thrust zones the diamictite lithology is camouflaged by intense shear deformation. The metamorphism is of greenschist facies with chlorite and pale mica formed in the matrix. The diamictite of the Koppang Fm. is best preserved in the eastern part of the Kvitvola Nappe Complex in the Engerdalen district (Nystuen 1980) (Figs 59.2 & 59.3).

Stratigraphy and sedimentology of the Moelv Fm. Moelv Fm. deposited directly on Precambrian basement

Structural setting of the Koppang Fm. The Koppang Fm., occurring in the Kvitvola Nappe Complex, is interpreted to have been deposited in a wide pericratonic shelf

The Moelv Fm. is developed as brown to grey diamictite in all localities where the formation rests directly on granitic basement rocks, in autochthonous position, in basement thrust sheets or in

616

J. P. NYSTUEN & J. T. LAMMINEN

structural windows (Figs 59.2 & 59.3). In these settings, the diamictite unit is up to c. 20 m in thickness, structureless, unsorted and with clasts up to about 1 m in diameter. The matrix is mud-rich sandstone. The clasts are dominated by different types of granitic rocks and rhyolitic porphyries; grey and red quartzite and sandstone, dolerite and vein quartz are also common (Nystuen & Sæther 1979; Nystuen & Ilebekk 1981). On the northern side of the Atnsjøen window, the 20-m-thick diamictite is dominated by granitic clasts of the same type as in the basement below; however, gabbro, which is also a prominent basement rock in this window structure, is not recorded among the clast population (Siedlecka & Ilebekk 1982). The diamictite resting on gabbro in Ormtjernskampen in the Valdres Nappe Complex (Ormtjernskampen West, Fig. 59.2) is dominated by gabbro clasts, but granite and quartzite clasts also occur in this 2-m-thick diamictite that is succeeded by laminated siltstone and quartz arenite formations correlated with the Ekre and Vangsa˚s formations, respectively, in the Hedmark Group (Nickelsen 1974; Nickelsen et al. 1985). A conglomerate rests on gabbro basement in the Feforkampen outlier of the Valdres Nappe Complex (Fig. 59.2). According to

Englund (1973a), the gabbroic basement rock is overlain by a thin quartz- and quartzite pebble conglomerate, followed by 5– 8-m-thick diamictite with clasts up to 1 m across of quartz diorite, granite and anorthosite. The present authors visited the Feforkampen locality during the summer of 2008 and found this unit to be a matrix- to clast-supported conglomerate dominated by quartzite boulders of up to c. 1.5 m. The basement –diamictite boundary in outcrops east of Lake Storsjøen and in the windows in the northern part of the Sparagmite region is a smooth unconformity with relief of up to some few metres. Parallel striations with direction from ESE to WNW are present on the ‘polished’ granite surface below the diamictite in the hill at Andra˚ east of Lake Storsjøen. The diamictite is preserved as erosional remnants, in thickness varying from some few centimetres (Tufsingdalen) to 5–20 m (Atnsjøen, Spekedalen and Øversjødalen) (Figs 59.2 & 59.3). At some localities, diamictite is overlain by lower Cambrian quartz arenite (Vangsa˚s Fm.) and greenish grey shale, followed by Middle Cambrian black alum shale. The uniform lithology of the diamictite, the polished and striated crystalline substratum and the wide regional extent of the

Fig. 59.4. Palaeogeographical reconstruction of South Norway at the time of the Neoproterozoic glaciation, showing the Western Baltoscandian Ice Sheet and the Hedmark and Valdres rift basins and the pericratonic Engerdalen Basin, together with the present outcrop area of the Moelv and Koppang glacial formations in the Sparagmite region. Note the reference locality of autochthonous diamictite of the Moelv Fm. at the Baltica basement east of Lake Storsjøen. The TIB domain and the Gothian and Sveconorwegian domain are explained in terms of geological age in Figure 59.1. The Western Gneiss Region (WGR) also discussed in the text comprises the present northwestern coastal regions of South Norway. The Engerdalen Basin was located west of the pre-Caledonian position of WGR.

NEOPROTEROZOIC GLACIATION OF SOUTH NORWAY

diamictite unit are features favouring an origin as basal till of a continental ice sheet. The clast assemblage is dominated by granitic and porphyritic rock types common for the TIB and associated sandstones, quartzites and dolerites to the east and NE of the present outcrop region for the Moelv diamictites. In the Valdres Nappe Complex the grounded ice sheet has picked up local bedrock types. The clast-supported conglomerate resting on top of the gabbro in Feforkampen is lithologically quite different from the other diamictites and a glacial origin is here considered uncertain.

Moelv Fm. in the Hedmark Basin The Moelv Fm. rests with depositional contact within the Hedmark Group on (1) coarse-grained fan-delta sandstones of the Ring Fm., (2) carbonate platform and shale beds of the Biri Fm., (3) fluvial sandstones of the Rendalen Fm., (4) shallow-marine sandstones of the Atna Fm. and (5) basalt flows of the Svarttjørnkampen Fm. (Figs 59.2 & 59.3). Except for some deep marine sub-basins, most of the Hedmark Basin was filled up before the onset of the glaciation, in the east with at least 4000-m-thick alluvial conglomerates and fluvial sandstones (Rendalen Fm.) and in the west by more than 4000 m of marine turbidites and black shale (Brøttum Fm.), submarine conglomerate fans (Biskopa˚sen Fm.) and coarsegrained fan-delta accumulations (Ring Fm.). Local basalt flows of the Svarttjørnkampen Fm. on top of the Rendalen Fm. and within and on top of the Atna Fm. in the northern and eastern part of the basin are interpreted as being related to fissure eruptions (Nystuen 1982, 1987; Furnes et al. 1983). Basalt clasts in marine canyon infill conglomerates of the Biskopa˚sen Formation located at the top of the Brøttum Fm. in the western part of the basin indicate that local volcanic eruptions may have taken place repeatedly at several places in the rift basin (Siedlecka et al. 2004). The pre-glacial basin infill culminated with the formation of carbonate platforms, and black mud in adjacent deep sub-basins (Biri Fm.). Acritarchs in the upper part of the Brøttum Fm. and in the Biri Fm. indicate a late Riphean to early Vendian age (Cryogenian to Ediacaran) (Vidal & Nystuen 1990). The lower boundary of the Moelv Fm. is a distinct erosional unconformity where the formation’s basal lithology is diamictite. A sand-filled wedge in underlying fluvial Rendalen sandstone is interpreted as a fossil ice-wedge (Nystuen 1976b). At localities where the formation rests on thick marine shales of the Biri Fm., the lower boundary appears gradual from shale to laminated grey silt-rich mudstone with scattered outsized stones of the Moelv Fm. The Moelv Fm. usually varies in thickness from 1 to 2 m up to 150 m. The laminated facies with outsized stones overlying the Biri shale is thinnest. The most frequently encountered thickness is in the order of 10 –15 m, and the dominant lithology is brownish, greenish or grey diamictite. In many localities, the diamictite passes upwards into laminated mudstone with outsized stones, a facies interval that may be up to c. 2 m in thickness. An upwards decrease in clast content marks the transition of this facies into the overlying Ekre Fm., a laminated green or red siltstone without any outsized stones. The thickness increases to 100–150 m in Fonna˚sfjell, Slemdalen and Brenna˚sen (Figs 59.2 & 59.3). In the eastern hillside of the mountain Fonna˚sfjell, the formation consists of structureless to faintly stratified light grey diamictite with clasts up to c. 1 m in diameter. In Slemdalen and Brenna˚sen, the c. 150-m-thick Moelv Fm. consists of structureless and poorly stratified brown diamictite beds, stratified sandstone beds with scattered pebbles and boulders, and laminated siltstone with outsized stones up to c. 1 m in diameter. The clast (larger than 2 cm) content of the diamictite facies ranges from c. 7.5 to 30% (Nystuen 1976a). The extrabasinal clast suite in eastern and northern outcrops is very similar to that of the Moelv diamictite located directly on crystalline basement

617

rocks. Recent radiometric datings and isotopic studies (U – Pb and Lu –Hf in zircons) of the igneous clast material have confirmed that the TIB, or its equivalents in terms of age and petrogenetic history, contributed significantly to the clastic material in the eastern Hedmark Basin with ages ranging from 1770 Ma to 1010 Ma, most being c. 1670 Ma. Only one clast with an age of c. 1540 Ma has been found, which means that clasts of Late ˚ ha¨ll & Connelly 2008) age are Gothian (c. 1590 –1520 Ma; A rare in the Moelv Fm. In the western part of the Hedmark Basin, the clast assemblage also includes fine-grained quartzite, grey and pale granite, gneisses and amphibolite (Løberg 1970; Englund 1973a). Many Sveconorwegian (Grenvillian) c.960 Ma granitoid clasts occur in the western and southwestern parts of the basin (Lamminen et al. 2009). Clasts corresponding to typical rock types dominating in the Gothian-Sveconorwegian basement adjacent to the southernmost outcrops of the Moelv Fm., such as coarse-grained red granodiorites, various types of gneisses, amphibolites and gabbros (Nordgulen 1999), are not present in the diamictite. Most clasts of crystalline rock types are angular to sub-rounded, whereas sandstone, quartzite and vein quartz clasts of pebble size are skewed towards the sub-rounded to well-rounded range (Nystuen 1976a). The lithology of formations in the Hedmark Group directly underlying the Moelv Fm. is reflected in the composition of the diamictite close to the lower boundary. The matrix is enriched in coarse-grained sand particles, granules and pebbles of quartz and feldspar when resting upon the coarse-clastic Rendalen and Ring formations. The basal zone of the diamictite carries small clasts of limestone and basalt at sites where Biri limestone and basalt are the underlying stratigraphic units, respectively. Limestone clasts are also abundant at localities where limestone is not directly underlying the diamictite, as at the type locality in Moelv. This may be due to the presence of a carbonate platform of the Biri Fm. that was located along the western margin of the Hedmark Basin at the time when the diamictite was formed (Løberg 1970; Bjørlykke et al. 1976). Clasts of siltstone and sandstone show glacial striations (Bjørlykke 1974; Nystuen & Sæther 1979). The diamictite deposits of the Moelv Fm. are unsorted to poorly sorted and display grain size curves very similar to those of Pleistocene diamictons. Fabric studies in the eastern part of the Hedmark Basin reveal an east – west-oriented peak of preferred long-axis orientation that has been inferred to represent an original depositional fabric formed by subglacial westward transport of debris (Nystuen 1976a). The structureless diamictite is interpreted as tillite sensu stricto, formed as basal till beneath a grounded ice sheet, as also suggested for the similar type of diamictite in settings directly on basement. The siliciclastic beds in the Rendalen and Ring formations, directly underlying the Moelv Fm., may have been poorly consolidated at the time of glaciation, thus supplying clastic grains to the till matrix. Successions consisting of structureless and faintly stratified diamictite beds, stratified sandstone and laminated siltstone with outsized stones have been interpreted as ice-margin deposits, formed at sites where water depth was close to the buoyancy depth of the ice sheet (Nystuen 1976a). Another possible depositional environment for these localized stratified successions may have been small subglacial water bodies. Laminated siltstone with outsized stones within successions of mixed lithology at the top towards the overlying Ekre Fm. is thought to represent glacial mud with icedropped debris formed in an iceberg zone. This facies probably formed during the final stage of glaciation, maybe during eustatic sea-level rise.

Moelv Fm. in the Valdres Basin In the Valdres Basin, the best outcrops of diamictite correlated with the Moelv Fm. are in the Ormtjernskampen area (Figs 59.2 &

618

J. P. NYSTUEN & J. T. LAMMINEN

59.3). In addition to the diamictite resting on gabbroic basement rock, described above from the outcrop in the western part of the Ormtjernskampen area, a phyllitic diamictite unit also occurs in the eastern part of the area on top of the 800-m-thick Ormtjernskampen Fm. of alluvial conglomerate and below a sandstone unit correlated with the Vangsa˚s Fm. in the Hedmark Group (Nickelsen 1974; Nickelsen et al. 1985) (Ormtjernskampen East, Fig. 59.2). This diamictite contains the same clast lithologies as the diamictite resting on the gabbroic crystalline basement in the outcrop Ormtjernskampen West (Fig. 59.2). At Mellene (Figs 59.2 & 59.3), a 0.5–3-m-thick diamictite, correlated with the Moelv Fm., occurs on top of the 3000-m-thick Olefjell Fm. of arkosic sandstone. As with the Moelv Fm. in the Hedmark Basin, the diamictite is succeeded by siltstone and sandstone, possible correlatives to the Ekre and Vangsa˚s formations, and Cambro-Ordovician strata (Loeschke & Nickelsen 1968). The diamictite unit recorded within the rift basin succession in the Valdres Basin reveals similar lithologies and stratigraphic characteristics to those of the Moelv Fm. in the Hedmark Basin, which supports the interpretation of this diamictite unit as a glacial deposit. The stratigraphic position of the Moelv Fm. in the Valdres Basin reflects the geological setting of glaciation in a continental rift basin with crystalline basement forming the basin margins, together with coarse-clastic marginal alluvial fans, and very thick fluvial accumulations in the central part, all settings in which glacial diamictite was deposited. Thus, in the Ormtjernskampen area, the glacial diamictite was deposited on top of an alluvial fan succession and on crystalline basement, and these two sites were likely separated by a fault. In the Mellene area, the glacial formation was deposited in a more central position of the rift basin.

extent and setting within a pericratonic basin (Nystuen 1980; Kumpulainen & Nystuen 1985; Siedlecka et al. 2004). The facies association in the underlying Hyllera˚sen Fm. implies that this carbonate unit was formed during warm-water to tropical conditions in a carbonate platform setting including evaporitic lagoonal sub-environments favourable for magnesite formation, anoxic sub-basins and sabkhas.

Chemostratigraphy No data are available from the glacial units themselves. The distribution of major and some trace elements through the Hedmark Group and the overlying Cambro –Silurian succession was presented by Englund (1973b) and Bjørlykke & Englund (1979). The stratigraphic resolution for the interval containing the Moelv Fm. is too low for chemostratigraphic correlation. Tucker (1983), in his study of the Biri Fm., published stable isotope analyses with an average d18O of –11.4‰ and d13C of þ1.4‰ relative to PDB, indicating that the limestone was formed primarily from aragonite in seawater of normal composition. The data do not give any stratigraphic trend that can be correlated by isotope curves obtained from other Neoproterozoic carbonate successions.

Palaeolatitude No data are available.

Palaeogeography Stratigraphy and sedimentology of the Koppang Fm. The Koppang Fm. occurs in the lower part of the Engerdalen Group in the Kvitvola Nappe Complex. The best outcrops of the succession are in the Engerdalen area (Fig. 59.2). The formation is a 5–15-m-thick diamictite and rests with primary depositional contact on a carbonate unit, the Hyllera˚sen Fm. (Fig. 59.3). The Hyllera˚sen Fm. is 15–30 m thick and consists predominantly of dolomite with thin siltstone laminae, but also contains magnesite with chert nodules (Nystuen 1969), black shale and laminated fine-grained sandstone, locally with dolomite pseudomorphs after gypsum crystals (Nystuen 1980). A unit of shallowmarine sandstone locally occurs below the carbonate formation. The diamictite in the Koppang Fm. is structureless, except close to its upper boundary at some localities, where it may be weakly laminated. Its lower boundary is erosional, and the diamictite is enriched with dolomite clasts (up to 90% in the basal zone), with the largest clast having a diameter of 80 cm. Extrabasinal clasts are sub-angular to sub-rounded red and grey granite, with boulders up to 1 m in diameter, rhyolitic porphyry, red sandstone, light grey quartzite and vein quartz pebbles. The clast assemblage resembles that of the Moelv Fm. in the eastern part of the Hedmark Basin. The diamictite is overlain by medium-grained, well-sorted sandstone inferred to be of shallow-marine origin (Nystuen 1980). The upper contact towards the overlying sandstone formation is lithologically sharp and often tectonically deformed. The carbonate–diamictite–sandstone succession can be traced along the sole of the Kvitvola Nappe Complex for c.140 km from west to east (Figs 59.2 & 59.3). The Hyllera˚sen Fm. at its base is commonly cut by the sole thrust of the nappe complex and has obviously acted as a detachment horizon during emplacement of the Kvitvola Nappe Complex. The carbonate–diamictite–sandstone tripartite succession is frequently strongly tectonized and deformed. A glacial origin is inferred for the Koppang Fm. from its diamictite lithology and clast content, erosional lower boundary with clasts derived from underlying formation, its laterally wide

Pre-Caledonian position of the Hedmark, Valdres and Engerdalen basins The Hedmark and Valdres rift basins and the pericratonic Engerdalen Basin were formed at the western Baltoscandian margin (present coordinates) (Siedlecka et al. 2004; Nystuen et al. 2008). Their pre-Caledonian structural setting and relative position to the Baltoscandian craton is crucial for the interpretation of the extent and character of the Neoproterozoic glaciation of western Baltica. Palaeogeography of the glaciated and depositional area during the Neoproterozoic glaciation in South Norway can first be inferred after palinspastic restoration of the allochthonous successions in the Osen-Røa, Kvitvola and Valdres nappe complexes. Oftedahl (1943) first quantified the shortening of the southernmost part of the Osen-Røa Nappe Complex, represented by the Cambro –Silurian decollement succession now preserved in the Oslo Graben, and concluded that the succession in the northern part of the lake Mjøsa area had been dislocated c. 150 km to the SSE during the Caledonian orogeny. Nystuen (1981) showed that the Osen-Røa Nappe Complex, riding on top of the window structures in northern part of the Sparagmite region, had to be restored to a position NW of the windows, giving a minimum thrust distance of Osen-Røa Nappe Complex rocks at the erosional nappe front within the Mjøsa-Trysil sector (Fig. 59.2) of c. 140 km. Morley (1986) restored hanging- and footwall cut-offs within the Osen-Røa Nappe Complex and obtained a minimum thrust distance of 130 km of rocks in the nappe complex at the northern part of Lake Mjøsa, about the same shortening as quantified by Oftedahl (1943). Morley (1986) restored the window structures to a position NW of the restored position of the Hedmark Basin, outside the present coastline of mid-Norway. This restoration model was further developed by Rice (2005), who included the window structures in a hypothetical ‘Jostedal platform’ of Baltica basement northwestward of the Hedmark Basin. However, the thrust position of the Osen-Røa Nappe Complex

NEOPROTEROZOIC GLACIATION OF SOUTH NORWAY

above the structural windows and their sedimentary cover succession shows that the windows must be restored to a position cratonward of the Hedmark Basin, and not the opposite as proposed in the models of Morley (1986) and Rice (2005). Hossack et al. (1985) restored the package of thrust sheets in the Valdres area to the NW by c. 315 km, west and NW of the original position of the Hedmark Basin. The Kvitvola Nappe Complex with the Engerdalen Basin, located on top of the Osen-Røa Nappe Complex, has been displaced an even greater distance, probably in the order of 400 –600 km, from the NW to the SE (Siedlecka et al. 2004). The Caledonian nappe pile in South Norway is located on top of the Western Gneiss Region of the southern Scandinavian Caledonides (Fig. 59.1). Consequently, the nappe complexes in discussion were likely derived from positions west of the Western Gneiss Region. In the gneiss region, Precambrian basement rocks with overlying and infolded Neoproterozoic to lower Palaeozoic sedimentary cover successions and Caledonian thrust sheets are strongly deformed and metamorphosed. Rice (2005) restored the rocks of the Western Gneiss Region to the NW as one single crustal block close to the position of the Valdres Basin and west of the Hedmark Basin. However, the heterogeneous structural architecture of the Western Gneiss Region indicates that the displacement history of this region and its cover of sedimentary successions and thrust sheets are very complicated (Tucker et al. 2004), and accordingly difficult to restore palinspastically. It is beyond the scope of this paper to discuss various palinspastic models of the southern Scandinavian Caledonides. However, irrespective of models, it must be concluded that during Neoproterozoic times when the Moelv and Koppang formations were deposited, a rather well-denudated Baltica craton extended several hundred kilometres further west of the outcrop east of lake Storsjøen where the Moelv diamictite is located autochthonously on striated basement rocks, a reference locality of crucial importance for any reconstruction of the palaeogeography of Neoproterozoic glaciation in South Norway. The Hedmark and Valdres rift basins and the Engerdalen shelf basin were located in outboard positions of the denuded Baltica.

Geochronological constraints No radiometric age data are available to unequivocally constrain the ages of the Moelv and Koppang formations. Rankama (1973) reported a Rb – Sr whole-rock age of 612 + 18 Ma on an argillaceous sample from the Ekre Fm., overlying the Moelv Fm. in the western part of the Hedmark Basin. Bingen et al. (2005) identified a 677 + 15 to 620 + 14 Ma old detrital zircon U – Pb age population in the Rendalen Fm. at Hanestad in the eastern part of the Hedmark Basin (Fig. 59.2). This is a very uncommon age range within Fennoscandian crystalline rocks, and the source for these zircon grains remains enigmatic. Within the range of this age population, a granite in the allochthonous Seve Nappe in northern Sweden gave a U –Pb titanite age of 637 + 3 Ma; the granite was considered to be related to terrane accretion in the westernmost part of Baltica during the Timanian orogeny (Rehnstro¨m et al. 2002). Recent detrital zircon U –Pb analyses from the uppermost parts of the Brøttum Fm. have yielded a wide range of Neoproterozoic ages. The same c. 620 Ma population also shows up in the Brøttum Fm. Because the Brøttum Fm. underlies the Moelv Fm., this confirms the findings of Bingen et al. (2005) that the Moelv Fm. appears to be younger than c. 620 Ma. Hannah et al. (2007) reported a c. 560 Ma Re –Os age for a black shale in the Biri Fm., beneath the Moelv Fm. in the Mjøsa area. However, the same authors also reported a c. 300 Ma age for the same formation at another locality further NE. The latter age was explained by thermal disturbance in the Re – Os isotopic system due to the Permian Oslo rift. These authors later investigated scatter in their Re –Os data and related it to oxidation

619

processes (Hannah et al. 2008). Because the Re –Os system is susceptible to disturbances that are difficult to detect, unlike the U – Pb system in zircon, the present authors think that a robust age for the Moelv Fm. still remains to be established (see below).

Discussion The western Baltoscandian ice sheet According to the tectonostratigraphic setting of the Moelv and Koppang glacial formations and the palinspastic restoration of the Caledonian nappe complexes carrying the Neoproterozoic basins (see above), the area covered with glacial ice must have had a width of at least 400–600 km from SE to NW, measured from the autochthonous reference outcrop of the Moelv Fm. at Storsjøen to the restored position of the Engerdalen Basin at the Baltoscandian margin (Figs 59.2 –59.5). This western Baltoscandian ice sheet also covered carbonate platforms of the Risba¨ck rift basin and the pericratonic Tossa˚sfja¨llet Basin along the Baltoscandian margin (Kumpulainen 2011; Kumpulainen & Greiling 2011). Clast lithologies of the Moelv and Engerdalen formations indicate that the glacial ice streams were sourced from areas located within the TIB and the Sveconorwegian domain in the central, and the western to south-western part of Fennoscandia, respectively. The Hedmark and Valdres rift basins probably formed along zones of weakness between these crustal blocks (Lamminen et al. 2009). Uplifted shoulders of the rift basins may have given rise to locally glaciated mountain ridges (Figs 59.4 & 59.5). Ice streams that originated in highland areas must have coalesced into extensive ice sheets that moved across lowland regions, now represented by denuded Baltica basement and the carbonate platform of the Engerdalen Basin, both types of substratum being covered by glaciogenic diamictite. Striations on the granitic surface beneath diamictite east of lake Storsjøen demonstrate the warm-based character of the glacier ice at the time. Lithoclasts and the textural and mineralogical composition of basal tills are generally dominated by the local substratum. This is also revealed by the concentration of carbonate clasts in the lower part of the Koppang Fm. where the diamictite rests directly on dolomite, and by basalt clasts in diamictite of the Moelv Fm. at sites where this formation has been deposited upon basalt of the Svarttjørnkampen Fm. (Figs 59.2 & 59.3). The absence of Gothian gneisses, which are typical in the autochthonous basement, among clasts in the Moelv Fm. at Mjøsa is thus in accordance with the allochthonous position of the glacial formation in the area. The sedimentological character of the carbonate units underlying the Moelv and Koppang formations, the Biri and Hyllera˚sen formations, respectively, indicates dominating warm and arid climatic conditions before the onset of glaciation. The presence of fossil ice wedges in alluvial sediments beneath the Moelv Fm. shows that periglacial climatic conditions prevailed in the area in front of the advancing ice sheet.

Correlation and age The Moelv and Koppang formations are generally correlated with the Mortensnes Fm. in Finnmark, the younger of the two Neoproterozoic glacial units traditionally referred to the ‘Varangerian glacial period’ in northern Norway, and to the La˚ngmarkberg and Lillfja¨llet glacial formations in Sweden (e.g. Bjørlykke & Nystuen 1981; Kumpulainen & Nystuen 1985; Siedlecka et al. 2004; Nystuen et al. 2008). Halverson et al. (2005) correlated the Mortensnes Fm. with the Gaskiers glaciation and the older glacial Smalfjord Fm. with the Marinoan glaciation, from d13C profiles obtained from carbonate deposits associated with the two formations. The Marinoan glacial units are dated to

620

J. P. NYSTUEN & J. T. LAMMINEN

Fig. 59.5. Simplified and composite section across the western Baltoscandian craton and the Hedmark and Valdres rift basins and the pericratonic Engerdalen Basin, showing the Western Baltoscandian Ice Sheet from the cratonic interior in the SE to the Baltoscandian Margin in the NW, with local areas of glaciation along mountain ridgers along the rift basins. Note the reference locality of autochthonous diamictite of the Moelv Fm. at the Baltica basement east of Lake Storsjøen.

c. 635 Ma (Hoffmann et al. 2004) and the Gaskiers and Squantum glacial deposits of Avalonia to c. 580 Ma (Myrow & Kaufman 1999; Thompson & Bowring 2000; Bowring et al. 2003; Thompson et al. 2007). The Marinoan glacial episode may have lasted at least 3 million, and most likely 12 million, years (Bodiselitsch et al. 2005). Duration and variation in age for the Gaskiers glacial event or events are also very uncertain. On the basis of d13C values in cap carbonates of the youngest of two diamictite units in Svalbard, Halverson et al. (2004) correlated the Wilsonbreen Fm. with the Marinoan glacial event and the inferred global extent of this glaciation, in accordance with the Snowball Earth hypothesis (Hoffman & Schrag 2002). In the British-Irish Caledonides, three stratigraphic levels of diamictite have been correlated as equivalents to the c. 700 Ma Sturtian, c. 635 Ma Marinoan and c. 580 Ma Gaskiers glacials, respectively (McCay et al. 2006). Bingen et al. (2005) suggested that the Moelv Fm. corresponds to the Gaskiers glacial event, constrained by the 620 + 14 Ma detrital zircon age obtained from the Rendalen Fm. at Hanestad. The Lillfja¨llet Fm. in the Tossa˚sfja¨llet Basin, Sweden, an alleged correlative to the Moelv Fm. (Nystuen et al. 2008), is penetrated by the Ottfja¨llet dolerite dyke swarm (Kumpulainen 1980, 2011). This dyke swarm is generally considered to be the result of preIapetus rifting at the western Baltoscandian craton, contemporaneous with the intrusion of the 616 + 3 Ma Egersund dolerite dyke swarm (Bingen et al. 1998) in southwestern Norway (e.g. Siedlecka et al. 2004; Nystuen et al. 2008). There are presently several data and arguments favouring a mid-Ediacaran (possible Gaskiers equivalent) age for Neoproterozoic glaciation in South Norway and Sweden, represented by glacial diamictite formations. The glacial formations are located in the uppermost part of Neoproterozoic successions, with the Moelv Fm. above a stratigraphic level in the uppermost part of the Brøttum Fm. that contains c. 620 Ma zircon grains, and are overlain by Ediacaran and/or Lower Cambrian formations. The glacial deposits also lack cap carbonates, typical of most Gaskiers glacial deposits according to Halverson et al. (2005). If all glacial diamictite units in central western Sweden and in South Norway are Gaskiers equivalents, then deposits of the allegedly global end-Cryogenian or Marinoan glacial period appear not to be represented within these western Baltoscandian

basin successions, at least not as glacial diamictites. In this connection, it should be emphasized that the various glacial formations in South Norway, Finnmark of North Norway, and Sweden are lithostratigraphically, not chronostratigraphically, correlated. A Gaskiers age is not presently unequivocally constrained for the formations, neither by radiometric age determinations, nor by d13C studies on carbonate units beneath the glacial units. If the Ottfja¨llet dolerite dykes were shown to have about the same age as the Egersund dyke swarm (c. 616 Ma), this would put the age of the glacial Lillfja¨llet Fm., and maybe also the Moelv Fm., to be older than Gaskiers. It should also be emphasized that the detrital zircon U –Pb age of c. 620 Ma is problematic in the sense that no likely source of these zircons is known, and this age appears to belong to a zircon U – Pb age population with a wide time range (Bingen et al. 2005). The meaning of this is not known. Improved constraining of the depositional age for Neoproterozoic glacial formations in central and southern Scandinavia may be obtained by radiometric age determination of the Svarttjørnkampen basalt directly underlying the Moelv Fm. and the Ottfja¨llet dolerites cutting the Lillfja¨llet Fm. Such studies are in progress. In addition, d13C studies of carbonate formations underlying the glacial formations (cf. Siedlecka et al. 2004; Nystuen et al. 2008) should be performed.

Conclusions The Neoproterozoic Moelv and Koppang glacial formations were mainly formed as basal till from a warm-based, grounded, western Baltoscandian ice sheet that covered large parts of the northwestern Baltica, and the sedimentary Hedmark, Engerdalen and Valdres basins. Accumulation areas of the glacial ice were in the zone of the TIB in the east and NE, and highland areas dominated by rocks with metamorphic and igneous ages corresponding to the Gothian and the Sveconorwegian (Grenvillian) orogeneses in the SE and SW. Local ice-marginal and/or subglacial waterbody successions were formed in deeper parts of the rift basins. Laminated mudstone with ice-dropped stones originated towards the end of the glaciation due to a rise in sea level and the change from grounded to floating ice sheet margins and icebergs. The glaciation may correspond to the c. 580 Ma Gaskiers glacial event of

NEOPROTEROZOIC GLACIATION OF SOUTH NORWAY

Avalonia, but this interpretation still needs to be better constrained by radiometric age determinations of igneous rocks associated with the glacial successions and d13C studies on carbonate formations underlying the glacial formations. We thank R. Ba¨ckmark and H. B. Totland for making drawings (Figs 59.2, 59.3 & 59.5). The manuscript has benefited from the criticism of an earlier version by H. Rice and by corrections of the English text by A. Read. We also thank editor G. Shields for corrections and suggestions for improvements to this contribution. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References ˚ ha¨ll, K.-I. & Connelly, J. N. 2008. Long-term convergence along A SW Fennoscandia: 330 m.y. of Proterozoic crustal growth. Precambrian Research, 163, 402–421. ˚ berg, G. & Simonsen, Andersen, T., Andersson, U. B., Graham, S., A S. 2009. Granitic magmatism by melting juvenile continental crust: new constraints on the source of Palaeoproterozoic granitoids in Fennoscandia from Hf isotopes in zircon. Journal of Geological Society, London, 166, 1 –15. Bodiselitsch, B., Koeberl, C., Master, S. & Reimold, W. U. 2005. Estimating duration and intensity of Neoproterozoic snowball glaciations from Ir anomalies. Science, 308, 239–242. Bingen, B., Demaiffe, D. & Van Breemen, O. 1998. The 616 Ma old Egersund basalt dike swarm, SW Norway, in the context of lateNeoproterozoic opening of the Iapetus Ocean. Norsk geologisk tidsskrift, 79, 69 – 86. Bingen, B., Griffin, W. L., Torsvik, T. H. & Saeed, A. 2005. Timing of Late Neoproterozoic glaciation on Baltica constrained by detrital zircon in the Hedmark Group, south-east Norway. Terra Nova, 17, 250– 258. Bingen, B., Andersson, J., So¨derlund, U. & Mo¨ller, C. 2008. The Mesoproterozoic in the Nordic countries. Episode, 31, 29 –34. Bjørlykke, K. 1966. Studies on the Latest Precambrian and Eocambrian rocks in Norway. 1. Sedimentary petrology of the Sparagmites of the Rena district, S. Norway. Norges geologiske undersøkelse, 238, 5 – 53. ¨ sterdalen. Norsk geoBjørlykke, K. 1969. Geologien i sentrale deler av O logisk tidsskrift, 49, 313– 318. Bjørlykke, K. 1974. Glacial striations on clast from the Moelv Tillite of the Late Precambrian of Southern Norway. American Journal of Science, 274, 443– 448. Bjørlykke, K. & Englund, J. O. 1979. Geochemical response to upper Precambrian rift basin sedimentation and lower Palaeozoic epicontinental sedimentation in South Norway. Chemical Geology, 27, 271– 295. Bjørlykke, K. & Nystuen, J. P. 1981. Late Precambrian tillites of South Norway. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 624–628. Bjørlykke, K., Englund, J.-O. & Kirkhusmo, L. 1967. Latest Precambrian and Eocambrian stratigraphy of Norway. Norges geologiske undersøkelse, 251, 5– 17. Bjørlykke, K., Elvsborg, A. & Høy, T. 1976. Late Precambrian sedimentation in the central sparagmite basin of South Norway. Norsk geologisk tidsskrift, 56, 233–290. Bowring, S. A., Myrow, P. M., Landing, E., Ramezani, J. & Grotzinger, J. 2003. Geochronological constraints on terminal Neoproterozoic events and the rise of metazoans. Geophysical Research Abstract, 5, 13219. Englund, J.-O. 1966. Sparagmittgruppens bergarter ved Fa˚vang, Gudbrandsdalen-En sedimentologisk og tektonisk undersøkelse. Norges geologiske undersøkelse, 238, 55 – 103. Englund, J.-O. 1973a. Stratigraphy and structure of the Ringebu-Vinstra district, Gudbrandsdalen; with a short analysis of the western part of the sparagmite region in Southern Norway. Norges geologiske undersøkelse, 293, 1 –58. Englund, J.-O. 1973b. Geochemistry and mineralogy of pelitic rocks from the Hedmark Group and the Cambro-Ordovician

621

sequence, Southern Norway. Norges geologiske undersøkelse, 286, 1– 60. Esmark, J. 1829. Reise fra Christiania til Trondhjem. Christiania, Oslo, 81. Furnes, H., Nystuen, J. P., Brunfelt, A. O. & Solheim, S. 1983. Geochemistry of Upper Riphean-Vendian basalts associated with the ‘sparagmites’ of southern Norway. Geological Magazine, 120, 349– 361. Gayer, R. A. & Greiling, R. O. 1989. Caledonian nappe geometry in north-central Sweden and basin evolution of the Baltoscandian margin. Geological Magazine, 126, 499–513. Gayer, R. A., Rice, A. H. N., Roberts, D., Townsend, C. & Welbon, A. 1987. Restoration of the Caledonian Baltoscandian margin from balanced cross-sections: the problem of excess continental crust. Transactions of the Royal Society of Edinburgh, 78, 197– 217. Gee, D. G., Juhlin, C., Pascal, C. & Robinson, P. 2010. Collision orogeny in the Scandinavian Caledonides (COSC). Geologiska Fo¨reningens i Stockholm Fo¨rhandlingar, 132, 29– 44. Halverson, G. P., Maloof, A. C. & Hoffman, P. F. 2004. The Marinoan glaciation (Neoproterozoic) in northeast Svalbard. Basin Research, 16, 297– 324. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Toward a Neoproterozoic composite carbonisotope record. Geological Society of America Bulletin, 117, 1181–1207. Hannah, J. L., Yang, G., Bingen, B., Stein, H. J. & Zimmerman, A. 2007.  560 Ma and 300 Ma Re –Os ages constrain Neoproterozoic glaciation and record Variscan hydrocarbon migration on extension of Oslo rift. Goldschmidt Conference Abstracts 2007, A378. Hannah, J. L., Yang, G., Xu, G., Zimmerman, A., Stein, H. J. & Egenhoff, S. O. 2008. Re –Os Isotopic disturbances at unconformities: challenges and opportunities. American Geophysical Union, Fall Meeting 2008, abstract #PP31C-1519. Heim, M., Skio¨ld, T. & Wolff, F. C. 1996. Geology, geochemistry and age of the ‘Tricolor’ granite and some other Proterozoic (TIB) granitoids at Trysil, southeast Trysil. Norsk geologisk tidsskrift, 76, 45– 54. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129– 155. Hoffman, K.-H, Conodon, D. J., Bowring, S. A. & Crowley, J. L. 2004. U– Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817– 820. Holtedahl, O. 1922. A tillite-like conglomerate in the ‘Eo-Cambrian’ Sparagmite of Southern Norway. American Journal of Science, 4, 165– 173. Hossack, J. R., Garton, M. R. & Nickelsen, R. P. 1985. The geological section from the foreland up to the Jotun thrust sheet in the Valdres area, south Norway. In: Gee, D. G. & Sturt, B. A. (eds) The Caledonide Orogen – Scandinavia and Related Areas. John Wiley & Sons, Chichester, 443–456. Ho¨gdahl, K., Andersson, U. B. & Eklund, O. (eds) 2004. The Transscandinavian Igneous Belt (TIB) in Sweden; A Review of its Character and Evolution. Geological Survey of Finland, Special Paper, 37, 123. Kumpulainen, R. 1980. Upper Proterozoic stratigraphy and depositional environments of the Tossa˚sfja¨llet Group, Sa¨rv Nappe, southern Swedish Caledonides. Geologiska Fo¨reningen i Stockholm Fo¨rhandlingar, 102, 531–550. Kumpulainen, R. A. 2011. The Neoproterozoic Lillfja¨llet Formation, southern Swedish Caledonides. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 629– 634. Kumpulainen, R. A. & Greiling, R. O. 2011. Evidence for Neoproterozoic glaciation in central Scandinavian Caledonides. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 623–628. Kumpulainen, R. & Nystuen, J. P. 1985. Late Proterozoic basin evolution and sedimentation in the westernmost part of Baltoscandia. In: Gee, D. G. & Sturt, B. A. (eds) The Caledonide Orogen – Scandinavia and Related Areas. John Wiley & Sons, Chichester, 213– 232.

622

J. P. NYSTUEN & J. T. LAMMINEN

Lamminen, J., Nystuen, J. P. & Andersen, T. 2009. U –Pb ages and Lu –Hf isotopes of granitoid clasts and basement rocks of the Hedmark Group: implications for structural setting of the Hedmark Basin at the Baltoscandian margin. NGF Vinterkonferansen 2009, Bergen 13 – 15 January 2009. Abstracts and Proceedings of the Geological Society of Norway, 60. Loeschke, J. & Nickelsen, R. P. 1968. On the age and tectonic position of the Valdres Sparagmite in Slidre (Southern Norway). Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, 131, 337– 367. Løberg, B. E. 1970. Investigations at the south-western border of the sparagmite basin (Gausdal Vestfjell and Fa˚berg Vestfjell), Southern Norway. Norges geologiske undersøkelse, 266, 160– 205. McCay, G. A., Prave, A. R., Alsop, G. I. & Fallick, A. E. 2006. Glacial trinity: Neoproterozoic Earth history within the British– Irish Caledonides. Geology, 34, 909–912. Morley, C. K. 1986. The Caledonian thrust front and palinspastic restorations in the southern Norwegian Caledonides. Journal of Structural Geology, 8, 753– 765. Myrow, P. M. & Kaufman, A. J. 1999. A newly discovered cap carbonate above Varanger-age glacial deposits in Newfoundland, Canada. Journal of Sedimentary Research, 69, 784– 793. Nickelsen, P. R. 1974. Geology of the Røssjøkollan – Dokkvatn area, Oppland. Norges geologiske undersøkelse, 314, 53– 100. Nickelsen, P. R., Garton, M. & Hossack, J. R. 1985. Late Precambrian to Ordovician sedimentology and stratigraphic correlation of the Valdres and Synnfjell thrust sheets in the Valdres area, central Norway. In: Gee, D. G. & Sturt, B. A. (eds) The Caledonide Orogen – Scandinavia and Related Areas. John Wiley & Sons, Chichester, 369–378. Nordgulen, Ø. 1999. Geologisk kart over Norge, berggrunnskart (bedrock map) HAMAR, M 1:250 000. Norges geologiske undersøkelse. Nystuen, J. P. 1969. On the paragenesis of chert and carbonate minerals in chert-bearing magnesitic dolomite from the Kvitvola nappe, Southern Norway. Norges geologiske undersøkelse, 258, 66 –78. Nystuen, J. P. 1976a. Facies and sedimentation of the Late Precambrian Moelv Tillite in the eastern part of the Sparagmite Region, southern Norway. Norges geologiske undersøkelse, 329, 1 – 70. Nystuen, J. P. 1976b. Late Precambrian Moelv Tillite deposited on a discontinuity surface associated with a fossil ice wedge, Rendalen, southern Norway. Norsk geologisk tidsskrift, 56, 29 –56. Nystuen, J. P. 1980. Stratigraphy of the Upper Proterozoic Engerdalen Group, Kvitvola Nappe, southeastern Scandinavian Caledonides. Geologiska Fo¨reningen i Stockholm Fo¨rhandlingar, 102, 551– 560. Nystuen, J. P. 1981. The Late Precambrian ‘Sparagmites’ of southern Norway: a major Caledonian allochthon – The Osen-Røa Nappe Complex. American Journal of Science, 281, 69– 94. Nystuen, J. P. 1982. Late Proterozoic basin evolution on the Baltoscandian Craton: the Hedmark Group, southern Norway. Norges geologiske undersøkelse, 375, 1 –74. Nystuen, J. P. 1985. Facies and preservation of glaciogenic sequences from the Varanger Ice Age in Scandinavia and other parts of the North Atlantic region. Palaeogeography, Palaeoclimatology, Palaeoecology, 51, 209–229. Nystuen, J. P. 1987. Synthesis of the tectonic and sedimentological evolution of the late Proterozoic-early Cambrian Hedmark Basin, the Caledonian Thrust Belt, southern Norway. Norsk geologisk tidsskrift, 67, 395– 418. Nystuen, J. P. 2008. Neoproterozoic Moelv Tillite and the Hedmark Basin, Mjøsa Area, South Norway. 33 IGC Excursion No 101, 33rd

International Geological Congress, Oslo, Norway, available at http://www.33igc.org/coco/LayoutPage.aspx, 59. Nystuen, J. P. & Sæther, T. 1979. Clast studies in the Late Precambrian Moelv Tillite and Osdal Conglomerate, Sparagmite Region, south Norway. Norsk geologisk tidsskrift, 59, 239– 251. Nystuen, J. P. & Ilebekk, S. 1981. Stratigraphy and Caledonian structures in the area between the Atnsjøen and Spekedalen windows, Sparagmite Region, southern Norway. Norsk geologisk tidsskrift, 61, 17 – 24. Nystuen, J. P., Andresen, A., Kumpulainen, R. & Siedlecka, A. 2008. Neoproterozoic basin evolution in Fennoscandia, East Greenland and Svalbard. Episodes, 31, 35– 43. Oftedahl, C. 1943. Overskyvninger i den norske fjellkjede. Naturen (Oslo), 5, 143–150. Rankama, K. 1973. The Late Precambrian glaciation, with particular reference to the Southern Hemisphere. Journal of Proceedings of the Royal Society of New South Wales, 106, 89– 97. Rehnstro¨m, E. F., Corfu, F. & Torsvik, T. H. 2002. Evidence of a Late Precambrian (637 Ma) deformational event in the Caledonides of Northern Sweden. Journal of Geology, 110, 591– 601. Rice, A. H. N. 2005. Quantifying the exhumation of UHP-rocks in the Western Gneiss Region, S.W. Norway: a branch-line – balanced cross section model. Austrian Journal of Earth Sciences, 98, 2– 21. Sæther, T. & Nystuen, J. P. 1981. Tectonic framework, stratigraphy, sedimentation and volcanism of the late Precambrian Hedmark Group, Østerdalen, south Norway. Norsk geologisk tidsskrift, 61, 193– 211. Siedlecka, A. & Ilebekk, S. 1982. Forekomster av tillitt pa˚ nordsiden av Atnsjøvinduet, Syd-Norge. Norges geologiske undersøkelse, 373, 33 – 37. Siedlecka, A., Nystuen, J. P., Englund, J.-O. & Hossack, J. 1987. LILLEHAMMER – berggrunnskart M. 1:250 000. Norges geologiske undersøkelse. Siedlecka, A., Roberts, D., Nystuen, J. P. & Olovyanishnikov, V. G. 2004. Northeastern and northwestern margins of Baltica in Neoproterozoic time: evidence from the Timanian and Caledonian Orogens. In: Gee, D. G. & Pease, V. (eds) The Neoproterozoic Timanide Orogen of Eastern Baltica. Geological Society, London, Memoirs, 30, 169–190. Thompson, M. D. & Bowring, S. A. 2000. Age of the Squantum ‘tillite’, Boston basin, Massachusetts: U–Pb zircon constraints on terminal Neoproterozoic glaciation. American Journal of Science, 300, 630– 655. Thompson, M. D., Grunow, A. M. & Ramezani, J. 2007. Late Neoproterozoic paleogeography of the Southeastern New England Avalon Zone: insights from U –Pb geochronology and paleomagnetism. Geological Society of America Bulletin, 119, 681–696. Tucker, M. E. 1983. Sedimentation of organic-rich limestones in the late Precambrian of southern Norway. Precambrian Research, 22, 295– 315. Tucker, R. D., Robinson, P. et al. 2004. Thrusting and extension in the Scandian hinterland, Norway: new U– Pb ages and tectonostratigraphic evidence. American Journal of Science, 304, 477– 532. To¨rnebohm, A. E.1896. Grunddragen af det centrala Skandinaviens bergbyggnad. Kongl. svenska vetenskaps-akademiens handlingar, 28, 1 – 210. Vidal, G. & Nystuen, J. P. 1990. Micropaleontology, depositional environment, and biostratigraphy of the Upper Proterozoic Hedmark Group, southern Norway. American Journal of Science, 290-A, 170–211.

Chapter 60 Evidence for late Neoproterozoic glaciation in the central Scandinavian Caledonides RISTO A. KUMPULAINEN1* & REINHARD O. GREILING2* 1

Department of Geological Sciences, Stockholm University, SE-106 91 Stockholm, Sweden

2

Institut fu¨r Angewandte Geowissenschaften, Abteilung Strukturgeologie und Tektonophysik, Karlsruher Institut fu¨r Technologie (KIT), Hertzstraße 16, D-76187 Karlsruhe, FR Germany *Corresponding authors (e-mail: [email protected]; [email protected]) Abstract: The La˚ngmarkberg Formation (Fm). rests on either the Neoproterozoic rift-related Risba¨ck Group or Palaeoproterozoic basement in central Scandinavia. It is always succeeded in stratigraphy by the marine Ga¨rdsjo¨n Fm. The La˚ngmarkberg Fm. is a thin, generally less than 50-m-thick, discontinuous, but persistent unit, primarily in the Lower Allochthon; subordinately also in the Autochthon, and Middle Allochthon of the central Scandinavian Caledonides. The formation is composed of a lower diamictite and an upper laminated, lonestone-bearing mudstone. Criteria supporting the glaciogenic origin of the formation are (i) a large regional extension of the formation (.5000 km2) and (ii) its fixed position in the regional stratigraphy, (iii) the ‘tillite-like’ appearance of the lower part of the formation, (iv) the lonestone-bearing mudstone in the upper part, (v) the presence of striated clasts and (vi) striated pavement. It is interpreted as having formed during one single glacial–deglacial cycle. Information on chemostratigraphy and realistic geochronology from this region is missing. The La˚ngmarkberg Fm. has commonly been correlated with the Moelv Fm. in southern Norway and the Mortensnes Fm. in northernmost Norway. The Lower Allochthon in this region also contains an older, potentially glacially influenced diamictite unit at the base of the Risba¨ck Group, which is also briefly discussed.

The La˚ngmarkberg Fm. composed of diamictite and mudstone is part of the Ja¨mtland Supergroup (Gee 1975), which comprises the Neoproterozoic to Lower Palaeozoic sedimentary succession of the Caledonian Lower Allochthon (Fig. 60.1a) in central Scandinavia. The formation occurs in small outcrops, which typically may be up to a few hundred square metres in size and extend up to a few hundred metres along strike. The formation is a thin, discontinuous, but regionally persistent unit resting on either the Risba¨ck Group or the Palaeoproterozoic basement and is always overlain by the Ga¨rdsjo¨n Fm. (Fig. 60.2). The preserved thickness of the unit varies from less than a metre commonly to less than 20 m, but ranges locally to more than 50 m. The majority of the La˚ngmarkberg Fm. outcrops occur within the imbricated Lower Allochthon and are indicated (by stars) in Figure 60.1. Palinspastic restoration of the imbricated thrust units to the original pretectonic (pre-Caledonian) basin(s) suggests that the minimum depositional area exceeded 5000 km2 (Kumpulainen & Nystuen 1985; Gayer et al. 1987; Gayer & Greiling 1989, Siedlecka et al. 2004; Nystuen et al. 2008). Only a few outcops of the formation are known between the main study area (Fig. 60.1b) and Laisvall in the north, where a diamictite unit of uncertain origin is described from the Autochthon (Ackerselet Fm., Willde´n 1980; Greiling et al. 1999a). The unit reappears again in the Akkajaure area further north (Stodt et al. 2011). Towards the south, only a few outcrops of the formation are known in the Lower Allochthon (Fig. 60.1a), before reaching southeastern Norway (Nystuen & Lamminen 2011). The Middle Allochthon (Gayer & Greiling 1989; Greiling & Zachrisson 1999a, b) hosts sections of the La˚ng¨ stersund markberg Fm. NE of Stalon (Fig. 60.1b) and south of O (Fig. 60.1a). This chapter focuses primarily on the La˚ngmarkberg Fm. between Stro¨ms Vattudal and Dikana¨s (Fig. 60.1b). However, diamictite units of uncertain origin, which occur locally at the base of the Risba¨ck Group, are also discussed. Historically, a ‘tillite-like sedimentary breccia’ was described by Asklund & Thorslund (1934) in Sjouta¨lven (old spelling, Sjougda¨lven; Fig. 60.1b) of the Lower Allochthon of west-central Sweden. Later, Asklund (1938) named this unit ‘Tillitabteilung/Tillit Fm.’ and correlated it with the Moelv Conglomerate of Holtedahl (1922)

in southern Norway. Subsequently, Kulling (1942) described about 20 diamictite sections in this region. Although unable to produce a proper description of the section at La˚ngmarkberget (Fig. 60.1b; 648320 N/158210 E), Kulling (1942) selected it to be the type section and named the unit the La˚ngmarkberg Fm. Recent mapping in the area reveals that this type section is complicated tectonically (Zachrisson 1997d ). Work in the late 1960s to mid-1990s revealed more outcrops, which amount today to more than 60. Some of the other sections are much better exposed than La˚ngmarkberget. For this reason the unit was renamed the Dabbsjo¨n Fm. (Gee et al. 1974) after a new type section at Dabbsjo¨n (Fig. 60.1b; 648410 N/158180 E). However, encouraged by the ‘rule of priority’, Gee et al. (1978) revived the name La˚ngmarkberg Fm. and the sections at Dabbsjo¨n remained reference sections. The work up to about the mid-1970s was summarized by Thelander (1981). Some additional sedimentological information derives from geological mapping of the areas north of Bijelite and Malgomaj (Greiling & Zachrisson 1999a, b; Greiling et al. 1999b).

Structural framework The Scandinavian Caledonides are a typical thrust-and-fold belt and are subdivided into the Caledonian Autochthon, Lower, Middle, Upper and Uppermost Allochthons (e.g. Gee et al. 1985). Only the Autochthon and Lower and Middle Allochthons are relevant for the discussion in this chapter. In the Autochthon of central Scandinavia, the sedimentary succession of sandstone and shale of Cambrian age was deposited on the passive Baltica margin (Willde´n 1980; Greiling et al. 1999a). The succession is subhorizontal and undeformed. The alteration in the Autochthon is diagenetic with peak temperatures below 200 8C (Warr et al. 1996). The Lower Allochthon is composed of the Neoproterozoic to Lower Palaeozoic sedimentary succession of the Ja¨mtland Supergroup (Gee 1975) and slices of the Palaeoproterozoic basement. The Ja¨mtland Supergroup represents (Fig. 60.2) a gradual development from an initial continental rift setting (Risba¨ck Group) to passive, Iapetus marginal setting (Ta˚sjo¨n Group) along the

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 623– 628. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.60

624

R. A. KUMPULAINEN & R. O. GREILING

Fig. 60.1. (a) Simplified geological map of the central Scandinavian Caledonides. (b) Simplified geological map of the Stro¨ms Vattudal–Dikana¨s areas, modified from Kumpulainen & Thelander (1978), Gee et al. (1985), Zachrisson (1997a, b, c, d) and the sources referred to in the text.

western (present coordinates) Baltica margin (Kumpulainen & Nystuen 1985; Gayer & Greiling 1989). Subsequent Caledonian orogenic shortening formed hinterland dipping duplexes and imbricate fans (Gee et al. 1978; Gayer & Greiling 1989; Zachrisson & Greiling 1993a, b; Febbroni 1997; Greiling & Zachrisson 1999a, b; Greiling et al. 1999b, c). In the interior parts of the fold belt, antiformal stacks of the Lower Allochthon are exposed in tectonic windows, for example Børgefjellet, Ba˚ngona˚ive and Nasafja¨ll (Fig. 60.1), (Greiling 1988; Greiling et al. 1993). The sedimentary sequence is faulted and sheared, with one or two cleavages in shaly rocks. Metamorphism is of sub-greenschist grade (anchimetamorphism) in the marginal fold belt (Warr et al. 1996) and up to biotite grade in the tectonic windows (Greiling et al. 1993; our data). Primary sedimentary structures and textures, as well as fossil assemblages, are widely preserved. Similar to the Lower Allochthon, the Middle Allochthon is formed of imbricate units with crystalline basement of Palaeo- and Mesoproterozoic age and metamorphosed units of

conglomerate, sandstone and shale of probable Neoproterozoic age. The deformation is polyphase, including a pervasive cleavage in the pelitic rocks (Greiling 1985; Greiling & Zachrisson 1999a, b). Primary sedimentary features are preserved only in competent, psammitic rocks.

Stratigraphy The Neoproterozoic to Lower Palaeozoic sedimentary succession, the Ja¨mtland Supergroup (Gee 1975), is subdivided into the ¨ nge Groups (Fig. 60.2). Risba¨ck, Sjouta¨lven, Ta˚sjo¨n and A ¨ The Neoproterozoic Risback Group (Gee et al. 1974) is dominated by alluvial siliciclastic rocks. It may reach a thickness of c. 1.5 km, but wedges out laterally to nothing. In its lower part, the Cryogenian to Cambrian Sjouta¨lven Group is composed of the glaciogenic La˚ngmarkberg Fm. (Kulling 1942; Thelander 1981), which is stratigraphically always succeeded by the

CENTRAL SCANDINAVIAN CALEDONIDES

625

Fig. 60.2. Ja¨mtland Supergroup stratigraphy (Gee 1975) with outlines of the geological development of the area.

¨ nge Ga¨rdsjo¨n Fm. (Gee et al. 1974). The overlying Ta˚sjo¨n and A groups are not discussed here. Sedimentary facies types in the rift-related Risba¨ck Group suggest a general change from a proximal alluvial facies, dominated by arkosic sandstone with minor (braided-stream-type) conglomerate, mudstone + occasional beds of poorly sorted matrix-supported conglomerate in the east to a marine facies of sandstone and shale including a dolomite unit, the Kalvberget Fm., in the west and NW (Kumpulainen & Nystuen 1985; Nystuen et al. 2008). In its type area the dolomite is microcrystalline, but contains scattered euhedral dolomite crystals (,1 mm). In other areas it is strongly recrystallized, and a network of quartz veins occurs also in Kalvberget. The La˚ngmarkberg Fm. forms a thin, discontinuous, but regionally persistent unit and rests on the Risba¨ck Group. Outside the Risba¨ck basin, the La˚ngmarkberg Fm. was deposited on top of the slightly undulating Palaeoproterozoic basement (Fig. 60.2). The La˚ngmarkberg Fm. is, in all outcrops if not disturbed by fault surfaces, succeeded by the Ga¨rdsjo¨n Fm. formed during several transgression –regression cycles in a shallow-marine setting. Regressions produced quartz-arenite units, whereas transgressions deposited mudstone-dominated units (e.g. Kumpulainen & Nystuen 1985; Greiling et al. 1999a). The Ga¨rdsjo¨n Fm. rocks are distinctly different, in lithology and facies, from those in the Risba¨ck Group, so mistaking these stratigraphic units in an outcrop is unlikely, also outside the Stro¨ms Vattudal –Dikana¨s area (Fig. 60.1b), despite the fact that this sedimentary succession is tectonically imbricated.

upper part is mudstone-dominated. The boundary from the lower unit to the upper unit may be gradational or sharp (Fig. 60.3). The thickness of the formation varies from less than a metre generally to c. 20 m (Fig. 60.3), but thicknesses exceeding 50 m have been encountered at Bergvattnet (Fig. 60.1b) and Jillesna˚le (Fig. 60.1a); there are no systematic changes of the thickness in any particular direction. The coarse-clastic lower part of the formation may be composed of up to three diamictite units that are poorly sorted, structureless and matrix-supported, although changes in the grain size of the matrix may produce an indistinct stratification. The poorly sorted matrix in the diamictite varies from clayey to coarse sandy. The proportion of matrix and framework clasts varies greatly. Stratified sandstone- or mudstone beds with lonestones up to c. 30 cm in diameter occur between diamictite beds. The size of the clasts in the diamictite varies from pebble to boulder, with the largest boulder measuring 2 m in diameter. Arkose and other feldspar-bearing sandstone, quartzite, mudstone

Glaciogenic deposits and associated strata The La˚ngmarkberg Fm. The description of the La˚ngmarkberg Fm. in this chapter is essentially a summary of the general characteristics of the formation based on Thelander (1981), Kulling (1942) and, north of BijeliteMalgomaj, on Greiling & Zachrisson (1999a, b). On the basis of colour and clast content, Kulling (1942) subdivided the La˚ngmarkberg Fm. into a lower diamictite and an upper diamictite. This subdivision has not been confirmed by later studies (Thelander 1981). The stratigraphy and facies in the formation vary greatly from one section to another (Fig. 60.3). However, generally (for exception, see Fig. 60.3, section 5), the lower part of the formation is dominated by coarse-clastic, matrix-supported diamictite beds, whereas the

Fig. 60.3. Sections through the La˚ngmarkberg Fm. in the Dabbsjo¨n – La˚ngvattnet areas. The locations are shown in Figure 60.1b: 1– 3, on the northeastern shore of Dabbsjo¨n (648410 N/158180 E), redrawn from Thelander (1981); 4, south of Blaikliden, south of Marsa˚n (65820 400 N/158470 700 E); 5, SE Grytsjo¨, south of the Grytsjo¨ –Blaikliden road (65800 2100 N/158360 4600 E); 6, east of Grytsjo¨, road section at north side of road Grytsjo¨ – Blaikliden (65800 3400 N/ 158370 1000 E); 7, western shore of La˚ngvattnet (658100 1100 N/168310 5300 E).

626

R. A. KUMPULAINEN & R. O. GREILING

and dolomite are the dominant clast types of sedimentary origin; gneiss clasts of sedimentary origin also occur. The majority of the crystalline rocks are represented by various types of granite, felsic porphyry and syenite. With few exceptions, the clast content in the lower part of the diamictite sections reflects the composition of the substratum. The proportion of crystalline clasts increases up-section and they always dominate in the lonestone-bearing, laminated upper mudstone unit. The proportion of crystalline clasts in the formation varies from 10 to 100%, whereas the proportion of sedimentary clasts varies from 0 to 90 %. Sub-angular to sub-rounded clasts dominate, but shapes range from angular to well-rounded. Some stones are faceted and a few display striations (Kulling 1942). Laminated mudstone with or without normal grading is a common component in these diamictite sections, particularly in the upper part of the formation. Occasionally, the mudstone unit contains lenses of muddy diamictite and some stratified gravel beds. Thin sandy beds are particularly common in the basal part of a mudstone unit. Lonestones range from sand-grade to boulders in size. The preserved thickness of this facies type is generally in the order of 5–10 m.

Vindela¨lven River Diamictite outcrops in the Lower Allochthon north of the Vindela¨lven River at Jillesna˚le rest on the Risba¨ck Group (Grambow 2001, unpublished diploma thesis, Heidelberg University). It is similar in composition to the La˚ngmarkberg Fm. diamictite, although its exact stratigraphic relationship remains uncertain. Basal, laminated mudstone, c. 5 m thick, with lonestones is overlain by more than 50 m of coarse, matrix-supported diamictite. This diamictite contains boulders of meta-arkose (26%), syenite (26%), rhyolite (19%), quartzite (18%), granite (10%) and meta-pelite (1%). Ten percent of the boulders are angular, 75% sub-angular and 15% rounded. No striated stones are reported.

dominates the group, whereas units of greenish grey to dark grey mudstone are subordinate. Dolomites of the Kalvberget Fm. up to c. 100 m thick are a discontinuous but significant rock unit in the upper part of the group. A unit of diamictite in the Flakatra¨sk and Vojmsjo¨n areas (Fig. 60.1), described by Kulling (1942) as ‘red conglomerate’, underlies the Risba¨ck Group composed of arkose and feldspathic quartzite (Febbroni 1997; Greiling & Zachrisson 1999b; Greiling et al. 1999c). At Vojmsjo¨n, a similar matrix-supported diamictite contains boulders of c. 5– 50 cm in diameter (Grambow 2001, unpublished diploma thesis, Heidelberg University). The boulders are composed of granitoid gneiss (44%), rhyolite (36%), vein quartz (10%), meta-arkose and meta-pelite (5%), epidote fragments (4%) and amphibolite (1%). Most of these boulders are angular or sub-angular; no striated boulders have been reported.

The Ga¨rdsjo¨n Fm. The Ga¨rdsjo¨n Fm. always succeeds the La˚ngmarkberg Fm. The Ga¨rdsjo¨n Fm. has been divided (Gee et al. 1974) into ten informal subunits, which may be named M1 to M10; five are sandstone members, representing sea-level lowstands, and the remaining five are mudstone-dominated members, representing sea-level highstands. Tectonic deformation disturbs the preservation of sedimentary structures, but in many sections they are common, and readily discernible in both rock types. The Ga¨rdsjo¨n Fm. is a common component in the Lower Allochthon and certain units, particularly a maroon mudstone member in the middle of the formation, and may be traced long distances along most of the thrust belt (e.g. Thelander 1982; Greiling et al. 1999a). Owing to the onlapping character of the Ga¨rdsjo¨n Fm., only some of the upper members are present in the Autochthon (Fig. 60.1b), such as the Storuman–Laisvall area, where they are referred to as the Sa˚vvovare Fm. (Willde´n 1980; Eliasson et al. 2003; Greiling et al. 1999b).

Ba˚ngona˚ive window Diamictite of uncertain origin and stratigraphic position is exposed in the Lower Allochthon of the Ba˚ngona˚ive window (Fig. 60.1a) NE of Ta¨rnaby (648430 4700 N/158240 1400 E) at the base of the Oltokken Fm. (Stephens 1977) interpreted as an equivalent to the La˚ngmarkberg and Ga¨rdsjo¨n formations combined (Greiling et al. 1993). The diamictite forms a lensoidal body, with a maximum thickness of c. 1 m, only a few metres stratigraphically above the crystalline substratum. The exposed lateral extent is in the order of 100 m. The diamictite is composed of a fine sandy matrix and enclosed angular clasts of quartzite, feldspathic quartzite and subordinate felsic crystalline rocks. The clast size varies from a few centimetres to decimetres.

Børgefjellet window Close to the eastern margin of the Børgefjellet window (Fig. 60.1a), a diamictite unit of uncertain stratigraphic position overlies the Proterozoic basement at the northern slope of the Sipmekfja¨ll hill (658430 4700 N/158240 1400 E; Greiling 1988). The unit, c. 1 m thick, occurs in a tectonic slice and extends laterally for a few hundred metres. The diamictite is characterized by a quartz-rich, shaly matrix, which is intensely cleaved and contains angular fragments of white vein quartz and quartzite up to c. 10 cm in diameter. The diamictite is overlain by white, coarse feldspathic quartzite, interpreted as an equivalent of the Ga¨rdsjo¨n Fm. (Gayer & Greiling 1989).

The Risba¨ck Group and diamictites older than the Risba¨ck Group In the Risba¨ck area, the type area of the Risba¨ck Group, a succession of maroon to grey, arkose and feldspathic sandstone

Boundary relations with overlying and underlying non-glacial units The La˚ngmarkberg Fm. The La˚ngmarkberg Fm. rests on three different rock types: (i) Proterozoic crystalline rocks, (ii) arkose and feldspathic sandstone of the Risba¨ck Group and (iii) dolomite of the Kalvberget Fm. The lower boundary of the La˚ngmarkberg Fm. at a regional scale conforms to bedding in the underlying Risba¨ck Group. At a local scale, grooves and linear depressions and disconformities are observed (Thelander 1981; Greiling & Kumpulainen 2004). La˚ngmarkberget (Fig. 60.1b) exposes a low-angle disconformity at the top of the Risba¨ck Group. In a minor outcrop SE of La˚ngmarkberget, bedding in the Risba¨ck Group is cut at a high angle by the formation boundary, which Thelander (1981) and Gee et al. (1990) suggested was a major unconformity. However, soft-sediment deformation occurs frequently in this part of the Risba¨ck Group and it is possible that the erosional surface only incidentally truncated an overturned soft-sediment fold in the Risba¨ck Group. Asklund (1960) reported a striated pavement below the La˚ngmarkberg Fm., but later that interpretation was questioned by Crowell (1964). However, still later, Thelander (1981) reported such structures in this region. In most cases and independent of the substratum, the boundary is sharp. A close correlation of clast lithologies in the basal parts of the diamictites and their substratum additionally indicates that the boundary must be erosional. Only locally, at La˚ngvattnet (Fig. 60.3), the Kalvberget Fm. dolomites display a gradual transition up to a diamictite (of uncertain origin). The diamictite, in turn, grades upwards into a finer-grained clastic sediment with coarse framework grains of quartz in a dolomite matrix, and then

CENTRAL SCANDINAVIAN CALEDONIDES

to the Ga¨rdsjo¨n Fm. quartzite (Fig. 60.3); the lonestone-bearing mudstone is missing here. With this exception, the upper contact of the La˚ngmarkberg Fm. with the Ga¨rdsjo¨n Fm. is sharp. Erosional features are encountered locally.

Diamictite units older than the Risba¨ck Group The diamictite units at the base of the Risba¨ck Group rest on crystalline basement rocks; the exact contact is not exposed, but was probably deposited in shallow depressions of the basement surface. The upper boundary of the diamictite is gradational to red arkose of the Risba¨ck Group. This diamictite is distinctly red, which implies staining by iron oxides, mostly hematite and goethite, similar to the overlying arkose and feldspathic sandstone of the Risba¨ck Group. However, magnetic susceptibility in the diamictite is exceptionally high, in the order of 5–10  1023 SI units (Febbroni 1997, our data). Such high susceptibilities point to an additional magnetite content of up to 1%. The magnetite is distributed both in the clasts (boulders) of crystalline basement rocks and in the matrix of the diamictite.

Chemostratigraphy No data are available.

Palaeolatitude and palaeogeography Restorations by Greiling & Smith (2000), Hartz & Torsvik (2002) or Cawood & Pisarevsky (2006) imply high southern latitudes for the present area at the time of La˚ngmarkberg Fm. deposition.

Geochronological constraints No data are available.

Discussion There are potentially two different glaciogenic diamictite units in this region, one on top of the Risba¨ck Group and the other below the Risba¨ck Group.

La˚ngmarkberg Fm. Although thin and discontinuous at a regional scale, the La˚ngmarkberg Fm. can be traced all across the Cryogenian to Cambrian succession from the Autochthon, Lower and Middle Allochthons of the central Scandinavia Caledonides and further north and south. The interpretations of the various facies types within the La˚ngmarkberg Fm. rely essentially on Hambrey & Harland (1981) and Hambrey (1994). The deposition of the La˚ngmarkberg Fm. has been interpreted (Thelander 1981) in the framework of one glacial–deglacial cycle. The various sections through the formation vary greatly in facies, although generally poorly sorted matrix-supported diamictite occupies the basal part of the sections. The massive units may have been deposited by meltout, dumping or redeposition subglacially or in a proglacial marine setting. The poorly stratified diamictite may have been deposited subglacially from a grounded ice-sheet. The stratified, sandy to gravelly beds in or between diamictite beds could have been deposited from glaciofluvial traction currents beneath temporarily floating ice or close to the subaqueous grounding line of a glacier. The upper, mudstone-dominated part displays distinct lamination originating from subaqueous deposition at some distance from floating ice or a grounded ice margin. Melting icebergs transported clastic debris, dropstones up to boulder size, to this muddy

627

environment. Occasional lensoidal bodies of diamictite with muddy matrix within this upper laminated unit are probable massflow deposits or were dumped from icebergs. The lower boundary of the formation is a regional, erosional, low-angle disconformity, which is substantiated by the presence (side views) of grooves and other minor linear irregularities. The diamictite units and their substratum are so indurated that the possible glacially striated contact surface is only rarely available. This also limits the findings of striated lonestones. The upper boundary is commonly sharp and erosional and overlain by the basal quartz arenite (occasionally conglomerate-bearing) of the Ga¨rdsjo¨n Fm. The La˚ngmarkberg Fm. has been correlated stratigraphically with the upper tillite horizon, the Mortensnes Fm. in northern Norway (Siedlecka et al. 2004) and with the Moelv Fm. in southeastern Norway (Asklund & Thosrslund 1934; Thelander 1981; Siedlecka et al. 2004; Nystuen et al. 2008; Nystuen & Lamminen 2011), but the scarce isotopic evidence derived from this particular region of Scandinavia precludes such correlations and remains speculative.

Possible glaciogenic diamictite at the base of the Risba¨ck Group The ‘red conglomerate’ at the base of the Risba¨ck Group was interpreted by Kulling (1942) as an ‘in water redeposited conglomerate, which originally may have been a till’. In addition, it is peculiar, because of its characteristic magnetite content (Febbroni 1997) which has not been observed in the overlying sedimentary succession. Magnetite is not stable under present atmospheric or shallowwater conditions (e.g. Piper 1987; Chan et al. 2005). It is therefore speculated that magnetite formation and/or stability in this sedimentary environment may be due to exceptional atmospheric/climatic conditions and perhaps related to a Snowball Earth situation (e.g. Hoffman & Schrag 2002). Therefore, the magnetite contents, if related to a Snowball Earth situation, may also be a further indication for a diamictite horizon at the base of the Risba¨ck Group in other areas. The red conglomerates at the base of the Risba¨ck Group may then correspond to the lower tillite horizon (Smalfjord Fm.) in northern Norway. R.O.G. acknowledges support from the Geological Survey of Sweden. Some of the data presented here were acquired during the PNASTINA project. Sorsele Kommun supported the work in the Jillesna˚le/Vindela¨lven areas. R.A.K. acknowledges support from Stockholm University. The manuscript was improved by the comments of E. Arnaud, an anonymous reviewer and G. A. Shields. This represents a contribution of the IUGS- and UNESCOfunded IGCP (International Geoscience Programme) Project #512.

References Asklund, B. 1938. Hauptzu¨ge der Tektonik und Stratigraphie der mittleren Kaledoniden in Schweden. Sveriges geologiska underso¨kning C, 417, 99. Asklund, B. 1960. Studies of the thrust region of the southern part of the Swedish mountain chain. International Geological Congress, XXI Session Norden, Guide to the Excursions, A24, C19. Asklund, B. & Thorslund, P. 1934. Fja¨llkedjerandens bergbyggnad i ˚ ngermanland. Sveriges geologiska underso¨knnorra Ja¨mtland och A ing C, 382, 110. Cawood, P. A. & Pisarevsky, S. A. 2006. Was Baltica right-way-up or upside down in the Neoproterozoic? Journal of the Geological Society London, 163, 753–759. Chan, M. A., Bowen, B. B., Parry, W. T., Ormo¨, J. & Komatsu, G. 2005. Red rock and red planet diagenesis: comparisons of Earth and Mars concretions. GSA Today, 15, 4 – 10. Crowell, J. C. 1964. Climate significance of sedimentary deposits containing disperced megaclasts. In: Nairn, A. E. M. (ed.) Problems in Palaeoclimatology. Interscience, London, 86– 99.

628

R. A. KUMPULAINEN & R. O. GREILING

Eliasson, T., Greiling, R. O. & Triumf, C.-A. 2003. Bedrock map 24H Sorsele SV, 1:50,000. Sveriges geologiska underso¨kning Ai, 188. Febbroni, S. 1997. Rilevamento geologico e dei valori della suscettivita magnetica nelle Caledonidi della Lapponia svedese. AZ Marmi, 131, 50. Gayer, R. A. & Greiling, R. O. 1989. Caledonian nappe geometry in north-central Sweden and basin evolution on the Baltoscandian margin. Geological Magazine, 126, 499– 513. Gayer, R. A., Rice, A. H. N., Roberts, D., Townsend, C. & Welbon, A. 1987. Restoration of the Caledonian Baltoscandian margin from balanced cross-sections: the problem of excess continental crust. Transactions of the Royal Society of Edinburgh: Earth Sciences, 78, 197– 217. Gee, D. G. 1975. A geotraverse through the Scandinavian Caledonides – ¨ stersund to Trondheim. Sveriges geologiska underso¨kning C, O 717, 66. Gee, D. G., Karis, L., Kumpulainen, R. & Thelander, T. 1974. A summary of the Caledonian front stratigraphy, northern Ja¨mtland/ southern Va¨sterbotten, central Swedish Caledonides. Geologiska Fo¨reningens i Stockholm Fo¨rhandlingar, 96, 389– 397. Gee, D. G., Kumpulainen, R. & Thelander, T. 1978. The Ta˚sjo¨n De´collement, central Swedish Caledonides. Sveriges geologiska underso¨kning C, 742, 35. Gee, D. G., Kumpulainen, R., Roberts, D., Stephens, M. B., Thon, A. & Zachrisson, E. 1985. Scandinavian Caledonides –Tectonostratigraphic Map, Scale 1:2,000,000. Sveriges geologiska underso¨kning Ba, 35. Gee, D. G., Kumpulainen, R. & Karis, L. 1990. Fja¨llrandens berggrund. In: Lundqvist, T., Gee, D. G., Kumpulainen, R., Karis, L. & Kresten, P. Beskrivning till berggrundskartan o¨ver Va¨sternorrlands la¨n. Sveriges geologiska underso¨kning Ba, 31, 206–237. Greiling, R. O. 1985. Strukturelle und metamorphe Entwicklung an der Basis grosser, weittransportierter Deckeneinheiten am Beispiel des Mittleren Allochthons in den zentralen Skandinavischen Kaledoniden (Stalon-Deckenkomplex in Va¨sterbotten, Schweden). Geotektonische Forschungen, 69, 129. Greiling, R. O. 1988. Ranseren, berggrunnskart 2025/3, 1:50,000, foreløpig utgave. Norges geologiske underso¨kelse. Greiling, R. O. & Zachrisson, E. 1999a. Bedrock map 23 G Dikana¨s NW, 1:50,000. Sveriges geologiska underso¨kning Ai, 122. Greiling, R. O. & Zachrisson, E. 1999b. Bedrock map 23 G Dikana¨s SW, 1:50,000. Sveriges geologiska underso¨kning Ai, 123. Greiling, R. O. & Smith, A. G. 2000. The Dalradian of Scotland: missing link between the Vendian of northern and southern Scandinavia? Physics and Chemistry of the Earth, 25, 495– 498. Greiling, R. O. & Kumpulainen, R. 2004. Spa˚r av va¨rldens sto¨rsta inlandsis i vilhelminafja¨llen? Geologiskt Forum, 43, 8– 11. Greiling, R. O., Gayer, R. A. & Stephens, M. B. 1993. A basement culmination in the Scandinavian Caledonides formed by antiformal stacking (Ba˚ngona˚ive, northern Sweden). Geological Magazine, 130, 471– 482. Greiling, R. O., Jensen, S. & Smith, A. G. 1999a. Vendian– Cambrian subsidence of the passive margin of western Baltica – application of new stratigraphic data from the Scandinavian Caledonian margin. Norsk Geologisk Tidsskrift, 79, 133– 144. Greiling, R. O., Zachrisson, E., Thelander, T. & Stra¨ng, T. 1999b. Bedrock map 23 G Dikana¨s NE, 1:50,000. Sveriges geologiska underso¨kning Ai, 124. Greiling, R. O., Zachrisson, E., Thelander, T. & Stra¨ng, T. 1999c. Bedrock map 23 G Dikana¨s SE, 1:50,000. Sveriges geologiska underso¨kning Ai, 125. Hambrey, M. J. 1994. Glacial Environments. University College London Press, London. Hambrey, M. J. & Harland, W. B. 1981. Part I Introduction. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 3 – 28. Hartz, E. H. & Torsvik, T. H. 2002. Baltica upside down: a new plate tectonic model for Rodinia and the Iapetus Ocean. Geology, 30, 255– 258.

Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129–155. Holtedahl, O. 1922. A tillite-like conglomerate in the ‘Eo-Cambrian’ Sparagmite of Southern Norway. American Journal of Science, 4, 165– 173. Kulling, O. 1942. Grunddragen av fja¨llkedjerandens bergbyggnad inom Va¨sterbottens la¨n. Sveriges geologiska underso¨kning C, 445, 320. Kumpulainen, R. & Thelander, T. 1978. Geological map of the area between Stro¨ms Vattudal and Malgomaj, northern Ja¨mtland– southern Va¨sterbotten, central Swedish Caledonides. Sveriges geologiska underso¨kning BRAP, 97001 (unpublished map 1:100,000). Kumpulainen, R. & Nystuen, J. P. 1985. Late Proterozoic basin evolution and sedimentation in the westermost part of Baltoscandia. In: Gee, D. G. & Sturt, B. A. (eds) The Caledonide Orogen – Scandinavia and Related Areas. John Wiley & Sons, Chichester, 213– 232. Nystuen, J. P. & Lamminen, J. T. 2011. Neoproterozoic glaciation of South Norway: from continental interior to rift and pericratonic basins in western Baltica. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 613–622. Nystuen, J. P., Andresen, A., Kumpulainen, R. & Siedlecka, A. 2008. Neoproterozoic basin evolution in Fennoscandia, East Greenland and Svalbard. Episodes, 31, 35 –43. Piper, J. D. A. 1987. Palaeomagnetism and the Continental Crust. Open University Press, Milton Keynes. Siedlecka, A., Roberts, D., Nystuen, J. P. & Olovyanishnikov, V. G. 2004. Northeastern and northwestern margins of Baltica in Neoproterozoic time: evidence from the Timanian and Caledonian Orogens. In: Gee, D. G. & Pease, V. (eds) The Neoproterozoic Timanide Orogen of Eastern Baltica. Geological Society, London, Memoir, 30, 169– 190. Stephens, M. B. 1977. Stratigraphy and relationship between folding, metamorphism and thrusting in the Ta¨rna-Bjo¨rkvattnet area, northern Swedish Caledonides. Sveriges geologiska underso¨kning C, 726, 1 –146. Stodt, F., Rice, A. H. N., Bjo¨rklund, L., Bax, G., Halverson, G. P. & Pharaoh, T. C. 2011. Evidence of Neoproterozoic glaciation in the Caledonides of NW Scandinavia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 603– 611. Thelander, T. 1981. The late Precambrian La˚ngmarkberg Formation, central Swedish Caledonides. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 615–619. Thelander, T. 1982. The Tornetra¨sk Formation of the Dividal Group, northern Swedish Caledonides. Sveriges geologiska underso¨kning C, 789, 1 –41. Warr, L. N., Greiling, R. O. & Zachrisson, E. 1996. Thrustrelated, very low-grade metamorphism in the marginal part of an orogenic wedge, Scandinavian Caledonides. Tectonics, 15, 1213– 1229. Willde´n, M. Y. 1980. Paleoenvironment of the autochthonous sedimentary rock sequence at Laisvall, Swedish Caledonides. Stockholm Contributions in Geology, 33, 1– 100. Zachrisson, E. 1997a. Bedrock map 22F Risba¨ck NW, 1:50,000. Sveriges geologiska underso¨kning Ai, 102. Zachrisson, E. 1997b. Bedrock map 22F Risba¨ck SW, 1:50,000. Sveriges geologiska underso¨kning Ai, 103. Zachrisson, E. 1997c. Bedrock map 22F Risba¨ck NE, 1:50,000. Sveriges geologiska underso¨kning Ai, 104. Zachrisson, E. 1997d. Bedrock map 22F Risba¨ck SE, 1:50,000. Sveriges geologiska underso¨kning Ai, 105. Zachrisson, E. & Greiling, R. O. 1993a. Bedrock map 23F Fatmomakke NE, 1:50,000. Sveriges geologiska underso¨kning Ai, 77. Zachrisson, E. & Greiling, R. O. 1993b. Bedrock map 23F Fatmomakke SE, 1:50,000. Sveriges geologiska underso¨kning Ai, 78.

Chapter 61 The Neoproterozoic glaciogenic Lillfja¨llet Formation, southern Swedish Caledonides RISTO A. KUMPULAINEN Department of Geological Sciences, Stockholm University, SE-106 91 Stockholm, Sweden (e-mail: [email protected]) Abstract: The Tossa˚sfja¨llet Group, which crops out in the Middle Allochthon of the Scandinavian Caledonides, contains the glaciogenic Lillfja¨llet Fm. Evidence that supports the interpretation of the Lillfja¨llet formations as glacially related include the presence of diamictite beds, the presence of lonestones, although few, in laminated units, one possible faceted clast, involution-like deformation structures, and the presence of sandstone wedges cutting 2.5– 3 m down into two different diamictite beds. Similar to the other glaciogenic formations in southern and central Scandinavia (Moelv and La˚ngmarkberg formations), the Lillfja¨llet Fm. rests on a unit of peritidal, sabkha-type dolomite of the Stora˚n Fm. Poor exposure contributes to the uncertainty about the thickness of the formation. In this account, the Lillfja¨llet Fm. stratigraphy has been divided into three subunits: (i) a lower diamictite-dominated unit, (ii) a middle unit composed of distinctly laminated grey sandy mudstone and (iii) an upper diamictite-dominated unit. There is no information available from the Tossa˚sfja¨llet Group concerning palaeolatitudes and chemostratigraphy. Poor isotopic evidence (40Ar/39Ar) from the Ottfja¨llet Dolerites, which cut the Tossa˚sfja¨llet Group, indicates that the succession is older than c. 665 Ma, so a reliable correlation with global glaciation events is not yet possible.

Glaciogenic diamictite and associated sandy mudstone occur in the Neoproterozoic Tossa˚sfja¨llet Group of the Sa¨rv Nappe of the Caledonian Middle Allochthon in west-central Sweden (Fig. 61.1; Gee et al. 1985). This glaciogenic unit, which was named the Lillfja¨llet Fm. by Kumpulainen (1980, 1981), is represented by less than a dozen, relatively small outcrops (stars in Fig. 61.1), arranged in a curvilinear manner across the nappe over a distance of c. 70 km. Along the southern margin of the nappe, close to the basal mylonites (Fig. 61.1), Stro¨mberg (1961) and Ro¨shoff (1975) described a possible glaciogenic diamictite, which was given the names Gro¨nstack and Ulvberget complexes, respectively. These outcrops are now included in the Lillfja¨llet Fm. The Lillfja¨llet Fm. has been correlated with other glaciogenic diamictites and related rocks in the Caledonides of southern and central Scandinavia. Descriptions of these deposits are given by Nystuen & Lamminen (2011) and Kumpulainen & Greiling (2011). A summary of other correlatives of these glaciogenic successions further north in Scandinavia is given by Stodt et al. (2011). This chapter is based essentially on the first descriptions by Kumpulainen (1980, 1981). The descriptions derive from two areas, the Lillfja¨llet area in the south and the Stor-Lo¨vsjo¨n area in the north (Fig. 61.1). The other exposures are more deformed and are not included here.

Structural relationships The Sa¨rv Nappe is the uppermost tectonic unit in the Scandinavian Caledonides that hosts well-preserved glaciogenic rocks of Neoproterozoic age. The high-quality preservation is a consequence of the intrusion of tens of thousands of dolerite dykes, the Ottja¨llet dolerite (Holmqvist 1894), into the c. 4.5-km-thick Tossa˚sfja¨llet Group. This dyke swarm reinforced the sedimentary succession and limited Caledonian deformation to only a few shear zones within the nappe. Owing to this, most of the original sedimentary structures survived. The whole nappe body, that is, the sedimentary succession and the dyke swarm together, was folded into a regional fold. The sedimentary bedding in the western limb of this fold is gently west-dipping but is slightly overturned in the Stor-Lo¨vsjo¨n area in the east (Kumpulainen 1980; Gilotti & Kumpulainen 1986). In the overturned sequence the bedding-dyke angle still remains close to 908. Approaching the basal shear zone, the bedding-dyke angle decreases gradually and reduces

over a distance of a few metres to zero in the basal mylonitic foliation. The arkose and dolerite dykes were altered to banded, felsic and mafic mylonite, respectively. The nappe was subjected to biotite-grade metamorphism, which recrystallized the rocks, and blurs the micro-textures. The tectono-sedimentary development of the Tossa˚sfja¨llet Group has been correlated with that of the other Neoproterozoic successions in southern and central Scandinavia (Nystuen et al. 2008). One of the features used to correlate these successions is the ‘carbonate –tillite’ couplet, which occurs in most of the successions in this region. Other correlative features include the gradual facies change from proximal to distal depositional sites and systematic change in thickness of the successions below the carbonate –tillite couplet. These Neoproterozoic sedimentary successions were involved in the Caledonian orogenic shortening and the various allochthons were translated towards ESE across the western Baltica margin to their present locations. Being the uppermost tectonic unit in central and southern Scandinavia to contain an assemblage of well preserved rocks, the Tossa˚sfja¨llet Group provides the most reliable information from the outermost possible succession more than 500 km WNW of its present location (Kumpulainen & Nystuen 1985; Gayer & Greiling 1989; Nystuen et al. 2008). A pre-Iapetus plate tectonic reconstruction of Baltica and adjacent crustal units (Greiling & Smith 2000) places Scotland close to the western (present coordinates) Baltica margin, suggesting a juxtaposition(?) of the Tossa˚sfja¨llet succession with the Dalradian succession (including the Port Askaig Fm.) in southwestern Scotland.

Stratigraphy The unfossiliferous, Neoproterozoic Tossa˚sfja¨llet Group (Fig. 61.2) was divided by Kumpulainen (1980) into the Lunndo¨rrsfja¨llen (alluvial), Kra˚khammaren (alluvial-marine), Stora˚n (peritidal), Lillfja¨llet (glaciogenic), and Lo¨van (marine/lacustrine-alluvial) formations. The stratigraphic position of the Lundo¨rrsfja¨llen Fm. is uncertain, because it is separated from the main nappe body by a fault zone (Gilotti & Kumpulainen 1986, fig. 5) and hence correlation between the successions is currently not possible. Otherwise the other formations may be traced across the nappe from SW to NE (Fig. 61.1). Gilotti & Kumpulainen (1986) identified in the Stor-Lo¨vsjo¨n area one more stratigraphic unit, an ‘unnamed unit’ (UNU). On the map (Fig. 61.3), the UNU exceeds 500 m in thickness and occupies the area between the

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 629– 634. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.61

630

R. A. KUMPULAINEN

Fig. 61.1. Generalized map of the main outcrop area of the Sa¨rv nappe in the southern Swedish Caledonides, distinguishing between pre-glacial rocks, on the left, and post-glacial rocks, on the right, separated by the Lillfja¨llet Fm. (dotted line) (modified from Kumpulainen 1980). The inset map shows the location of the study area within the Caledonian Orogen in Scandinavia.

poorly exposed Stora˚n Fm. in the west and Lillfja¨llet Fm. diamictite beds in the east. Bedding in this unit conforms to that of the regional structure (Gilotti & Kumpulainen 1986, figs 4 & 5), justifying the interpretation that the UNU is part of the Sa¨rv Nappe stratigraphy. This new stratigraphic unit warrants a modification of the original definition of the Lillfja¨llet Fm. The new interpretation of the Lillfja¨llet Fm. stratigraphy (Fig. 61.4) places the whole Lillfja¨llet section in the base of the formation (on Fig. 61.3) between the Stora˚n Fm. and the UNU. However, to confirm this hypothesis requires a drilling operation. Resting on top of the UNU, the diamictite units in the Stor-Lo¨vsjo¨n area are placed in the upper part of the formation. Hence, in the present account the UNU is tentatively interpreted as a member within the Lillfja¨llet Fm. (Figs 61.2 & 61.4). In this chapter, the discussion concerns the development of the Lillfja¨llet Fm. and its relation to the Stora˚n (dolomite) and Lo¨van (sandstone and conglomerate) formations (Figs 61.2 & 61.3).

Glaciogenic and associated deposits The Lillfja¨llet Fm. On the northern slope of the Lillfja¨llet hill (628330 N/28500 E), five diamictite units and four sandy mudstone and sandstone units (Figs 61.1 & 61.5) occur with a collective thickness of c. 120 m. The thickness of the individual diamictite units (informally D1 to D5) varies from c. 0.05 m to c. 50 m, whereas the sandy mudstone units are less than 2 m thick, except for the uppermost, which is c. 16 m thick. The diamictite is massive, and stratification has been observed only at the base of the D1 and D4 units. The matrix in the various diamictite units is sand-dominated, poorly sorted, and contains, particularly in the lowermost diamictite, a significant proportion of sand-grade clastic dolomite debris. The

Fig. 61.2. Stratigraphy of the Tossa˚sfja¨llet Group (modified from Kumpulainen 1980). The location of the Lunndo¨rrsfja¨llen Fm. in the Sa¨rv Nappe stratigraphy is uncertain as it may compose its own tectonic subunit within the nappe. s, silt; m, medium sand; g, gravel.

clast content in the lowermost diamictite bed is c. 25% dolomite, 38% other sedimentary rocks and 38% igneous rocks (granite, felsic porphyry and occasional greenstone). Higher up, the content of sedimentary clasts (including subordinate dolomite) varies between c. 10% and 25%, whereas crystalline clasts vary between 75% and 90%. The diamictite units are interrupted by units of horizontally laminated sandy mudstone and weakly stratified dolomitic sandstone. In the thick mudstone unit between D4 and D5, lamination disappears gradually towards the upper part of the unit, where a few lonestones are found, the largest being c. 0.4  0.6 m in size. The lower and upper boundaries of D1 are not exposed. The first mudstone unit grades over a short interval into D2. The top of D2 is undulating and sharply overlain by the second mudstone – sandstone. The base and the top of the thin D3 are sharp, as are the base and the top of D4; the top of D4 is cut erosively by a thin conglomerate at the base of the uppermost mudstone unit. The base of D5 is sharp. No striated stones have been observed in this section, which may be because these rocks, including the clasts, split across, rather than along the matrix –clast boundary. The UNU in the Stor-Lo¨vsjo¨n area is more than 500 m thick and is composed of grey, distinctly parallel-laminated, sandy mudstone, in which neither lonestones nor cross-bedded laminae have been observed. In the same area (628580 N/138210 E), the UNU grades gradually upwards to the lower of the two diamictite beds (Fig. 61.4). These two diamictite beds, c. 5–8 m (lower) and c. 5 m (upper), are separated by a unit of laminated mudstone. The succession here displays a stretching lineation, which to some extent makes examination difficult. The diamictite units are dominated by igneous clasts;

¨ LLET FORMATION NEOPROTEROZOIC GLACIOGENIC LILLFJA

Fig. 61.3. Geological map of the Stor-Lo¨vsjo¨n area. The Tossa˚sfja¨llet Group stratigraphy occurs in the steeply standing succession around Stor-Lo¨vsjo¨n (modified from Gilotti & Kumpulainen 1986). The Stora˚n Fm. occurs west of Buavan as a poorly exposed narrow belt resting on top of the shallow-marine Kra˚khammaren Fm.; neither the lower nor the upper boundary of the Stora˚n Fm. were observed in this area. The next belt to the east is composed of laminated sandy mudstone, the ‘unnamed unit’ (UNU). The UNU grades upwards to diamictite in the uppermost part of the Lillfja¨llet Fm. In this account the UNU is interpreted as a member of the Lillfja¨llet Fm. Key: 1, Kra˚khammaren Fm. (sandstone and shale); 2, Stora˚n Fm. (dolomite, shale and sandstone); 3, UNU (sandy mudstone); 4, Lillfja¨llet Fm. (diamictite); 5, lower part of the Lo¨van Fm. (mudstone to sandstone); 6, upper part of the Lo¨van Fm. (gravelly sandstone); 7, sandstone of uncertain affiliation (Lundo¨rrsfja¨llen Fm.). The inferred lower dolomite-dominated unit in the lower part of the Lillfja¨llet Fm. is not shown on this map.

the largest observed, which is composed of syenite, measures c. 1.6  0.8 m in size. These two diamictite beds are separated by a unit of thinly laminated sandstone and mudstone with some minor soft-sediment deformation structures. The upper boundary of the mudstone with the upper diamictite is not exposed. The most spectacular details exposed at this locality are clastic wedges cutting into the top of each of the two diamictite beds; at least four in the lower and five in the upper have been observed. One wedge from each unit is described in more detail (Fig. 61.6). The bowl-shaped upper end of the wedge in the lower diamictite bed (Fig. 61.6a) is c. 0.8 –0.9 m wide. Its thickness decreases rapidly downwards and at c. 1 m depth it is less than 0.2 m wide. Its total depth is almost 3.5 m. The grain size in the wedge decreases downwards from very poorly sorted coarse sandy to gravely infill to fine-grained sandstone. Other bowl-shaped depressions occur on top of this particular diamictite unit and appear to form a semi-continuous bed on top of the lower diamictite. Laminated mudstone and a bed of deformed coarse sand succeed this diamictite. No lonestones have been observed in this laminated unit. The other clastic wedge, which cuts the top of the upper diamictite, is more wedge-shaped (Fig. 61.6b). Its upper end is less than 0.4 m wide and extends downwards at least 3 m. The infill of this wedge is vertically laminated. Similar to the other wedge, the grain size here also decreases downwards from coarse to fine sand. The upper end of the wedge is sharply cut by a poorly sorted

631

Fig. 61.4. Stratigraphy of the Lillfja¨llet and adjacent formations (not to scale). The Kra˚khammaren Fm. subjacent to the Stora˚n Fm. and the upper part of the Lo¨van Fm. are not shown here. The two sections, Lillfja¨llet section and the Stor-Lo¨vsjo¨n section, respectively, indicate in this diagram what stratigraphic intervals they represent in the Tossa˚sfja¨llet Group, respectively. The UNU was probably never deposited in the Lillfja¨llet area.

coarse-grained sandstone bed, c. 0.1–0.2 m thick. Within this sandstone bed and resting on the diamictite, there is a faceted, flat-based clast (c. 0.1  0.25 cm) with a convex upper side. This diamictite unit is also overlain by a laminated and small-scale cross-laminated sandy mudstone (0.6 m thick) and a soft-sediment deformed sandstone bed (0.2 m thick). One small lonestone of sand grade occurs in the laminated mudstone c. 0.4 m above the diamictite top.

Boundary relationships to the underlying and overlying non-glacial units In the Lillfja¨llet section, the glaciogenic Lillfja¨llet Fm. (Kumpulainen 1980) rests on thinly bedded dolomite of the Stora˚n Fm. (Fig. 61.5). The exact boundary is hidden in a c. 1–2 m large unexposed interval between the formations. The Stora˚n Fm. contains casts after evaporitic anhydrite that was replaced by chalcedony during diagenesis. The chalcedony was then recrystallized to microcrystalline quartz by the Caledonian metamorphism (Kumpulainen 1980, fig. 11). The Stora˚n Fm. also contains occasional sandstone and mudstone beds. In this section, the Lillfja¨llet Fm. is overlain by coarse-grained sandstones of the Lo¨van Fm., but the boundary area is unexposed and corresponds to a stratigraphic thickness of a few tens of metres. In the Stor-Lo¨vsjo¨n area (Fig. 61.3), a narrow belt of locally derived dolomite debris suggests the presence of the Stora˚n Fm. striking parallel and next to the underlying Kra˚khammaren Fm.

632

R. A. KUMPULAINEN

Fig. 61.5. (a) Profile along the northern Lillfja¨llet hill side, where a dolerite dyke embraces lenses of Stora˚n and Lillfja¨llet formations. (b) Logged section of the sedimentary succession in the hill side. No pattern in the top of the upper mudstone indicates massive mudstone. (c) Detail of (b). Modified from Kumpulainen (1980).

Neither the lower nor the upper boundary of the Stora˚n Fm. has been observed along this belt. The next exposed rock unit to the east of the Stora˚n Fm. is the sandy mudstone unit, UNU. The unexposed ground on the map between these two belts of rocks is wide enough to accommodate a hypothetical lower diamictitedominated unit corresponding to the Lillfja¨llet section as suggested

above in the new interpretation of the Lillfja¨llet Fm. stratigraphy (Figs 61.2 & 61.4; see ‘Discussion’). The upper boundary of the Lillfja¨llet Fm. is located on top of the sandstone bed 1 m above the second diamictite (Fig. 61.6b). A c. 4-m-thick unit of laminated sandy mudstone resting sharply on top of the Lillfja¨llet Fm. is referred to the base of the Lo¨van Fm. This mudstone bed grades upwards to thinly bedded sandstone (Fig. 61.4), which coarsens gradually into trough-cross-bedded coarse-grained sandstone and conglomerate.

Chemostratigraphy No data are available from this succession.

Palaeolatitudes No data are available from Tossa˚sfja¨llet Group. Data from other successions (Greiling & Smith 2000; Hartz & Torsvik 2002; Cawood & Pisarevsky 2006) imply high southern latitudes for this area at the time of deposition of the Lillfja¨llet Fm.

Geochronology Although no data are available from the Tossa˚sfja¨llet sedimentary succession itself, the group is cut by (Fig. 61.4a) the Ottja¨llet dolerite dyke swarm, which has given an 40Ar/39Ar age of 665 + 10 Ma (Claesson & Roddick 1983). These dolerites are commonly correlated with extensive dyke swarms in other parts of the Scandinavian Caledonides, providing ages in the interval c. 610 –550 Ma (Svenningsen 2001; Paulsson & Andre´asson 2002) and also with the Egersund dolerite dykes in southwestern Norway (c. 616 Ma; Bingen et al. 1998). New dating of the Ottfja¨ll dolerites is necessary to better correlate successions from one region to another in Scandinavia. Fig. 61.6. Two clastic wedges cutting tops of two diamictite beds, lower (a) and upper (b), respectively, on the western shore of Storavan. Note that the framework clasts in both diamictites are parallel to the wedges, being probably the result of the approximately east–west-trending stretching of the nappe body and simultaneous rotation of the clasts. This implies also that the wedges have been stretched and elongated. The original depth of the wedges is estimated at c. 2.5– 3 m.

Discussion A striking feature of the Neoproterozoic sedimentary successions in the central and southern Scandinavian Caledonides is that peritidal, evaporitic, dolomite extends over the whole region and is succeeded by presumably glaciogenic sediments attesting to a

¨ LLET FORMATION NEOPROTEROZOIC GLACIOGENIC LILLFJA

significant climate change in Neoproterozoic times (Nystuen & Lamminen 2011; Kumpulainen & Greiling 2011). The extensive dolomite unit additionally supports the interpretation that the Neoproterozoic landscape was close to peneplained and just above the sea level of that time. There are a range of criteria that have been described to support the glaciogenic origin of a certain rock unit or a stratigraphic section (e.g. Hambrey & Harland 1981; Hambrey 1994). A number of these criteria have been described from the Lillfja¨llet Fm., suggesting that it is of glaciogenic origin. In the Lillfja¨llet section, the glaciogenic evidence includes (i) the presence of diamictite and (ii) the presence of lonestones, that is, probable ice-rafted debris, although they are few (see below). The StorLo¨vsjo¨n area additionally offers (iii) periglacial clastic wedges, (iv) periglacial involution-like structures, and (v) a flat stone on top of the upper diamictite (Fig. 61.6b) may be glacially faceted and subsequently reworked by wind action. No striated pavements and striated stones have been found to date. Soft-sediment deformation in the stratified units in Lillfja¨llet is very subordinate (Fig. 61.5b,c). The absence of lamination in the upper part of the thick sandy mudstone unit may be due to partial liquefaction, possibly due to rapid deposition of the next diamictite unit and rapid increase of the pore-water pressure. The interpretation of the origin of the large lonestone in the massive upper part of this sandy mudstone is not straightforward, because the possible deformation structures around that boulder are obliterated. The interpretation that the Lillfja¨llet succession was overrun by five consecutive glaciers is not justified. More likely, the Lillfja¨llet succession was formed sub-aqueously, the diamictites being meltout tills or redeposited tills (debris flows), rather than having been in contact with glacier ice. Also in the Stor-Lo¨vsjo¨n area, the two diamictite units are massive and may have been deposited sub-aqueously; the basal contact of the lower unit is gradual and may indicate that this diamictite unit was dumped on its laminated muddy substratum. The base of the upper diamictite is not exposed, providing no further evidence as to how the upper diamictite was formed. The tops of both of the diamictite units are sharp, and both are cut by sandstone wedges forming a network on both upper surfaces. The lower top surface additionally displays bowl-shaped structures, possible involution structures (cf. French 1976; Miller 1996). These features support the interpretation that, although they may have been formed sub-aqueously, they were later subject to sub-aerial conditions. The clastic wedges in the Stor-Lo¨vsjo¨n area indicate permafrost conditions. Wedge (a), in the lower diamictite, is a probable periglacial ice-wedge cast: a soil wedge (cf. Washburn 1980). With no evidence of erosion, this frost-reworked land surface was submerged before the deposition of the laminated mudstone. Wedge (b) (Fig. 61.6b) is different in origin, although it too formed sub-aerially. The vertically laminated sand inside this wedge indicates that the wedge is a sand wedge (Pe´we´ 1959), which formed in permafrost areas with very low precipitation and without snow cover. It was not filled originally with ice, but with dry sand as the frost crack opened. The vertical lamination is produced after a series of openings and fillings. After the formation of wedge (b), erosion removed some of the land surface including a sand-filled depression that is commonly located above a sand wedge (Berg & Black 1966). A poorly sorted gravelly sandstone was formed possibly by winnowing on the erosional surface. Subsequent submergence of the land surface allowed the deposition of a laminated mudstone and the succeeding sandstone in the top of the Lillfja¨llet Fm. Permafrost-related sand wedges may penetrate several metres into the ground (Berg & Black 1966). In the Stor-Lo¨vsjo¨n area, the estimated original depth of the wedges is c. 2.5– 3 m, which is probably too deep for a wedge to be formed in a perennially active frozen layer. This suggests that the glaciogenic Stor-Lo¨vsjo¨n succession was

633

subjected twice to permafrost conditions, due to temporary emergence of the depositional site. The major difference in stratigraphy between our two sub-areas is the presence of the UNU in the Stor-Lo¨vsjo¨n area and its apparent absence in the Lillfja¨llet area. This could be for tectonic or depositional reasons. The Sa¨rv nappe is composed of coherent tectonic lenses separated by shear zones. No such deformation zone is observed to cut the stratigraphy in the Lillfja¨llet area, which may be due to limited exposure, but the inferred absence of such a deformation zone is corroborated by the fact that the dyke swarm preserves its parallel trend over Lillfja¨llet. A shear zone would be identified by an offset in the dykes’ trend. Hence, tectonic thinning or cutting is less probable here. The UNU is a more than 500-m-thick monotonous unit of grey parallel-laminated sandy mudstone, where neither lonestones nor cross-laminated units have been observed. The distinct, parallel lamination suggests that this unit was deposited in standing body of water, probably a pro-glacial lake, where currents were too weak to produce traction deposits. Also, no mass-flow deposits have been observed. The thickness of the unit requires a significant accumulation space or a regular subsidence of the basin floor. Evidence described above suggests that the Lillfja¨llet Fm. is glaciogenic, and its depositional development may be summarized as follows. The lower part of the formation was deposited on top of a peritidal dolomite unit, the Stora˚n Fm., as described from the Lillfja¨llet section (Figs 61.4 & 61.5). The various diamictite units were deposited either as sub-aqueous meltout tills or debris-flow units. The diamictite deposition does not require repeated ice advance and retreat events. A paucity of lonestones (and possible dropstones) suggests that icebergs were not able to travel these waters. The present interpretation also suggests that a corresponding lower diamictite-dominated section was deposited on top of the Stora˚n Fm. in the Stor-Lo¨vsjo¨n area (Fig. 61.3); this section could be thinner or thicker than that in the Lillfja¨llet section. For a confirmation of this hypothesis, drilling is required in this area. The UNU occupies the middle part of the Lillfja¨llet Fm. stratigraphy (Figs 61.2 & 61.4). It occurs as a thick unit in the StorLo¨vsjo¨n area, but appears to be absent in the Lillfja¨llet section. This thickness change is interpreted here as a difference in the sedimentary system, a large probable pro-glacial lake, where the StorLo¨vsjo¨n area was the depocentre of the UNU and the Lillfja¨llet section was located outside that depocentre. The fine-grained clastic material in this unit was provided by glacial melt waters. The scarcity of dropstones suggests that the glacial margin at that time was located on land or otherwise isolated from this depositional area. The Lillfja¨llet Fm. has been correlated previously (Kumpulainen 1980) with the glaciogenic Moelv Fm. in southern Norway, the La˚ngmarkberg Fm. in central Scandinavia and the glaciogenic deposits in northern Norway. Since the group carries no fossils, the Ottfja¨llet dolerite dyke swarm remains the only significant basis for a correlation of the Lillfja¨llet Fm. with other successions in Scandinavia. Earlier dating methods for mafic rocks have given very variable ages, and the 40Ar/39Ar age of 665 + 10 Ma (Claesson & Roddick 1983) may contain a principal methodological error. Speculating that the various dyke swarms that produce ages of c. 600–615 Ma could be correlated with the Ottfja¨llet dolerites, then the Lillfja¨llet Fm. must be older than that age interval, meaning that this probably glaciogenic formation can be correlated with neither the Gaskiers glaciation event (c. 580 Ma) nor the Marinoan glaciation event (c. 635 Ma). Modern dating techniques are required. The author wishes to thank T. Thelander and J. P. Nystuen for discussions about glacial deposits. N. Johansson provided valuable geographical information from the Stor-Lo¨vsjo¨n area. The manuscript was improved by the comments of E. Arnaud, an anonymous reviewer and G. A. Shields. The author is grateful for support from Stockholm University. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

634

R. A. KUMPULAINEN

References Bingen, B., Demaiffe, D. & van Breemen, O. 1998. The 616 Ma old Egersund basalt dike swarm, SW Norway, in the context of lateNeoproterozoic opening of the Iapetus Ocean. Norsk Geologisk Tidsskrift, 79, 69 –86. Berg, T. E. & Black, R. F. 1966. Preliminary measurements of the growth of nonsorted polygons, Victoria Land, Antarctica. American Geophysical Union, National Academy of Sciences – National Research Council Publication, 1418, 61– 108. Cawood, P. A. & Pisarevsky, S. A. 2006. Was Baltica right-way-up or upside down in the Neoproterozoic? Journal of the Geological Society London, 163, 753– 759. Claesson, S. & Roddick, J. C. 1983. 40Ar/39Ar data on the age and metamorphism of the Ottfja¨llet Dolerites, Sa¨rv Nappe, Swedish Caledonides. Lithos, 16, 61– 73. French, H. M. 1976. The Periglacial Environment. Longman Group, New York. Gayer, R. A. & Greiling, R. O. 1989. Caledonian nappe geometry in north-central Sweden and basin evolution on the Baltoscandian margin. Geological Magazine, 126, 499– 513. Gee, D. G., Kumpulainen, R., Roberts, D., Stephens, M. B., Thon, A. & Zachrisson, E. 1985. Scandinavian Caledonides –Tectonostratigraphic Map, Scale 1:2,000,000. Sveriges geologiska underso¨kning Ba, 35. Gilotti, J. A. & Kumpulainen, R. 1986. Strain softening induced ductile flow in the Sa¨rv thrust sheet, Scandinavian Caledonides. Journal of Structural Geology, 8, 441– 455. Greiling, R. O. & Smith, A. G. 2000. The Dalradian of Scotland: missing link between the Vendian of northern and southern Scandinavia? Physics and Chemistry of the Earth, 25, 495– 498. Hambrey, M. J. 1994. Glacial Environments. University College London Press, London. Hambrey, M. J. & Harland, W. B. 1981. Part I Introduction. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 3 – 28. Hartz, E. H. & Torsvik, T. H. 2002. Baltica upside down: a new plate tectonic model for Rodinia and the Iapetus Ocean. Geology, 30, 255– 258. Holmqvist, P. J. 1894. Om diabasen pa˚ Ottfja¨llet. Geologiska Fo¨reningens i Stockholm Fo¨rhandlingar, 16, 175– 192. Kumpulainen, R. 1980. The Upper Precambrian stratigraphy and depositional environments of the Tossa˚sfja¨llet Group, Sa¨rv Nappe, southern Swedish Caledonides. Geologiska Fo¨reningens i Stockholm Fo¨rhandlingar, 102, 531–550. Kumpulainen, R. 1981. The Late Precambrian Lillfja¨llet Formation in the southern Swedish Caledonides. In: Hambrey, M. J. &

Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 620– 623. Kumpulainen, R. & Nystuen, J. P. 1985. Late Proterozoic basin evolution and sedimentation in the westermost part of Baltoscandia. In: Gee, D. G. & Sturt, B. A. (eds) The Caledonide Orogen – Scandinavia and Related Areas. John Wiley & Sons, Chichester, 213– 232. Kumpulainen, R. A. & Greiling, R. O. 2011. Evidence for late Neoproterozoic glaciation in the central Scandinavian Caledonides. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 623–628. Miller, J. M. G. 1996. Glacial sediments. In: Reading, H. G. (ed.) Sedimentary Environments, Facies and Stratigraphy. Blackwell Science, Oxford, 454– 484. Nystuen, J. P. & Lamminen, J. T. 2011. Neoproterozoic glaciation of South Norway: from continental interior to rift and pericratonic basins in western Baltica. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 613–622. Nystuen, J. P., Andresen, A., Kumpulainen, R. & Siedlecka, A. 2008. Neoproterozoic basin evolution in Fennoscandia, East Greenland and Svalbard. Episodes, 31, 35 –43. Pe´we´, T. L. 1959. Sand-wedge polygons (tessellations) in the McMurdo Sound region, Antarctica – in progress report. American Journal of Science, 257, 545– 552. Paulsson, O. & Andre´asson, P. G. 2002. Attempted break-up of Rodinia at 850 Ma; geochronological evidence from the Seve – Kalak Superterrane, Scandinavian Caledonides. Journal of the Geological Society London, 159, 751–761. Ro¨shoff, K. 1975. A probable glaciogenic sediment in the Sa¨rv Nappe, central Swedish Caledonides. Geologiska Fo¨reningens i Stockholm Fo¨rhandlingar, 97, 192– 195. Stodt, F., Rice, A. H. N., Bjo¨rklund, L., Bax, G., Halverson, G. P. & Pharaoh, T. 2011. Evidence for late Neoproterozoic glaciation in the Caledonides of NW Scandinavia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 603–611. Stro¨mberg, A. 1961. On the tectonics of the Caledonides in the southwestern part of the county of Ja¨mtland, Sweden. Bulletin of the Geological Institutions of the University of Uppsala, 39, 1– 92. Svenningsen, O. M. 2001. Onset of seafloor spreading in the Iapetus Ocean at 608 Ma; precise age of the Sarek Dyke Swarm, northern Swedish Caledonides. Precambrian Research, 110, 251– 254. Washburn, A. L. 1980. Cryogeology; A Survey of Periglacial Processes and Environments. Edward Arnold, London.

Chapter 62 The Port Askaig Formation, Dalradian Supergroup, Scotland EMMANUELLE ARNAUD1* & IAN J. FAIRCHILD2 1

School of Environmental Science, University of Guelph, Guelph, Ontario, N1G 2W1, Canada

2

School of Geography, Earth and Environmental Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK *Corresponding author (e-mail: [email protected]) Abstract: The Port Askaig Formation (Fm.) is a thick glaciogenic succession within the Dalradian Supergroup that consists of over 700 m of variably dolomitic diamictite, conglomerate, sandstone mudstone and minor dolomite, and is bounded by mixed siliciclastic –carbonate successions of the Islay (Lossit) and Bonahaven formations. These strata are exposed in the metamorphic Caledonides of Scotland, although excellent preservation of sedimentary structures can be found at several sites. An extensional setting for this succession has been proposed based on stratigraphic and structural arguments. The available chemostratigraphic data include the Chemical Index of Alteration, d13C and Sr-isotope values. Palaeomagnetic analyses have been shown to be subject to post-depositional Caledonian overprinting. There is also continued debate over the regional palaeogeographical reconstructions of the Scottish promontory for this time period. The succession is chronologically poorly constrained with U– Pb analyses of stratigraphically much higher or lower deposits. The thick succession is thought to record glacially influenced marine sedimentation and reworking of unstable sediments in a tectonically active setting with evidence of ice-margin fluctuations. Alternative palaeoenvironmental interpretations that focus on glacial terrestrial processes and emphasize climatic influence instead of tectonic activity have also been proposed. The overlying carbonate is a lithologically diverse coastal complex and so does not fit the Neoproterozoic norm. Research has to date focused on the stratigraphic and sedimentological aspects of this succession, as well as some of the broader palaeogeographical and structural features of the Dalradian basin. Future efforts should focus on the chronological, structural and palaeogeographical constraints of this succession.

The Port Askaig Fm. is exposed at several sites in Scotland, including Islay, the Garvellach Islands, Schichallion, Braemar, Muckle Fergie Burn and Fordyce (Spencer & Pitcher 1968; Spencer 1971; Litherland 1980) as well as in Donegal, Mayo and Connemara in Ireland (Howarth 1971; Tanner & Shackleton 1979; Max 1981). The type section is located at Port Askaig on the island of Islay, with several good along-dip outcrops over several kilometres of shoreline and nearby moorland (Spencer 1971). The best outcrops can be seen on the Garvellach Islands, with lateral and vertical exposures of hundreds of metres to several kilometres, though only the lowermost three members of the formation are exposed there (Fig. 62.1). These exposures are of phenomenal quality and, compared with other Neoproterozoic successions, relatively easy to access. They provide abundant sedimentological information as well as an excellent sense of lateral and vertical facies variability and enable relatively detailed palaeoclimatic reconstructions. Other sites in Scotland and Ireland tend to have limited outcrop exposure, complicated structural relationships, and higher degree of metamorphism. The Port Askaig Fm. is also known as the Port Askaig Tillite in the literature, but this genetic term is avoided here in accordance with modern stratigraphic practice. The Port Askaig Fm. outcrops are significant in providing the first record of glaciogenic deposits now known to be Precambrian (Thomson 1871). Subsequent Survey mapping and other works carried out until the early 1930s (references in Spencer 1971) provided more geological context, but with little consciousness of their international significance. Bailey (1916) was the first to use way-up indicators in these rocks and to demonstrate the simplicity of the structure. Later, the stratigraphy and origin of these deposits became the focus of several detailed studies (Kilburn et al. 1965; Spencer 1971; Eyles & Eyles 1983; Eyles 1988; Arnaud & Eyles 2006). The most comprehensive study was Spencer’s (1971) work, in which the variable sedimentology and regional stratigraphy were documented at various sites in Scotland and Ireland. The majority of other studies have focused on the excellent exposures on the Garvellach Islands. Additional papers have focused on specific aspects of this thick succession such as provenance (Anderton 1980; Fitches et al. 1996; Evans et al. 1998; Cawood et al. 2003), the sandstone intrusions or

wedges (Eyles & Clark 1985), the Great Breccia (Arnaud & Eyles 2002; Benn & Prave 2006), the giant cross-bedded sandstone (Arnaud 2004) and the origin of associated carbonate strata (Fairchild 1980a). Anderton (1982, 1985), Yardley et al. (1982), Harris et al. (1978, 1993) and Prave (1999) provide valuable information on the depositional setting and tectonic evolution of the Dalradian Supergroup, while discussing ongoing controversies. Harris et al. (1993) also provides a useful set of lithostratigraphic columns of the Dalradian Supergroup at selected sites in Ireland, Scotland and Shetland. Field-trip guidebooks are available (itineraries I, II and III in Hambrey et al. 1991; Arnaud & Shields 2005). Sedimentological studies of the overlying Bonahaven Fm. (Spencer & Spencer 1972; Fairchild 1977, 1980a, b, 1985b) have recently been supplemented with chemostratigraphic studies of the bounding formations that constrain their global context (Brasier & Shields 2000; McCay et al. 2006; Prave et al. 2009). The underlying carbonates have been described on the Garvellachs (Spencer 1971), but are less well exposed and less studied on Islay, although an updated map of northern Islay has been published (British Geological Survey 1994).

Structural framework The Port Askaig Fm. and associated strata are part of the thick Dalradian Supergroup exposed within the metamorphic Caledonides between the Great Glen Fault and the Highland Boundary Fault. Regionally, the Dalradian Supergroup rocks were affected by several phases of deformation and folding associated with the Caledonian orogeny (Treagus 1987). Slides or thrust faults that lie sub-parallel to bedding and are thought to have last moved in Ordovician times, and Devonian-age granitic intrusions are also present in the region (Harris et al. 1993). The outcrops at Port Askaig are exposed in the relatively structurally simple NNEtrending Islay anticline (Bailey 1916; Fairchild 1980c). On the Garvellach Islands, the beds dip uniformly to the south and SE (c. 358). Tertiary-age dolerite dykes cross-cut the strata. The Argyll Group, which includes the Port Askaig Fm. at its base (Fig. 62.2), is thought to have accumulated in an extensional

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 635– 642. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.62

636

E. ARNAUD & I. J. FAIRCHILD

Fig. 62.1. Map showing location of Port Askaig Fm. outcrops on the Garvellach Islands (modified from Spencer 1971). Detailed map of outcrops on Islay can be found in Spencer (1971) and Hambrey et al. (1991). Note the lithology ‘sandstone’ includes from 0 to 90þ% dolomite at different horizons.

basin experiencing increasing tectonic activity and development of fault-bounded sub-basins prior to the opening of the Iapetus ocean (Anderton 1982; Harris et al. 1978, 1993). Evidence that basin extension started during accumulation of Port Askaig Fm. sediments includes rapid lateral thickness changes over short distances (Anderton 1982, 1985). For example, the Port Askaig Fm. appears to thin rapidly westwards on Islay to the limit of its outcrop (British Geological Survey 1994) and the overlying Bonahaven Fm. thins significantly west across the Bolsa fault in this region (Fairchild 1980c; Anderton 1985). There is also a coarse conglomerate facies of reworked limestone in the Lossit Limestone in the western part of its outcrop (British Geological Survey 1994). It is within this context that the Ordovician-age Grampian slides were proposed to be Proterozoic synsedimentary listric faults that were later re-activated, based on their association with rapid thickness changes at various stratigraphic levels within the Dalradian Supergroup (Soper & Anderton 1984; Anderton 1985, 1988). Most recently, some of the diamictite units as well as repeated horizons of soft-sediment deformation structures have been interpreted as additional indicators of early extensional tectonic activity in Argyll times (Arnaud & Eyles 2002, 2006). The rocks have undergone deformation and metamorphism as a result of the Caledonian orogeny, with open folding and upright cleavage developed in the least affected areas of Islay, Garvellachs and Northern Donegal (Spencer 1971; Fairchild 1980b). Locally there are high-strain areas (Borradaile 1979; Fairchild 1980c) particularly nearing bounding thrusts, re-activated faults and some major lithological boundaries (Anderton 1985; Fairchild 1985b; Treagus & Treagus 2002). Pressure-solution effects are widespread more generally, with limestone and diamictite much more strongly deformed than dolostone or quartz sandstone (Fairchild 1985b). In the most-studied locations at Port Askaig and in the

Garvellach Islands, sedimentary structures are well preserved, although noticeably deformed in cleaved lithologies.

Stratigraphy The Port Askaig Fm. is found within the predominantly marine succession of the Argyll Group in the Dalradian Supergroup (Fig. 62.2), being underlain by the Islay (Lossit) Limestone (mixed dolomitic-siliciclastic at its top) and overlain by the mixed dolomitic –siliciclastic Bonahaven Fm. The Port Askaig Fm. is over 700 m thick. Spencer (1971) defined five members based on predominant facies and clast lithology (Fig. 62.3). Members I to III are best exposed on the Garvellach Islands, whereas Members IV and V are only exposed on the island of Islay. The lowermost member (Member I, Beannan Buidhe) consists primarily of stacked carbonate-rich diamictite with commonly discontinuous sandstone and conglomerate interbeds. Member I also contains a thick diamictite bed called the Great Breccia, with very large clasts in a muddy sandstone matrix and a series of folded and boudinaged interbedded units of variable lithologies (mudstone, sandstone, dolomite and diamictite) commonly referred to as the ‘Disrupted Beds’. Member II (An Tamhanachd) consists of diamictite associated with sandstone and mudstone interbeds with a base defined by the appearance of extrabasinal clasts. Member III (Creagan Loisgte) consists of thick packages of laterally continuous sandstone interbedded with diamictite. Member IV (Ruahd Phort Beag) is again dominated by diamictite, whereas Member V (Con Tom) is dominated by sandstone with minor diamictite. Significant lateral changes to the west on Islay were mentioned in the previous section. Within the Port Askaig Fm., the succession on the island of Islay is thin at the base with the unique facies of the overlying Disrupted Beds being the first unequivocal lithostratigraphic tie to the Garvellachs, but which is separated from the underlying carbonates by ,6 m of sandstone and dolomitic conglomerate correlated with the Great Breccia by Spencer (1971). The remainder of Members I, II and III are broadly similar between Islay and the Garvellachs.

Glaciogenic deposits and associated strata The Port Askaig Fm.

Fig. 62.2. Generalized stratigraphy and tectonic setting of the Port Askaig Fm. and associated strata. P, Port Askaig Fm.; I, Islay Limestone; B, Bonahaven Fm.; J, Jura Quartzite; S, Scarba conglomerate; *earthquake-induced liquefaction features (modified from Arnaud & Eyles 2006).

The following is a summary of a relatively thick, complex and superbly exposed succession based on work by Kilburn et al. (1965), Spencer (1971), Eyles & Clark (1985), Eyles (1988), Arnaud (2002), Arnaud & Eyles (2002, 2006), Arnaud (2004) and Benn & Prave (2006). The Port Askaig Fm. consists of a thick succession of diamictite (44 distinct units), interbedded with sandstone and minor units of mudstone and conglomerate. Diamictite are massive to stratified with clasts up to several metres in diameter floating in a siltstone to silty sandstone matrix. Some diamictite beds exhibit coarse-tail inverse grading, and many contain sandstone stringers or inclusions of variable lithologies (Arnaud & Eyles 2006). Some of the diamictite units

PORT ASKAIG FORMATION, SCOTLAND

637

Spencer 1971; Arnaud & Eyles 2002; Benn & Prave 2006 for detailed descriptions). Sandstone in the Port Askaig Fm. exhibits a variety of characteristics including a wide range of textures (fine to very coarse), sorting (poorly to well sorted), lithologies (dolomitic to quartzitic) and structures (massive, horizontally bedded, cross-bedded and deformed). Thick packages of sandstone within Member III are particularly notable as they contain thick to very thick sets of cross-bedded sandstone (individual set thickness averages 3 m, maximum is 11 m; Arnaud 2004). The lowermost package of these giant cross-bedded sandstone exhibits a preferred southerly palaeocurrent direction, whereas cross-bedded sandstone in the rest of the succession is more variable (Spencer 1971; Arnaud 2004). Other interbeds within the Port Askaig Fm. include finely laminated mudstone (rhythmites), (clastic) dolomite and massive to stratified conglomerate. Conglomerate often overlies and loads into diamictite, although some also occurs interbedded with sandstone and mudstone. Outsized clasts in laminated mudstones or diamictites occur at various horizons, but only limited examples displaying clear deflection of underlying laminae occur, especially in the Disrupted Beds (Spencer 1971; Hambrey et al.1991). Petrographic study of the carbonate in the Port Askaig Fm. indicates that it appears to be largely detrital in origin, with that in the rhythmite facies being secondary (Fairchild 1985a; Hambrey et al. 1991, p. 27, 34). Deformation structures have been documented at multiple distinct stratigraphic horizons. Several of these, namely sandstone wedges, sandstone dykes and sandstone downfold structures, have been examined in detail (Spencer 1971; Eyles & Clark 1985). The sandstone wedges are predominantly ,10 cm wide, have sharp irregular outer geometries and commonly penetrate up to several metres downward into diamictite units or into carbonate units within the underlying Islay Limestone. Some exhibit branching or polygonal patterns on bedding plane surfaces. Sandstone dykes are more tabular in form and intrude diamictite and siltstone. Sandstone downfold structures, refered to as load casts or ball and pillow structures by Arnaud & Eyles (2006), are also common on top of diamictite units or within interbedded sandstone and mudstone. These features, together with horizons of convolute and contorted bedding and pseudonodules, were shown to be found most commonly within Member II sediments on the Garvellach Islands (Arnaud & Eyles 2006).

The Lossit Limestone

Fig. 62.3. Stratigraphic log of the Port Askaig Fm. based on outcrops on the Garvellach Islands (Members I to III) and on the island of Islay (Members IV and V). GB, Great Breccia; DB, Disrupted Beds; XB, Giant cross-bedded sandstone (modified from Arnaud & Eyles 2002). Note the dolomite interbeds are considered detrital in origin (see text for discussion).

contain lenses of bedded siltstone sandstone and/or conglomerate (Spencer 1971). Clast and matrix lithology change up-section from predominantly intrabasinal and dolomitic at the base to extrabasinal and siliciclastic at the top. Pink granitoid extrabasinal clasts appear to have come from Palaeoproterozoic plutonic rocks of the Svecofennian-Makkovik-Ketilidian province (Evans et al. 1998). Beds are typically several metres to 10 m in thickness with sharp conformable to erosional basal contacts. Some basal contacts are gradational with associated laminated mudstone (Arnaud 2002; Arnaud & Eyles 2006). The Great Breccia is a particularly thick diamictite unit (up to 50 m), with megaclasts ranging from several metres to over 100 m in diameter (see

Strata underlying the Port Askaig Fm. have been known as the Islay Limestone and on the Garvellachs a succession of 72 m of predominantly mixed dolomitic –siliciclastic sediments with some stromatolitic and other limestone is present (Spencer 1971). There are some pure dolomicrite and dolomitic stromatolites, but for the most part the facies contain both terrigenous debris and reworked intrabasinal carbonate (Hambrey et al. 1991). The most interesting structures are crystal pseudomorphs originally illustrated by Spencer (1971). Similar structures from the Irish Dalradian were interpreted as glendonites, which are pseudomorphs after ikaite (Johnston 1995). On Islay, the top 300 m of the Islay Limestone have been redesignated as the Lossit Limestone Fm. (British Geological Survey 1994). Its topmost (Persabus) member (c. 80 m thick) consists of interbedded dolostone (locally stromatolitic or intraclastic), quartzite, slate and mixed lithologies and overlies a c. 70 m pure (Kiells) limestone member, locally oolitic, which in turn overlies several thick slate and limestone units.

The Bonahaven Fm. The lithostratigraphy and geological context of the overlying mixed carbonate –siliciclastic Bonahaven Fm. has been documented by Spencer & Spencer (1972), who divided it into four

638

E. ARNAUD & I. J. FAIRCHILD

members (Fairchild 1977, 1980a, b, 1985b; Hambrey et al. 1991). On the east coast of Islay, the sub-arkosic arenite at the top of the Port Askaig Fm. is succeeded by c. 65 m of sandy facies with mudstone and local dolomite at the top (Member 1), 25 m of quartzite (Member 2), c. 200 m of dolomitic sandstone, oolitic facies and mudrock (Member 3) and 55 m of heterolithic siliclastic rocks with a central 12 m of pure dolostone (Member 4). In turn this is succeeded by several kilometres of pure quartzite (Jura Quartzite). The structures at the base of the Bonahaven Fm., taken by previous workers to be desiccation cracks, have been reinterpreted as interstratal dewatering features by Tanner (1998). Primary carbonate is first encountered near the top of Member 1, where displacive dolomite fabrics in terrigenous mudstone with genuine desiccation cracks (Fairchild 1977, 1980b) are suggestive of dolocrete. The main carbonate unit (Member 3) consists predominantly of three dolomite facies (stromatolitic, bimodally cross-stratified intraclastic and/or oolitic sandstone, and mudstone with thin graded and/or wave-rippled sand) (Fairchild 1980a).

Boundary relations with overlying and underlying non-glacial units In eastern Islay, basal sandstone, clastic dolostone and diamictite of the Port Askaig Fm. variably rest erosionally on the underlying dolomite (at times stromatolitic) and thin-bedded quartzite and shale, which in turn rest on pure limestone (Spencer 1971; British Geological Survey 1994). Over 90 m of Member I exposed on the Garvellachs is missing here on Islay. Further west, Port Askaig Fm. diamictite units are less poorly exposed, but commonly are limited by sedimentary limestone breccias (British Geological Survey 1994). On the Garvellach Islands, diamictite sharply and erosionally overlies the Islay Limestone in places (Spencer 1971). In other areas of the Garvellach Islands (Dun Chonnuill & Garbh Eileach), the contact is gradational as shown by the appearance of rare clasts and the presence of siltstone (typical of the siltstone matrix of the overlying diamictite) interbeds in the upper part of the Islay limestone (Spencer 1971; Arnaud 2002; Arnaud & Eyles 2006). The upper contact of the Port Askaig Fm. is seen on Islay, where it is conformable with the Bonahaven Fm., but it is difficult to define where the last evidence of glacial phenomena occur as the dominant facies near the top is quartzite, with only rare thin diamictite horizons. The highest occurrence of chess-board albite, which is diagnostic of the granitic pebbles of the Port Askaig Fm. (Spencer 1971), is within a channelled conglomerate horizon, located c. 25 m above the base of the Bonahaven Fm. (Hambrey et al. 1991, p. 39, 40). The first occurrence of dolomite is found near the top of member 1 of the Bonahaven Fm. where it occurs in tidal flat facies (Fairchild 1977, 1980c) and is succeeded by more extensive dolomite facies in Member 3, with local undolomitized oolitic and micritic limestone (Fairchild 1980a).

Chemostratigraphy Panahi & Young (1997) carried out a study of major and trace element geochemistry, specifically utilizing the Chemical Index of Alteration (CIA). Most of the 21 samples were taken from the matrix of diamictite exposed on Garbh Eileah, Garvellach Islands. Analysis showed a decrease in CIA values from the base of the Port Askaig Fm. (values range from 68 to 77) to the upper part of Member III (values range from 60 to 68). The CIA values are an indication of the extent of weathering based on the relative proportions of alkali and alkaline earth elements and thus the decrease was interpreted as indicating a change in sediment source. The high CIA values reflect erosion of underlying sedimentary rocks, which had already experienced weathering, whereas the lower CIA values at the top of the section indicate incorporation of sediments from

relatively unweathered basement rocks (Panahi & Young 1997). In addition, analysis of the trace element geochemistry suggests erosion of shale developed on a post-Archaean crystalline basement, although a specific source area could not be identified. Brasier & Shields (2000) provided the first isotopic chemostratigraphic constraints, with reliable Sr-isotope values as low as 0.7067 being obtainable from the Sr-rich facies of pure Lossit (Islay) Limestone underlying the Port Askaig Fm. These Sr-isotope values are comparable with facies underlying the earliest evidence of Neoproterozoic glaciation in other regions. Thomas et al. (2004) verified this result at a slightly lower horizon in the Islay limestone (Storakaig limestone of Ballygrant) (0.706651 –0.706902) using a slightly more careful sample preparation protocol. Sawaki et al. (2010) carried out a careful chemostratigraphic study on both Islay and the Garvellachs. The samples closest to the onset of glaciation, which also passed a stringent test for preservation (Mn/Sr , 0.2) were found 40 m below the base of the Port Askaig Fm. on Garbh Eilach with values as low as 0.70640. These are close to the lowest values found in East Greenland immediately prior to glaciation (Fairchild et al. 2000) and suggest that the Garbh Eilach section is similarly complete. McCay et al. (2006) described evidence for a third Dalradian glacial from Ireland and also built on earlier d13C results of Brasier & Shields (2000). Values of þ5‰PDB in a lower (Ballygrant) limestone were followed by a decline to weakly positive to negative values in the Lossit Limestone (Brasier & Shields 2000). McCay et al. (2006) and Prave et al. (2009) show that in the Garvellachs section, both dolomitic and limestone facies show a change upwards from negative values (–4 to –6) in the strata in which Sawaki et al. (2010) measured the lightest Sr-isotope signatures, to weakly positive values. In terms of the carbonate rocks overlying the Port Askaig Fm., all these publications show that the main ferroan dolomitic part (Member 3) of the Bonahaven Fm. displays negative values, and Brasier & Shields (2000) document values in the Member 4 dolomite horizon exceeding þ10‰.

Other characteristics (e.g. economic deposits, biomarkers) Mining of zones of epigenetic Pb–Zn mineralization formerly occurred on Islay in the Ballygrant and Lossit limestones (British Geological Survey 1994). The Port Askaig Fm. is locally rich in detrital magnetite and in the Disrupted Beds there are occasional massive layers of magnetite. Spencer (1971) reported possible organic traces in Member 1 (Spencer & Spencer 1972) of the Bonahaven Fm. A detailed description of sole structures interpreted as representing chains of faecal pellets was made by Brasier & McIlroy (1998), although Brasier and Shields (2000) conceded that they could well be of inorganic origin. Fairchild (1977) described clear 0.1-mm-sized mica spheres within a 2 m stratigraphic interval of carbonaceous mudstones near the top of Member 1 of the Bonahaven Fm. Petrographic evidence indicates that they were delicate enough to collapse when mud desiccated, yet were mineralized (perhaps by glauconite) when eroded as intraclasts. They were interpreted as an unusual form of preservation of acritarch fossils. However, their great similarity with Triassic mica spheres from SW England was noted and the latter were subsequently interpreted as tektites by Walkden et al. (2002); this seems a more likely explanation for the Islay occurrence. Only one other example of impact-related phenomena has been reported in the Neoproterozoic of the British Isles (Amor et al. 2008).

Palaeolatitude and palaeogeography Earlier palaeomagnetic work by Tarling (1974) and UrrutiaFucugauchi & Tarling (1983) based on large sample suites on the Garvellachs appeared to be consistent with low palaeolatitudes,

PORT ASKAIG FORMATION, SCOTLAND

although difficulties were encountered due to the unknown age of the deposits and the inability to eliminate the possibility of Caledonian overprinting. Stupavsky et al. (1982) specifically addressed the overprinting issues using clasts (79 specimens from 36 cores) and matrix (2 to 3 specimens from 20 cores) of the diamictite units as well as siltstone (2 to 3 specimens from 6 cores) from Garbh Eileach, Garvellach Islands. Remanent magnetization measurements were used to calculate remanence angular standard deviation in order to evaluate within-specimen homogeneity and reliability of specimens that were sampled. Specimens were also subjected to alternating field and thermal demagnetization. All samples yielded similar results, thus failing the conglomerate test, and suggesting that these deposits have been remagnetized by Ordovician-age overprinting (Stupavsky et al. 1982). The Dalradian Supergroup is generally associated with the Proto-Iapetus Ocean, although palaeogeographical reconstructions are uncertain, in part because of an incomplete understanding of Dalradian basin development (Soper 1994; Tanner & Bluck 1999; Prave 1999; Dalziel & Soper 2001; Demspter et al. 2002; Hutton & Alsop 2004; Tanner et al. 2005). Some have suggested the Dalradian experienced orogenesis prior to rifting and opening up of the Iapetus Ocean in Argyll-Southern Highland. Such an orogenic event would suggest an affinity with Gondwana and a palaeogeographical location either off NW Gondwana (Bluck & Dempster 1991) or off Amazonia (Dalziel 1994). Others have suggested the Dalradian experienced prolonged rifting and extension throughout its history consistent with a palaeogeographical location on the margin of Laurentia (Dalziel & Soper 2001). Others still have suggested the Dalradian was an extension of Baltica (Greiling & Smith 2000) based on the similarity between granitic clasts in the Port Askaig Fm. and Scandinavian intrusions. Recent work by Cawood et al. (2003) demonstrates the similar characteristics of the detrital zircon populations throughout the Dalradian and their close match to Laurentian sources, with the Port Askaig Fm. clasts having closest matches in the North Atlantic Borderlands (Makkovik, Ketilidian and Svecofennian provinces). In this regard, the distinctive provenance represents along-basin rather than cross-basin transport.

Geochronological constraints Radiometrically, the Port Askaig Fm. is relatively poorly constrained. A maximum age of c. 806 Ma comes from the underlying Grampian Shear Zone located at the base of the Dalradian Supergroup (Noble et al. 1996). This finding is based on three dates (806 + 3, 808 þ 11/–9 and 804 þ 13/–12 Ma) obtained from U –Pb isotope analyses of primary monazite in pegmatite and neocrystalline monazite associated with mylonitic host rocks. Samples were collected from the Grampian Shear Zone at Lochindorb and A’Bhuidheanaich, East of the Great Glen Fault and Inverness, Scotland. A minimum age for the Port Askaig Fm. is provided by two dates from the Tayvallich volcanic rocks, which are stratigraphically 8 km above the Port Askaig Fm. at the top of the Argyll Group in SW Scotland (Fig. 62.2; Prave 1999). Halliday et al. (1989) present various data from U –Pb, 207Pb/206Pb and Sm –Nd isotopic analyses of zircons from a keratophyre sampled from a small laccolithic body on the Tayvallich Peninsula. They conclude that the most likely age of the keratophyre is 595 þ /24 based on their Pb/Pb analyses. Field relationships between the small laccolithic body and the Tayvallich volcanic rocks are somewhat unclear, and Dempster et al. (2002) have suggested that this age may be younger than the Tayvallich volcanic rocks. In an attempt to refine this age, Dempster et al. (2002) analysed 14 zircons from a felsic tuff collected at Port a’ Bhuailteir on the Tayvallich Peninsula. Concordia diagrams show a mean 206 Pb – 238U age of 601.4 þ /23.7 Ma (n ¼ 13/14; 2s; MSWD ¼ 0.82).

639

Lithological comparisons with other diamictite-bearing successions in Scotland and Ireland and throughout the North Atlantic, as well as chemostratigraphic studies, have led to various regional and global stratigraphic correlation schemes (Spencer 1975; Hambrey 1983; Prave 1999; Brasier & Shields 2000; Halverson et al. 2005; McCay et al. 2006), with the most recent work suggesting that the Port Askaig Fm. likely represents the oldest of several Neoproterozoic glacial periods within the Dalradian basin and the North Atlantic region. A particularly distinctive characteristic is the low Sr-isotope ratios of the underlying limestone (Brasier & Shields 2000; Thomas et al. 2004; Sawaki et al. 2010). However, this has recently been called into question based on new Re-Os dates that provide a maximum age of 659.6 + 9.6 Ma for the Port Askaig Fm. (Rooney et al. 2011).

Discussion Although there is broad agreement that the Port Askaig Fm. records environmental conditions during one of the Neoproterozoic glacial periods, several palaeoenvironmental models have been proposed with differences hinging largely on the interpretation of diamictite units as directly deposited by ice (Kilburn et al. 1965; Spencer 1971; Benn & Prave 2006) or as primarily deposited in a marine setting influenced by ice-rafting and tectonic instability (Eyles 1988; Arnaud & Eyles 2006). In the models that emphasize a subglacial or ice marginal origin, diamictite units are interpreted as tills recording multiple grounded ice advances based on the lateral extent of diamictite, the discontinuous nature of some of the interbeds in the lower members, the presence of large extrabasinal clasts and faceted clasts, the sandstone wedges that are interpreted as periglacial, and the siltstone with outsized clasts that are interpreted as varves with ice-rafted debris. In these models, the Great Breccia is thought to record glaciotectonic deformation of sediments based on its similarity to deformed chalk rafts in tills of Norfolk (UK) and its correlative erosional unconformity on Islay. The sandstone wedges are thought to be periglacial because of their similarity to those formed by repeated freeze and thaw. The associated interbeds are thought to record either terrestrial or marine depositional conditions preserved between successive ice advances. Benn & Prave (2006) took this argument further by suggesting that the Great Breccia and associated Disrupted Beds recorded proglacial and subglacial phases of a single glaciotectonic deformation cycle. Although they acknowledged that the sedimentary characteristics of the Great Breccia were consistent with either a glaciotectonic or non-glacial sediment gravity flow origin, they preferred a glacial origin based on their interpretation of several features within the associated Disrupted Beds as indicative of subglacial deformation; namely laminae that resemble glaciotectonic laminae, deformation indicative of shear, an increase up-section in the number of extrabasinal clasts, and evidence of increasing upwards cumulative strain in diamictite. In the tectonically influenced glaciomarine models, many diamictite units are thought to record sediment instability associated with basin development, whereas diamictite units in the uppermost part of the succession are thought to record reworking of sediments and rainout of fine-grained sediments and ice-rafted debris in a glacially influenced basin (Arnaud & Eyles 2006). This interpretation is based on the presence of coarse-tail inverse grading, gradational basal contacts and the close association with other sediment gravity flow deposits (Boulton 1972; Nardin et al. 1979; Mulder & Alexander 2001; Arnaud & Eyles 2006). The Great Breccia is interpreted as a catastrophic subaqueous landslide associated with local tectonic activity (Arnaud & Eyles 2002) based on mapping that revealed (i) the Great Breccia to be a ‘composite graded sequence’, (ii) an intimate association with undeformed subaqueous sediment gravity flow and traction current deposits and (iii) a similarity to published studies of allochtonous carbonate

640

E. ARNAUD & I. J. FAIRCHILD

megabreccia (Arnaud & Eyles 2002). The stratigraphic horizons of deformation structures, including the overlying Disrupted Beds, are interpreted as seismites indicative of local tectonic instability based on their form, geographical extent, and tectonic setting in which they are found (Arnaud & Eyles 2006). Stratigraphic analysis of the sedimentary facies and indicators of glacial and tectonic activity suggest that repeated ice-margin fluctuation and tectonically quiet conditions occurred during deposition of Member III sediments (Arnaud & Eyles 2006). The sandstone, mudstone and conglomerate interbeds record shallow marine conditions affected by sediment gravity flows and traction currents (Arnaud & Eyles 2006). The giant cross-beds are thought to result from the migration of large dunes under strong tidal currents considering the Dalradian basin is thought to be narrow at this time (Eyles 1988; Arnaud 2004). In terms of the dolomite interbeds, there is no specific evidence of chemical deposition and much evidence of detrital dolomite (Fairchild 1985a; Hambrey et al. 1991). The superb exposures in the Garvellach Islands have resulted in numerous detailed studies, yet the resulting depositional models proposed for the Port Askaig Fm. have some significant differences in their climatic and palaeoenvironmental implications. A full discussion of these models is beyond the scope of this paper and the reader is referred to the original works for more details. Although recent developments in glacial geology have allowed some of the earlier interpretations to be discounted, there are still instances where the exact nature of glacial influence over these deposits is debatable and difficult to establish unequivocally. The carbonate succession underlying the Port Askaig Fm. on Islay is consistent with marine regression from offshore shale to coastal facies. On the Garvellachs, the mixed carbonatesiliciclastic sediments bear local probable ikaite pseudomorphs near the top, suggestive of cool marine conditions. Facies resembling modern carbonate tropical platforms are limited to local intraclastic dolostones in parts of the Persabus Member at the top of the Lossit Limestone, whereas more distinctive oolitic limestone facies occur below in the Kiells member. The relative stratigraphic positionings of these occurrences between the Garvellachs and Islay cannot be resolved because of the lack of marker horizons between the two. The mixed carbonate-siliciclastic succession that overlies the Port Askaig Fm. is thought to record shallow marine sedimentation. No climatically distinctive facies occur in the basal part of the Bonahaven Fm. The most prominent carbonate (Member 3 of the Bonahaven Fm.) records a lagoonal and tidal complex (Fairchild 1980a). It is generally agreed that the Bonahaven Fm. is not a typical cap-carbonate sequence (Halverson et al. 2005; McCay et al. 2006). It is hard to pinpoint the last glacial influence (the highest characteristic conglomerate is argued to be reworked and represents a tidal channel deposit, Hambrey et al. 1991) and the facies succession does not resemble that of carbonate successions found elsewhere. Member 3 carbonates have only one exact facies counterpart in the geological record – the upper Canyon Fm. of East Greenland – and the latter occurs some distance above the upper of two glacial deposits in that area (Fairchild 1989). A difference of viewpoint exists about the significance of the negative C-isotope anomalies below the Port Askaig Fm. and in the overlying Bonahaven Fm. Prave et al. (2009) argue that the C-isotope stratigraphy recorded in the Dalradian succession as a whole bears close comparison with global trends and that the published signatures can be taken at face value as primary signatures. The alternative view is that wherever impure ferroan dolomites are present, it would be expected that negative deviations from marine signals would be present. This applies to some (but not all) of the pre-glacial facies of the Lossit Limestone on the Garvellachs and to the samples of Member 3 of the Bonahaven Fm. that have been studied so far. The exact facies equivalents of the Bonahaven rocks in East Greenland show coherent variations in negative d13C

values over vertical distances of 10 –20 m (Fairchild 1991). There are limestone facies in the Bonahaven Fm. that could be studied to help resolve this issue. The extraordinarily high d13C signature in the pure dolostone of Member 4 is an enigma; a similar strong signal appears rather soon after glaciation in deposits of NW Canada, known as the Keele peak (McCay et al. 2006). Whereas Halverson et al. (2005) chose to trust the pre-Port Askaig d13C negative anomaly to correlate with late Cryogenian glaciations elsewhere, it is now clear that such an anomaly is present below both early and late Cryogenian glacial deposits (Prave et al. 2009). Most authors believe that the low Sr-isotope ratio is a more specific feature with which to establish an early- to midCryogenian age for the Port Askaig Fm. The minimum value obtained a short distance below the Port Askaig Fm. of 0.70640 (Sawaki et al. 2010) is close to the lowest value of 0.7063 found in East Greenland immediately prior to glaciation (Fairchild et al. 2000), and suggest that the Garbh Eilach section is similarly complete. There are difficulties in reconstructing stratigraphic profiles through the entire thickness of the Dalradian succession and there is also evidence of diachroneity of widely distributed glacial deposits thought previously to be correlative (e.g. Kendall et al. 2006; Fanning & Link 2004), but Prave et al. (2009) were optimistic that there are sufficient diagnostic chemostratigraphic results to allow the global stratigraphic context of the Port Askaig Fm. as early- to mid Cryogenian (Sturtian in their terminology) to be confirmed. Rooney et al.’s (2011) findings will ensure continued debate and suggest additional work is needed to constrain the timing of glaciation. The authors would like to thank A. Spencer and G. Shields for helpful comments on earlier drafts of this paper. E. A.’s research on Neoproterozoic glaciogenic successions is supported by the Natural Sciences and Engineering Research Council of Canada. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Amor, K., Hesselbo, S. P., Porcelli, D., Thackrey, S. & Parnell, J. 2008. A Precambrian proximal ejecta blanket from Scotland. Geology, 36, 303–306. Anderton, R. 1980. Distinctive pebbles as indicators of Dalradian provenance. Scottish Journal of Geology, 16, 143–152. Anderton, R. 1982. Dalradian deposition and the late Precambrian – Cambrian history of the North Atlantic region: a review of the early evolution of the Iapetus Ocean. Journal of the Geological Society, London, 139, 421–431. Anderton, R. 1985. Sedimentation and tectonics in the Scottish Dalradian. Scottish Journal of Geology, 21, 407–436. Anderton, R. 1988. Dalradian slides and basin development: a radical interpretation of stratigraphy and structure in the SW and Central Highlands of Scotland. Journal of the Geological Society, London, 145, 669– 678. Arnaud, E. 2002. Sedimentological analysis of Neoproterozoic glaciogenic successions in Norway and Scotland. Unpublished PhD thesis, School of Geography and Geology, McMaster University. Arnaud, E. 2004. Giant cross-beds in the Neoproterozoic Port Askaig Formation, Scotland: implications for snowball Earth. Sedimentary Geology, 165, 155–174. Arnaud, E. & Eyles, C. H. 2002. Catastrophic mass failure of a Neoproterozoic glacially-influenced continental margin, the Great Breccia, Port Askaig Formation, Scotland. Sedimentary Geology, 151, 313– 333. Arnaud, E. & Shields, G. 2005. The sedimentary record of a Neoproterozoic glaciation. (IGCP# 512 International Field workshop guidebook), International Association of Sedimentologists Conference on Glacial Sedimentary Processes and Products, University of Wales, Aberystwyth, 29 August– 3 September 2005, 60. Arnaud, E. & Eyles, C. H. 2006. Neoproterozoic environmental change recorded in the Port Askaig Formation, Scotland: climatic and tectonic controls on sedimentation. Sedimentary Geology, 183, 99 – 124.

PORT ASKAIG FORMATION, SCOTLAND

Bailey, E. B. 1916. The Islay Anticline (Inner Hebrides). Quaterly Journal of the Geological Society of London, 72, 132– 164. Benn, D. I. & Prave, A. R. 2006. Subglacial and proglacial glacitectonic deformation in the Neoproterozoic Port Askaig Formation, Scotland. Geomorphology, 75, 266–280. Bluck, B. J. & Dempster, T. J. 1991. Exotic metamorphic terranes in the Caledonides: tectonic history of the Dalradian block, Scotland. Geology, 19, 1133– 1136. Borradaile, G. J. 1979. Strain study of the Caledonides in the Islay region, SW Scotland: implications for strain histories and deformation mechanisms in greenschists. Journal of the Geological Society, London, 136, 77 –88. Boulton, G. S. 1972. Modern arctic glaciers as depositional models for former ice sheets. Journal of the Geological Society, London, 128, 361– 393. Brasier, M. D. & McIlroy, D. 1998. Neonereites uniserialis from c. 600 Ma year old rocks in western Scotland and the emergence of animals. Journal of the Geological Society, London, 155, 5 –12. Brasier, M. D. & Shields, G. 2000. Neoproterozoic chemostratigraphy and correlation of the Port Askaig glaciation, Dalradian Supergroup of Scotland. Journal of the Geological Society, London, 157, 909– 914. BRITISH GEOLOGICAL SURVEY. 1994. North Islay. Scotland Sheet 27. Solid and drift geology. 1:50000 provisional series. British Geological Survey, Keyworth. Cawood, P. A., Nemchin, A. A., Smith, M. & Loewy, S. 2003. Source of the Dalradian Supergroup constrained by U– Pb dating of detrital zircon and implications for the East Laurentian margin. Journal of the Geological Society, London, 160, 231–246. Dalziel, I. W. D. 1994. Precambrian Scotland as a Laurentia-Gondwana link: Origin and signficance of cratonic promontories. Geology, 22, 589– 592. Dalziel, I. W. D. & Soper, N. J. 2001. Neoproterozoic extension on the Scottish Promontory of Laurentia: paleogeographic and tectonic implications. Journal of Geology, 109, 299–317. Dempster, T. J., Rogers, G. et al. 2002. Timing of deposition, orogenesis and glaciation within the Dalradian rocks of Scotland: constraints from U– Pb zircon ages. Journal of the Geological Society, London, 159, 83– 94. Evans, J. A., Fitches, W. R. & Muir, R. J. 1998. Laurentian clasts in a Neoproterozoic tillite from Scotland. Journal of Geology, 106, 361– 366. Eyles, C. H. 1988. Glacially and tidally-influenced shallow marine sedimentation of the Late Precambrian Port Askaig Formation, Scotland. Palaeogeography, Palaeoclimatology, Palaeoecology, 68, 1 –25. Eyles, C. H. & Eyles, N. 1983. Glaciomarine model for upper Precambrian diamictites of the Port Askaig Formation, Scotland. Geology, 11, 692– 696. Eyles, N. & Clark, B. M. 1985. Gravity-induced soft-sediment deformation in glaciomarine sequences of the Upper Proterozoic Port Askaig Formation, Scotland. Sedimentology, 32, 789–814. Fairchild, I. J. 1977. Phengite spherules from the Dalradian Bonahaven Formation, Islay, Scotland: glauconitized microfossils? Geological Magazine, 114, 355– 364. Fairchild, I. J. 1980a. Sedimentation and origin of a late Precambrian ‘dolomite’ from Scotland. Journal of Sedimentary Petrology, 50, 423– 446. Fairchild, I. J. 1980b. Stages in Precambrian dolomitization, Scotland: cementing v. replacement textures. Sedimentology, 27, 631–650. Fairchild, I. J. 1980c. The structure of NE Islay. Scottish Journal of Geology, 16, 189–197. Fairchild, I. J. 1985a. Comment on ‘Glaciomarine model for upper Precambrian diamictites of the Port Askaig Formation, Scotland’. Geology, 13, 89 –90. Fairchild, I. J. 1985b. Petrography and carbonate chemistry of some Dalradian dolomitic metasediments: preservation of diagenetic textures. Journal of the Geological Society, London, 142, 167– 185. Fairchild, I. J. 1989. Dolomitic stromatolite-bearing units with storm deposits from the Vendian of East Greenland and Scotland: a case of facies equivalence. In: Gayer, R. A. (ed.) The Caledonide Geology of Scandinavia. Graham & Trotman, London, 275– 283.

641

Fairchild, I. J. 1991. Origins of carbonate in Neoproterozoic stromatolites and the identification of modern analogues. Precambrian Research, 53, 281– 299. Fairchild, I. J., Spiro, B., Herrington, P. M. & Song, T. 2000. Controls on Sr and C isotope compositions of Neoproterozoic Sr-rich limestones of East Greenland and North China. In: Grotzinger, J. P & James, N. P. (eds) Carbonate Sedimentation and Diagenesis in the Evolving Precambrian World. SEPM Special Publication, 67, 297– 313. Fanning, C. M. & Link, P. K. 2004. U–Pb SHRIMP ages of Neoproterozoic (Sturtian) glaciogenic Pocatello Formation, southeastern Idaho. Geology, 32, 881– 884. Fitches, W. R., Pearce, N. J. G., Evans, J. A. & Muir, R. J. 1996. Provenance of late Proterozoic Dalradian tillite clasts, Inner Hebrides, Scotland. In: Brewer, S. T. (ed.) Precambrian Crustal Evolution in the North Atlantic Region. Geological Society, Special Publications, London, 112, 367– 377. Greiling, R. O. & Smith, A. G. 2000. The Dalradian of Scotland: Missing link between the Vendian of Northern and Southern Scandinavia? Physics and Chemistry of the Earth, Part A, 25, 495– 498. Halliday, A. N., Graham, C. M., Aftalion, M. & Dymoke, P. 1989. Short paper: The depositional age of the Dalradian Supergroup: U–Pb and Sm–Nd isotopic studies of the Tayvallich Volcanics, Scotland. Journal of the Geological Society, London, 146, 3– 6. Halverson, G. P., Hoffman, P. F., Schrag, D. P., Maloof, A. C. & Rice, A. H. N. 2005. Toward a Neoproterozoic composite carbonisotope record. Geological Society of America Bulletin, 117, 1181–1207. Hambrey, M. J. 1983. Correlation of Late Proterozoic tillites in the North Atlantic region and Europe. Geological Magazine, 120, 209–232. Hambrey, M. J., Fairchild, I. J., Glover, B. W., Stewart, A. D., Treagus, J. E. & Winchester, J. A. 1991. The Late Precambrian Geology of the Scottish Highlands and Islands. Geologists Association Guide No. 44. The Geologists’ Association, London, England. Harris, A. L., Baldwin, C. T., Bradbury, H. J., Johnson, H. D. & Smith, R. A. 1978. Ensialic basin sedimentation: the Dalradian Supergroup. In: Bowes, D. R. & Leake, B. L. (eds) Crustal Evolution in Northwestern Britain and Adjacent Regions. Seel House Press, Liverpool, 115–138. Harris, A. L., Haselock, P. J., Kennedy, M. J., Mendum, J. R., Long, C. B., Winchester, J. A. & Tanner, P. W. G. 1993. The Dalradian Supergroup in Scotland, Shetland, and Ireland. In: Gibbons, W. & Harris, A. L. (eds) A Revised Correlation of Precambrian Rocks in the British Isles. Geological Society, Special Report, London, 33– 53. Howarth, R. J. 1971. The Port Askaig tillite succession (Dalradian) of Co. Donegal. Proceedings of the Royal Irish Academy, 71, 1 –36. Hutton, D. H. W. & Alsop, G. I. 2004. Evidence for a major Neoproterozoic orogenic unconformity within the Dalradian Supergroup of NW Ireland. Journal of the Geological Society, London, 161, 629– 640. Johnston, J. D. 1995. Pseudomorphs after ikaite in a glaciomarine sequence in the Dalradian of Donegal, Ireland. Scottish Journal of Geology, 31, 3 –9. Kendall, B. S., Creaser, R. A. & Selby, D. 2006. Re– Os geochronology of postglacial black shales in Australia: constraints on the timing of the Sturtian glaciation. Geology, 34, 729–732. Kilburn, C., Pitcher, W. S. & Shackleton, R. M. 1965. The stratigraphy and origin of the Port Askaig boulder bed series (Dalradian). Geological Journal, 4, 343– 360. Litherland, M. 1980. The stratigraphy of the Dalradian rocks around Loch Creran, Argyll. Scottish Journal of Geology, 16, 105–123. Max, M. D. 1981. E20: Dalradian Tillite of northwestern Ireland. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 640–642. McCay, G. A., Prave, A. R., Alsop, G. I. & Fallick, A. E. 2006. Glacial trinity: Neoproterozoic Earth history within the British-Irish Caledonides. Geology, 34, 909– 912. Mulder, T. & Alexander, J. 2001. The physical character of subaqueous sedimentary density flows and their deposits. Sedimentology, 48, 269– 299. Nardin, T. R., Hein, F. J., Gorsline, D. S. & Edwards, B. D. 1979. A review of mass movement processes, sediment and acoustic

642

E. ARNAUD & I. J. FAIRCHILD

characteristics and contrasts in slope and base of slope systems v. canyon fan basin floor systems. In: Doyle, L. J. & Pilkey, O. H. (eds) Geology of Continental Slopes. Society of Economic Paleontologists and Mineralogists, Special Publication, 61 –73. Noble, S. R., Hyslop, E. K. & Highton, A. J. 1996. High precision U –Pb monazite geochronology of the c. 806 Ma Grampian Shear Zone and the implications for the evolution of the Central Highlands of Scotland. Journal of the Geological Society, London, 153, 511– 514. Panahi, A. & Young, G. M. 1997. A geochemical investigation into the provenance of the Neoproterozoic Port Askaig tillite, Dalradian Supergroup, western Scotland. Precambrian Research, 85, 81– 96. Prave, A. R. 1999. The Neoproterozoic Dalradian Supergroup of Scotland: an alternative hypothesis. Geological Magazine, 136, 609– 617. Prave, A. R., Fallick, A. E., Thomas, C. W. & Graham, C. M. 2009. A composite C-isotope profile for the Neoproterozoic Dalradian Supergroup of Scotland and Ireland. Journal of the Geological Society, London, 166, 845–857. Rooney, A. D., Chew, D. M. & Selby, D. 2011. Re-Os geochronology of the Neoproterozoic-Cambrian Datradian Supergroup of Scotland and Ireland: implications for Neoproterozoic stratigraphy, glaciations and Re-Os systematics. Precambrian Research, 185, 202– 214. Sawaki, Y., Kawai, T. et al. 2010. 87Sr/86Sr chemostratigraphy of Neoproterozoic Dalradian carbonates below the Port Askaig glaciogenic Formation, Scotland. Precambrian Research, 179, 150– 164. Soper, N. J. 1994. Neoproterozoic sedimentation on the northeast margin of Laurentia and the opening of Iapetus. Geological Magazine, 131, 291– 299. Soper, N. J. & Anderton, R. 1984. Did the Dalradian slides originate as extensional faults? Nature, 307, 357–360. Spencer, A. M. 1971. Late Precambrian glaciation in Scotland. Memoirs of the Geological Society of London, 6, 1 –100. Spencer, A. M. 1975. Late Precambrian glaciation in the North Atlantic region. In: Wright, A. E. & Moseley, F. (eds) Ice Ages, Ancient and Modern. Seel House Press, Liverpool, 217– 240. Spencer, A. M. & Pitcher, W. S. 1968. Occurrence of the Port Askaig Tillite in north-east Scotland. Proceedings of the Geological Society of London, 1650, 195– 198. Spencer, A. M. & Spencer, M. O. 1972. The late Precambrian/Lower Cambrian Bonahaven Dolomite of Islay and its stromatolites. Scottish Journal of Geology, 8, 269–282. Stupavsky, M., Symons, D. T. A. & Gravenor, C. P. 1982. Evidence for metamorphic remagnetisation of upper Precambrian tillite in the

Dalradian Supergroup of Scotland. Transactions of the Royal Society of Edinburgh, Earth Sciences, 73, 59 –65. Tanner, P. W. G. 1998. Interstratal dewatering origin for polygonal patterns of sand-filled cracks; a case study from late Proterozoic metasediments of Islay, Scotland. Sedimentology, 45, 71 –89. Tanner, P. W. G. & Shackleton, R. M. 1979. Structure and stratigraphy of the Dalradian rocks of the Bennabeola area; Connemara, Eire. In: Harris, A. L., Holland, C. H. & Leake, B. E. (eds) The Caledonides of the British Isles – Reviewed. Geological Society, London, Special Publications, 8, 243 – 256. Tanner, P. W. G. & Bluck, B. J. 1999. Current controversies in the Caledonides. Journal of the Geological Society, London, 156, 1137– 1141. Tanner, P. W. G., Alsop, G. I. & Hutton, D. 2005. Discussion on evidence for a major Neoproterozoic orogenic unconformity within the Dalradian Supergroup of NW Ireland. Journal of the Geological Society, London, 162, 221–224. Tarling, D. H. 1974. A paleomagnetic study of Eocambrian tillites in Scotland. Journal of the Geological Society, London, 130, 163– 177. Thomas, C. W., Graham, C. M., Ellam, R. M. & Fallick, A. E. 2004. 87 Sr/86Sr chemostratigraphy of Neoproterozoic Dalradian limestones of Scotland and Ireland: constraints on depositional ages and time scales. Journal of the Geological Society, London, 161, 229– 242. Thomson, J. 1871 (for 1870). On the occurrence of pebbles and boulders of granite in schistose rocks in Islay, Scotland. 40th Meeting British Association, Liverpool, Transactions, 88. Treagus, J. E. 1987. The structural evolution of the Dalradian of the Central Highlands of Scotland. Transactions of the Royal Society of Edinburgh, 78, 1– 15. Treagus, S. H. & Treagus, J. E. 2002. Studies of strain and rheology of conglomerates. Journal of Structural Geology, 24, 1541–1567. Urrutia-Fucugauchi, J. & Tarling, D. H. 1983. Paleomagnetic properties of Eocambrian sediments in Northwestern Scotland: implications for world wide glaciation in the Late Precambrian. Palaeogeography, Palaeoclimatology, Palaeoecology, 41, 325– 344. Walkden, G., Parker, J. & Kelley, S. 2002. A late Triassic impact ejecta layer in southwestern Britain. Science, 298, 2185–2188. Yardley, B. W. D., Vine, F. J. & Baldwin, C. T. 1982. The plate tectonic setting of NW Britain and Ireland in late Cambrian and early Ordovician times. Journal of the Geological Society of London, 139, 455– 463.

Chapter 63 The Neoproterozoic glaciogenic deposits of Scotland and Ireland A. R. PRAVE1 * & A. E. FALLICK2 1

Department of Earth Science, University of St Andrews, St Andrews, KY16 9AL, UK

2

Scottish Universities Environmental Research Centre, East Kilbride, G75 0QF, UK *Corresponding authors (e-mail: [email protected])

Abstract: Of the three major Neoproterozoic supracrustal units in the Scottish and Irish Highlands (the Torridonian, Moine and Dalradian Supergroups), only the latter contains evidence of Neoproterozoic glaciations. The Dalradian is siliciclastic-dominated and constitutes much of the Scottish–Irish Highlands between the Great Glen and Highland Boundary Fault Zones, and their correlatives in Ireland. At the time of writing, three stratigraphically distinct glacial intervals in the Dalradian have been documented in the literature. The oldest is the Port Askaig Formation (Fm.) at the base of the Argyll Group (see Arnaud & Fairchild 2011). It ranges from several tens to many hundreds of metres in thickness and occurs in numerous localities in Scotland and the north of Ireland. A second glacial is recorded in the middle part of the Argyll Group (Easdale Subgroup) and consists of localised sedimentary breccias as well as pelites and schists containing dropstone/lonestone units inferred to be ice-rafted debris; these rocks are patchily preserved and typically a few metres or less in thickness. It is sharply overlain by a variably developed carbonate unit that is marked by a 1 –7-m-thick, lightcoloured, basal dolostone or dolomitic limestone interpreted as a cap carbonate. This succession is best preserved in Donegal, Ireland, as the Stralinchy–Reelan glacial and Cranford cap-carbonate sequence. A correlative cap carbonate, the Whiteness Limestone, has been identified in the Shetland Islands. The third and youngest glacial is represented by locally preserved dropstone and polymict diamictite beds ranging in thickness from several to a few tens of metres in thickness in the lower Southern Highland Group. These include the MacDuff and Loch na Cille Boulder Beds in, respectively, NE and SW Scotland, and the Inishowen Beds in Donegal, Ireland.

The late Proterozoic geological framework of the Scottish –Irish Highlands consists largely of three (meta)sedimentary successions (each many kilometres thick): the Torridonian, preserved NW of the Moine Thrust; the Moine, in the upper plate of the Moine Thrust zone southeastward to the Great Glen Fault; and the Dalradian, which is sandwiched between the Great Glen and Highland Boundary Fault Zones (Fig. 63.1). Only the Dalradian contains a record of Neoproterozoic glaciations. A purported glacial deposit was inferred for a unit in the lower part of the Torridonian (Davison & Hambrey 1996) but this has been shown to be incorrect (Stewart 1997, 2002). Along most of the Moine outcrop belt the rocks are too strongly deformed to obtain sedimentological information, but even in areas of low strain no evidence has been found for glaciogenic deposits. This is not surprising given that both the Moine and Torridonian pre-date Cryogenian time: the Moine was deposited prior to c. 840 Ma, the age of cross-cutting igneous intrusions and magmatic overgrowths on detrital zircons (e.g. Vance et al. 1998; Millar 1999; Kirkland et al. 2008), and deposition of the Torridonian occurred before c. 950 Ma, as based on Rb – Sr and Pb/Pb diagenetic ages and U –Pb ages on the youngest detrital zircons (Turnbull et al. 1996; Rainbird et al. 2001). The Dalradian is a mainly metasedimentary succession subdivided into four major units (Harris et al. 1994): the Grampian, Appin, Argyll and Southern Highland Groups (Figs 63.1 & 63.2). Greenschist- to amphibolite-facies metamorphism and penetrative tectonic fabrics associated with the Neoproterozoic Knoydartian (c. 840 –720 Ma; e.g. Piasecki & van Breeman 1983; Vance et al. 1998; Tanner & Evans 2003) and Ordovician Grampian (c. 465 Ma; Oliver et al. 1998) tectonothermal events in many places obscure original sedimentary features and render problematic any assessment of depositional processes. Areas of low strain are geological rosetta stones from which to establish stratigraphic relationships and make palaeoenvironmental interpretations, and to extrapolate those to the more deformed portions of the Dalradian outcrop belt. It needs to be noted that in much of the older literature on the Dalradian, the term ‘Boulder Bed’ appears. This term was applied to schists and pelites containing polymict, variably shaped clasts in matrix-supported textures with poor sorting. Many of these deposits are not glaciogenic in

origin; research is currently being undertaken by the authors to document those that are. The best-known, longest-studied ancient glacial deposit in the Dalradian is the Port Askaig Fm. at the base of the Argyll Group (Fig. 63.2). It is present in numerous localities scattered across the Dalradian outcrop belt and is more than 500 m thick at its type locality on the Inner Hebridean island of Islay (Fig. 63.1). The seminal study of Spencer (1971), building upon the efforts of earlier workers, confirmed its glaciogenicity. Another, but younger glaciogenic interval to be well documented is the 12-m-thick MacDuff Boulder Bed, exposed near the eponymous town along the Banffshire coastline in NE Scotland (Figs 63.1 & 63.2). Sutton & Watson (1954) initially, and Stoker et al. (1999) subsequently, showed that this unit contains ice-rafted dropstones. Condon & Prave (2000) reported on other ice-rafted-debris beds a few decimetres to several metres thick along the Inishowen Peninsula of Donegal, Ireland (Figs 63.1 & 63.2). They proposed correlations of these beds to the MacDuff Boulder Bed, as well as another Boulder Bed, the Loch na Cille Boulder Bed, on the Tayvallich Peninsula in southwestern Scotland (Fig. 63.1); all of these occurrences are in the lower Southern Highland Group. McCay et al. (2006) documented the Stralinchy –Reelan glacial and Cranford Limestone cap-carbonate sequence in the middle of the Argyll Group (near the top of the Easdale Subgroup) in Donegal (Figs 63.1 & 63.2). This work confirmed that the Dalradian contained three, stratigraphically distinct glacial intervals: (i) the Port Askaig Fm. (base of the Argyll Group); (ii) the glacial –cap carbonate couplet of the Stralinchy –Reelan and Cranford Limestone formations (Easdale Subgroup); and (iii) the MacDuff –Inishowen – Loch na Cille Beds (lower Southern Highland Group). Subsequently, Prave et al. (2009a) interpreted the Whiteness Limestone in the Shetland Islands, some 200 km north of the Scottish mainland, as a cap-carbonate correlative with the Cranford Limestone.

Structural framework The rather severe tectonothermal overprinting of the Dalradian rocks during early Palaeozoic Caledonian orogenesis resulted in

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 643– 648. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.63

644

A. R. PRAVE & A. E. FALLICK

Fig. 63.1. Highly generalized geological map of the Scottish– Irish Dalradian Highlands highlighting the outcrop belts of the Argyll and Southern Highlands Groups. Small inset map shows simplified tectonic template of the Highlands. GGF, Great Glen Fault; HBF, Highland Boundary Fault; MT, Moine Thrust. These fault zones demarcate the three Proterozoic Supergroups of the Highlands: the Torridonian occurs NW of the Moine Thrust, the Moine is between the Moine Thrust and Great Glen Fault, and the Dalradian (grey-shading) is bounded by the Great Glen and Highland Boundary Fault Zones. Main map, locations discussed in the text: C, Cranford; LC, Loch na Cille; M, MacDuff; P, Port Askaig; R, Reelan; S, Stralinchy.

structural disruption of the Dalradian outcrop belt by a series of so-called tectonic ‘slides’. Most of these are interpreted as extensional collapse structures reactivated along original reverse or thrust faults (Harris et al. 1994), but several are thought to represent original basin-bounding faults (e.g. Soper & Anderton Ma

582

Ediacaran

MacDuff–Inishowen–Loch na Cille Boulder Beds

Crinan & Tayvallich Sgps

601 Ma Tayvallich volcanics

Cranford–Whiteness Limestones (cap carbonates) ‘lower’ Ardrishaig–Craignish Phy. Stralinchy Congl.

IRD

635

Easdale Slates

Scarba Congl. Jura – Slieve Tooey Quartzites Bonahaven Dolostone

Port Askaig Formation

713

Islay – Glencolumbkille Limestones

Cryogenian

Easdale Subgroup

‘upper’ Ardrishaig–Craignish & Port Ellen Phyllites

Ballachulish & Blair Islay Atholl Subgroups Subgroup

U O R G R E P U S

Argyll Group

N A I D

Appin Gp

A R L A

Grampian Gp

D ?

542

Southern Highland Gp

P

Cambrian Leny Limestone

1984). This structural and metamorphic overprinting makes assessment of original depositional settings problematic. Nevertheless, the Dalradian Supergroup is often interpreted as recording deposition in a succession of rift basins (Anderton 1982, 1985; Robertson & Smith 1999; Dalziel & Soper 2001; Arnaud & Eyles 2006). This view, however, has been questioned, and the exact nature of basinal genesis remains an outstanding problem to resolve (Prave 1999). The terminal Neoproterozoic geodynamic evolution of the Dalradian succession is, however, relatively uncontroversial and records syn-depositional extensional tectonism and continental rifting (Anderton 1979, 1982, 1985, 1988; Soper & Anderton 1984). This phase of extensional tectonics began prior to deposition of the middle glacial –cap carbonate couplet (Stralinchy –Reelan and Cranford formations) and progressed to active rifting through the lower Southern Highland Group. Antecedent to this, the lithostratigraphic framework of the underlying parts of the Dalradian is not readily attributable to deposition in rift basins and any such interpretation should be viewed with a healthy scepticism (Prave 1999).

? ? (nature of base, whether unconformable or stratigraphic, remains unresolved)

Stratigraphy Group and subgroup correlations across the Dalradian outcrop belt are achievable but, because of variable exposure quality and pervasive deformational fabrics, finer-resolution correlations between many areas are seldom attainable. Consequently, correlations from one locality to the next rely largely on the recognition of packages of associated lithologies, rather than single units or stratal surfaces (such as those utilized in sequence stratigraphy). Nevertheless, over 150 years of mapping has provided a reliable geological framework from which to construct a stratigraphy for the Dalradian’s glaciogenic units. Like many long-studied successions, the Dalradian is rife with stratigraphic nomenclature. To avoid getting bogged down in a morass of names, discussion is restricted to the stratigraphic units directly pertinent to the glaciogenic rocks and their associated strata (see Fig. 63.2).

metamorphic rocks having 840–800 Ma deformational ages (U-Pb zircon & monazite)

Fig. 63.2. Simplified stratigraphic framework of the Dalradian Supergroup. The three recognized glaciogenic intervals in the Dalradian (shaded) and their inferred correlation to geochronological ages and subdivisions of Neoproterozoic Earth history are shown, as are other stratigraphic units mentioned in the text. IRD, ice-rafted debris. See text for details.

Glaciogenic deposits and associated strata The following focuses on the two younger Neoproterozoic glacial units and their associated strata. The oldest glacial, the Port Askaig, is discussed by Arnaud & Fairchild (2011).

NEOPROTEROZOIC GLACIOGENIC DEPOSITS OF SCOTLAND AND IRELAND

The Stralinchy – Reelan formations and the Cranford and Whiteness cap carbonates McCay et al. (2006) showed that the Stralinchy Conglomerate and its basinal equivalent containing isolated dropstones, the Reelan Fm., together with the overlying Cranford Limestone constitute a glacial –cap carbonate succession in the middle part of the Argyll Group in Donegal, Ireland (Figs 63.1 & 63.2). The glacial rocks occur as lenses (from 0 to several tens of metres in thickness) of mono- to poly-mict, poorly sorted sedimentary breccia (Stralinchy Conglomerate) and laterally equivalent thin-bedded pelite and limestone containing pebble- to cobble-sized dropstones and lonestones (Reelan Fm.). Clasts consist of quartzo-feldspathic, quartzitic, schistose and carbonate lithologies. The sedimentary breccias alter from clast-supported to diamictic in texture, clasts range in size from granule to metre-sized blocks, and matrix composition varies along strike from siliciclastic- to carbonatedominated. Lonestones and dropstones pierce and distort laminae in decimetre-thick intervals of finely compositionally layered pelite and limestone; on-lap and splash-up structures are associated with some of the larger clasts. Sitting sharply on these glaciogenic units is the Cranford Limestone. Its base is marked by a 1– 7-m-thick, laterally continuous tan-coloured dolomicrite, that is, a cap carbonate. At the type locality, the cap is overlain by 200–300 m of limestone-rhythmite and rhythmite breccia. In basinal settings, the Cranford Limestone shales out (metamorphosed to phyllite; Alsop & Hutton 1990) and only the basal cap dolostone remains. As documented by Prave et al. (2009a), an inferred correlative cap carbonate, the Whiteness Limestone, is present in the Shetland Islands (these islands lie 200 km north of mainland Scotland). Similar to the Cranford Limestone (and to cap carbonates elsewhere; e.g. Hoffman & Schrag 2002; Fairchild & Kennedy 2007), the Whiteness Limestone has a basal, 2– 5-m-thick, lightcoloured dolostone to dolomitic limestone overlain by 20– 40 m of thin-bedded limestone-pelite rhythmites. These, in turn, pass upward into a tens-of-metres-thick pelitic interval, itself overlain by many hundreds of metres of thick-bedded calcitic marbles. The basal dolostone sits sharply on a thick (1– 2 km) sequence of siliciclastic units devoid of carbonate rocks. Within 5– 10 m of the base of the cap, the underlying schistose rocks become conspicuously finer-grained and darker-coloured. It is within this interval that rare quartzo-feldspathic lonestone clasts can be found. The absence of any physical or textural evidence for deposition of the lonestones by traction currents, sediment-gravity-flow or volcanic ejecta processes, and their stratigraphic restriction to subjacent to the base of the Whiteness cap, led Prave et al. (2009a) to interpret the lonestones as ice-rafted debris, that is, a glacially influenced phase of sedimentation pre-dating deposition of the cap carbonate. The presence of glaciogenic rocks and cap carbonates at the southwestern (Donegal) and northeastern (Shetland) edges of the Dalradian outcrop belt has focused research efforts on mainland Scotland to document correlative glacial –cap carbonate couplets there, or find a reason for their absence.

Boundary relations with overlying and underlying non-glacial units In many places across the Dalradian outcrop belt, Caledonian strain compromises contacts, but in areas of low strain, and where the glaciogenic units consist of sedimentary breccia or diamictite (e.g. the Stralinchy type locality in Donegal), the contact is sharp and erosive. Where the glaciogenic units consist of dropstone or lonestone intervals, defining contacts relies on finding dispersed clasts and/or the contact with the overlying cap carbonate to confirm being in or near glacial strata. The base of the cap

645

carbonate is everywhere sharp. Commonly, only the basal cap dolostone is present and in these localities the top also appears sharp. Where the entire cap-carbonate sequence is developed (the eponymous type localities of the Cranford and Whiteness Limestones), the basal cap displays a gradational contact into overlying carbonate rhythmites and/or a thin calc-pelite unit, itself transitioning into carbonate rhythmites. Thick siliciclasticdominated intervals overlie the cap sequences.

MacDuff, Inishowen and the Loch na Cille Boulder Beds The MacDuff Boulder Bed occurs in the core of a syncline of the lower Southern Highland Group rocks along the coast adjacent to the eponymous Banffshire town (Fig. 63.1); it is c. 12 m thick and has long been recognized as glaciogenic in origin (Sutton & Watson 1954; Stoker et al. 1999). It was speculated to be as young as Ordovician in age (Hambrey 1983; Molyneux 1997), but this is wrong: regional mapping leaves no doubt that it is much older and in the Southern Highland Group (e.g. Read 1923). The MacDuff Boulder Bed is part of a thick succession of siliciclastic rocks interpreted as a deep-marine fan complex (e.g. Trewin 1987). What distinguishes it is the presence of several dropstone intervals in which variably shaped clasts, consisting of quartzite, gneiss, igneous and pale-coloured carbonate rocks ranging from angular to rounded and from granule to boulder in size, can be seen to pierce laminae, be on-lapped by subsequent laminae and, in some places, show splash-up structures. They are concentrated in discrete, decimetre-thick units indicating that they originated during periods of rainout, rather than as part of some continuous background sedimentation process, and an origin as ice-rafted debris, not tills (suggested by Hambrey & Waddams 1981), is certain (Sutton & Watson 1954; Stoker et al. 1999). Stoker et al. (1999) also detailed aspects of seven decimetre-thick diamictite beds. These contain a variety of clasts of varying sizes and shapes in a mud matrix and were interpreted as resedimented glaciomarine deposits. Condon & Prave (2000) documented glaciogenic deposits in the Southern Highland Group along the northern Inishowen Peninsula (Fig. 63.2), Donegal, Ireland. These consist of five dropstone units, each many centimetres to several decimetres thick, spaced irregularly through a c. 500-m-thick arkosic psammite-pelite succession. The dropstone units are characterized by granule- to pebble-sized clasts of quartzite, gneiss, schist and igneous rocks; in many places the clasts can be observed to pierce laminae and be on-lapped by successive laminae. These units are separated by decimetre- to metre-thick intervals of fine-grained pelite devoid of lonestones or gritty horizons. The arkosic nature of these strata is noteworthy because, for hundreds of metres above and below this interval, units are quartzitic in composition. Condon & Prave (2000) and Condon et al. (2002) interpreted the dropstone beds as ice-rafted debris and speculated that the arkosic composition of the encasing strata was due to cold-climate weathering processes. Condon et al. (2002) also documented an iceberg dump structure, a poorly sorted, coarse-grained dome-shaped sedimentary feature having a flat basal surface moulded onto the top of a pelite bed. It is c. 1–2 m in length along its longest exposed axis and c. 10– 20 cm thick along its crest before tapering off irregularly to its thinned edges. A distinctive unit occurs in the lower Southern Highland Group in SW Scotland along the Tayvallich Peninsula (Fig. 63.2), the Loch na Cille Boulder Bed (Elles 1934). It occurs as discontinuous lenses, typically several metres thick, and consists of matrixsupported polymict clasts, including felsite and mafic volcanites, granitoid, sedimentary and metamorphic rock fragments (Alsop et al. 2000). It was recognized and mapped separately because of its distinctiveness from the associated hundreds of metres of volcanic and pelitic rocks. Its matrix is finer-grained, lighter-coloured and more quartzo-feldspathic than the encasing rocks, and its

646

A. R. PRAVE & A. E. FALLICK

clast content is polymict, rather than monomict volcanic detritus. These observations, and compatible stratigraphic position of the Loch na Cille Boulder Bed with the MacDuff and Inishowen beds, indicate that it, too, is a glaciogenic deposit.

Boundary relations with overlying and underlying non-glacial units The lowest observed, glacially influenced deposits in the Southern Highland Group rocks occur abruptly. They are eye-catching because they are compositionally distinct, both in clast content and matrix, from encasing strata whether those are sedimentary, as in the case of the MacDuff and Inishowen rocks, or volcaniclastic, as in the Loch na Cille Beds. The upper contacts of the Inishowen and Loch na Cille Beds are placed at the last occurrence of ice-rafted debris; the MacDuff Boulder Bed has no top, it is the youngest unit in the core of an exposed syncline. The impression one gets from all three units is that glaciogenic deposition was intermittent, discrete events superimposed on the overall background sedimentation.

Chemostratigraphy

confirm that negative C-isotopic excursions are antecedent to two stratigraphically and temporally distinct glaciations and urge caution when attempting to use such excursions as chronostratigraphic markers (see Halverson et al. 2005). The inferred cap carbonates exhibit a trend commonly attributed to post-Marinoan-equivalent successions (e.g. Kennedy et al. 1998), namely values declining from c. –3‰ in the basal cap, reaching a nadir at c. – 6‰ in overlying limestone rhythmites, and then recovering back towards positive values up-section. It should be noted that the basal Whiteness Limestone on the Shetland Islands has initial values between 1‰ and 2‰. A more detailed C-isotopic study of the Dalradian is given in Prave et al. (2009b).

Other characteristics (e.g. economic deposits, biomarkers) There are no economic deposits associated with the Dalradian’s glacial rocks. The cap carbonates have, in many places, been quarried, but those operations were small and not commercially viable. Likewise, no depositional biomarkers have been found in any of these rocks, not surprising given the metamorphic grade.

Palaeolatitude and palaeogeography

Figure 63.3 provides a summary diagram of our C-isotopic data associated with the Dalradian’s glaciogenic intervals. Dramatic declines, from values of þ2‰ to þ6‰ to as low as –10‰ to –12‰, and recoveries mark the carbonate strata predating both the Port Askaig and Stralinchy-Reelan formations. These data

INISHOWEN–MACDUFF–LOCH NA CILLE GLAC. Marinoan–equivalent cap carbonates Whiteness Limestone Cranford Limestone

There are no reliable depositional palaeomagnetic data on the Dalradian rocks; this renders palaeolatitudinal placements, at best, speculative. Generalized palaeogeographies envisage east to west (current-day coordinates) transitions from shallow-marine shelf to deeper-marine settings in rift basins along the eastern margin of Laurentia for most of Dalradian time (Anderton 1982, 1985; Harris et al. 1994). Although several nicely detailed local studies have been undertaken (Klein 1970; Anderton 1976; Fairchild 1980), much additional work is needed to help refine reconstructions. Complicating matters are the wildly varying estimates for the magnitude of pre-Devonian displacement along the Great Glen Fault Zone; restoration of the Dalradian block can vary by as much as 1000 km (e.g. Dewey & Strachan 2003). Most reconstructions incorporate the Dalradian as part of eastern Laurentia, the position of which within a disassembling Rodinia is often depicted as moving along a broadly low- to mid-latitude arc during Cryogenian time (e.g. Dalziel 1997; Pisarvesky et al. 2008).

STRALINCHY – REELAN GLACIAL Geochronological constraints

Ardrishaig Phyllite (& correlatives)

pre–glacial decline and recovery (Trezona anomaly equivalent?)

Geochronological constraints are frustratingly sparse for the Dalradian: Easdale Subgroup carbonate rocks

Bonahaven Formation

PORT ASKAIG GLACIAL pre–glacial decline and recovery

-15

-10

Islay–Glencolumbkille– Limestones

-5

0

5

10

δ C‰ (V-PDB) 13

Fig. 63.3. C-isotopic trends for part of the Dalradian Supergroup with inferred correlations to known Neoproterozoic chemostratigraphic events. Data are from McCay et al. (2006), Prave et al. (2009a) and Prave & Fallick (unpublished). See text for discussion.

† The only direct depositional age for the Dalradian is from an ash bed in the Tayvallich Volcanics near the base of the Southern Highland Group. It has a U –Pb zircon age of 601 + 4 Ma (Dempster et al. 2002) and occurs more than 500 m beneath the youngest Dalradian glacial beds (in this instance, the Loch na Cille Boulder Bed). The top of the Dalradian, well above the youngest glacial rocks, passes stratigraphically into the Cambrian Leny Limestone (Tanner 1995). † The nature of the base of the Dalradian, whether unconformable or depositional on rocks containing U –Pb zircon and monazite deformational ages ranging from c. 840–800 Ma (Highton et al. 1999; Noble et al. 1996), remains unresolved. If the base defines an unconformity, then these are maximum ages for the initiation of Dalradian sedimentation; if the base is transitional (albeit tectonized), then the lower parts of the Dalradian could be much older. Regardless, the earliest record of glaciation in the Dalradian, the Port Askaig Fm. at the base of the Argyll Group, occurs some 10 km above these rocks. † The youngest concordant detrital zircon U –Pb ages for the Argyll Group are c. 1000 Ma (Cawood et al. 2003). These are

NEOPROTEROZOIC GLACIOGENIC DEPOSITS OF SCOTLAND AND IRELAND

not robust constraints for depositional ages, but do confirm a post-early Neoproterozoic age for the initiation of Argyll Group sedimentation. † Time lines for the Dalradian have been proposed using chronostratigraphic constraints established for global climatic events and secular variations in the isotopic compositions of Neoproterozoic oceans. The proposed linkage by McCay et al. (2006) of the three Dalradian glacials to Neoproterozoic glaciations known from elsewhere is viable and testable. In this framework, the Port Askaig Fm. would be chronostratigraphically tied to the mid-late Neoproterozoic Sturtian glaciation(s). We recognize that the age and number of glaciations during this climatic phase remain contentious, but Brasier & Shields (2000) have argued, based on C and Sr isotopes, that the Port Askaig is best correlated to the c. 713 Ma Gubrah glacial in Oman (Bowring et al. 2007). This, then, would be the age of (or close to) initiation of Argyll Group sedimentation. The Cranford –Whiteness Limestones are interpreted as Marinoan-equivalent cap carbonates (McCay et al. 2006; Prave et al. 2009a), thereby making this part of the Easdale Subgroup 635 Ma, that is the age for the end of the Marinoan glaciation (Hoffmann et al. 2004; Condon et al. 2005). The MacDuff – Inishowen –Loch na Cille beds, which occur above the 601 Ma Tayvallich Volcanics and their correlatives, can be reasonably linked to the 582 Ma Gaskiers glaciation (Bowring et al. 2003, 2007). The best these data permit is to indicate that the post-Easdale Subgroup portion of the Dalradian belongs to the Ediacaran Period and that the underlying Dalradian is Cryogenian in age (Fig. 63.2). Such broad age constraints are unsatisfying. The irony is that, although it is one of the most-studied and mapped regions in the world, the Scottish –Irish Highlands is one of the most poorly constrained, geochronologically.

Discussion Two main data sets have enabled construction of a glacial-based stratigraphic framework for the Scottish –Irish Dalradian. The first is the presence of discrete glacially influenced deposits ranging from thick diamictites to thin, discontinuous ice-rafteddebris beds (in areas of higher metamorphic grade, these are recognized as fine-grained schists with lonestones). Excepting the Port Askaig Fm., with its impressive, laterally extensive diamictite beds, the record of Neoproterozoic glaciation in the Dalradian is modest and mostly represented by thin ice-rafted-debris intervals. The second data set comes from carbonate units that are lithologically and C-isotopically compatible with Neoproterozoic cap carbonates. The Scottish – Irish Dalradian succession contains evidence for three distinct glaciations, from oldest to youngest, (i) the Port Askaig Fm. at the base of the Argyll Group (note that no associated cap carbonate has been recognized), (ii) the Stralinchy – Reelan formations (and correlative ice-rafted-debris beds) and the Cranford and Whiteness cap carbonates in the Easdale Subgroup, and (iii) the MacDuff –Inishowen –Loch na Cille Boulder Beds and ice-rafted-debris units in the lower Southern Highland Group. Geochronological constraints are sparse for the Dalradian, but those that exist, combined with C-isotope chemostratigraphy, enable placing the base of the Ediacaran Period in the Easdale Subgroup and the underlying Dalradian units in the Cryogenian. Numerous colleagues have provided insight and help over the years and we would like to particularly acknowledge I. Alsop, D. Benn, D. Condon, G. McCay, G. Oliver, C. Rose, J. Soper, R. Strachan and G. Thomas. We thank The Carnegie Trust for The Universities of Scotland, The Russell Trust and The Schroder Foundation (Islay) for financial support. This represents a contribution of the IUGSand UNESCO-funded IGCP (International Geoscience Programme) Project #512.

647

References Alsop, G. I. & Hutton, D. W. 1990. A review and revision of Dalradian stratigraphy in central and southern Donegal, Ireland. Irish Journal of Earth Sciences, 10, 181–198. Alsop, G. I., Prave, A. R., Condon, D. J. & Phillips, C. A. 2000. Cleaved clasts in Dalradian conglomerates: possible evidence for Neoproterozoic compressional tectonism in Scotland and Ireland? Geological Journal, 35, 87– 98. Anderton, R. 1976. Tidal shelf sedimentation: an example from the Scottish Dalradian. Sedimentology, 23, 429–458. Anderton, R. 1979. Slope, submarine fans, and syn-depositional faults: sedimentology of parts of the Middle and Upper Dalradian in the SW Highlands of Scotland. In: Harris, A. L., Holland, C. H. & Leake, B. E. (eds) The Caledonides of the British Isles – Reviewed. Geological Society, London, Special Publications, 8, 483–488. Anderton, R. 1982. Dalradian deposition and the late Precambrian – Cambrian history of the North Atlantic region: a review of the early evolution of the Iapetus Ocean. Journal of the Geological Society, London, 139, 421– 431. Anderton, R. 1985. Sedimentation and tectonics in the Scottish Dalradian. Scottish Journal of Geology, 21, 407– 436. Anderton, R. 1988. Dalradian slides and basin development: a radical reinterpretation of stratigraphy and structure in the SW and Central Highlands of Scotland. Journal of the Geological Society, London, 145, 669–678. Arnaud, E. & Eyles, C. H. 2006. Neoproterozoic environmental change recorded in the Port Askaig Formation, Scotland: climatic and tectonic controls on sedimentation. Sedimentary Geology, 183, 99– 124. Arnaud, E. & Fairchild, I. J. 2011. The Port Askaig Formation, Dalradian Supergroup, Scotland. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 635– 642. Bowring, S., Myrow, P., Landing, E., Ramezani, J. & Grotzinger, J. 2003. Geochronological constraints on terminal Neoproterozoic events and the rise of metazoans. Geophysical Research Abstracts, 5, 13219. Bowring, S. A., Grotzinger, J. P., Condon, D. J., Ramezani, J., Newall, M. & Allen, P. A. 2007. Geochronologic constraint of the chronostratigraphic framework of the Neoproterozoic Huqf Supergroup, Sultanate of Oman. American Journal of Science, 307, 1097–1145. Brasier, M. D. & Shields, G. 2000. Neoproterozoic chemostratigraphy and correlation of the Port Askaig glaciation, Dalradian Supergroup of Scotland. Journal of the Geological Society, London, 157, 909– 914. Cawood, P. A., Nemchin, A. A., Smith, M. & Loewy, S. 2003. Source of the Dalradian Supergroup constrained by U –Pb dating of detrital zircon and implications for the East Laurentian margin. Journal of the Geological Society, London, 160, 231–246. Condon, D. J. & Prave, A. R. 2000. Two from Donegal: Neoproterozoic glacial episodes on the NE margin of Laurentia. Geology, 28, 951– 954. Condon, D. J., Prave, A. R. & Benn, D. 2002. Neoproterozoic glacial – rainout intervals: observations and implications. Geology, 30, 35– 38. Condon, D., Zhu, M., Bowring, S., Wang, W., Yang, A. & Jin, Y. 2005. U–Pb ages from the Neoproterozoic Doushantuo Formation, China. Science, 308, 95– 98. Dalziel, I. W. D. 1997. Neoproterozoic-Paleozoic geography and tectonics: review, hypothesis, environmental speculation. Geological Society of America Bulletin, 109, 16– 42. Dalziel, I. W. D. & Soper, N. J. 2001. Neoproterozoic extension on the Scottish Promontory of Laurentia: paleogeographic and tectonic implications. Journal of Geology, 109, 299–317. Davison, S. & Hambrey, M. J. 1996. Indications of glaciation at the base of the Proterozoic Stoer Group (Torridonian), NW Scotland. Journal of the Geological Society, London, 153, 139–149. Dempster, T. J., Rogers, G. et al. 2002. Timing of deposition, orogenesis and glaciation within the Dalradian rocks of Scotland: Constraints from U –Pb zircon ages. Journal of the Geological Society, London, 159, 83 – 94.

648

A. R. PRAVE & A. E. FALLICK

Dewey, J. F. & Strachan, R. A. 2003. Changing Silurian –Devonian relative plate motion in the Caledonides: sinistral transpression to sinistral transtension. Journal of the Geological Society, London, 160, 219– 229. Elles, G. L. 1934. The Loch na Cille Boulder Bed and its place in the Highland succession, Geological Society of London Quarterly Journal, 91, 111– 147. Fairchild, I. 1980. Sedimentation and origin of a late Precambrian ‘dolomite’ from Scotland. Journal of Sedimentary Petrology, 50, 423–446. Fairchild, I. & Kennedy, M. J. 2007. Neoproterozoic glaciation in the Earth System. Journal of the Geological Society, London, 164, 895– 921. Hambrey, M. J. 1983. Correlation of Late Proterozoic tillites in the North Atlantic region and Europe. Geological Magazine, 120, 209– 232. Hambrey, M. J. & Waddams, P. 1981. Glaciogenic boulder-bearing deposits in the Upper Dalradian MacDuff Slates, northeastern Scotland. In: Hambrey, M. J. & Harland, W. W. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 571– 575. Harris, A. L., Haselock, P. J., Kennedy, M. & Mendum, J. R. 1994. The Dalradian Supergroup in Scotland, Shetland and Ireland. In: Gibbons, W. & Harris, A. L. (eds) A Revised Correlation of Precambrian Rocks in British Isles. Geological Society, London, Special Report, 22, 33 –53. Highton, A. J., Hyslop, E. K. & Noble, S. R. 1999. U– Pb zircon geochronology of migmatisation in the northern-central Highlands: evidence for pre-Caledonian (Neoproterozoic) tectonometamorphism in the Grampian block, Scotland. Journal of the Geological Society, London, 156, 1195–1204. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129–155. Hoffmann, K. H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. A U–Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817–820. Kennedy, M. J., Runnegar, B., Prave, A. R., Hoffmann, K. H. & Arthur, M. 1998. Two or four Neoproterozoic glaciations? Geology, 26, 1059–1063. Kirkland, C. L., Strachan, R. A. & Prave, A. R. 2008. Detrital zircon signature of the Moine Supergroup, Scotland: contrasts and comparisons with other Neoproterozoic successions within the circum-North Atlantic region. Precambrian Research, 163, 332–350. Klein, G. De V. 1970. Tidal origin of a Precambrian quartzite: lower finegrained quartzite (Middle Dalradian) of Islay, Scotland. Journal of Sedimentary Petrology, 40, 973– 985. McCay, G. A., Prave, A. R., Alsop, G. I. & Fallick, A. E. 2006. Glacial trinity: Neoproterozoic Earth history within the British– Irish Caledonides. Geology, 34, 901–912. Millar, I. L. 1999. Neoproterozoic extensional basic magmatism associated with the West Highland granite gneiss in the Moine Supergroup of NW Scotland. Journal of the Geological Society, London, 156, 1153– 1162. Molyneux, S. G. 1997. An upper Dalradian microfossil reassessed. Journal of the Geological Society, London, 155, 741–743. Noble, S. R., Hyslop, E. K. & Highton, A. J. 1996. High precision U –Pb monazite geochronology of the c.806 Ma Grampian Shear Zone and implications for the evolution of the Central Highlands of Scotland. Journal of the Geological Society, London, 153, 511–514. Oliver, G. J. H., Chen, F., Buchwald, R. & Hegner, E. 1998. Fast tectonometamorphism and exhumation in the type area of the Barrovian and Buchan zones. Geology, 28, 459– 462. Piasecki, M. A. J. & van Breemen, O. 1983. Field and isotope evidence for a c.750 Ma tectonothermal event in Moine rocks in the Central

Highland region of the Scottish Caledonides. Transactions of the Royal Society of Edinburgh: Earth Sciences, 73, 119– 134. Pisarvesky, S. A., Murphy, J. B., Cawood, P. A. & Collins, A. S. 2008. eeLate Neoproterozoic and early Cambrian palaeogeography: models and problems. Journal of the Geological Society, London, 294, 9– 31. Prave, A. R. 1999. The Neoproterozoic Dalradian Supergroup of Scotland: an alternative hypothesis. Geological Magazine, 136, 609– 617. Prave, A. R., Strachan, R. A. & Fallick, A. E. 2009a. Global C cycle perturbations recorded in marbles: a record of Neoproterozoic Earth history within the Shetland Islands, Scotland. Journal of the Geological Society, London, 166, 129–135. Prave, A. R., Fallick, A. E., Thomas, C. W. & Graham, C. M. 2009b. A composite C-isotopic profile for the Neoproterozoic Dalradian Supergroup of Scotland and Ireland. Journal of the Geological Society, London, 166, 845–857. Rainbird, R. H., Hamilton, M. A. & Young, G. M. 2001. Detrital zircon geochronology and provenance of the Torridonian, NW Scotland. Journal of the Geological Society, London, 158, 15 –27. Read, H. H. 1923. The geology of the country around Banff, Huntly and Turriff. Memoir Geological Survey Scotland, 86 & 96, 240. Robertson, S. & Smith, M. 1999. The significance of the Geal Charn– Ossian Steep Belt in basin development and inversion in the Central Scottish Highlands. Journal of the Geological Society, London, 156, 1175– 1182. Soper, N. J. & Anderton, R. 1984. Did the Dalradian slides originate as extensional faults? Nature, 307, 357–360. Spencer, A. M. 1971. Late Pre-cambrian Glaciaion in Scotland. Geological Society, London, Memories, 6, 100. Stewart, A. D. 1997. Discussion on: Indications of glaciation at the base of the Proterozoic Stoer Group (Torridonian), NW Scotland. Journal of the Geological Society, London, 154, 375– 376. Stewart, A. D. 2002. The Later Proterozoic Torridonian Rocks of Scotland: their Sedimentology, Geochemistry and Origin. Geological Society, London, Memoirs, 24, 130. Stoker, M. S., Howe, J. A. & Stoker, S. J. 1999. Late Vendian? Cambrian glacially influenced deepwater sedimentation, MacDuff Slate Formation (Dalradian), NE Scotland. Journal of the Geological Society, London, 156, 55 –61. Sutton, J. & Watson, J. V. 1954. Ice-borne boulders in the MacDuff Group of the Dalradian of Banffshire. Geological Magazine, 91, 391– 398. Tanner, P. W. G. 1995. New evidence that the Lower Cambrian Leny Limestone at Callender, Perthshire, belongs to the Dalradian Supergroup, and a reassessment of the ‘exotic’ status of the Highland Border Complex. Geological Magazine, 132, 473– 483. Tanner, P. W. G. & Evans, J. A. 2003. Late Precambrian U– Pb titanite age for peak regional metamorphism and deformation (Knoydartian orogeny) in the western Moine, Scotland. Journal of the Geological Society, London, 160, 555–564. Trewin, N. H. 1987. MacDuff, Dalradian turbidite fan and glacial deposits. In: Trewin, N. H., Kneller, B. C. & Gillen, C. (eds) Geology of the Aberdeen Area. Geological Society of Aberdeen. Scottish Academic Press, Edinburgh, 79 –88. Turnbull, M. J. M., Whitehouse, M. J. & Moorbath, S. 1996. New isotopic age determinations for the Torridonian, NW Scotland. Journal of the Geological Society, London, 153, 955– 964. Vance, D., Strachan, R. A. & Jones, K. A. 1998. Extensional versus compressional settings for metamorphism: garnet chronometry and pressure– temperature –time histories in the Moine Supergroup, NW Scotland. Geology, 26, 927– 930.

Chapter 64 Neoproterozoic glacial deposits of Tasmania CLIVE R. CALVER Mineral Resources Tasmania, P.O. Box 56, Rosny Park, Tasmania, Australia 7018 (e-mail: [email protected]) Abstract: In Tasmania, Neoproterozoic glaciogenic deposits were laid down in one or more epicratonic basins, probably situated at the eastern margin of the Australian– Antarctic craton. Rifting and volcanism took place in the late Cryogenian to early Ediacaran. On King Island, north of Tasmania, the Cottons Breccia consists of 50–200 m of diamictite, conglomerate and sandstone. Limestone and dolostone clasts are abundant in the diamictite, although carbonate is unknown in the underlying successions. The Cottons Breccia is overlain by 10 m of laminated dolostone and limestone with a negative, upward-decreasing d13C profile. Rift volcanics and shallow intrusives higher in the sequence are dated at c. 575 Ma. In NW Tasmania, two diamictite units are found in the Togari Group. The Julius River Member, 200 m thick, contains dominantly dolostone clasts and overlies a shallow-marine dolostone unit with vase-shaped microfossils and C-isotopes consistent with a mid-Cryogenian age. Some clasts in the Julius River Member contain a stromatolite (Baicalia cf. B. burra) very similar to a form that is abundant in the middle part of the Burra Group, Adelaide rift basin. The Julius River Member is immediately overlain by black shale and impure carbonate dated by Re–Os at 641 + 5 Ma. The younger diamictite in the Togari Group is the Croles Hill Diamictite, 70 m thick, with predominantly volcanic clasts, underlain by a shale and mafic-volcaniclastic succession and overlain by thin mudstone followed by thick rift tholeiites. At one locality this diamictite is underlain by a rhyodacite flow dated at 582 + 4 Ma. In southern Tasmania, diamictites are found in the Wedge River Beds and in the Cotcase Creek Formation (Fm.) (Weld River Group). Laminated siltstone with dropstones is associated with the diamictites in the Cotcase Creek Fm. The southern Tasmanian deposits are poorly constrained in age. Chronometric and other evidence suggests correlation of the Julius River Member, Cottons Breccia and Croles Hill Diamictite with the Sturt, Elatina and Gaskiers glacial phases, respectively. However, a glacial origin for the Julius River Member and Croles Hill Diamictite remains uncertain.

Proterozoic rocks of King Island and western Tasmania comprise a continental fragment that lay at the eastern margin of the Australian/east Antarctic craton in the Neoproterozoic (Burrett & Berry 2000; Li et al. 2008). This fragment probably rifted at c. 580 Ma and accreted back to the craton in a late phase of the Cambrian, Delamerian/Ross Orogeny (Berry et al. 2008). Alternatively, King Island (and perhaps western Tasmania as well) comprised a rifted margin that never completely separated from the craton (Meffre et al. 2000, 2004). Cryogenian to Ediacaran successions unconformably overlie Mesoproterozoic to early Neoproterozoic sediments and metasediments in many places, but Cambrian and later tectonism, as well as significant lateral variability in the Cryogenian –Ediacaran successions themselves, have made correlation across Tasmania problematic. Consequently, three regions with known or probable Neoproterozoic glaciogenic deposits are described separately below: King Island, NW Tasmania and southern Tasmania.

King Island: the Cottons Breccia Introduction The Cottons Breccia (Jago 1974) occurs near the base of the Grassy Group on King Island (Fig. 64.1). It consists of up to 200 m of diamictite and minor conglomerate and sandstone, and crops out over a strike length of about 8 km along or near the SE coast of the island from Cottons Flat to north of Cumberland Creek (Fig. 64.1b). In a few places there is excellent exposure on the coast, although the basal part of the unit is relatively poorly exposed inland. Waterhouse (1916) first described the unit as a lithified ‘glacial till’; Carey (1947) suggested it was an Adelaidean tillite. Jago (1974) named and described the Cottons Breccia in some detail, but left its mode of origin unresolved. No type section was nominated. Further description and discussion were provided by Jago (1981) and Waldron & Brown (1993), the latter authors favouring a debris-flow origin. Calver & Walter (2000) gave evidence, including d13C stratigraphy of the overlying dolostone, for correlation of the Cottons Breccia with the Elatina

glacials of South Australia. Meffre et al. (2004) and Calver et al. (2004) isotopically dated the overlying volcanics and shallow intrusives respectively, the latter authors specifically in the context of correlation and age constraints for the Cottons Breccia. Direen & Jago (2008) contended that the Cottons Breccia is a nonglaciogenic mass flow deposit laid down in a developing rift.

Structural framework The Grassy Group unconformably overlies a thick, Mesoproterozoic to early Neoproterozoic siliciclastic shelf succession (Black et al. 2004). The Grassy Group was deposited in an epicratonic rift basin that may have evolved into an east-facing passive margin soon after deposition of the Cottons Breccia (Direen & Crawford 2003; Meffre et al. 2004; Direen & Jago 2008). The thick mafic volcanics (rift tholeiites and picrites) that begin c. 100 m above the Cottons Breccia have been interpreted as analogous to the mafic volcanic packages (‘seaward dipping reflector sequences’) at the base of Mesozoic –Cenozoic volcanic continental margins (Direen & Crawford 2003; Meffre et al. 2004). The distribution of these volcanics can be traced as an offshore magnetic anomaly that extends southward to offshore western Tasmania, but it is not continuous with the similar-aged Spinks Creek Volcanics (next section), suggesting the Grassy Group and Togari Group filled separate structural basins. King Island was probably weakly deformed in the Cambrian, Tyennan Orogeny (Delamerian/Ross Orogeny correlate). Further deformation may have occurred in the Devonian, Tabberabberan Orogeny, followed by local granitoid intrusion in the Early Carboniferous. In the Cottons Flat –Cumberland Creek area, the Grassy Group dips moderately east, and a weak, subvertical cleavage is developed in shales. Minor cross-faulting results in different parts of the succession being exposed along the coast. Mineral assemblages in the volcanics indicate low greenschistfacies metamorphism (Meffre et al. 2004). Near Grassy, however, the sequence has been strongly contact metamorphosed and metasomatized by the Early Carboniferous Sandblow Granite (Fig. 64.1b).

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 649– 657. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.64

650

C. R. CALVER

Fig. 64.1. (a) Distribution of Proterozoic rocks and major tectonic elements in Tasmania. (b) Simplified geological map of SE King Island with localities mentioned in text. (c) Generalized stratigraphic section through Grassy Group in Bold Head–Cumberland Creek area.

Stratigraphy The Grassy Group (Fig. 64.1c) unconformably overlies the Fraser Fm., a thick succession of siliceous siltstone and mudstone of Mesoproterozoic or early Neoproterozoic age (Black et al. 2004; Direen & Jago 2008). The basal unit of the Grassy Group is the Robbins Creek Fm., consisting of laminated siltstone and shale with minor lavas and a thin basal conglomerate (Calver 2008a, b). The overlying Cottons Breccia shows considerable lithological variation, both laterally and vertically, and varies irregularly in thickness from c. 50 m in the north (Cumberland Creek) to c. 200 m in the south (City of Melbourne Bay) (Jago 1974; Calver 2008b). The Cottons Breccia is overlain by the Cumberland Creek Dolostone (c. 10 m thick), which passes up gradationally into green and black, then red, shale (the Yarra Creek Shale: Calver et al. 2004). The Yarra Creek Shale varies in thickness from c. 10 m to 150 m, apparently in response to movement on penecontemporaneous faults (Direen & Jago 2008; Calver 2008a). A differentiated sill of intermediate composition (Grimes Intrusive Suite) intrudes the lower Yarra Creek Shale and older units, and is locally amygdaloidal. The Yarra Creek Shale is conformably overlain by the tholeiitic City of Melbourne Volcanics consisting of peperites and volcanic breccias overlain by pillow lavas. The picritic Shower Droplet Volcanics follow, then the tholeiitic Grahams Road Volcanics (Waldron & Brown 1993; Meffre et al. 2004; Calver 2008b). The top of the Grassy Group is unknown and lies offshore. Aeromagnetic mapping shows that the volcanics continue offshore and may be several kilometres thick in total (Direen & Crawford 2003).

Glaciogenic deposits and associated strata The Cottons Breccia is dominantly massive or crudely stratified diamictite with minor conglomerate, sandstone and laminated pebbly siltstone (Jago 1974). Clasts in the diamictite are predominantly sub-angular to sub-rounded, and the maximum size in any outcrop is typically 300–500 mm, but commonly less

(c. 100 mm). The larger clasts are commonly oriented with long axes parallel to stratification. At the Gut (Fig. 64.1b) a 15-mthick interval contains carbonate boulders up to 3 m long, but this horizon cannot be traced any distance laterally. Clast lithologies comprise abundant carbonate (including cream and grey dolostone, dark limestone, oolitic limestone and dolostone conglomerate), abundant fine-grained siliciclastics (mainly finegrained quartzarenite, siltstone, sandstone and mudstone), and rare chert, red jasper, basic volcanics and dacite (Jago 1974; Waldron & Brown 1993; Calver & Walter 2000). Common striated clasts were noted by Solomon (1969) but have not been confirmed by subsequent workers (Jago 1974; Waldron & Brown 1993). Diamictite in the lower part of the Cottons Breccia has clasts of predominantly carbonate and a grey muddy carbonate matrix, while in the upper part, siliciclastic clasts become more abundant and the matrix is a red-brown mudstone (Calver & Walter 2000). Jago (1974) noted graded bedding (30 –300 mm) in southern outcrops and a number of lenses of well-sorted sandstone and siltstone up to 300 mm thick and 10 m long, within the diamictite. Between Cottons Flat and the Gut there is a middle unit, up to 20 m thick, of green mafic-volcaniclastic sandstone, with shards visible in thin section. At Cottons Flat, diamictite in the lower part of the formation includes minor intervals of well-laminated siltstone crowded with dropstones up to 100 mm in size (Jago 1974, figs 23, 24; P. Hoffman, pers. comm.). At the Gut, there is an upward transition from diamictite through 2 m of closed-framework conglomerate, then 0.8 m of plane-laminated red sandstone, pebbly at the base, which in turn is conformably and gradationally overlain by the Cumberland Creek Dolostone. The Cumberland Creek Dolostone is a pale grey to pale pinkishgrey, fine-grained laminated dolostone (weathering to a pale yellow-brown), passing up into thinly interbedded dolostone and shale, then shale with thin, pale grey, fine-grained limestone beds, c. 10 m thick in total. Thin sections of the dolostone show dolomicrite or turbid dolomicrosparite, with scattered rounded micritic peloids 0.1–0.25 mm in diameter, and irregular microfenestrae (Calver & Walter 2000). Sharp-crested intrastratal anticlines, 100– 200 mm in amplitude, are found here and there in

NEOPROTEROZOIC GLACIAL DEPOSITS OF TASMANIA

the lower, dolomitic part of the unit. The overlying Yarra Creek Shale is massive or plane-laminated, and pale yellow-brown to red except for a middle interval with beds of black shale that have the characteristic microfabric of benthic microbial mats (Calver & Walter 2000).

Boundary relations with overlying and underlying non-glacial units The base of the Cottons Breccia is not exposed, but its variable thickness suggests erosional incision into the underlying Robbins Creek Fm. (Calver 2008a). The top of the Cottons Breccia is, at least locally, a conformable and gradational contact with the overlying Cumberland Creek Dolostone (previous section). However, at a locality north of Conglomerate Creek, the contact with the dolostone is abrupt and disconformable (Jago 1974).

Chemostratigraphy Carbonate d13C and d18O data for six samples of the Cumberland Creek Dolostone were presented by Calver & Walter (2000). The unit undergoes a steep upward decline in d13C, from –1.9‰ at the base to –5.0‰ near the top. O-isotopic compositions are moderately to strongly depleted ( –7.9 to –16.7‰ VPDB).

Other characteristics At Grassy, the Grassy Group hosts economically important calcic scheelite skarns in the contact aureole of the Sandblow Granite (Danielson 1975).

651

NW Tasmania: Julius River Member and Croles Hill Diamictite Introduction In the Smithton Synclinorium of NW Tasmania, two diamictite units (the Julius River Member and the Croles Hill Diamictite) are found in a Cryogenian to Early Cambrian succession known as the Togari Group (Calver 1998; Everard et al. 2007). The Julius River Member is found within the upper part of the Black River Dolomite, and crops out at widespread localities within the Smithton Synclinorium (Fig. 64.2). It was first briefly described by Longman & Matthews (1962). Griffin & Preiss (1976) described outcrops now recognized as belonging to the Julius River Member, and the stromatolites contained in some of the clasts. The Julius River Member was described under the informal name ‘Trowutta Breccia’ by Jago (1981). The unit was formally defined by Everard et al. (2007), with a type section, c. 200 m thick, on the Arthur River (Fig. 64.2). Calver (1998) reported C and Sr chemostratigraphic data for the Togari Group and suggested correlation of the Julius River Member with the Sturtian glacial deposits of the Adelaide Geosyncline. Kendall et al. (2007) supported this correlation with a Re – Os age determination on black shale immediately overlying the Julius River Member. The Croles Hill Diamictite (Everard et al. 2007) is part of the Kanunnah Subgroup, which overlies the Black River Dolomite (Fig. 64.2). Up to 250 m thick, the formation is best developed in the south of the Synclinorium, with significant localities also near Forest in the NE and at Robbins Passage in the NW of the Synclinorium (Fig. 64.2; Brown 1989, p. 25; Calver et al. 2004). Everard et al. (2007) formally defined the Croles Hill Diamictite, with a type section on the Arthur River (Fig. 64.2).

Palaeolatitude and palaeogeography Structural framework A palaeomagnetic study by T. Raub (pers. comm.) shows a probable primary magnetization in the upper Cottons Breccia and basal Cumberland Creek Dolostone, with both polarities preserved and directed approximately to present-day north and south. These units appear to have been deposited within 58 of the equator.

Geochronological constraints Calver & Walter (2000) argued for correlation between the Cottons Breccia and the Elatina (‘Marinoan’) glacials of the Adelaide Geosyncline on the basis of, inter alia, the strong lithologic and d13C-chemostratigraphic resemblance between the Cumberland Creek Dolostone and the Nuccaleena Fm. (Williams et al. 2011). Correlation of the Yarra Creek Shale with the lower Brachina Fm. of the Adelaide Geosyncline is also supported by lithostratigraphy, including the presence of microbialite black shale beds both in the Yarra Creek Shale and low in the Brachina Fm. and equivalents (Logan et al. 1999; Calver 2000). There are two radiometric minimum age constraints on the Cottons Breccia: (i) a Nd –Sm isochron age of 579 + 16 Ma on the Shower Droplet Volcanics and Grahams Road Volcanics (Meffre et al. 2004), (n ¼ 5, 2s error limits, MSWD ¼ 1.6; initial 1Nd of 4.2); (ii) a U –Pb (SHRIMP on zircon) age of 574.7 + 3.0 Ma on the Grimes Intrusive Suite (Calver et al. 2004) (n ¼ 26, MSWD ¼ 0.73; precision limits at 95% (2s) confidence level). Calver et al. (2004) considered the Cottons Breccia to be not very much older than these dates, assuming the Cumberland Creek Dolostone and Yarra Creek Shale were deposited during the post-glacial transgression, and given that the Grimes intrusives are locally vesicular (and therefore probably shallow) and that the basaltic volcanics were extruded while the Yarra Creek Shale was still unconsolidated.

The Togari Group unconformably overlies a thick early Neoproterozoic siliciclastic shelfal succession, the Rocky Cape Group. The part of the Togari Group below the Julius River Member records epicratonic, shallow-marine, predominantly carbonate sedimentation. In the Kanunnah Subgroup above the Julius River Member, basalts (dominantly rift tholeiites) and associated volcaniclastics are abundant. The NE-trending Roger River Fault was an east-side-down growth fault at this time (Everard et al. 2007). NW –SE extension is also shown by the approximately coeval intrusion of the Tayatea Dyke Swarm (588 – 600 Ma) into basement east of the Smithton Synclinorium (Brown 1989). Rifting was evidently aborted at this location as the Kanunnah Subgroup is conformably followed by thick upper Ediacaran shallow-marine carbonates of the Smithton Dolomite. The succession was gently folded in the Cambrian Tyennan Orogeny and again in the Devonian, Tabberabberan Orogeny. The Smithton Synclinorium (a large, gently north-plunging synclinal structure) is probably largely a Devonian structure. Dips are mainly gentle to moderate except along the western margin, where steep dips are associated with east-directed thrusting (Everard et al. 2007). The rocks are undeformed to weakly deformed except locally in the south, where pelitic lithologies are moderately cleaved. Low-grade regional metamorphism (of prehnite-pumpellyite grade, Griffin & Preiss 1976) affects the Togari Group.

Stratigraphy The Black River Dolomite (c. 800 m thick), or an impersistent basal siliciclastic unit, the Forest Conglomerate, rests with gently angular unconformity or paraconformity on the lower

652

C. R. CALVER

Fig. 64.2. (a) Bedrock geological map of Smithton Synclinorium (for location see Fig. 64.1a). (b) Stratigraphic sections at localities shown as numbered asterisks in (a).

Neoproterozoic Rocky Cape Group (Black et al. 2004; Everard et al. 2007). The Julius River Member comprises a 200-m-thick unit of diamictite in the upper part of the Black River Dolomite. The Black River Dolomite is conformably overlain by the Kanunnah Subgroup, c. 1200 m of shale, siltstone, volcaniclastic sandstone and basalt. Thicknesses given above are characteristic of the eastern and central parts of the Synclinorium: in the western part, the units tend to be considerably thinner. Early workers (up to and including Saito et al. 1988) erroneously correlated the dolostone underlying the Julius River Member with the Smithton Dolomite, now recognised as being higher in the stratigraphy (Brown 1989; see next section). The Kanunnah Subgroup is laterally variable. In its type section (Arthur River, Fig. 64.2) it begins with poorly exposed siltstone and mudstone (c. 120 m), followed by the Croles Hill Diamictite (70 m), basalt (Spinks Creek Volcanics, 250 m), then more, poorly exposed siltstone and shale (c. 500 m), which in turn is

conformably overlain by the Smithton Dolomite (1500 m) (Fig. 64.2b). The sediments of the Kanunnah Subgroup (other than the Croles Hill Diamictite) are collectively known as the Keppel Creek Fm. That part underlying the Croles Hill Diamictite will be referred to here as the lower Keppel Creek Fm.

Glaciogenic deposits and associated strata: Julius River Member That part of the Black River Dolomite underlying the Julius River Member consists of fine-grained uniform dolostone, with minor intercalated chert, limestone and black shale. Locally, stromatolites or peloidal, intraclastic and oolitic fabrics are preserved, suggesting shallow-water conditions. The Julius River Member, in the incompletely exposed type section, is an open-framework diamictite with sparse to abundant,

NEOPROTEROZOIC GLACIAL DEPOSITS OF TASMANIA

angular to subrounded clasts up to 500 mm in size (Calver 1998; Everard et al. 2007). Griffin & Preiss (1976) gave 2 m as the maximum clast size, while Jago (1981) considered the great majority of clasts to be ,10 mm. The clasts are predominantly pale grey, fine-grained dolostone, with minor chert and mudstone. Rare limestone and basalt clasts have been noted elsewhere (Everard et al. 2007). Diagenesis (dolomitization, cementation) of the dolostone clasts was complete at the time of incorporation into the diamictite (Griffin & Preiss 1976; Calver 1998). Some of the dolostone clasts contain well-preserved stromatolites attributable to Baicalia cf. B. burra (Griffin & Preiss 1976). The great majority of clasts could have been derived from the underlying Black River Dolomite, although in situ Baicalia has not been recorded. The matrix of the diamictite is a dark grey, dolomitic silty mudstone, diffusely laminated in places (Everard et al. 2007). Poorly developed, metre-scale graded bedding was observed in one outcrop by Griffin & Preiss (1976). A complete intersection of the Julius River Member is available in the fully cored Forest-1 drill hole (Fig. 64.2, locality 4). This consists of 67 m of dark grey diamictite with predominantly dolostone clasts, overlain by 8 m of pale grey, fine-grained (microsparitic) limestone, then 23 m of diamictite with predominantly limestone clasts. The upper contact is not exposed in the type section, but in Forest-1, the Julius River Member is overlain by c. 5 m of black, uniform, impure fine-grained dolostone, then interbedded black pyritic shale, grey shale and minor limestone. These units comprise an uppermost, un-named member of the Black River Dolomite. The top of this member (and base of the Kanunnah Subgroup) is transitional, and is taken as the point at which beds of volcaniclastic sandstone become common, 41 m above the top of the Julius River Member. A probable correlate of the Julius River Member is found in the upper part of the Success Creek Group in western Tasmania (Calver 1996).

Boundary relations with overlying and underlying non-glacial units: Julius River Member The basal contact of the Julius River Member upon dolostone is abrupt in Forest-1. Longman & Matthews (1962) regarded the base of the diamictite as disconformable. Calver (1998) suggested a disconformity at this level based on comparison of d13C profiles between Forest-1 and Arthur River. However, Everard et al. (2007, p. 44) describe an apparently transitional contact, over a few tens of metres, in outcrop near the Arthur River. Longman & Matthews (1962) stated that the diamictite grades up into dolomitic siltstone, in turn gradationally overlain by banded black and grey siltstone. In Forest-1, the upper contact is conformable with the overlying black dolostone at the base of the uppermost, un-named member of the Black River Dolomite (previous section).

Glaciogenic deposits and associated strata: Croles Hill Diamictite In the Arthur River section, the lower Keppel Creek Fm. consists of poorly exposed, grey-green and black, thin-bedded mudstone and siltstone. The contact with the overlying Croles Hill Diamictite is not exposed. At Robbins Passage (Fig. 64.2, locality 1), black, organic-rich siltstone and mudstone are overlain by a rhyodacite flow c. 30 m thick, in turn overlain by the Croles Hill Diamictite, although the contact is not exposed. In the Forest-1 drill hole (Fig. 64.2, locality 4), the lower Keppel Creek Fm. is relatively thick (540 m) and is dominantly volcaniclastic sandstone in graded beds that are probably turbidites (Brown 1989). There is lesser interbedded shale and siltstone. Minor basalt flows, basaltic breccia and conglomerate were also intersected in this interval.

653

The Croles Hill Diamictite is an open-framework diamictite with angular to rounded clasts up to 300 mm, exceptionally up to 1 m, in diameter (Everard et al. 2007). Clasts are predominantly of various basaltic rocktypes, with some felsic volcanic rocks, and less common lithic sandstone, gabbro, dolostone, siltstone, mudstone and chert. The matrix is a dark grey-green, less commonly red, silty mudstone. Locally the matrix is laminated, and the clasts occur as relatively sparse, small (,20 mm) lonestones and possible dropstones (i.e. that interrupt and deflect thin lamination: Calver et al. 2004, fig. 4; Everard et al. 2007, p. 54). The Croles Hill Diamictite is relatively thin, and locally absent, to the west of the Roger River Fault, which is interpreted to have been an east-side-down normal fault active during Kanunnah Subgroup deposition (Everard et al. 2007). At Robbins Passage, felsic volcanic clasts are abundant, and a second rhyodacite flow occurs within the Croles Hill Diamictite, which here is c. 100 m thick. As mapped, the Croles Hill Diamictite is shown as being directly overlain by basalt of the Spinks Creek Volcanics over wide areas (e.g. Everard et al. 1996). However, well-exposed sections generally show a thin intervening unit of red to grey shale. At Robbins Passage, red mudstone (1 m) conformably overlies the diamictite, then 10 m of unexposed section is succeeded by at least 200 m of Spinks Creek Volcanics. In the Forest-1 drill hole, the Croles Hill Diamictite (89 m) is sharply overlain by pebble conglomerate and sandstone grading up into laminated red shale, 23 m thick in total, overlain in turn by basalt (Spinks Creek Volcanics).

Boundary relations with overlying and underlying non-glacial units: Croles Hill Diamictite Upper and lower contacts are abrupt and apparently conformable in the Forest-1 drill hole, and the Croles Hill Diamictite is also conformably overlain by red mudstone at Robbins Passage (previous section). A significant stratigraphic break or disconformity at the base of the Croles Hill Diamictite may be implied by the highly variable thickness of the lower Keppel Creek Fm. (e.g. Fig. 64.2b), and by the apparent absence of evidence of a c. 635 Ma glaciation from the section (see Discussion).

Chemostratigraphy In the Arthur River section, the lower half of the Black River Dolomite is characterized by d13C between þ3 and þ5‰. d18O is relatively high, averaging about – 2‰ with a maximum of 0‰. There is a fall to negative d13C (– 1 to – 2.5‰) in the upper part of the formation, in samples above and below the Julius River Member. A corresponding fall is seen in the Forest-1 drill hole, in both carbonate and organic d13C. A single analysis of the black dolostone immediately above the Julius River Member is anomalous with respect to this trend (d13Ccarb þ1.6‰). The limestone unit within the Julius River Member in Forest-1 is –4‰ in d13C and 0.7063 –0.7066 in 87Sr/86Sr. Sr content of the limestone is high (900 –1200 ppm) and these values are considered little altered (Calver 1998). No chemostratigraphy has been undertaken on the Croles Hill Diamictite or enclosing Kanunnah Subgroup, which lack primary carbonates. The Smithton Dolomite (c. 1 km stratigraphically higher than the Croles Hill Diamictite) is variable in d13C ( – 3 to þ5‰) and has 87Sr/86Sr, interpreted as little altered, ranging from 0.7079 at the base to 0.7085 in the upper part (Calver 1998).

Palaeolatitude and palaeogeography Although Tasmania’s position relative to the Australian craton in the Cryogenian is uncertain, some affinity, and perhaps proximity

654

C. R. CALVER

to the Adelaide Geosyncline is suggested by the presence in both places of mid-Cryogenian carbonates with Baicalia burra followed by diamictites, then equivalent-aged black shales (see below) (Griffin & Preiss 1976; Calver & Walter 2000; Kendall et al. 2007). No palaeomagnetic data relevant to the Julius River Member are available. The Spinks Creek Volcanics, above the Croles Hill Diamictite, are rift tholeiites, probably the same or similar in age to the volcanics of the Grassy Group on King Island. Direct palaeomagnetic investigation of the Croles Hill Diamictite has so far been unsuccessful. However, the Spinks Creek Volcanics and related intrusives exhibit a primary magnetization that is mostly east-directed (significantly different from the Cottons Breccia: see previous section), and suggestive of a palaeolatitude within 158 of the equator (T. Raub, pers. comm.).

Geochronological constraints That part of the Black River Dolomite below the Julius River Member has d13C values consistent with correlation with the Burra Group of the Adelaide Geosyncline (Hill & Walter 2000). The stromatolite Baicalia cf. burra, found as clasts in the Julius River Member, is closest to a form common in the middle Burra Group (Torrensian) of the Adelaide Geosyncline (Griffin & Preiss 1976). The Burra Group is younger than 777 Ma (Preiss 2000). Vase-shaped microfossils are present in chert in the Black River Dolomite below the Julius River Member (Saito et al. 1988), also consistent with a broadly pre-Sturtian age (Porter & Knoll 2000). 87Sr/86Sr of the limestone within the diamictite in Forest-1 is relatively low, implying a Cryogenian or older age (e.g. Walter et al. 2000). Calver (1998) and Calver & Walter (2000) suggested correlation of the Julius River Member with the Sturtian glacials of the Adelaide Geosyncline on chemostratigraphic and lithostratigraphic grounds. Black shale overlying the Julius River member in Forest-1 yielded a Re – Os date of 640.7 + 4.7 Ma (2s, n ¼ 19, MSWD ¼ 0.91, Model 1) (Kendall et al. 2007), identical (given 2s uncertainties) to the 643 + 2 Ma Re –Os age for the Sturtian post-glacial Tindelpina Shale Member (Kendall et al. 2006). The rhyodacite underlying the Croles Hill Diamictite at Robbins Passage has been dated by U– Pb (SHRIMP on zircon) at 582.1 + 4.1 Ma (n ¼ 46, MSWD ¼ 0.94, precision limits at 95% (2s) confidence level: Calver et al. 2004). The 87Sr/86Sr trend in the Smithton Dolomite suggests a middle to upper Ediacaran age for that unit (Calver 1998). The Kanunnah Subgroup and Smithton Dolomite are unfossiliferous except for rare stromatolites.

Southern Tasmania: Wedge River Beds and Cotcase Creek Fm.

Fig. 64.3. Bedrock geological map of Jubilee region, southern Tasmania (for location see Fig. 64.1a). Possible glaciogenic rocks are found in the Wedge River Beds and the Cotcase Creek Fm. Localities 1 and 2 (see text) are shown as numbered asterisks.

of the Group (the Cotcase Creek Fm.). The unit is poorly exposed and difficult to access because of thick forest cover.

Introduction

Structural framework

Two possibly glaciogenic units of uncertain relative age are found in separate tectonic units in southern Tasmania. The Wedge River Beds (Corbett 1970) is a deformed conglomeratic succession cropping out over c. 2 km2 just north of the Sentinel Range, at the eastern edge of the Tyennan region (Locality 1, Fig. 64.3). Jago (1981) briefly described the Wedge River Beds, noting the presence of possible dropstones. Turner (1989, pp. 171– 172) suggested correlation of the Wedge River Beds with the basal, conglomeratic part of the Weld River Group. Further description was provided by Turner (in Calver et al. 1990). The Weld River Group (Calver 1989) is a Neoproterozoic, dolostone-dominated succession that underlies much of the Weld and Huon River valleys in southern Tasmania, and is part of a tectonic unit known as the Jubilee region (Fig. 64.1a). A number of diamictite units are found in the uppermost formation

The Wedge River Beds lie at the eastern margin of a large area of deformed Proterozoic rocks, known as the Tyennan region (Fig. 64.1a). The Wedge River Beds overlie, with inferred unconformity, a quartzarenite-phyllite succession typical of most of the Tyennan region, but both successions are strongly deformed and the unconformity does not correspond to a significant deformational episode (Turner, in Calver et al. 1990). The deformation, multiple cleavages and low-grade regional metamorphism affecting the Wedge River Beds are due to the Cambrian, Tyennan Orogeny (Turner et al. 1998). The Weld River Group may be broadly correlative with the Togari Group of NW Tasmania (previous section) and records epicratonic, shallow-marine, predominantly carbonate sedimentation. However, no rift volcanics are known in the Weld River Group. By the Cambrian, Tasmania probably comprised

NEOPROTEROZOIC GLACIAL DEPOSITS OF TASMANIA

an attenuated east-facing continental margin that collided with an oceanic fore-arc c. 515–510 Ma (Crawford & Berry 1992; Meffre et al. 2000), producing extensive thrusting, deformation and regional metamorphism in the Tyennan and Jubilee regions, though less intense in the latter (Fig. 64.1a). This event, the Tyennan Orogeny, also involved obduction of forearc lithologies (ultramafics) and melange units (Ragged Basin Complex, Fig. 64.3). In the Weld Valley area, major folds plunge NW and are upright or overturned to the NE. However, ooids and stromatolites in the Weld River Group dolostones appear largely unstrained, though pelites and diamictites are weakly cleaved. Regional metamorphism is slight (sub-greenschist facies) (Calver 1989).

Stratigraphy Mapped thickness of the Wedge River Beds is c. 300 m (Turner et al. 1985), but the contacts are not exposed and the unit is poorly exposed and strongly deformed. To the NW, the Wedge River Beds are overlain, possibly conformably, by a succession of phyllite with minor quartzarenite and silicified dolomite (Turner et al. 1985; Turner in Calver et al. 1990). To the NE, the Wedge River Beds are unconformably overlain by the Middle Cambrian, Island Road Fm. The Weld River Group unconformably overlies the predominantly siliciclastic Clark Group, which probably correlates with the early Neoproterozoic Rocky Cape Group of NW Tasmania. The lower part of the Weld River Group (as exposed in the type section NE of Mt Anne) consists of basal conglomerate and sandstone (Annakananda Fm., 25 m), massive, fine-grained dolostone (Gomorrah Dolomite, 800 m), then oolitic dolograinstone (Devils Eye Dolomite, .1300 m). The Cotcase Creek Fm. occupies the western part of the Weld Valley and part of the Huon Valley, and is faulted against the older parts of the Weld River Group. It consists of alternating dolostone and diamictite with minor sandstone and shale. No stratigraphic sequence within the formation can be worked out because of poor exposure, paucity of evidence for stratigraphic way-up and faulting, although Calver (1989) argued that it may be some 2 km thick. The many mapped diamictite occurrences (Turner et al. 1985; Calver et al. 2007) are probably structural repetitions of a smaller (but unknown) number of stratigraphic units. The top of the Cotcase Creek Fm. is unknown; it is variably faulted against the lower(?) Cambrian Ragged Basin Complex and unconformably overlain by sediments of probable Middle Cambrian age (Calver 1989; Calver et al. 1990, 2007).

Glaciogenic deposits and associated strata: Wedge River Beds The Wedge River Beds consist of pebbly and cobbly metasandstone, with some bouldery (up to 700 mm) intervals, forming unbedded units up to 12 m thick. Clasts are mainly quartzarenite similar to the underlying succession, and are deformed to discoidal shapes due to deformation. Clasts were originally well-rounded to angular. The matrix is a poorly sorted siliceous sandstone or siltstone, displaying metamorphic recrystallization in thin section (Turner, in Calver et al. 1990). The Wedge River Beds also includes graded sandstone beds (,100 mm thick) with rare sole marks, thin-bedded siltstone and sandstone containing possible dropstones (Jago 1981), and minor dark grey phyllite.

655

underlying Proterozoic quartzarenite-dominated succession on the Sentinel Range. North of the Wedge River beds, and possibly conformably overlying them, is a succession of phyllite with minor quartzarenite and silicified dolomite (Turner et al. 1985; Turner in Calver et al. 1990).

Glaciogenic deposits and associated strata: Cotcase Creek Fm. The Cotcase Creek Fm. consists of massive dolostone with several units of diamictite, mudstone and sandstone. The diamictite units range in thickness from a few metres to over 400 m. They are typically massive, with sparse (5 –30%) clasts, predominantly of dolostone, in a matrix of impure, black sandy dolomitic mudstone. Clasts are angular to sub-rounded and usually pebble to cobble size (rarely boulder-size). The dolostone clasts includes dolomicrite, oolitic dolograinstone and coarsely recrystallized dolostone. Dolomitization and diagenesis occurred prior to reworking and incorporation into the diamictite. A small proportion (c. 5%) of the clasts is of siliciclastic composition (quartzarenite, siltstone, mudstone and chert) and there is very rare, altered basalt. Diamictites of the Cotcase Creek Fm. closely resemble the Julius River Member (previous section). In several places at the western edge of the Cotcase Creek Fm. (probably at or near the top of the formation), there are exposures of laminated dolomitic siltstone with sparse (a few %), pebblesized lonestones, including probable dropstones, of dolostone and quartzite. These are usually less than 30 mm in size; rarely up to 150 mm (Calver et al. 2007). Loose aggregates of granules and small pebbles resemble till clasts (Calver 1989). Good exposures of this facies, and of diamictite and dolostone of the Cotcase Creek Fm., are found on the Huon River at locality 2 (Fig. 64.3).

Boundary relations with overlying and underlying non-glacial units: Cotcase Creek Fm. Contacts between diamictite and the enclosing succession are poorly exposed. In at least two places there is a conformable upward succession from mudstone, through diamictite, then dolostone conglomerate, then massive dolostone (Calver et al. 2007). At another locality, diamictite grades up into 1 m of laminated dark grey dolostone, overlain by massive pale dolostone (Calver et al. 1990). The sparsely pebbly laminated siltstone is observed to conformably overlie dolostone south of Locality 2 (Calver et al. 2007), but no exposure of its upper contact is known.

Chemostratigraphy Reconnaissance C-isotope data for the Weld River Group are presented by Calver (1995). The lower part of the Weld River Group (including the Gomorrah and Devils Eye Dolomites) tends to be moderately 13C-depleted (d13C average – 1.2 + 1.5‰, n ¼ 17), whereas the dolostones of the Cotcase Creek Fm. are moderately enriched (2.3 + 2.9‰, n ¼ 14). The data argue against simple equivalence of the Weld River Group with the Black River Dolomite.

Boundary relations with overlying and underlying non-glacial units: Wedge River Beds

Palaeolatitude and palaeogeography

The basal contact is not exposed, but clast lithologies imply an unconformable, or at least erosional, relationship with the

No relevant palaeomagnetic data are available for the southern Tasmanian successions.

656

C. R. CALVER

Geochronological constraints Age constraints are poor. Quartzarenite-rich successions that underlie both the Wedge River Beds and the Weld River Group are younger than 1400 Ma based on detrital zircon dating, and may be c. 1000– 750 Ma if equivalence with the Rocky Cape Group is accepted (Black et al. 2004). The Wedge River Beds are unconformably overlain by fossiliferous Middle Cambrian rocks (Turner, in Calver et al. 1990). The Weld River Group is overlain, with inferred unconformity, by sediments of probable Middle Cambrian age. Calver et al. (2007) suggested a Cryogenian age for the Weld River Group, but only on the basis of the presence of apparently glaciogenic rocks.

Discussion In spite of recent dissenting opinion (Direen & Jago 2008), evidence favouring a glacial origin of the Cottons Breccia is here considered to be strong. Features consistent with a glacial influence include the unsorted, open-framework fabric of the diamictite, its massive to poorly stratified nature, the presence of undeformed and relatively well-sorted sandstone and siltstone lenses, and the presence of dropstones in laminated siltstone beds. Jago (1974, 1981) cites the lack of exotic clast lithologies as evidence against a glacial origin, but it should be noted that a source for the abundant and lithologically diverse carbonate clasts is unknown on King Island. The stratigraphic association with an overlying cap carbonate (Cumberland Creek Dolostone), lithologically and isotopically very similar to the Nuccaleena Fm. of the Adelaide Geosyncline (Williams et al. 2011), is consistent with correlation of the Cottons Breccia with a unit of established glacial origin (Elatina Fm.) on mainland Australia. There is no strong sedimentological evidence for a glacial origin for the Julius River Member. Griffin & Preiss (1976) suggested that the absence of striated or faceted clasts, and the restricted range of clast lithologies, favour an origin as a submarine mass flow. However, three-dimensional clast morphologies and surface features can rarely be examined satisfactorily in the Julius River Member. There is good lithostratigraphic and chemostratigraphic evidence for correlation of the Julius River Member with the Sturtian glacials of the Adelaide Geosyncline (Calver 1998; Calver & Walter 2000; Hill & Walter 2000). This correlation is strongly supported by equivalent Re – Os dates from black shales directly overlying both units (previous section; Kendall et al. 2007), suggesting that the Julius River Member was deposited synchronously with glaciogenic rocks on the mainland. In the case of the Croles Hill Diamictite, there is circumstantial evidence for an origin as subaqueous debris flows, namely the limited, dominantly volcanic clast composition, tectonically unstable environment and other evidence for active faulting. No cap carbonate has been found. On the other hand, a glacial origin is suggested by possible small dropstones in laminated intervals, and by equivalence or near equivalence in age to the 580 Ma Gaskiers Fm. of Newfoundland (Bowring et al. 2003). Calver et al. (2004) correlated the Cottons Breccia with the Croles Hill Diamictite (,582 Ma), which would imply an age of c. 580 Ma for both these units, and also for the Elatina Fm. (and base of the Ediacaran System) since the Elatina-Cottons correlation appears well-founded (Calver & Walter 2000). However, dates on putative ‘Marinoan’ (i.e. Elatina) glacial correlatives on other continents are c. 635 Ma. Evidence for correlation of the Elatina Fm. with the c. 635 Ma Ghaub Fm. of Namibia (Hoffmann et al. 2004), primarily from d13C chemostratigraphy, appears to be strong (e.g. Halverson et al. 2005). Any lithological correlative of the c. 635 Ma glacials is missing in the Togari Group. The correlative (and/or evidence of associated large base-level changes: e.g. Hoffman et al. 2007) should lie between the Julius River Member and the Croles Hill Diamictite. The only obvious

stratigraphic break in this interval (as observed in the fully cored Forest-1 drill hole) is the possible disconformity at the base of the Croles Hill Diamictite. Similarly, if the Elatina Fm. (and Cottons Breccia) are c. 635 Ma, then a condensed section or c. 60 Ma stratigraphic break ought to be present in the Yarra Creek Shale on King Island, since the overlying volcanics are c. 575 Ma (Calver 2008a). More dating is needed to better constrain the ages of the Cottons Breccia and Elatina Fm. Jago (1981) flagged a possible glacial influence in the Wedge River Beds because of ‘possible dropstones’ in thin-bedded siltstone and sandstone. Turner (in Calver et al. 1990) noted an absence of faceted or striated clasts, and suggested that the conglomeratic units were probably gravity flow deposits because of their unbedded nature and the presence of graded sandy interbeds. However, the strong tectonic deformation of the Wedge River Beds renders interpretation difficult. A submarine debris-flow origin for the diamictites of the Cotcase Creek Fm. is favoured by the almost wholly intrabasinal provenance of the clasts. No faceting or striation of clasts has been observed, but would be difficult to identify as most clasts are less resistant to weathering than the matrix. However, the laminated siltstone with dropstones up to 150 mm in size, and possible till clasts, suggest a glaciomarine imprint, and by association some of the diamictites may also be glacial. No firm conclusions can be drawn regarding correlation of the southern Tasmanian diamictites. This paper is published with the permission of the Director, Mineral Resources Tasmania. T. Raub and P. Hoffman are thanked for permission to include unpublished data. This represents a contribution of the IUGS- and UNESCOfunded IGCP (International Geoscience Programme) Project #512.

References Berry, R. F., Steele, D. A. & Meffre, S. 2008. Proterozoic metamorphism in Tasmania: implications for tectonic reconstructions. Precambrian Research, 166, 387–396. Black, L. P., Calver, C. R., Seymour, D. B. & Reed, A. 2004. SHRIMP U– Pb detrital zircon ages from Proterozoic and Early Palaeozoic sandstones and their bearing on the early geological evolution of Tasmania. Australian Journal of Earth Sciences, 51, 885– 900. Bowring, S., Myrow, P., Landing, E., Ramezani, J. & Grotzinger, J. 2003. Geochronological constraints on terminal Neoproterozoic events and the rise of metazoans. Geophysical Research Abstracts, 5, 13219. Brown, A. V. 1989. Geological Atlas 1:50,000 Series, Sheet 21 (7916S). Smithton. Geological Survey Explanatory Report, Department of Mines, Tasmania. Burrett, C. & Berry, R. 2000. Proterozoic Australia –Western United States (AUSWUS) fit between Laurentia and Australia. Geology, 28, 103– 106. Calver, C. R. 1989. The Weld River Group: a major upper Precambrian dolomite sequence in southern Tasmania. Papers and Proceedings of the Royal Society of Tasmania, 123, 43 –53. Calver, C. R. 1995. Ediacarian isotope stratigraphy of Australia. PhD thesis, Macquarie University (unpublished). Calver, C. R. 1996. Reconnaissance isotope chemostratigraphy of Neoproterozoic carbonates in western Tasmania. Tasmanian Geological Survey Record, 1996/10. Calver, C. R. 1998. Isotope stratigraphy of the Neoproterozoic Togari Group, Tasmania. Australian Journal of Earth Sciences, 45, 865– 874. Calver, C. R. 2000. Isotope stratigraphy of the Ediacarian (Neoproterozoic III) of the Adelaide Rift Complex, Australia, and the overprint of water column stratification. Precambrian Research, 100, 121– 150. Calver, C. R. 2008a. Tasmanian Neoproterozoic glacial deposits and their age constraints. In: Gallagher, S. J. & Wallace, M. W. (eds) Neoproterozoic Extreme Climates and the Origin of Early Metazoan Life. Selwyn Symposium of the GSA Victoria Division,

NEOPROTEROZOIC GLACIAL DEPOSITS OF TASMANIA

September 2008, Geological Society of Australia Extended Abstracts, 91, 45 – 48. Calver, C. R. 2008b. Digital Geological Atlas 1:25000 series sheet 2456. Grassy. Mineral Resources, Tasmania. Calver, C. R., Turner, N. J., McClenaghan, J. & Brown, A. V. 1990. Tasmanian Department of Mines, Geological Atlas 1:50,000 series. Explanatory Report, Sheet 8112S, Pedder, Tasmania. Calver, C. R. & Walter, M. R. 2000. The late Neoproterozoic Grassy Group of King Island, Tasmania: correlation and palaeogeographic significance. Precambrian Research, 100, 299– 312. Calver, C. R., Black, L. P., Everard, J. L. & Seymour, D. B. 2004. U –Pb zircon age constraints on late Neoproterozoic glaciation in Tasmania. Geology, 32, 893– 896. Calver, C. R., Forsyth, S. M. & Everard, J. L. 2007. Geology of the Maydena, Skeleton, Nevada, Weld and Picton 1:25 000 scale map sheets. Tasmanian Geological Survey Record, 2006/04. Carey, S. W. 1947. Occurrence of Tillite on King Island. Report of the Australian Association for the Advancement of Science, 52, 349. Corbett, K. D. 1970. Sedimentology of an upper Cambrian flyschparalic sequence (Denison Group) on the Denison Range, Southwest Tasmania. PhD thesis, University of Tasmania (unpublished). Crawford, A. J. & Berry, R. F. 1992. Tectonic implications of Late Proterozoic – Early Palaeozoic igneous rock associations in western Tasmania. Tectonophysics, 214, 37 – 56. Danielson, M. L. 1975. King Island scheelite deposits. In: Knight, C. L. (ed.) Economic Geology of Australia and Papua-New Guinea. Australian Institute of Mining and Metallurgy, Monograph, 5, 592– 599. Direen, N. G. & Crawford, A. J. 2003. The Tasman Line: where is it, what is it, and is it Australia’s Rodinian breakup boundary? Australian Journal of Earth Sciences, 50, 491– 502. Direen, N. G. & Jago, J. B. 2008. The Cottons Breccia (Ediacaran) and its tectonostratigraphic context within the Grassy Group, King Island, Australia: a rift-related gravity slump deposit. Precambrian Research, 165, 1– 14. Everard, J. L., Seymour, D. B., Brown, A. V. & Calver, C. R. 1996. Geological atlas 1:50,000 series, sheet 7915N. Trowutta. Mineral Resources Tasmania, Tasmania Development and Resources. Everard, J. L., Seymour, D. B., Reed, A. R., McClenaghan, M. P., Green, D. C. & Calver, C. R. 2007. Regional geology of the Southern Smithton Synclinorium. Explanatory Report for the Roger, Sumac and Dempster 1:25 000 scale geological map sheets, far northwestern Tasmania. 1:25 000 Scale Digital Geological Map Series Report 2. Griffin, B. J. & Preiss, W. V. 1976. The significance and provenance of stromatolitic clasts in a probable late Precambrian diamictite in northwestern Tasmania. Papers and Proceedings of the Royal Society of Tasmania, 110, 111– 127. Halverson, G. P., Hoffman, P., Schrag, D. P., Maloof, A. C. & Rice, A. 2005. Towards a Neoproterozoic composite carbon isotope record. Geological Society of America Bulletin, 117, 1181– 1207. Hill, A. C. & Walter, M. R. 2000. Mid-Neoproterozoic (830–750 Ma) isotope stratigraphy of Australia and global correlation. Precambrian Research, 100, 181– 211. Hoffman, P. F., Halverson, G. P, Domack, E. W., Husson, J. M., Higgins, J. A. & Schrag, D. P. 2007. Are basal Ediacaran (635 Ma) post-glacial ‘cap dolostones’ diachronous? Earth and Planetary Science Letters, 258, 114– 131. Hoffmann, K.-H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. U– Pb zircon date from the Neoproterozoic Ghaub Fm., Namibia: constraints on Marinoan glaciation. Geology, 32, 817– 820. Jago, J. B. 1974. The Origin of the Cottons Breccia, King Island, Tasmania. Transactions of the Royal Society of South Australia, 98, 13 – 28.

657

Jago, J. B. 1981. Possible Late Precambrian (Adelaidean) tillites of Tasmania. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, 549– 554. Kendall, B., Creaser, R. A. & Selby, D. 2006. Re– Os geochronology of postglacial black shales in Australia: constraints on the timing of ‘Sturtian’ glaciation. Geology, 34, 729– 732. Kendall, B., Creaser, R. A, Calver, C. R. & Evans, D. A. D. 2007. Neoproterozoic paleogeography, Rodinia breakup, and Sturtian glaciation: constraints from Re– Os black shale ages from southern Australia and northwestern Tasmania. Geological Society of America Abstracts with Programs, 39, 335. Li, Z. X., Bogdanova, S. V. et al. 2008. Assembly, configuration, and break-up history of Rodinia: a synthesis. Precambrian Research, 160, 179–210. Logan, G. A., Calver, C. R., Gorjan, P., Summons, R. E., Hayes, J. M. & Walter, M. R. 1999. Terminal Proterozoic mid-shelf benthic microbial mats in the Centralian Superbasin and their environmental significance. Geochimica et Cosmochimica Acta, 63, 1345– 1358. Longman, M. J. & Matthews, W. L. 1962. The Geology of the Bluff Point and Trowutta quadrangles. Tasmania Department of Mines Technical Reports, 6, 48 – 54. Meffre, S., Berry, R. F. & Hall, M. 2000. Cambrian metamorphic complexes in Tasmania: tectonic implications. Australian Journal of Earth Sciences, 47, 971– 985. Meffre, S., Direen, N. G., Crawford, A. J. & Kamenetsky, V. 2004. Mafic volcanic rocks on King Island, Tasmania: evidence for 579 Ma break-up in east Gondwana. Precambrian Research, 135, 177–191. Porter, S. M. & Knoll, A. H. 2000. Testate amoebae in the Neoproterozoic era: evidence from vase-shaped microfossils in the Chuar Group, Grand Canyon. Paleobiology, 26, 360–385. Preiss, W. V. 2000. The Adelaide Geosyncline of South Australia and its significance in Neoproterozoic continental reconstruction. Precambrian Research, 100, 21 –63. Saito, Y., Tiba, T. & Matsubara, S. 1988. Precambrian and Cambrian cherts in northwestern Tasmania. Bulletin of the National Science Museum, Tokyo, Series C, 14, 59 – 70. Solomon, M. 1969. The nature and possible origin of the pillow lavas and hyaloclastite breccias of King Island, Australia. Quarterly Journal Geological Society, London, 124, 153– 169. Turner, N. J. 1989. The Adamsfield District. In: Burrett, C. F. & Martin, E. L. (eds) Geology and Mineral Resources of Tasmania. Geological Society of Australia Special Publication, 15, 168–174. Turner, N. J., Calver, C. R., McClenaghan, M. P., McClenaghan, J., Brown, A. V. & Lennox, P. G. 1985. Pedder. Geological Atlas 1:50,000 series, sheet 81125S. Turner, N. J., Black, L. P. & Kamperman, M. 1998. Dating of Neoproterozoic and Cambrian orogenies in Tasmania. Australian Journal of Earth Sciences, 45, 789–806. Waldron, H. M. & Brown, A. V. 1993. Geological setting and petrochemistry of Eocambrian–Cambrian volcano-sedimentary rock sequences from southeast King Island, Tasmania. Mineral Resources Tasmania Report 1993/28, 28. Walter, M. R., Veevers, J. J., Calver, C. R., Gorjan, P. & Hill, A. C. 2000. Dating the 840– 544 Ma Neoproterozoic interval by isotopes of strontium, carbon and sulfur in seawater, and some interpretative models. Precambrian Research, 100, 371–433. Waterhouse, L. L. 1916. Notes on the Geology of King Island. Annual Report to the Secretary for Mines, Tasmania, for 1915, 88– 93. Williams, G., Gostin, V. A., McKirdy, D. M., Preiss, W. V. & Schmidt, P. W. 2011. The Elatina glaciation (late Cryogenian), South Australia. In: Arnaud, E., Halverson, G. P. & ShieldsZhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 713– 721.

Chapter 65 Neoproterozoic glacial deposits of the Kimberly Region and northwestern Northern Territory, Australia MAREE CORKERON Biogeoscience, Queensland University of Technology, Gardens Point Campus, GPO Box 2434, Brisbane, Qld, 4001, Australia (e-mail: [email protected]) Abstract: Neoproterozoic glaciogenic formations are preserved in the Kimberley region and northwestern Northern Territory of northern Australia. They are distributed in the west Kimberley adjacent to the northern margins of the King Leopold Orogen, the Mt Ramsay area at the junction of the King Leopold and Halls Creek orogens, and east Kimberley, adjacent to the eastern margin of the Halls Creek Orogen. Small outlier glaciogenic deposits are preserved in the Litchfield Province, Northern Territory (Uniya Fm.) and Georgina Basin, western Queensland (Little Burke Fm.). Glaciogenic strata comprise diamictite, conglomerate, sandstone and pebbly mudstone and characterize the Walsh, Landrigan and Fargoo/Moonlight Valley formations. Thin units of laminated dolomite sit conformably at the top of the Walsh, Landrigan and Moonlight Valley formations. Glaciogenic units are also interbedded with the carbonate platform deposits of the Egan Fm. and Boonall Dolomite. d13C data are available for all carbonate units. There is no direct chronological constraint on these successions, and dispute over the regional correlation of the Neoproterozoic succession has been largely resolved through biostratigraphic, chemostratigraphic and lithostratigraphic analysis. However, palaeomagnetic results from the Walsh Fm. are inconsistent with sedimentologically based correlations. Two stratigraphically defined glaciations are preserved in northwestern Australia: the ‘Landrigan Glaciation’, characterized by SW-directed continental ice-sheet movement and correlated with late Cryogenian glaciation elsewhere in Australia and the world; and, the ‘Egan Glaciation’, a more localized glaciation of the Ediacaran Period. Future research focus should include chronology, palaeomagnetic constraint and tectonostratigraphic controls on deposition.

Neoproterozoic glaciogenic rocks are well preserved in the Kimberley region of northwestern Australia and westernmost Northern Territory (NT) (Fig. 65.1). Five Kimberley deposits are identified from three areas: the Walsh Fm. from the Mount House Group (west Kimberley), the Landrigan Fm. from the Kuniandi Group and the Egan Fm. from the Louisa Downs Group (Mt Ramsay area), and the Fargoo and Moonlight Valley formations from the Duerdin Group (east Kimberley and the western NT). A further postulated glaciogenic deposit is preserved in the Boonall Dolomite of the Albert Edward Group. Distribution of the Neoproterozoic successions is confined to erosive remnants on the upturned margins of the Kimberley Basin adjacent to the bounding orogens. Grey & Blake (1999) postulated that the Kimberley successions may have once been contiguous with the Centralian Superbasin of central and southern Australia. Glaciogenic rocks were first identified in the Kimberley region by Guppy et al. (1958) and mapped in the western (Mount House area), central (Mount Ramsay area) and east Kimberley by Harms (1959) (Fig. 65.1). The Kimberley and western NT (Keep River and Skinner Point deposits) glaciogenic successions were jointly mapped in the 1960s by the Bureau of Mineral Resources (BMR, now Geoscience Australia, GA) and the Geological Survey of Western Australia (GSWA) (see reports by Dow & Gemuts 1969; Roberts et al. 1972; Derrick & Playford 1973; Gellatly et al. 1975). Dow (1965) first assessed the veracity of glacial deposits from the east Kimberley and concluded that ‘their glacial origin is confirmed by the presence of very large polished and striated erratics, the great diversity of megaclasts, their regional extent, and the presence of striated, grooved, and polished quartzite bedrock directly beneath the tillite’. Despite differing stratigraphic nomenclatures used for the three Kimberley areas, there was consensus among the mapping geologists that the glacial successions were probably correlatives (Dow & Gemuts 1969; Plumb 1981). Dow & Gemuts (1969) made lithological and stratigraphic descriptions of the east Kimberley succession, and correlated the succession with the Mount Ramsay and Mount House successions (Fig. 65.1). Roberts et al. (1972) presented a lithological description of the Mount Ramsay succession, measuring stratigraphic

sections through the Kuniandi Group, and the Egan and Yurabi formations in the Louisa and O’Donnell Synclines. The Mount House succession is described in explanatory notes accompanying the Lennard River and Lansdowne map sheets (Derrick & Playford 1973; Gellatly et al. 1975). Sweet (1977) documented comparisons with Neoproterozoic successions in the Victoria River region (northwestern Northern Territory) and Edgegoose et al. (1989) reported Proterozoic glaciogenic sediments from the Uniya Fm., Litchfield Province in the Northern Territory (Fig. 65.1). Corkeron (2002) reviewed the sedimentology of the glaciogenic units from the Kimberley and Victoria River areas. Edgegoose et al. (1989) first reported glaciogenic strata and pavements in the Uniya Fm. of the Litchfield Province and correlated them with the Kimberley successions. The Little Burke Tillite and lowermost dolomite of the Mount Burnie Beds in northwestern Queensland was described by de Keyser (1972).

The Walsh Fm., west Kimberley The Walsh Fm. is discontinuously distributed throughout the Mount House and Traine River areas in west Kimberley (Fig. 65.2). These rocks belong to the Mount House Group, which overlies the late Palaeoproterozoic Kimberley Group with slight angular unconformity (Gellatly et al. 1975) and is overlain by Cenozoic soils. The Walsh Tillite type section is located east of Mt Clifton on the exposed banks of the Throssell River (Fig. 65.2). The 1:250 000 geological map sheet descriptions (Derrick & Playford 1973; Gellatly et al. 1975; Griffin et al. 1993) and limited references by Plumb (1981) and Brookfield (1994) represent almost all that has been published on the Mount House Group. Unpublished descriptions of the Walsh Fm. are available in Corkeron (2002, 2007b). Additionally, Perry & Roberts (1968) described the striated glacial pavements associated with the Walsh Fm. and Williams (1979) and Corkeron (2007a) detail the dolomicrite unit at the top of the Walsh Fm., including its geochemistry. Correlation of the group is discussed in Dow & Gemuts (1969), Coats & Preiss (1980), Grey & Corkeron (1998), Li (2000) and Corkeron (2007a).

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 659– 672. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.65

660

M. CORKERON

Fig. 65.1. Main tectonic units of northwestern Australia. (a) Distribution of glaciogenic strata in Neoproterozoic sedimentary units are shown by white diagonal fill. Boxed areas refer to Figures 65.2– 65.5. Box 1, Mount House area (Fig. 65.2); Boxes 2i and 2ii, Mt Ramsay area (Figs 65.3 & 65.4, respectively); Box 3, east Kimberley area (Fig. 65.5). (b) Stratigraphy and preferred correlation of Kimberley Neoproterozoic successions. Adapted from Corkeron (2007a), Tyler & Hocking (2002), NT Geological Suvey (2006) and Plumb (1981).

KIMBERLEY NEOPROTEROZOIC GLACIAL DEPOSITS

661

Fig. 65.2. (a) Distribution of the Mount House Group and Walsh Fm. in the Mount House area. Site 1, Walsh Fm. type section; site 2, ‘anticline’ study site; sites 3 and 4, striated pavement. (b) Stratigraphic log of the Walsh Fm. from the type section. C-isotope stratigraphy for the uppermost dolomicrite from type section and ‘anticline’ section. Dolomicrite unit thickness normalized against type section. Simplified from Corkeron (2007a).

Structural framework The Mount House Group is generally flat lying and unconformable with the underlying gently warped Kimberley Basin succession (Gellatly et al. 1975). The southern and southwestern extent of the group was affected by SW-directed thrusting and large-scale folding associated with the youngest deformation period affecting the King Leopold Orogen (Tyler & Griffin 1990). To the SW of Mount House Station, the glaciogenic strata are moderately to steeply folded with SE-plunging fold hinges. There is no available literature on basin evolution and history of the Mount House Group.

Stratigraphy The Mount House Group is distributed across the Mount House region (Lansdowne and Lennard River 1:250 000 map sheets; Fig. 65.2). At the type section the Walsh Fm. is c. 10 m thick and consists of a basal diamictite unit overlain by thin units of fissile pebbly mudstone and laminated dolomicrite (Fig. 65.2). It is conformably overlain by massive, medium- to coarse-grained

lithic sandstone of the Traine Fm. Fissile siltstone of the Throssell Shale overlies the Traine Fm., and is in turn gradationally overlain by the sandstones and siltstones of the Estaughs Fm. The muddy sandstone of the Estaughs Fm. at the top of the succession is relatively resistant to weathering and forms a prominent mesa (Mt Clifton) c. 250 m high. The entire Mount House Group is up to c. 540 m thick (Derrick & Playford 1973).

Glaciogenic deposits and associated strata The Walsh Fm. is the basal unit of the Mount House Group. Polished and striated pavements underlie several Walsh Fm. outcrops. Pavements are typically on quartzite basement, and beautifully preserved scour and chatter marks suggest palaeo-ice-flow to the south and SW (Perry & Roberts 1968; Corkeron 2008). The Walsh Fm. varies in thickness from 5 to 60 m (Gellatly et al. 1975). Exposure of the Walsh Fm. in the Mount House region is discontinuous, preserved in scattered outcrops near the confluence of the Traine and Hann Rivers beneath resistant scarps of overlying Traine Fm. (Gellatly et al. 1975). The type section west of Glenroy

662

M. CORKERON

Station (Fig. 65.2) is the best preserved section and displays the greatest facies variation. In the Adcock River valley, the Walsh Fm. is recessive in the flood-plain areas but preserved in the hinge of large-scale anticlines on the western side of the valley (Fig. 65.2). Elsewhere in the Lennard River and Lansdowne map areas it is mapped as pebble and boulder scree within soil cover (Gellatly et al. 1975). Two broad lithological units, the lower diamictite unit and the upper dolomite unit characterize the formation (Corkeron 2007a) (Fig. 65.2). At the type section the diamictite is c. 10 m thick and forms 3–4-m-thick discrete lobes. Diamictite is characterized by clusters of sand- to boulder-sized clasts of quartzite, dolomite, mudstone, jasper and various igneous and metamorphic rock types supported by a fissile (and locally indurated) siltstone-dolomicrite matrix. Quartzite boulders are sub-angular to sub-rounded and some are polished and striated. Red fine-grained sandstone rafts up to 3 m wide and 0.5 m thick are scattered throughout the matrix in the lower beds. Pebble-sized quartzite clasts are preserved within the sandstone rafts and occasionally sit at the sandstone/matrix contact (Gellatly et al. 1975). The quartz content in the matrix is generally high at the base, but the matrix becomes more clay-rich toward the top. Up-section, the massive diamictite grades to a stratified diamictite that shows a reduction in both matrix grain size and framework clast size. Clast content decreases further up-section as diamictite grades into a fissile laminated grey and purple siltstone facies that is recessive but laterally continuous. Soft-sediment deformation is indicated by slump folds in diamictite, roll-over of sandstone lenses within the diamictite, ball and pillow structures and injection of diamictite matrix into sandstone lenses. A discrete horizon of imbricated sandstone rafts is preserved toward the base of the diamictite. In the Adcock River area, diamictite is distributed over tens of square kilometres and is at least tens of metres thick, although accurate estimation of thickness is precluded by folding. Here massive diamictite is overlain by a massive pebbly sandstone and conglomerate facies. The conglomerate is clast-supported, comprising granules and pebbles in a sparse sandy matrix. Lenses of a stratified conglomerate facies, tens of metres long and up to 10 m thick, are interbedded with the massive conglomerate facies. Typically, pebble lags line the base of the lenses and grade upward to pebble-granule conglomerate and coarse sandstone. Large metre-scale boulders with polished and striated surfaces rest on top of the conglomerate beds. A cream dolomite unit lies at the top of the Walsh Fm. and is described in Gellatly et al. (1975) and Plumb (1981), with facies and geochemical analyses outlined in Corkeron (2007a). At the base of the dolomicrite unit laminae are deformed into intact open synforms and brecciated antiforms. There is a gradational change over several centimetres to undeformed planar dolomicrite laminae (Corkeron 2007a). Depositional interpretation. The lowermost diamictite of the Walsh

Fm. records deposition of glacially scoured debris either as grounding line deposits or as subglacial outwash deposits in a glaciomarine, grounding-line fan complex (Corkeron 2002). Debris flows may have redeposited the initial sediment mass, causing slump folding and injection of semi-liquefied diamictite by rapid loading of underlying semiconsolidated sediment. Meltwater or thermohaline currents are indicated by stratified sandy interbeds. The fining-up trend suggests ongoing reworking of sediments by ocean currents, associated with gradual relative sea-level rise from glacial melting. The widespread distribution of mudstone towards the top of the formation is more consistent with a marine setting below storm wave base than localized lagoon or lacustrine settings. An interpretation of low-energy deep-water deposition, related to suspension settling + weak turbidity current activity from the water column for the laminated dolomicrite at the top of the formation, is consistent with marine transgression associated with deglaciation.

Boundary relations with underlying and overlying non-glaciogenic units The Walsh Fm. overlies the late Palaeoproterozoic Carson Volcanics, Warton Sandstone, King Leopold Sandstone and Hart Dolerite with slight angular unconformity (Gellatly et al. 1975). Locally, polished and striated pavements underlie several Walsh Fm. outcrops. The dolomite unit at the top of the formation is conformably overlain by dolomitic sandstone of the Traine Fm. The lowermost several metres of the Traine Fm. contain numerous cross-bed parallel tabular cavities after dolomite rip-up clasts. There is no evidence for a glaciogenic influence in the Traine Fm. or any of the overlying formations.

Chemostratigraphy Corkeron (2007a) undertook systematic stable isotope analysis of the dolomite strata from two sections. At the type section, d13CPDB shows an up-section trend from – 2.9‰ at the base to –4.7‰ near the top of the formation (Fig. 65.2b). At a second location (Anticline locality; Corkeron 2007a) d13C ranged stratigraphically from –4 to –4.9‰. Sr-isotope analysis was trialled on one sample from dolomite from the Walsh Fm. but the high 87Sr/86Sr value measured (0.719015) is regarded as secondary.

Palaeolatitude and palaeogeography Li (2000) identified a high-temperature magnetic component carried entirely by hematite for the Walsh cap dolomite that gave a bedding-corrected mean direction for five sites of declination D ¼ 148.68 and inclination I ¼ – 63.38 (a95 ¼ 9.88), indicating a palaeopole position of 21.58S, 282.48E (dp ¼ 12.28, dm ¼ 15.48) and a palaeolatitude of 45 + 128. A complication in interpreting this result is that the measured palaeolatitude for the Kimberley region is c. 308 higher than that predicted from the end Cryogenian Elatina Fm. palaeopole. By comparison with the known apparent polar wander path (APWP) for East Gondwana, Li (2000) concluded that the Walsh Fm. palaeopole records the position of northern Australia between 770–750 Ma and that the Walsh Fm. correlated with mid-Cryogenian glaciation (Sturtian) of South Australia. (See ‘Glacigenic correlation’ below for further discussion).

Geochronological constraints There is no radiometric control on the Walsh Fm. and no reliable minimum age for the Mt House Group. Plumb (1981) gave a Rb –Sr whole-rock age of 671 + 72 Ma from the Throssell Shale, recalculated from data obtained by Bofinger (1967).

The Landrigan and Egan formations, Mount Ramsay area The Mount Ramsay succession comprises the mostly siliciclastic Kuniandi and Louisa Downs groups. The Kuniandi Group is preserved in the McKinnon Syncline to the north of the Mt Ramsay area and on the western limb of the Louisa Syncline (Fig. 65.3). The Louisa Downs Group disconformably overlies the Kuniandi Group and crops out extensively on the margins of the O’Donnell and Louisa Synclines (Fig. 65.4). The Landrigan Fm. is the basal unit of the Kuniandi Group (maximum thickness c. 350 m at the type section; Roberts et al.

KIMBERLEY NEOPROTEROZOIC GLACIAL DEPOSITS

663

Fig. 65.3. (a) Distribution of the Kuniandi Group and Landrigan Fm. in the Mt. Ramsay area. Type section at site 2. APV, Antrim Plateau Volcanics. (b) Stratigraphic logs of the Landrigan Fm. from sites 3 (log i) and 1 (log ii), including C-isotope trends from cap dolomicrite unit (from Corkeron 2007a).

1972) and comprises intercalated sandstone, siltstone, diamictite and conglomerate and is overlain by a thin unit (c. 9 m thick) of thinly bedded dolomite. Laterally, facies in the dolomite vary from sandy dolomite to thinly laminated dolomite (Roberts et al. 1972; Plumb 1981; Corkeron 2007). In the McKinnon Syncline the dolomite overlies striated pavements on basement (Roberts et al. 1972). The Egan Fm. is the lowermost unit of the Louisa Downs Group, comprising dolomite and limestone with interbedded siliciclastic diamictite, conglomerate and sandstone. Roberts et al. (1972) provided a stratigraphic outline and lithological descriptions of the Louisa Downs Group based on measured sections throughout the O’Donnell and Louisa Synclines. They interpreted a depositional history encompassing a glacial epoch (Egan Glaciation) followed by marine sedimentation. Brief lithological descriptions are given in the context of both interregional and continent-wide stratigraphic

correlation by Plumb & Gemuts (1976), Coats & Preiss (1980) and Plumb (1981). Corkeron & George (2001) described the facies relationships between calcium-carbonate shelf deposits and glaciogenic deposits represented by diamictite, conglomerate and sandstone facies within the Egan Fm. Correlation of the Egan Fm. with the Centralian Superbasin using stromatolite biostratigraphy (Grey & Corkeron 1998), litho- and chemo-stratigraphy (Corkeron & George 2001; Corkeron 2007a) suggests that it records the youngest Proterozoic glaciation in Australia.

Structural framework A structural framework for the Louisa Basin, comprising the Kuniandi and Louisa Downs Groups located at the junction of the King Leopold and Halls Creek Orogens, is not well

664

M. CORKERON

Fig. 65.4. (a) Distribution of the Louisa Downs Group and Egan Fm. in the Mt. Ramsay area. Type section at site 1. Simplified from Corkeron & George (2001). APV, Antrim Plateau Volcanics. (b) Stratigraphic logs of the Egan Fm. from type section site 1 (log i), and O’Donnell South, site 2 (log ii), including C-isotope trends (from Corkeron 2007a).

documented. The Louisa Basin is superimposed on the broad cratonic Kimberley Basin (Plumb & Gemuts 1976; Plumb 1981). Compressional tectonism associated with the King Leopold Orogeny post-dates Neoproterozoic sedimentation and pre-dates Antrim Plateau Volcanics (early-mid Cambrian) (Thorne & Tyler 1996). The western limits of both the O’Donnell and Louisa Downs synclines are affected by tight folding with steeply inclined bedding related to a later more geographically restricted deformation event (D4 of the King Leopold Orogen; Tyler & Griffin 1990). In the Kuniandi Range the Kuniandi Group is extensively folded. In particular, the Landrigan and Stein formations are deformed by isoclinal folding with NW-plunging fold axes. The Kuniandi Range forms the western limb of the Louisa Syncline, with NW-trending fold axis. The eastern limb of the Louisa

Syncline preserves only outcrops of the overlying Louisa Downs Group. Outcrops of the Kuniandi Group are also preserved as upturned limbs of the west– east-oriented McKinnon Syncline to the north of Kuniandi Range (Roberts et al. 1972). To the east of the Mount Ramsay area, residual Antrim Plateau Volcanics form a resistant cap on the underlying units of the upper Louisa Downs Group.

Stratigraphy The c. 1000-m-thick Kuniandi Group unconformably overlies rocks of the Kimberley Group, the Halls Creek Group and the Lamboo Complex (Roberts et al. 1972). The Landrigan Fm., the

KIMBERLEY NEOPROTEROZOIC GLACIAL DEPOSITS

diamictite and conglomeratic basal unit of the Kuniandi Group, is conformably overlain by the Stein Fm., a dark purple unit of interbedded sandstone and siltstone, followed by the Wirara Fm., a green-purple laminated siltstone and the Mount Bertram Sandstone, a mauve quartzarenite (Fig. 65.1b). There is significant lateral variability in the apparent thickness of the Landrigan Fm., although Plumb (1981) interpreted the much greater thickness of diamictite in the northern Kuniandi Range to be the result of folding. The Louisa Downs Group comprises dolomite, diamictite, conglomerate and sandstone at the base, overlain by a thick package of rippled sandstone and black, green and purple mudstone and siltstone with interbeds of conglomerate and greywacke (Fig. 65.1b). The Egan Fm. is the basal glaciogenic unit, conformably overlain by sandstones of the Yurabi Fm. The thick monotonous shales of the McAlly Shale are conformable with the underlying Yurabi Fm. and overlying Tean Fm. (conglomerate and sandstone). The Lubbock Fm. is the uppermost conformable formation comprising almost 2000 m of medium-grained quartz wackestone and siltstone (Tyler et al. 1998).

Glaciogenic deposits Landrigan Fm. The diamictite unit at the base of the Landrigan Fm. ranges in thickness from 350 m in the north of the Kuniandi Range (Roberts et al. 1972) to ,10 m in the south. Locally, indurated and resistant metre-thick sandstone interbedded with pebbly orthoconglomerate form the base of the formation. Massive diamictite gradationally overlies the sandstone/conglomerate facies and is interbedded with massive and stratified pebble conglomerate. Ferruginous siltstone, sandstone and conglomerate form scattered lenses throughout the diamictite. In places, large-scale trough cross-bedding in conglomerate preserves channel fill up to 0.5 m thick. Mudstone, massive and stratified diamictite, and minor conglomerate and sandstone facies form a lithofacies association at the top of the formation. Where preserved, the laminated dolomite lies conformably upon massive diamictite or mudstone. A purple muddy to sandy matrix comprises c. 50% of the diamictite with the framework clasts ranging from granules (c. 30%), to pebbles/cobbles (c. 15%) and boulders (c. 5%) (maximum boulder size c. 5 m in diameter; Corkeron 2002). Clasts are dominantly white, grey or pink quartzite, with rare indurated mudstone, dolomite and very rare chloritic volcanic rocks. Clasts are rounded to sub-angular with polished, striated or faceted surfaces most commonly seen only on the megaclasts (Roberts et al. 1972). Clast distribution is typically heterogeneous with cobble- and boulder-rich zones, v. zones dominated by granules and small pebbles. Tight east – west folding imparts a pervasive slaty cleavage to the diamictite matrix. The uppermost unit of the Landrigan Fm. is laterally variable, ranging from thinly laminated cream dolomicrite to interlaminated pink mudstone and dolomicrite, and sandy dolomite. Thickness ranges from 2 to 9 m. In the northern Kuniandi Range, thinly laminated, recessive and nodular beds up to 5 cm thick are interbedded with very thin beds of maroon/pink dolomitic siltstone that grade up from clast-poor diamictite. In the central Kuniandi Range, massive graded beds of sandy dolomite fine upward to thinly laminated carbonate mudstone. These sand-carbonate mud couplets form stacked beds with scoured bases of decimetre-scale thickness and are interpreted as turbidites (Corkeron 2002). The sandy dolomite unit lies with sharp but conformable contact with underlying diamictite. In the southern Kuniandi Range the uppermost unit comprises c. 9 m of laminated dolomicrite lying with sharp but apparent conformity with the underlying diamicite (Fig. 65.3b). A 0.5-m-thick unit of gently folded dolomite with brecciated fold hinges lies 2 m above the basal contact. There is an upward transition from folded laminae to gently warped then planar laminations toward the top of the unit.

665

The Egan Fm. The sedimentary description and interpretations of the Egan Fm. were adapted from Corkeron & George (2001) and Corkeron (2007). The Egan Fm. is informally divided into three apparently conformable units: a lower carbonate unit (LCU), a siliciclastic diamictite-conglomerate unit (DGU), and an upper carbonate unit (UCU). The LCU is up to 10 m thick, comprising sandy dolomite, lime mud and cream oolitic and stromatolitic dolomite. Rare mud cracks and calcite pseudomorphs of selenitic gypsum are also preserved. The diamictite/conglomerate unit overlies the LCU, although the contact is poorly exposed within a recessive interval. In places the DGU directly overlies Palaeoproterozoic basement. Here, polished and striated pavements, with groove and chatter marks, are preserved. The unit is laterally extensive with lateral facies variability. Massive diamictite, massive and bedded conglomerate and sandstone facies characterize the unit. Diamictite varies from ,1 m to tens of metres in thickness and typically comprises a range of intra- and extrabasinal sub-rounded to subangular clasts of variable size (gravels to boulders) within a red dolomitic matrix. Some quartzite clasts are faceted and striated. In the northern O’Donnell Syncline, diamictite tends to grade upward to bedded conglomerate and sandstone. In the southern Louisa Syncline, thick beds of massive diamictite grade up to pink dolomicrite containing rare clasts. A cobble lag overgrown by a microbial bed (c. 20 cm thick) marks the top of the DGU. The upper carbonate unit conformably overlies the microbial bed and is laterally continuous across the Mt Ramsay area. Laminated and thinly bedded dolomitic sandstone grades up-section to sandy dolomite. Herringbone cross-bedding is preserved in the lower strata. The facies transition also records a compositional transition from siliciclastic dominated sedimentation to carbonate sedimentation. The top c. 30 m of the Egan Fm. is characterized by a return to carbonate-dominated deposits. Facies variability in the UCU is noted on the eastern limb of the Louisa Syncline, where thinly bedded sandy dolomite is absent. Here the UCU is only c. 4 m thick, comprising thinly laminated to massive peloidal dolomite.

Depositional interpretation Striated basement beneath diamictite of the Landrigan Fm. in the McKinnon Syncline indicates ice grounding and lodgment till deposition with westward directed palaeo-ice flow (Roberts et al. 1972). Ubiquitous striated and outsized clasts (up to c. 5 m in diameter) and associated diamictite and conglomerate facies support an interpretation of glaciogenic deposition. Interbedded conglomerate and sandstone facies are consistent with either fluvioglacial outwash or meltwater outwash into a glaciomarine setting. A rapid fining-up transition to siltstone and dolomicrite indicates marine transgression. Fine-grained strata record hemipelagic and turbidity current deposition (Corkeron 2002). These features are consistent with either a deep marine or lacustrine setting. The remainder of the Kuniandi Group records sustained marine deposition (Stein and Wirara formations; Mt Bertram Sandstone; Roberts et al. 1972), supporting a marine setting for the hemipelagic dolomicrite and turbidites at the top of the Landrigan Fm. Facies relations and sedimentary features in the LCU of the Egan Fm. suggest a carbonate platform depositional environment, intermittently exposed with evaporative conditions. Similar facies are observed in the UCU and likewise a carbonate shelfal environment is interpreted. Additionally, localized thick bioherms of stromatolites form fringing or small barrier reefs and are interbedded with oolitic and intraclastic sandy grainstones (Corkeron & George 2001). In contrast to the carbonate facies, the DGU displays abundant features consistent with glaciogenic depositional processes. Thin beds of massive diamictite with striated and faceted clasts overlie striated pavements and are interpreted as lodgement till.

666

M. CORKERON

Elsewhere, thicker packages of massive diamictite overlain by dolomicrite are interpreted as rainout till. Interbedded conglomerate and sandstone are consistent with fluvioglacial outwash deposits. There is no evidence for freeze–thaw repetition.

Boundary relations with overlying and underlying non-glacial units On the western limb of the Louisa Syncline, diamictite and sandstone of the Landrigan Fm. unconformably overlie Bow River Granite of the Lamboo Complex (1857 + 5 Ma; Page & Sun 1994) and various members of the Halls Creek Group. To the south of Kuniandi Range diamictite is absent and dolomite of the Landrigan Fm. lies directly on basement (Roberts et al. 1972). In the McKinnon Syncline, the Landrigan Fm. unconformably overlies the Carson Volcanics and Pentecost Sandstone of the Kimberley Group on which glacially striated pavements are preserved (Roberts et al. 1972). Roberts et al. (1972) recorded glaciated pavements at two sites in the McKinnon Syncline where the Landrigan Fm. lies on polished, grooved and striated bedrock. Grooves and striae show consistent east – west orientation, with overall palaeo-ice flow to the west (Roberts et al. 1972). The upper contact of the Landrigan Fm. is obscured by weathering of the uppermost beds. However, an apparent sharp but conformable contact separates the Landrigan Fm. from siliciclastic beds of either the Stein Fm. in the central Kuniandi area, or the Wirara Fm. in the north and south of the range (Corkeron 2007a). In the McKinnon Syncline, green laminated siltstone of the Wirara Fm. overlies the Landrigan Fm. with apparent conformity (Roberts et al. 1972). The Louisa Downs Group disconformably overlies sedimentary rocks of the Kuniandi Group on the western limb of the Louisa Syncline. A cobble/boulder horizon locally marks the contact between the Louisa Downs and Kuniandi groups. Elsewhere, the basal unit, the Egan Fm. unconformably overlies Palaeoproterozoic rocks of the Kimberley Group, Mesoproterozoic Crowhurst and Glidden Groups, Neoproterzoic Colombo Sandstone and granites of the Lamboo Complex (Tyler et al. 1998). Well preserved striated pavements on Palaeoproterozoic sandstone underlie diamictite in various locations, but particularly Pavement Hill. Ubiquitous grooves, crescent gouges and fractures and chattermarks are reliable indicators of palaeo-ice flow, which trend southward (Corkeron 2008). Medium-scale landscape structures such as roche moutonne´es are also preserved. There is a sharp but conformable contact between carbonate strata of the upper carbonate unit of the Egan Fm. and rippled sandstone of the overlying Yurabi Fm. Locally, thin interbeds of intraclastic sandstone mark the basal contact of the Yurabi Fm.

Chemostratigraphy d13C values range from –4.1‰ to –4.7‰ as measured from a 4-m-thick unit of cream dolomite at the top of the Landrigan Fm. located near Stein Creek in the south of the Kuniandi Range (Fig. 65.3b). d18O values for all samples cluster between –6.8 and – 8.2‰ (Corkeron 2007a). A composite chemostratigraphy of the Egan Fm., produced from the correlation of the three stratigraphic units (LCU, DGU and UCU), shows transition in d13C from þ1.5‰ in the LCU to 2.7‰ at the base of DGU, then to 0.4‰ at the top of the DGU (Corkeron 2007a). Directly overlying the glaciogenic strata, d13C is as low as –2‰, although a single sample has been measured at – 9‰. The profile for the remainder of the UCU shows a gradual trend toward 0‰, followed by a rise to þ2‰, at the top of the formation (Fig. 65.4b). Sr-isotope analysis from the Landrigan and Egan formations did not yield reliable results (Corkeron 2002).

Biostratigraphy The Egan Fm. contains the stromatolite biostratigraphic marker Tungussia julia in various forms ranging from incipient mat to 3-m-scale branching bioherms (Grey & Corkeron 1998). This stromatolite forms the basis for biostratigraphic correlation of the Egan Fm. with the Julie Fm. and basal Bonney Sandstone of central and southern Australia, respectively.

Palaeolatitude and palaeogeography Attempts at palaeomagnetic analysis of strata from the Kuniandi and Louisa Downs Groups have proven largely unsuccessful (Z-X Li 2002, pers. comm.).

Geochronological constraints There is no reliable direct geochronolgical control on either the Landrgian or Egan formations. A minimum age of 513 + 12 Ma is given by the Milliwindi Dyke, a feeder to the Antrim Pleateau Volcanics (Hanley & Wingate 2000) that unconformably overlie the Louisa Downs Group at the top of the Neoproterozoic succession.

The Fargoo and Moonlight Valley formations and the Boonall Dolomite, east Kimberley The Duerdin and Albert Edward groups along with the early Neoproterozoic Ruby Plains Group comprise marine siliciclastic strata of the Wolfe Creek Basin (Blake et al. 1999). Glaciogenic strata are preserved in the Duerdin Group (Dow & Gemuts 1969; Coats & Preiss 1980; Kennedy 1996; Corkeron 2007a, 2008) and glaciogenic material is preserved in the Boonall Dolomite of the Albert Edward Group. The Duerdin Group is preserved on the eastern margin of the Halls Creek Orogen, with outcrop extending NE from the Western Australia-Northern Territory border near Kununurra to SE of Halls Creek (Fig. 65.5). The Albert Edward Group lies east of the Duerdin Group but is confined to the central and southern regions of the east Kimberley district. The Fargoo Fm., Frank River Sandstone and Moonlight Valley Fm. comprise basal siliciclastic units of the Duerdin Group. A thin cap dolomicrite unit lies conformably at the top of the Moonlight Valley Fm. The Boonall Dolomite is a relatively thin (c. 50 m) formation within the Albert Edward Group, conformably bounded by mudstones of the Elvire Fm. below and Timperley Shale above. Sedimentology of the east Kimberley succession was well documented by Dow & Gemuts (1969), who identified and described the glaciogenic strata of the Duerdin Group. The Fargoo and Moonlight Valley formations are interpreted as having glaciogenic affinities. Dow & Gemuts (1969) suggested that these two units represent two phases of one glacial epoch which they named the Moonlight Valley Glaciation. More recent studies by Kennedy (1996) and Corkeron (2007a) focused on the relationship between glaciogenic strata and the thin dolomite beds that directly overlie them. Corkeron (2008) discussed ice-controlled sediment transport directions and palaeogeographical implications for the east Kimberley.

Structural framework Within the Wolfe Creek Basin, Corkeron (2008) identified a depocentre in the central Purnululu area during Duerdin Group deposition. Isostatic flexure associated with ice-sheet advance and retreat as well as ice scour, are probable sources for depositional

KIMBERLEY NEOPROTEROZOIC GLACIAL DEPOSITS

667

Fig. 65.5. (a) Distribution of the Duerdin and Albert Edward Groups in the east Kimberley. Distribution of Fargoo and Moonlight Valley formations at Moonlight Valley (b), Purnululu (c) and Palm Springs (d) study sites. Sites 1, 2, 3 and 4 relate to the Moonlight Valley Fm. type section, Purnululu section, Palm Springs section and Boonall Dolomite section, respectively (see Fig. 65.6) (from Corkeron 2007a).

topography. It is unclear as to the tectonostratigraphic mechanism that maintained accommodation during the deposition of the Albert Edward Group. Thorne & Tyler (1996) ascribed reactivation along the Halls Creek Orogen to the late Neoproterozoic based on folding of the Duerdin and Albert Edward Groups. However, Corkeron (2008) speculated that ice-loading reactivated the Osmond Range Fault (a Palaeoproterozic suture zone of the Kimberley Block; Fig. 65.5) as a growth fault could have enhanced subsidence at Purnlulu. Dow & Gemuts (1969) noted gentle folding of the Duerdin Group beneath the Albert Edward Group, thus marking an unconformity between the two groups. Subsequent deformation of the Duerdin and Albert Edward Group, represented by shallowly dipping bedding, was minimal.

Stratigraphy The Fargoo and Moonlight Valley formations, separated by the Frank River Sandstone, comprise the basal glaciogenic package of the lower Duerdin Group. Dow & Gemuts (1969) interpreted a probable disconformity between the Moonlight Valley Fm. and

underlying Frank River Sandstone. Sandstone overlain by progressively finer mudstone characterizes the upper part of the Duerdin Group. An unconformable relationship between the Duerdin and Albert Edward groups reflects a period of erosion and gentle folding (Dow & Gemuts 1969). The Albert Edward Group is a thick marine succession comprising conformable units of sandstone and mudstone with a thin dolomite unit, the Boonall Dolomite, near the base of the group. The top of the Albert Edward Group is an angular unconformity with the Antrim Plateau Volcanics. Dow & Gemuts (1969) interpreted the distribution of the Fargoo Tillite to include the central and northern areas of the east Kimberley, from directly west of the Dixon Range, northward to Skinner Point and Keep River National Park in the Northern Territory. In the southern area of the east Kimberley, near Palm Springs, the previously mapped Moonlight Valley Fm. was reinterpreted by Corkeron (2008) as Fargoo Tillite, overlain by Frank River Sandstone, with Moonlight Valley Fm. only forming the uppermost c. 10 m. In the central region of east Kimberley, near Purnululu National Park, the Frank River Sandstone and Moonlight Valley Fm. are mapped as relatively thin packages (Dow & Gemuts

668

M. CORKERON

1963; Tyler et al. 1998). Corkeron (2008) reinterpreted the distribution of both formations to include strata originally mapped as the Ranford Fm. (Dow & Gemuts 1963). Thus, the Frank River Sandstone in this area is c. 400 m thick, with recessive Moonlight Valley Fm. lying above it and marked by a resistant dolomite bed. The distribution of the Albert Edward Group is confined to the central and southern east Kimberley.

Glaciogenic deposits Fargoo Fm. Glacially scoured basement beneath the Duerdin

Group (Fargoo Fm.) is found in the northern part of the east Kimberley (Perry & Roberts 1968; Dow & Gemuts 1969). Here the Fargoo Fm. is preserved as rounded and weathered ridges of mostly massive diamictite with a characteristic green matrix. Further south, conglomerate and sandstone are interbedded with stratified and massive diamictite. Massive clast-rich and graded diamictite facies form the bulk of the Fargoo Fm. Matrix is either grey to green dolomicrite to quartz siltstone. Granule- to boulder-sized clasts, sub-rounded to sub-angular, are common and locally cluster in lenses several metres long. Faceted, polished and striated surfaces are common on large indurated sandstone and granite clasts in the diamictite. In southern east Kimberley the matrix content of massive diamictitie decreases to 20– 30%, compared with up to 90% in the north. Stromatolitic dolomite clasts derived from underlying strata are abundant in the south. In central east Kimberley conglomerate and sandstone facies form lenses with both gradational and sharp contacts with massive diamictite facies. Sharp contacts are associated with low-angle erosion surfaces. Metre-thick, and up to 2-m-wide lenses of conglomerate and sandstone facies display tabular and trough crossbeds and are normally graded. Frank River Sandstone. Sandstone, conglomerate and siltstone

facies characterise the Frank River Sandstone. Sandstone is most common, including lithic-feldspathic sandstone, feldspathic sandstone, quartz sandstone and silty quartz sandstone. Conglomerate, dolomitic sandstone and siltstones are minor. The general composition of the sandstones and conglomerates are similar to the matrix and clast composition, respectively, of the underlying Fargoo Fm. The Frank River Sandstone is best preserved in the central east Kimberley where it is c. 400 m thick. Here, thickly bedded, massive or graded sandstone facies sharply overlie diamictite of the Fargoo Fm. Stratified sandstone is interbedded with dolomitic sandstone and siltstone in an overall fining-up package. In the northern east Kimberley, the Frank River Sandstone forms locally developed lenticular channel-fills up to 20 m wide and 5 m thick. These lenses comprise conglomerate grading up to trough and tabular cross-stratified sandstone. In southern part of the east Kimberley, it is up to 20 m thick and comprises a basal conglomerate with overlying asymmetrically rippled sandstone and pebbly sandstone. Conglomerate facies show large lateral variation in thickness and are commonly discontinuous over tens to hundreds of metres. The formation has an erosional base, whereas the upper contact is recessive and concealed. An enigmatic carbonate rock unit at the top of the Frank River Sandstone in the southern east Kimberley area is distinct from facies found elsewhere. The unit comprises interbedded buff laminated dolomicrite that is locally brecciated with tepee structures, dolomitic sandstone, and dark grey, massive carbonate breccia containing bedding-parallel, commonly brecciated chert nodules. The laminated dolomite shows features typical of shallow carbonate platform deposits, while the brecciated carbonate that crosscuts laminated material is interpreted to be the product of karst dissolution during localized subaerial exposure (Corkeron 2008). Moonlight Valley Fm. The Moonlight Valley Fm. comprises a

clast-poor diamictite, or pebbly mudstone, sharply overlain by

interbedded siltstone and dolomicrite. The diamictite has a maroon, highly friable, micritic quartz siltstone matrix. Clasts are rare and mostly rounded or tabular pebbles of quartz, dolomite and siltstone. In the northern east Kimberley area, recessiveweathering Moonlight Valley Fm. sharply overlies facies of the Frank River Sandstone forming a 40-m-thick laterally persistent unit traceable over several kilometres. Elsewhere, the diamictite is concealed by scree derived from the resistant overlying dolomicrite bed. Thinly laminated, homogeneous, pink to yellow dolomicrite everywhere overlies pebbly mudstone of the Moonlight Valley Fm. Locally the dolomicrite is interbedded with laminated siltstone, which increases gradually up-section. In the central and southern areas, basal laminae of the dolomicrite are folded and disrupted to form edge-wise breccia with cement-filled void spaces. Boonall Dolomite. No glaciogenic rocks were identified in the

Albert Edward Group by early workers (Dow & Gemuts 1969). However, Grey & Corkeron (1998), Corkeron & George (2001) and Corkeron (2007a) proposed that the Boonall Dolomite correlates with the Egan Glaciation on biostratigraphic, lithostratigraphic and chemostratigraphic grounds. Recently identified striated clasts from within the Boonall Dolomite support glaciogenic affinities (M. Corkeron unpublished data). The Boonall Dolomite is about 40 m thick and lies conformably between the Elvire Fm. below and the Timperely Shale above (Fig. 65.5). These bounding formations are characteristic of the overwhelmingly sicliciclastic marine units of the Albert Edward Group. The Boonall Dolomite comprises interbedded laminated buff dolomite and flaggy dolomitic siltstone at its base. Intraclastic, peloidal, oolitic and stromatolitic grainstone are common in the upper part of the formation. Lenticular, metre-scale beds of graded and massive conglomerate are discontinuously distributed above the laminated dolomite, towards the top of the formation. Pebbles and cobbles in the conglomerate are sub-rounded, subangular and angular, poorly sorted and compositionally dominated by dolomite suggesting limited transportation (Corkeron 2007a). Striations are preserved on flat cobble surfaces. Depositional interpretation. Glacially scoured basement and allied lithofacies associations within the lower Duerdin Group record glaciogenic depositional processes (including lodgment, rainout and meltwater reworking) consistent with a record of ice advance and retreat. The regional distribution of diamictite/conglomerate facies containing faceted, polished and striated boulders from the Fargoo Fm. is evidence for lodgment till deposition and grounded continental ice sheets at sea level. Dow & Gemuts (1969) envisaged ice flow directly eastward from a topographic high related to the Halls Creek Orogen, into in a marine basin to the east of the Orogen. Alternatively, Corkeron (2008) suggested grounding lodgment deposits directly overlying glacially striated pavements. The abundance of carbonate framework clasts and high carbonate content of the matrix is attributed to erosion of the underlying Bungle Bungle Dolomite, Eliot Range Dolomite and Olympio Fm. on a regional scale. Palaeo-ice flow was toward the SSW (Corkeron 2008). Basal meltout deposits accumulated during glacial retreat in the form of channel deposits and rippled sandstones of the Frank River Sandstone (Corkeron 2008). This formation records fluviomarine transition and marine shelf sands in the Purnululu area, with sediment being shed northward off the flanks of an isostatically induced topographic high to the south. Karstified dolomite at the top of the Frank River Sandstone is localized to this southern high. Post-glacial transgression resulted in basin-wide flooding associated with the deposition of the Moonlight Valley Fm. Dow & Gemuts (1969) interpreted the Moonlight Valley Fm. as marine deposits beneath an ice-covered sea. They envisage ice-capped land to the NE (Victoria River Plateau) as the source of southwesterly directed sediment flow. Both the mud matrix and pebbly

KIMBERLEY NEOPROTEROZOIC GLACIAL DEPOSITS

component of this formation are partially derived from ice-rafted debris. This transgression also accommodated the basin-wide deposition of laminated dolomicrite of at the top of the formation (Corkeron 2007a, 2008). Overlying formations in the Duerdin and Albert Edward groups record extensive siliciclastic marine deposition. A single interval of shallow carbonate platform deposition characterized by oolitic shoals, stromatolite bioherms and intraclastic grainstones is recorded by the Boonall Dolomite. Thin siliciclastic conglomerate lenses interbedded within the dolomite strata are cautiously interpreted as glaciogenic in origin as indicated by the presence of striated pebbles and cobbles. Localized brecciation and scour directly beneath the conglomerate may be derived from sediment gravity flow associated with glaciomarine outwash. A thick package of marine shales and sandstones overlies the Boonall Dolomite.

Boundary relations with overlying and underlying non-glacial units The Duerdin Group rests with major unconformity on the Helicopter Siltstone (Victoria River Basin) to the north and the Ruby Plains Group (Wolfe Creek Basin) to the south (Tyler 2000). Basal contacts are erosive, as indicated by glacial scour and erosional topography and incorporation of underlying rock types as diamictite clast and matrix. Striated pavements underyling Duerdin Group strata are reported from east of Moonlight Valley adjacent to the Ord River (Spring Creek Station) (Dow 1965; Perry & Roberts 1968) and in Keep River National Park (Corkeron 2008). Pavement comprising indurated Palaeoproterozoic sandstone is typically polished showing gouges, chatter marks and striations. These abrasive features indicate an ice-flow direction to the SSW (Corkeron 2008). Diamictite of the Fargoo Fm. lies directly upon the pavements. Dow & Gemuts (1969) recognized a ‘probable unconformity’ beneath the Moonlight Valley Fm., locally identified in the southern area as a karst surface. Marine siliciclastic strata conformably overlie the Moonlight Valley Fm. An unconformity between Duerdin and Albert Edward Group is marked by slight folding of the Duerdin Group (Dow & Gemut 1969) not observed in the Albert Edward Group. The Boonall Dolomite is apparently conformable with underlying and overlying shale. The Antrim Plateau Volcanics form an unconformable sheet across the Neoproterozoic succession.

Chemostratigraphy Corkeron (2007a) measured d13C of the laminated and brecciated carbonate beneath the Moonlight Valley Fm. These samples showed enriched 13C values of 3.3‰ and 1.8‰ (Fig. 65.6a). Laminated dolomicrite above the pebbly mudstone of Moonlight Valley Fm. show decreasing upward d13C trends from – 2.9‰ to –3.4‰ (Williams 1979; Kennedy 1996; Corkeron 2007a). In the Boonall Dolomite, d13C values (Corkeron 2007a) are c. 3.3‰ at the base and decrease up-section to c. – 1.5‰ before returning to a peak of 4.3‰ and stabilizing at c. 2.0‰ at the top of the section (Fig. 65.6b).

Palaeolatitude and palaeogeography No useful palaeomagnetic data have been derived from the Duerdin or Albert Edward Group. Interpretation of palaeolatitude is dependent on correlation with equivalent deposits elsewhere for which reliable palaeopoles exist. As for the Walsh and Landrigan formations, correlation with the Elatina Fm., with a wellconstrained palaeolatitude of 2.7 + 3.78 (Schmidt & Williams

669

1995) or 8.6 + 3.48 (Sohl et al. 1999) suggests a palaeolatitude for the Kimberley region at the end Cryogenian of 12 –168 (+38) (Sohl et al. 1999; Evans 2000).

Geochronological constraints There is no direct geochronological control on any of the rocks in the Duerdin or Albert Edward groups. The Duerdin Group unconformably overlies Palaeo- and Mesoproterozoic rocks of the Kimberley and Victoria River basins and Halls Creek and King Leopold orogens. In the Halls Creek area, the Duerdin Group unconformably overlies the Eliot Range Dolomite (top of the Ruby Plains Group), and is interpreted as early Neoproterozoic in age based on stromatolite biostratigraphy (Grey & Blake 1999). The upper unit of the Albert Edward Group is unconformably overlain by the Antrim Plateau Volcanics. Whole-rock Rb –Sr dating was undertaken on shales from the Neoproterozoic strata (Bofinger 1967) with ages ranging from c. 750 to 600 Ma, but these ages are considered unreliable (Coats & Preiss 1980; Plumb 1981; Dickin 1995).

Discussion The Walsh, Landrigan and Fargoo/Moonlight Valley formations are characterized by diamictite comprising very poorly sorted siliciclastic and carbonate clasts (sand to boulder grade), supported in a very fine silty carbonate matrix. Diamictite overlies striated basement and is interbedded with conglomerate and sandstone lenses. All three formations fine upwards to siltstone and are overlain by a thin carbonate unit characterized by very thinly laminated dolomicrite (Corkeron 2007a). In contrast, the Egan Fm. comprises carbonate strata bounding a siliciclastic unit of highly variable thickness. The siliciclastic rocks are similar, however, to those of the Walsh, Landrigan and Fargoo/Moonlight Valley formations in that they also comprise poorly sorted, mixed clasts in a silty carbonate matrix (Corkeron & George 2001). Facies analysis of the Walsh, Landrigan and Fargoo/Moonlight Valley formations shows that each is divisible into comparable facies associations. Massive diamictite with local stratified diamictite and lenticular conglomerate and sandstone dominate the formations. Glacially influenced deposition is interpreted for this facies assemblage in all three formations (Corkeron 2007a). The fine-grained components (siltstone and pebbly mudstone) overlying diamictite in the Walsh, Landrigan and Moonlight Valley formations are interpreted as marine transgression deposits (Kennedy 1996; Corkeron 2008). In the Egan Fm., the siliciclastic unit displays a similar facies assemblage to the diamictite units of the other formations. The Egan DGU is also interpreted as having a glacial affinity (Corkeron & George 2001) with marked lateral facies variation attributed to basin geometry. The dolomicrite capping the Walsh, Landrigan and Fargoo/ Moonlight Valley formations is dominated by stratified carbonate with minor stratified dolomitic sandstone. Deposition from suspension settling and distal turbidity currents is interpreted for all three formations to have occured in a marine environment, below wave base (Kennedy 1996; Corkeron 2007a). In the Egan Fm. there are two distinct carbonate units (Lower and Upper Carbonate Units; Corkeron & George 2001) in which the facies assemblages are different from those capping the Walsh, Landrigan and Moonlight Valley formations. Significant microbial influence in the precipitation and binding of the carbonate facies of the Egan Fm. is evident. Carbonate grains such as ooids, peloids, oncoids and intraclasts are abundant. Sedimentary structures include stromatolites, trough cross-beds and rare mudcracks. The depositional setting, therefore, was most likely a shallow, moderate- to high-energy, warm water environment (Corkeron &

670 M. CORKERON

Fig. 65.6. (a) Glaciogenic stratigraphy of the lower Duerdin Group. Examples from three sites: Texas/Mabel Downs area (Fig. 65.5b), Purnululu area (Fig. 65.5c) and Palm Springs area (Fig. 65.5d). (b) Boonall Dolomite (Albert Edward Group) stratigraphy. Simplified from Corkeron (2007). C-isotope trends for carbonate units as shown.

KIMBERLEY NEOPROTEROZOIC GLACIAL DEPOSITS

George 2001). This setting contrasts significantly with the lowenergy, deep-water environment interpreted for the cap dolomicrite overlying the Walsh, Landrigan and Moonlight Valley formations. Of note, the UCU at the Egan Fm. type section, characterized by thinly bedded micritic dolomite, more closely resembles the deep-water facies of the cap dolomicrites. This lithological similarity was the basis for the Coats & Preiss (1980) correlation of the Walsh and Moonlight Valley Tillites with the Egan Fm. Corkeron & George (2001), however, interpreted these strata as outer-shelf, below-wave-base lateral equivalents of the more shoreward UCU facies typical elsewhere. The Boonall Dolomite is sedimentologically similar to the Egan Fm. It is also characterized by carbonate platform strata containing minor siliciclastic conglomeratic lenses. Carbonate facies are analogous to those found in the Egan Fm. (Corkeron 2007a). Striated pebbles and cobbles in the conglomerate argue for a glaciogenic influence or the reworking of local glacial deposits.

Chemostratigraphy Two distinct groups of carbonate rocks from the Kimberley Neoproterozoic successions can be defined by their stable isotope values. Stable d13C values around –3 to –4‰ distinguish the cap dolomicrites of the Landrigan, Walsh and Moonlight Valley formations. In contrast, d13C is characteristically  –2‰ but erratic in the Egan Fm. and Boonall Dolomite. Moreover, the stratigraphic range of d13C for the former group is consistently stable around –3 to –4‰.

Glaciogenic correlation Correlation within the Kimberley successions has been controversial, with several conflicting regional correlations proposed. Dow & Gemuts (1969) originally suggested that the Walsh, Landrigan and Fargoo– Moonlight Valley formations correlate with the Sturt Tillite and equivalent (mid Cryogenian) glacials of South Australia and that the Egan Fm. correlates with the endCryogenian Elatina Fm. In contrast, Coats & Preiss (1980) (see also Williams 1979) considered the Egan Fm. and Walsh and Fargoo –Moonlight Valley formations to be equivalent of the Elatina Fm., whereas the Landrigan Fm. represented the sole Sturtian equivalent in the Kimberley region. Grey & Corkeron (1998) reinterpreted the Egan Fm. to be mid-Ediacaran (i.e. postElatina) in age based on biostratigraphic evidence. Li (2000) proposed an alternative interregional correlation with a mid-Cryogenian age for the Walsh Fm. Palaeomagnetic analysis of the ‘cap carbonate’ overlying Walsh Fm. diamictite indicates a palaeolatitude of 458 + 128. Comparison of this palaeolatitude with the known APWP for the late Proterozoic suggests a minimum age of c. 750 Ma, and hence the interpretation of the Walsh Fm. as being equivalent to the ‘Sturtian’ glacial deposits of South Australia (Li 2000). However, timing of the end of the Sturtian glaciation in the Adelaide Rift Complex was recently reinterpreted by Kendall et al. (2006) using Re –Os data. Dating on black shales from the Tindelpina Shale Member overlying ‘Sturtian’ glacial deposits gives an age of 643.0 + 2.4 Ma, whereas the Areyonga Fm. in the Amadeus Basin is constrained to an age older than 657.2 + 5.4 Ma from the overlying Aralka Fm. (Kendall et al. 2006). These dates suggest that mid-Cryogenian glaciation in Australia is younger than at least 685 Ma (Kendall et al. 2006). Consequently, the rationale for correlating the Walsh Fm. (and Landrigan and Fargoo/Moonlight Valley formations) with ‘Sturtian’ glaciation based on a palaeomagnetic ‘best fit’ with the Australian APWP at 750 Ma is no longer valid. The lateral persistence of the Fargoo/Moonlight Valley, Landrigan and Walsh formations, the genetic consistency and

671

uniformity of facies associations, and the consistency of C-isotope values and glacial features in underlying strata, support the interpretation that these deposits are most likely correlatives and probably genetically related (Corkeron 2008). Glacial pavements beneath the Walsh, Landrigan and Fargoo formations attest to ice movement across a broad continental land mass. This interpretation supports the Kimberley correlation originally proposed by Dow & Gemuts (1969), which recognized a widespread lower glaciation – the Landrigan Glaciation. Furthermore, up-section d13C depletion in dolomites in the Ngalia Basin (Mount Doreen Fm.), the Amadeus Basin (Olympic Fm.) and the Adelaide Rift Complex (Nuccaleena Fm.) is consistent with chemostratigraphic correlation of the Walsh/Landrigan/Moonlight Valley formations with end Cryogenian glaciation in Australia (Kennedy 1996; Corkeron 2007a). The depositional system associated with the Egan Fm. distinguishes this carbonate deposit from typical ‘cap carbonate’ rocks and argues against correlation of this formation with the Walsh and Moonlight Valley formations as proposed by Coats & Preiss (1980). Furthermore, the stable isotope characteristics of the Egan Fm. clearly distinguish these carbonate rocks from the cap dolomicrite suite. Whereas the lateral extent of carbonate rocks and diamictite associated with the Egan glaciation is restricted to the Mount Ramsay area, carbonate rocks of the Boonall Dolomite in the east Kimberley are a probable correlative. The Egan Fm. is biostratigraphically correlated with the Julie and Wonoka formations of central and southern Australia (Grey & Corkeron 1998). The Egan Fm. therefore records an Ediacaran glacial event pre-dating the appearance of Ediacaran fauna, and it is interpreted as a possible equivalent of the Gaskiers glaciation (Grey & Corkeron 1998; Corkeron & George 2001; Corkeron 2007a). W. Preiss and G. Williams are thanked for their editorial review and helpful advice to improve this chapter. G. Halverson was particularly helpful with editorial matters, advice and patience. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Blake, D. H., Tyler, I. M., Griffin, T. J., Sheppard, S., Thorne, A. M. & Warren, R. G. 1999. Geology of the Halls Creek 1:100 000 sheet area (4461), Western Australia. Geological series explanatory notes, Australian Geological Survey Organisation, 36. Bofinger, V. M. 1967. Geochronology in the East Kimberley area of Western Australia. PhD thesis, Australian National University. Brookfield, M. E. 1994. Problems in applying preservation, facies and sequence models to Sinian (Neoproterozoic) glacial sequences in Australia and Asia. Precambrian Research, 70, 113–143. Coats, R. P. & Preiss, W. V. 1980. Stratigraphic and geochronological reinterpretation of Late Proterozoic glaciogenic sequences in the Kimberley region, Western Australia. Precambrian Research, 13, 181– 208. Corkeron, M. 2002. Neoproterozoic glacial events in the Kimberley region, Western Australia: sedimentology and regional correlation in the context of continental- and global-scale Neoproterozoic glaciation. PhD thesis, The University of Western Australia. Corkeron, M. 2007a. ‘Cap carbonates’ and Neoproterozoic glaciogenic successions from the Kimberley region, north-west Australia. Sedimentology, 54, 871–903. Corkeron, M. 2007b. The Sedimentary Record of Neoproterozoic Glaciation in the Kimberley Region of Northwestern Australia. Field handbook, IGCP512 Neoproterozoic Ice Ages. Corkeron, M. 2008. Deposition and palaeogeography of a glaciogenic Neoproterozoic succession in the east Kimberley, Australia. Sedimentary Geology, 204, 61– 82. Corkeron, M. L. & George, A. D. 2001. Glacial incursion on a Neoproterozoic carbonate platform in the Kimberley region, Australia. Geological Society of America Bulletin, 113, 1121– 1132.

672

M. CORKERON

de Keyser, F. 1972. Proterozoic tillite at Duchess, nortwestern Queensland. Bulletin of the Bureau of Mineral Resources, Australia, 125, 1–6. Derrick, G. M. & Playford, P. E. 1973. Lennard River, Western Australia – 1:250 000 Map Sheet, Geology Series notes. Western Australian Geological Survey 120. Dickin, A. P. 1995. Radiogenic Isotope Geology. Cambridge University Press, Cambridge. Dow, D. B. 1965. Evidence of a Late Pre-Cambrian glaciation in the Kimberley region of Western Australia. Geological Magazine, 102, 407–414. Dow, D. & Gemuts, I. 1963. Dixon Range, Western Australia – 1:250 000 Map Sheet, Geology Series map. Western Australian Geological Survey. Dow, D. B. & Gemuts, I. 1969. Geology of the Kimberley Region, Western Australia: The East Kimberley. Bureau of Mineral Resources, Geology and Geophysics, Bulletin, 106. Edgegoose, C. J., Fahey, G. M. & Fahey, J. E. 1989. Wingate Mountains, Northern Territory – 1:250 000 Geological Series explanatory Notes. Department of Mines and Energy, Northern Territory Geological Survey 5069. Evans, D. A. D. 2000. Stratigraphic, geochronological, and paleomagnetic constraints upon the Neoproterozoic climatic paradox. American Journal of Science, 300, 347– 433. Gellatly, D. C., Derrick, G. M. & Plumb, K. A. 1975. The Geology of the Lansdowne 1:250 000 Sheet Area, Western Australia. Australian Government Publishing Service. Grey, K. & Corkeron, M. 1998. Late Neoproterozoic stromatolites in glaciogenic successions of the Kimberley region, Western Australia: evidence for a younger Marinoan glaciation. Precambrian Research, 92, 65 –87. Grey, K. & Blake, D. H. 1999. Neoproterozoic (Cryogenian) stromatolites from the Wolfe Creek Basin, east Kimberly, Western Australia: correlation with the Centralian Superbasin. Australian Journal of Earth Science, 46, 329–341. Griffin, T. J., Tyler, I. M. & Playford, P. E. 1993. Lennard River, Western Australia 1:25000 Geology Series map. Geological Survey of Western Australia. Guppy, D. J., Lindner, A. W., Rattigan, J. H. & Casey, J. N. 1958. The geology of the Fitzroy Basin Western Australia. Bureau of Mineral Resources Australia Bulletin, 36. Hanley, L. M. & Wingate, M. T. D. 2000. SHRIMP zircon age for an Early Cambrian dolerite dyke: an intrusive phase of the Antrim Plateau Volcanics of northern Australia. Australian Journal of Earth Sciences, 47, 1029– 1040. Harms, J. E. 1959. The geology of the Kimberley Division, Western Australia and of and adjacent area of the Northern Territory. MSc thesis, University of Adelaide. Kendall, B., Creaser, R. A. & Selby, D. 2006. Re– Os geochronology of postglacial black shales in Australia: constraints on the timing of ‘Sturtian’ glaciation. Geology, 34, 729–732. Kennedy, M. J. 1996. Stratigraphy, sedimentology, and isotopic geochemistry of Australian Neoproterozoic postglacial cap dolomites: deglaciation, d13C excursions, and carbonate precipitation. Journal of Sedimentary Research, 66, 1050–1064. Li, Z. X. 2000. New palaeomagnetic results from the ‘cap dolomite’ of the Neoproterozoic Walsh Tillite, northwestern Australia. Precambrian Research, 100, 359– 370.

NT GEOLOGICAL SURVEY. 2006. Geological regions map. Available at: http://www.nt.gov.au/dpifm/Minerals_Energy/Geoscience/ index.cfm?header=Mapping%20Data). Page, R. & Sun, S.-S. 1994. Evolution of the Kimberley region, W.A. and adjacent Proterozoic inliers – new geochronological contraints. Geological Society of Australia Abstracts, 37, 332–333. Perry, W. J. & Roberts, H. G. 1968. Late Precambrian glaciated pavements in the Kimberley region, Western Australia. Journal of the Geological Society of Australia, 15, 51– 56. Plumb, K. A. 1981. Late Proterozoic (Adelaidean) tillites of the Kimberley-Victoria River region, Western Australia and Northern Territory. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 504–514. Plumb, K. A. & Gemuts, I. 1976. Precambrian geology of the Kimberly region, Western Australia. 25th International Geological Congress Excursion Guide 44C. Roberts, H. G., Gemuts, I. & Halligan, R. 1972. Adelaidean and Cambrian stratigraphy of the Mount Ramsay 1:250,000 sheet area, Kimberley region, Western Australia. Bureau of Mineral Resources Report 150. Schmidt, P. W. & Williams, G. E. 1995. The Neoproterozoic climatic paradox: equatorial palaeolatitude for Marinoan glaciation near sea level in South Australia. Earth and Planetary Science Letters, 134, 102– 124. Sohl, L. E., Christie-Blick, N. & Kent, D. V. 1999. Paleomagnetic polarity reversals in Marinoan (c. 600 Ma) glacial deposits of Australia: implications for the duration of low-latitude glaciation in Neoproterozoic time. Geological Society of America Bulletin, 111, 1120– 1139. Sweet, I. P. 1977. The Precambrian Geology of the Victoria River Region, Northern Territory. Bureau of Mineral Resources, Geology and Geophysics Bulletin, 168. Thorne, A. M. & Tyler, I. M. 1996. Mesoproterozoic and Phanerozoic sedimentary basins in the northern Halls Creek Orogen: constraints on the timing of strike-slip movement on the Halls Creek Orogen. Geological Survey of Western Australia, Annual Review, 1995/ 1996, 156– 168. Tyler, I. M. 2000. Geological map of the Halls Creek Orogen, east Kimberley region (1:500 000). In: Tyler, I. M., Griffin, T. J., Sheppard, S. & Thorne, A. M. (eds) The Geology of the King Leopold and Halls Creek Orogens. Geological Survey of Western Australia, Bulletin, 143. Tyler, I. M. & Griffin, T. J. 1990. Structural development of the King Leopold Orogen, Kimberley region, Western Australia. Journal of Structural Geology, 12, 703–714. Tyler, I. M. & Hocking, R. M. 2002. A revised geological framework for Western Australia. Geological Survey of Western Australia extended abstracts, 2002, 1 –7. Tyler, I., Griffin, T. & Sheppard, S. 1998. Geology of the Dockrell 1:100 000 Sheet. Geological Survey of Western Australia 1:100 000 Geological Series Explanatory Notes. Williams, G. E. 1979. Sedimentology, stable-isotope geochemistry and palaeoenvironment of dolomites capping late Precambrian glacial sequences in Australia. Journal of the Geological Society of Australia, 26, 377–386.

Chapter 66 Mid-Ediacaran ice-rafting in the Adelaide Geosyncline and Officer Basin, South Australia VICTOR A. GOSTIN*, DAVID M. MCKIRDY, LYNN J. WEBSTER & GEORGE E. WILLIAMS Discipline of Geology and Geophysics, School of Earth and Environmental Sciences, University of Adelaide, SA 5005, Australia *Corresponding author (e-mail: [email protected]) Abstract: Sedimentary features characteristic of ice-rafting are present in the Bunyeroo Formation (Wilpena Group) of the Adelaide Geosyncline and in the coeval Dey Dey Mudstone (Ungoolya Group) of the eastern Officer Basin (Figs 66.1 & 66.2), providing evidence of a mid-Ediacaran glacial climate in South Australia. The Acraman asteroid impact, a negative shift in marine d13C, and a major acritarch turnover coincided with this frigid epoch.

The Adelaide Geosyncline (Preiss 1987) (or ‘Adelaide Rift Complex’) is well known for its sedimentary record of Neoproterozoic glaciation. The classic Cryogenian Sturt and Elatina glaciations in South Australia are reviewed elsewhere in this volume (Preiss et al. 2011; Williams et al. 2011). Additionally, there is growing evidence from the Adelaide Geosyncline and eastern Officer Basin of later, Ediacaran frigid climate and possible glaciation in South Australia (Gostin et al. 2010; Jenkins 2011). Here we summarize the sedimentological evidence for a glacial influence during deposition of the Ediacaran Bunyeroo Formation (Fm.) (Wilpena Group) and Dey Dey Mudstone (Ungoolya Group).

2000), but perhaps just prior to renewed rifting in the basin (Foden et al. 2001). Along with the rest of the Adelaidean, the Bunyeroo Fm. was deformed during the 514–490 Ma Delamerian orogeny (Foden et al. 2006). Strata in the Delamerian orogen were generally metamorphosed to lower greenschist facies, although the Central Flinders Zone is sub-greenschist facies (Preiss 2000). The Officer Basin was mainly platformal during most of the Adelaidean, with the southern part of the basin being strongly analogous to the Stuart Shelf (Preiss 1993).

Stratigraphy

Isolated pebbles of siltstone, sandstone and quartzite, clusters of quartz granules, rare rectangular patches of poorly sorted, coarsegrained quartzose sandstone up to c. 170 cm2 in area, and small (,2 mm long) ovoid pellets of poorly sorted quartz silt in an ultrafine matrix have been found in mudstone c. 80 m above the base of the Bunyeroo Fm. Similar small (200 –700 mm long) pellets occur in red mudstone and laminated siltstone –mudstone of the lower Dey Dey Mudstone (Webster 2001). Following the terminology of Gilbert (1990), the anomalously coarse-grained deposits in the Bunyeroo Fm. are interpreted as dropstones, dumps and frozen aggregates, respectively, and the siltstone pellets in the Bunyeroo Fm. and Dey Dey Mudstone as till pellets (Ovenshine 1970; van der Meer 1993), all dispersed by floating ice. Ice-rafted detritus in the Bunyeroo Fm. has only been found ,10 m stratigraphically below and above the Acraman impact ejecta horizon (Young 1995; Gostin et al. 2010). The till pellets in the Dey Dey Mudstone have been identified 11 –68 m below the Acraman impact ejecta horizon in the Munta 1 drill hole and 4.4 m below the ejecta horizon in the Murnaroo 1 drill hole (Fig. 66.1) (Gostin et al. 2010). Hence a glacial climate prevailed when the Acraman impact occurred.

The Bunyeroo Fm. is a c. 400-m-thick unit of mostly red, laminated and fissile, hemipelagic mudstone that occurs widely throughout the Adelaide Geosyncline (Fig. 66.1). Its lower contact with the ABC Range Quartzite is interpreted as an exposure surface (Christie-Blick 1995) and it is conformably overlain by the calcareous Wonoka Fm. (Fig. 66.2). The Bunyeroo Fm. begins with an immature gritty sandstone (Wilcolo Member) and grades upward into outer marine-shelf mudstone (Preiss 2000). A 25-cm-thick impact ejecta horizon of dacitic fragments (Gostin et al. 1986; Wallace et al. 1996) and associated Ir anomaly (Gostin et al. 1989), which are related to the c. 90-kmdiameter Acraman impact structure in the Mesoproterozoic Gawler Range Volcanics on the Gawler Craton c. 300 km to the west (Fig. 66.1) (Williams 1986, 1994; Schmidt & Williams 1996), occur 80 m above the base of the Bunyeroo Fm. Anomalous coarse-grained deposits interpreted as ice-rafted detritus (Gostin et al. 2010) lie above and below the impact ejecta horizon. The Dey Dey Mudstone (c. 200 –600 þ m) in the eastern Officer Basin (Fig. 66.2) is correlative with the Bunyeroo Fm. and its lower part comprises red brown mudstone with graded siltstone –mudstone laminae (Calver & Lindsay 1998; Arouri et al. 2000). The distal Acraman impact ejecta horizon has a maximum thickness of 7 mm in the lower Dey Dey Mudstone and provides a high-quality marker horizon between the two basins (Wallace et al. 1989, 1996; Williams & Wallace 2003; Hill et al. 2004, 2007; Williams & Gostin 2005).

Structural framework The Bunyeroo Fm. was deposited during a thermal subsidence phase in the development of the Adelaide Geosyncline (Preiss

Glaciogenic deposits and associated strata

Chemostratigraphy Calver (2000) measured d13C compositions and atomic H/C ratios on kerogens isolated from the Bunyeroo Fm. in Bunyeroo and Brachina gorges and in drill hole SCYW1A (Stuart Shelf). Absolute d13Corg values are variable between the sections, but in all cases the Acraman impact ejecta horizon and ice-rafted detritus occur within a trend of decreasing d13Corg in the lower half of the formation ( –24 to –30‰ in Brachina Gorge; Calver 2000), which is followed by a comparable rise in the upper half of the formation. This negative isotope excursion is replicated in the Dey Dey

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 673– 676. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.66

674

V. A. GOSTIN ET AL.

Fig. 66.1. Map of South Australia showing the Acraman impact structure and localities of ice-rafted detritus and the Acraman impact ejecta horizon in mid-Ediacaran strata in the Adelaide Geosyncline and Officer Basin. Triangles indicate ice-rafted detritus and ejecta seen in close stratigraphic proximity in outcrop. Solid circles indicate ejecta seen in outcrop, and open circles ejecta recorded in drill core. 1, Bunyeroo Gorge; 2, Brachina Gorge; 3, Parachilna Gorge; 4, Munta 1 well; 5, Murnaroo 1 well; 6, Observatory Hill 1 well. The Ediacaran GSSP is located 5 km east of Brachina Gorge. Ejecta localities modified from Williams & Wallace (2003) and Hill et al. (2004).

Mudstone in Observatory Hill 1 and Munta 1 (Munyarai Trough, Officer Basin; Fig. 66.1) where it coincides with a sharp increase in the relative abundance of ethylcholestane (McKirdy et al. 2006; Webster et al. 2007) and a coronene anomaly (Hallmann et al. 2010).

Fig. 66.2. Ediacaran stratigraphy of the Adelaide Geosyncline and eastern Officer Basin. The asterisks mark the stratigraphic position of the Acraman impact ejecta horizon (AIEH). The earliest Ediacaran metazoan fossils appear at the top of the Wonoka Fm. (Jenkins 1995). Deep canyons are incised from a correlative stratigraphic level within the Wonoka Fm. and Munyarai Fm. The scale bars are approximate. Modified from Calver & Lindsay (1998) and Williams & Gostin (2000).

of the Bunyeroo Fm. and equivalent strata gave an age of 593 + 32 Ma (Compston et al. 1987), whereas chemostratigraphic correlations suggested an age of c. 580 Ma (Walter et al. 2000).

Palaeolatitude and palaeogeography Biostratigraphy The Ediacaran succession in South Australia (Wilpena Group) was deposited in low palaeolatitudes (Schmidt et al. 2009; Schmidt & Williams 2010). A stable, high-temperature (c. 680 8C) palaeomagnetic component identified for red beds from the Bunyeroo Fm. provided a positive tectonic-fold test (99% level of confidence) and indicates a palaeolatitude of 15.7 + 6.58 for site-mean results and 15.0 + 2.28 for sample-mean results (Schmidt & Williams 1996).

In the eastern Officer Basin the interval of the lower Dey Dey Mudstone that spans the Acraman impact ejecta horizon and contains ice-rafted detritus also coincides with a major change in palynoflora. There, cyanobacteria and low-diversity, leiospheric acritarchs disappear and are shortly thereafter replaced by a diverse acritarch assemblage dominated by large acanthomorphs (Grey et al. 2003; Grey 2005).

Geochronological constraints

Discussion and conclusions

The Bunyeroo Fm. occurs, by definition (Knoll et al. 2006), in the middle of the Ediacaran Period. The age calibration of the Ediacaran is contentious due to few direct age constraints on the upper Adelaidean succession and controversial correlation of the Wilpena Group with other better dated late Neoproterozoic successions. A U –Pb zircon age of 657+17 Ma on a detrital zircon in the Marino Arkose Member in the Upalinna Subgroup (late Cryogenian) provides a loose maximum age constraint on the base of the Ediacaran Period (Preiss 2000). Rb – Sr whole-rock shale dating

Ediacaran glaciation has been identified elsewhere in Australia (Calver et al. 2004; Corkeron 2007) and in the United States, China, Norway, Scotland and Brazil (Fairchild & Kennedy 2007; Alvarenga et al. 2007). However, in the absence of accurate radiometric ages for all these various glacial deposits it is premature to argue that they record a single glacial episode of worldwide extent. Within the Ediacaran marine sedimentary record of South Australia, evidence of a glacial climate near sea level at the time

MID-EDIACARAN ICE-RAFTING, SOUTH AUSTRALIA

of the Acraman impact (Gostin et al. 2010) is followed in turn by anomalous concentrations of coronene, a polyaromatic hydrocarbon resulting from biomass combustion (Hallmann et al. 2010), a prominent negative C-isotopic excursion (Calver & Lindsay 1998; Calver 2000; Hill et al. 2006; Webster et al. 2007), the diversification of planktonic microfossils (Grey et al. 2003; Grey 2005), unusual biomarker signatures (McKirdy et al. 2006; Webster et al. 2007) and, ultimately, by the rise of the Ediacara biota (Narbonne 2005). Release from the combined environmental stresses of a frigid, glacial climate near sea level and the Acraman impact may therefore have been a catalyst for the subsequent dramatic evolution of the biosphere during the later Ediacaran (Gostin et al. 2010). We thank P. Haines, N. Lemon, R. Jenkins, K. Grey and A. Hill for helpful discussions and G. Halverson for his editorial suggestions. D. M. Mc. K. and L. J. W. also thank Primary Industry and Resources South Australia for financial support. L. J. W. acknowledges support from an Australian Postgraduate Award and an AAPG Grant-in-Aid. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Alvarenga, C. J. S. de, Figueiredo, M. F., Babinski, M. & Pinho, F. E. C. 2007. Glacial diamictites of Serra Azul Formation (Ediacaran, Paraguay belt): evidence of the Gaskiers glacial event in Brazil. Journal of South American Earth Sciences, 23, 236–241. Arouri, K., Conaghan, P. J., Walter, M. R., Bischoff, G. C. O. & Grey, K. 2000. Reconnaissance sedimentology and hydrocarbon biomarkers of Ediacarian microbial mats and acritarchs, lower Ungoolya Group, Officer Basin. Precambrian Research, 100, 235–280. Calver, C. R. 2000. Isotope stratigraphy of the Ediacarian (Neoproterozoic III) of the Adelaide Rift Complex, South Australia, the overprint of water column stratification. Precambrian Research, 100, 121– 150. Calver, C. R. & Lindsay, J. F. 1998. Ediacarian sequence and isotope stratigraphy of the Officer Basin, South Australia. Australian Journal of Earth Sciences, 45, 513–532. Calver, C. R., Black, L. P., Everard, J. L. & Seymour, D. B. 2004. U –Pb zircon age constraints on late Neoproterozoic glaciation in Tasmania. Geology, 32, 893– 896. Christie-Blick, N., Dyson, I. & Von der Borch, C. 1995. Sequence stratigraphy and the interpretation of Neoproterozoic Earth history. Precambrian Research, 73, 3 –26. Compston, W., Williams, I. S., Jenkins, R. J. F., Gostin, V. A. & Haines, P. W. 1987. Zircon age evidence for the Late Precambrian Acraman ejecta blanket. Australian Journal of Earth Sciences, 34, 435– 445. Corkeron, M. 2007. ‘Cap carbonates’ and Neoproterozoic glacigenic successions from the Kimberley region, north-west Australia. Sedimentology, 54, 871– 903. Fairchild, I. J. & Kennedy, M. J. 2007. Neoproterozoic glaciation in the Earth System. Journal of the Geological Society, London, 164, 895– 921. Foden, J., Barovich, K., Jane, M. & O’Halloran, G. 2001. Sr-isotopic evidence for Late Neoproterozoic rifting in the Adelaide Geosyncline at 586 Ma: implications for a Cu ore forming fluid flux. Precambrian Research, 106, 291– 308. Foden, J., Elburg, M. A., Dougherty-Page, J. & Burtt, A. 2006. The timing and duration of the Delamerian Orogeny: correlation with the Ross Orogen and implications for Gondwana assembly. Journal of Geology, 114, 189– 210. Gilbert, R. 1990. Rafting in glacimarine environments. In: Dowdeswell, J. A. & Scourse, J. D. (eds) Glacimarine Environments: Processes and Sediments. Geological Society, London, Special Publications, 53, 105–120. Gostin, V. A., Haines, P. W., Jenkins, R. J. F., Compston, W. & Williams, I. S. 1986. Impact ejecta horizon within late Precambrian shales, Adelaide Geosyncline, South Australia. Science, 233, 198– 200.

675

Gostin, V. A., Keays, R. R. & Wallace, M. W. 1989. Iridium anomaly from the Acraman impact ejecta horizon: impacts can produce sedimentary iridium peaks. Nature, 340, 542– 544. Gostin, V. A., McKirdy, D. M., Webster, L. J. & Williams, G. E. 2010. Ediacaran ice-rafting and coeval asteroid impact, South Australia: insights into the terminal Proterozoic environment. Australian Journal of Earth Sciences, 57, 859– 869. Grey, K. 2005. Ediacaran palynology of Australia. Memoirs Australasian Association of Palaeontologists, 31, 439. Grey, K., Walter, M. R. & Calver, C. R. 2003. Neoproterozoic biotic diversification: snowball Earth or aftermath of the Acraman impact? Geology, 31, 459–462. Hallmann, C., Grey, K., Webster, L. J., McKirdy, D. M. & Grice, K. 2010. Molecular signature of the Neoproterozoic Acraman impact event. Organic Geochemistry, 41, 111– 115. Hill, A. C., Grey, K., Gostin, V. A. & Webster, L. J. 2004. New records of Late Neoproterozoic Acraman ejecta in the Officer Basin. Australian Journal of Earth Sciences, 51, 47 –51. Hill, A. C., Webster, L. J. & McKirdy, D. M. 2006. The Ediacaran Acraman impact event: did it affect the long-term carbon cycle? 16th Annual V.M. Goldschmidt Conference, August 2006, Melbourne, Abstract S8-05. Hill, A. C., Haines, P. W., Grey, K. & Willman, S. 2007. New records of Ediacaran Acraman ejecta in drillholes from the Stuart Shelf and Officer Basin, South Australia. Meteoritics and Planetary Science, 42, 1883–1891. Jenkins, R. J. F. 1995. The problems and potential of using animal fossils and trace fossils in terminal Proterozoic biostratigraphy. Precambrian Research, 73, 51 –69. Jenkins, R. J. F. 2011. Billy Springs glaciation, South Australia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 693– 699. Knoll, A. H., Walter, M. R., Narbonne, G. M. & Christie-Blick, N. 2006. The Ediacaran Period: a new addition to the geologic time scale. Lethaia, 39, 13– 30. McKirdy, D. M., Webster, L. J., Arouri, K. R., Grey, K. & Gostin, V. A. 2006. Contrasting sterane signatures in Neoproterozoic marine rocks of Australia before and after the Acraman asteroid impact. Organic Geochemistry, 37, 189–207. Narbonne, G. M. 2005. The Ediacara biota: Neoproterozoic origin of animals and their ecosystems. Annual Review of Earth and Planetary Sciences, 33, 421–442. Ovenshine, A. T. 1970. Observations of iceberg rafting in Glacier Bay, Alaska, and the identification of ancient ice-rafted deposits. Geological Society of America Bulletin, 81, 891– 894. Preiss, W. V. (compiler) 1987. The Adelaide Geosyncline. Late Proterozoic Stratigraphy, Sedimentation, Palaeontology and Tectonics. Geological Survey of South Australia Bulletin, 53, 438. Preiss, W. V. 1993. Neoproterozoic. In: Drexel, J. F., Preiss, W. V. & Parker, A. J. (eds) The Geology of South Australia, Volume 1, The Precambrian. Geological Survey of South Australia Bulletin, 54, 171– 203. Preiss, W. V. 2000. The Adelaide Geosyncline in South Australia and its significance in Neoproterozoic continental reconstruction. Precambrian Research, 100, 21 –63. Preiss, W. V., Gostin, V. A., McKirdy, D. M., Ashley, P. M, Williams, G. E. & Schmidt, P. W. 2011. The glacial succession of Sturtian age in South Australia: the Yudnamutana Subgroup. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 701–712. Schmidt, P. W. & Williams, G. E. 1996. Palaeomagnetism of the ejectabearing Bunyeroo Formation, late Neoproterozoic, Adelaide fold belt, and the age of the Acraman impact. Earth and Planetary Science Letters, 144, 347–357. Schmidt, P. W. & Williams, G. E. 2010. Ediacaran palaeomagnetism and apparent polar wander path for Australia: no large true polar wander. Geophysical Journal International, 182, 711–726. Schmidt, P. W., Williams, G. E. & McWilliams, M. O. 2009. Palaeomagnetism and magnetic anisotropy of late Neoproterozoic strata, South Australia: implications for the palaeolatitude of late

676

V. A. GOSTIN ET AL.

Cryogenian glaciation, cap carbonate and the Ediacaran System. Precambrian Research, 174, 34 –52. van der Meer, J. J. M. 1993. Microscopic evidence of subglacial deformation. Quaternary Science Reviews, 12, 553– 587. Wallace, M. W., Gostin, V. A. & Keays, R. R. 1989. Discovery of the Acraman impact ejecta blanket in the Officer Basin and its stratigraphic significance. Australian Journal of Earth Sciences, 36, 585– 587. Wallace, M. W., Gostin, V. A. & Keays, R. R. 1996. Sedimentology of the Neoproterozoic Acraman impact-ejecta horizon, South Australia. AGSO Journal of Australian Geology and Geophysics, 16, 443– 451. Walter, M. R., Veevers, J. J., Calver, C. R., Gorjan, P. & Hill, A. C. 2000. Dating the 840– 544 Ma Neoproterozoic interval by isotopes of strontium, carbon, and sulfur in seawater, and some interpretative models. Precambrian Research, 100, 371– 433. Webster, L. J. 2001. Terminal Proterozoic biomarker assemblages in the Centralian Superbasin before and after the Acraman meteorite impact. BSc Honours thesis, University of Adelaide, Adelaide. Webster, L. J., McKirdy, D. M. & Grey, K. 2007. Biogeochemical anatomy of the Acraman bolide impact. 23rd International Meeting on Organic Geochemistry, Torquay, abstract P109-MO. Williams, G. E. 1986. The Acraman impact structure: source of ejecta in late Precambrian shales, South Australia. Science, 233, 200– 203.

Williams, G. E. 1994. Acraman: a major impact structure from the Neoproterozoic of Australia. In: Dressler, B. O., Grieve, R. A. F. & Sharpton, V. L. (eds) Large Meteorite Impacts and Planetary Evolution. Geological Society of America Special Paper, 293, 209– 224. Williams, G. E. & Gostin, V. A. 2000. Mantle plume uplift in the sedimentary record: origin of kilometre-deep canyons within late Neoproterozoic successions, South Australia. Journal of the Geological Society, London, 157, 759–768. Williams, G. E. & Wallace, M. W. 2003. The Acraman asteroid impact, South Australia: magnitude and implications for the late Vendian environment. Journal of the Geological Society, London, 160, 545– 554. Williams, G. E. & Gostin, V. A. 2005. Acraman–Bunyeroo impact event (Ediacaran), South Australia, and environmental consequences: twenty-five years on. Australian Journal of Earth Sciences, 52, 607– 620. Williams, G. E., Gostin, V. A., McKirdy, D. M., Preiss, W. V. & Schmidt, P. W. 2011. The Elatina glaciation (late Cryogenian), South Australia. In: Arnaud, E., Halverson, G. P. & ShieldsZhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 713–721. Young, T. 1995. The Bunyeroo Formation and its possible cold-water marine setting. BSc Honours thesis, University of Adelaide, Adelaide.

Chapter 67 Neoproterozoic glacial deposits of central Australia ANDREW C. HILL1, PETER W. HAINES2 & KATHLEEN GREY2 * 1

Centro de Astrobiologı´a (CSIC-INTA), Instituto Nacional de Te´cnica Aeroespacial, Ctra de Ajalvir, km 4, 28850 Torrejo´n de Ardoz, Madrid, Spain 2

Geological Survey of Western Australia, 100 Plain Street, East Perth, WA 6004, Australia *Corresponding author (e-mail: [email protected])

Abstract: There are two distinct stratigraphic levels of Neoproterozoic glacigenic deposits in central Australia, both Cryogenian in age, spread over an area greater than 2.5  106 km2. They were deposited in a once continuous intracratonic sag basin and are now preserved in four major structural basins: the Officer, Amadeus, Ngalia and Georgina Basins. In all four basins there are units that correlate with the (older) Sturt Tillite and equivalent glacial deposits in the Adelaide Rift Complex – the Sturt glaciation (Preiss et al. 2011) – and with the (younger) Elatina Formation (Fm.) and equivalents – the Elatina glaciation (Williams et al. 2008, 2011). The clearest evidence for glacial activity is the occurrence of diamictites that contain clasts of lithologically diverse origin which are often striated, faceted and polished, and the occurrence of dropstones. Most glacial deposits were deposited in shallow marine to fluvio-lacustrine palaeoenvironments. In all basins, the Elatina glacial deposits are overlain (at least locally) by dolostone units that mark the onset of post-glacial transgression and contain unique sedimentary and geochemical features. The cap dolomite units are distinct from dolomite beds within glaciogenic sediments, and those that occur near the top of Sturt glacial units in the Amadeus (Areyonga Fm.) and eastern Officer Basins (Chambers Bluff Tillite). None of the central Australian glacial units have direct geochronological constraints. There are, however, radiometric dates for a Sturt glacial unit in the Adelaide Rift Complex (Wilyerpa Fm.) and post-glacial shales in the Amadeus Basin (Aralka Fm.), Stuart Shelf (Tapley Hill Fm.) and Adelaide Rift Complex (Tapley Hill Fm.) that indicate a c. 660 Ma age for the Sturt glaciation in Australia (Kendall et al. 2006, 2007; Fanning & Link 2008). The age of the Elatina glaciation in Australia is constrained only by the age of the Sturt glaciation and the presence of the Ediacara fauna in overlying strata of all the basins except the Ngalia Basin. Consequently, correlations have been mainly established by means of lithostratigraphy, chemostratigraphy, palynology, and to a lesser extent, stromatolite biostratigraphy, mainly on the successions above and below the glacial units. Results from each of the above techniques show a remarkable consistency, and indicate that the two major Cryogenian glacial episodes are of similar age across Australia.

The four central Australian basins considered here contain numerous glacial units (Table 67.1), but this diversity reflects the historic need to set up separate stratigraphic nomenclatures for isolated outcrop areas because of correlation difficulties. The distribution of known units and the location of type sections or areas are shown in Figure 67.1 and the latter are listed in Table 67.1. The widespread extent of glacial successions in Australia was recognized many years ago (e.g. Mawson 1949; Dunn et al. 1971). A seminal paper, attempting to correlate successions in what is here referred to as the Cryogenian period, was published by Preiss et al. (1978), building on previous work in the Amadeus (Prichard & Quinlan 1962; Wells et al. 1965, 1966, 1967, 1970), Ngalia (Wells 1972; Wells et al. 1972) and Georgina (Smith 1963a, b, 1972) basins. Each of these basins was also tied to successions in the Officer Basin, Adelaide Rift Complex and Kimberley Region. A large volume of literature has been published since then, mostly delineating the stratigraphic successions as a whole rather than concentrating on the glaciogenic units themselves. Significant references include Walter (1980), Coats & Preiss (1980), Jackson & van de Graaff (1981), Preiss & Forbes (1981), Wells & Moss (1983), Preiss (1993), Williams (1994), Walter et al. (1994, 1995), Grey et al. (1999, 2005), Lindsay (2002), Eyles et al. (2007), and Haines et al. (2008). It is difficult to assess the relative stratigraphic positions and correlation of units in the central Australian basins without comparing them to, and correlating with, the much more continuous sections of the Adelaide Rift Complex, because most advances in understanding central Australian glacial units have been a consequence of the detailed analysis of the latter area (Preiss 1987, 1993, 2000; Williams et al. 2008). In the central Australian basins, Sturt glacial units are usually, but not universally, distinguishable from Elatina glacial units because of their chemically

reduced character – diamictites and interbedded siltstones of Sturt glacial units are typically grey or green in outcrop, and often dark grey to black in drill holes, in contrast to the commonly oxidized, red-brown colour of Elatina glacial units (Haines et al. 2008).

Structural framework Central Australian Neoproterozoic sedimentary basins (Officer, Amadeus, Ngalia and Georgina basins) cover an area in excess of 2.5  106 km2 and overlie Palaeoproterozoic and Mesoproterozoic continental crust (Fig. 67.1; Walter et al. 1995). The basins formed by widespread intracratonic subsidence between 850 and 800 Ma (Zhao et al. 1994), probably in response to continental rifting of Australia –Antarctica from China and/or other continents during the break-up of Rodinia (Li et al. 2008). Uniformity of initial basin sediments and their large areal extent indicates a single depositional system, the Centralian Superbasin (Walter et al. 1995), that subsequently fragmented into the present-day Officer, Amadeus, Ngalia and Georgina basins during the Late Neoproterozoic –Early Cambrian Petermann Orogeny (Wade et al. 2005) and Palaeozoic Alice Springs Orogeny (Haines et al. 2001). Reliable correlations, based on lithostratigraphy, sedimentology, stromatolite biostratigraphy, palynology and chemostratigraphy are found at several Neoproterozoic levels across Australia (Calver & Lindsay 1998; Hill & Walter 2000; Gorjan et al. 2000; Walter et al. 2000; McKirdy et al. 2001; Grey 2005; Hill 2005; Grey et al. 2005). The Centralian Superbasin initially formed as a sag basin, possibly a failed arm of the Adelaide Rift Complex. By late Ediacaran times, the original basin had fragmented. Both the Amadeus and Officer Basins developed an asymmetric cross-section, with a

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 677– 691. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.67

678

Table 67.1. Neoproterozoic glacial deposits and cap dolomites of central Australia. Period

Cryogenian

Location

Unit

Western Officer Basin NW Boondawari Formation

Thickness (m)

,800

Central and NE

Wahlgu Formation

122 –479

Cryogenian

NE (south of Musgrave Province)

Pirrilyungka Formation

1238

Cryogenian

NE (south of Musgrave Province)

Lupton Formation

250

Cryogenian(?)

SW

Turkey Hill Formation

Cryogenian

Eastern Officer Basin Northeastern Chambers Bluff Tillite

Cryogenian(?)

Northeastern

Elatina Formation equivalent(?)

Hundreds of metres

,450

,30

Diamictite, sandstone, conglomerate, siltstone, mudstone, dolomitic siltstone, dolomite

Sandy diamictite; minor mudstone, sandstone, conglomerate and dolomite, including uppermost 1.5 m in Empress 1/1A that resembles ‘upper marker cap dolomite’ in the Amadeus Basin Diamictite; minor conglomerate, sandstone and mudstone Diamictite, conglomerate, sandstone and mudstone

Sandstone, mudstone, diamictite (diamictite may be Permian)

Diamictite, sandstone; minor dolomite, partly stromatolitic

Red-brown, granule-bearing sandstone

Stratigraphic relationships

Type section

Depositional setting

Correlatives

References

Williams (1987, 1992, 1994), Williams & Tyler (1991), Walter et al. (1994), Grey et al. (1999, 2005) Eyles & Eyles (1998), Stevens & Apak (1999), Grey et al. (2005), Haines et al. (2008)

Boondawarri Creek east of Boondawarri Soak; Lat., 238310 3200 S, long. 1218290 5700 E

Glacio-marine

Elatina Fm and equivalents

Empress 1/1A, lat. 278030 1300 S, long. 1258090 2400 E

Fluvioglacial or glacio-marine

Elatina Fm and equivalents

Lower contact not exposed; disconformity with overlying Wahlgu Fm Unconformity with underlying Towsend Quartzite or Lefroy Fm, and unconformity with overlying Cambrian Table Hill Volcanics Basal contact not exposed; unconformity with overlying Permian Patterson Fm

Vines 1, lat. 268420 0500 S, long. 1288150 0800 E

Glacial

Sturt Tillite and equivalents

Haines et al. (2008)

Lupton Hills; lat. 26832.50 S, long. 12882.60 E

?Glacial in part

?

Jackson & van de Graaff (1981), Grey et al. (1999, 2005)

Miller Soak; lat. 2889.40 S, long. 124816.50 E

Glacial in part (glacial component may be Phanerozoic)

?

Jackson & van de Graaff (1981), Grey et al. (2005)

Unconformity with underlying Cryogenian sediments, and disconformity(?) with overlying Wantapella Volcanics not exposed

9.6 km N/NW of Chambers Bluff, lat. 278010 2100 S, long. 1338100 1400 E

Glacio-lacustrine, fluvio-glacial and glacio-marine

Sturt Tillite and equivalents

Preiss (1987, 1993), Morton (1997), Eyles et al. (2007), Preiss et al. (2011)

No type section nominated

Glacial in part(?)

Elatina Fm and equivalents(?)

Preiss (1993), Preiss et al. (2011)

Disconformity with underlying Mundadjini Fm or unconformity with basement, and unconformity with overlying McFadden Fm Disconformity with underlying Steptoe Fm (karstified surface), and disconformity with overlying Lungkarta Fm

A. C. HILL ET AL.

Cryogenian

Lithology

Ediacaran(?)

Cryogenian

Northeastern

Nuccaleena Formation equivalent(?)

Amadeus Basin North and Areyonga NE Formation, including ‘lower marker cap dolomite’

,5

not exposed

No type section nominated

?

Nuccaleena Fm and equivalents(?)

Preiss (1993), Preiss et al. (2011)

,250

Diamictite, sandstone, siltstone, conglomerate and carbonate

Unconformity with underlying Johnnys Creek beds or Bitter Springs Fm (karstified surfaces), and conformity/ disconformity with overlying Aralka Fm

Ellery Creek, 1.5 km south of contact with basement; lat. 238470 S, long. 1338040 E

Subglacial, ice-margin, shallow-marine ice proximal

Sturt Tillite and equivalents

Prichard & Quinlan (1962), Preiss et al. (1978), Wells (1981), Lindsay (1989), Indigo Oil & Sirgo Exploration (1990), Walter et al. (1994, 1995) Preiss et al. (1978), Wells (1981), Shaw & Wells (1983), Field (1991), Freeman et al. (1991), Williams et al. (2007), Skotnicki et al. (2008) Wells et al. (1967), Preiss et al. (1978), Wells (1981), Shaw & Wells (1983), Lindsay (1989), Field (1991), Freeman et al. (1991), Kennedy (1996) Ranford et al. (1965), Wells et al. (1966), Wells (1981), Lindsay (1989)

Cryogenian

North and NE

Pioneer Sandstone, including ‘upper marker cap dolomite’

,170

Feldspathic sandstone, conglomerate, dolomite and chert; minor lonestones

Disconformity with underlying Aralka Fm, and conformity with overlying Pertatataka Fm; paraconformity between sandstone and overlying ‘upper marker cap dolomite’

Ellery Creek, 1.0-1.3 km south of contact with basement; lat. 238470 2400 S, long. 1338040 1200 E

Glacial outwash fan

Elatina Fm and Nuccaleena Fm equivalents

Cryogenian

NE

Olympic Formation, including ‘upper marker cap dolomite’

,200

Mudstone, siltstone, diamictite, sandstone and dolomite

Disconformity with underlying Aralka Fm, and conformity with overlying Pertatataka Fm; paraconformity between sandstone and overlying ‘upper marker cap dolomite’

8 km SE of Ringwood Homestead; lat. 238540 S, long., 1348590 E

Periglacial: fluvial to marginal marine

Elatina Fm and Nuccaleena Fm equivalents

Cryogenian

South

Inindia beds

,2200

Lower: quartz sandstone, siltstone, oolitic chert and some stromatolitic dolomite; Upper: sandstone and diamictite

Northern shore of Lake Amadeus; lat. 248380 S, long. 1308520 E

Lower part, marine; upper part fluvial

Elatina Fm and Sturt Tillite equivalents

Cryogenian

West

Boord Formation

,425

Lower: breccia, conglomerate and sandstone; Middle: diamictite; Upper: oolitic and stromatolitic carbonate and siltstone

Unconformity with underlying Bitter Springs Fm or Pinyinna beds, and unconformity with overlying Winnall beds Interfingers eastwards with Carnegie Fm, and disconformity with underlying Bitter Springs Fm

Boord Ridges, lat. 238480 S, long. 1288430 E

Fluvial, glacial and shallow/marginal marine

Elatina Fm and equivalents

AUSTRALIAN NEOPROTEROZOIC GLACIAL DEPOSITS

Stromatolitic dolomite

Wells et al. (1964, 1970), Wells (1981)

(Continued) 679

680

Table 67.1. Continued Period

Location

Unit

Ngalia Basin North

Cryogenian

North

Mount Doreen Formation (Mount Davenport Diamictite Member)

Ediacaran

North

Mount Doreen Formation (Wanapi Dolomite Member)

Cryogenian

Southern Georgina Basin West Boko Formation

Cryogenian

Central-west

Naburula Formation

Mount Cornish Formation

Lithology

,8

Diamictite, mudstone-siltstone and cap dolomite

,77

Diamictite, sandstone and minor dolomite

,4.3

,30

,365

Dolomite

Diamictite

Diamictite and siltstone; minor sandstone, arkose and dolomite; possible ‘lower marker cap dolomite’

Stratigraphic relationships

Type section

Depositional setting

Correlatives

References

Unconformity with underlying basement, Patmungala beds or Vaughan Springs Quartzite, and conformity with overlying Rinkabeena Shale Disconformity with underlying Rinkabeena Shale or Vaughan Springs Quartzite, and paraconformity with overlying Wanapi Dolomite Member Paraconformity with underlying Mount Davenport Diamictite Member and conformity with overlying Newhaven Shale Member

Eastern Naburula Hills, lat. 228170 3000 S, long. 1318190 3000 E

Subglacial, shallow-marine ice proximal

Sturt Tillite and equivalents

Eastern Naburula Hills, lat. 228170 S, long. 1318190 E

Fluvioglacial

Elatina Fm and equivalents

Preiss et al. (1978), Preiss & Forbes (1981), Wells (1981), Wells & Moss (1983), Young et al. (1995) Wells et al. (1972), Preiss et al. (1978), Wells (1981), Wells & Moss (1983), Young et al. (1995)

Eastern Naburula Hills, lat. 228170 S, long. 1318190 E

Shallow marine, or lacustrine (Walter & Bauld, 1983); deeper-water marine (Kennedy, 1996)

Nuccaleena Fm, Wonnadinna Dolostone, lower Boondawari Fm

Wells et al. (1972), Preiss et al. (1978), Wells (1981), Wells & Moss (1983), Young et al. (1995), Kennedy (1996)

34 km W of Barrow Creek; lat. 21831.10 S, long. 133833.50 E

Non-marine moraine (till)

Elatina Fm and equivalents

Haines et al. (1991, 2007), Dunster et al. (2007)

4 km WSW to 4.8 km SSW of Mt Cornish; lat. 228490 S, long. 1368270 E

Glacial and periglacial

Sturt Tillite and equivalents

Preiss et al. (1978), Walter (1980, 1981), Freeman (1986), Dunster et al. (2007)

Unconformity with underlying Amesbury Quartzite or basement, and disconformity with overlying Mopunga Group; ?local conformity with overlying Oorabra Arkose Inferred disconformity with underlying Yackah beds, and disconformity with overlying Oorabra Arkose

A. C. HILL ET AL.

Cryogenian

Thickness (m)

Central-west

Oorabra Arkose

,1165

Arkose, siltstone, shale, conglomerate; minor dolomite and sandstone

Cryogenian

Central

Yardida Tillite

,650

Diamictite, siltstone, locally capped with dolomite and dolomitic shale; minor quartz sandstone and arkose

Cryogenian

Central

Sun Hill Arkose

,300

Arkose, arkosic sandstone, shale, siltstone, dolomitic siltstone, conglomerate and arkosic dolomite

Base not exposed, and unconformity with overlying Early Cambrian Sylvester Sanstone

Cryogenian

Central

Black Stump Arkose

,700

Arkose, sandstone, mudstone and siltstone

Ediacaran

Central

Wonnadinna Dolostone

,460

Dolomite and sandy dolomite; minor arkosic dolomite, siltstone and shale

Cryogenian

East

Little Burke Tillite

,34

Disconformity with underlying Yardida Tillite, and paraconformity with overlying Wonnadinna Dolostone Paraconformity with underlying Black Stump Arkose, and disconformity with overlying Gnallan-a-Gea Arkose Unconformity with basement and overlying Early Cambrian Mount Birnie beds

Diamictite and cap dolomite

Disconformity with underlying Mount Cornish Fm, and disconformity with overlying Mopunga Group. Possible local conformity with underlying Boko Formation Inferred disconformity with underlying Yackah beds, and disconformity with overlying Black Stump Arkose

0.8 km NNE of Grant Bluff, Elua Range; lat. 22841.50 S, long. 1358460 E

Proximal fluvial glacial outwash

Elatina Fm and equivalents

Walter (1980), Preiss & Forbes (1981), Freeman (1986), Haines et al. (2007), Dunster et al. (2007)

Composite section: Outcrop in Field River Anticline and area 6 km SSE of Aroota Bore, drillholes Hay River 5 (lat. 238110 900 N, long. 1378540 1200 E), 6 (lat. 238100 4200 N, long. 1378520 4800 E), and section at end of seismic traverse 3 Sun Hill and low hills to N, 19 km SE of Glenormiston Homestead; lat. 2381.90 S, long. 138855.20 E Composite section: N flank Field River Anticline N of drillhole Hay River 7 (lat. 23860 1500 N, long. 1378510 1800 E)

Glacial and periglacial

Sturt Tillite and equivalents

Walter (1980, 1981), Dunster et al. (2007)

Glacial outwash

Elatina Fm and equivalents

Smith (1972), Walter et al. (1995), Dunster et al. (2007)

Proximal glacial outwash

Elatina Fm and equivalents

Walter (1980, 1981), Preiss & Forbes (1981), Walter et al. (1995), Dunster et al. (2007)

Composite section: drillhole Hay River 8 (lat. 23880 1200 N, long. 1378410 000 E) and nearby outcrop

Intertidal to deep marine

Preiss et al. (1978), Walter (1980, 1981), Dunster et al. (2007)

9 km NE of township of Duchess, on eastern bank of Little Burke River; lat. 21818.50 S, long. 1398560 E

Marine till

Nuccaleena Fm, Wanapi Dolomite Member, lower Boondawari Fm Elatina and Nuccaleena Fm equivalents

AUSTRALIAN NEOPROTEROZOIC GLACIAL DEPOSITS

Cryogenian

de Keyser (1972), Plumb (1981), Dunster et al. (2007), Raub (2011)

References in bold type are the source(s) for type sections.

681

682

A. C. HILL ET AL.

126o

122o

(a)

130o

Northern Territory

Western Australia

1

134o 23o

NW OFFICER BASIN 25o

GEORGINA BASIN

NGALIA BASIN

Musg rave P rovinc e

CENTRAL OFFICER BASIN

AMADEUS BASIN

4

Adelaide Rift Complex

6

3

2

OFFICER BASIN

South Australia

5

EASTERN OFFICER BASIN

500 km

27

o

o

29

100 km

KEY Major sub-basins Approximate extent of glacial outcrop

(c)

Type section/area

135

o

Major thrust faults

(b)

6,7

o

129

132o

135

141o

138o

Northern Queensland Territory

16o

o

18o

NGALIA BASIN o

Arunta Province

23

1,2

5

20o

3 o

24

1 4 AMADEUS BASIN Mus

o

25

grav

Western Australia

eP

3

rovin

ce

6 4

100 km

100 km

Northern Territory

7

SOUTHERN GEORGINA 2 BASIN

22o

5 24o

o

26

Fig. 67.1. Neoproterozoic basins of central Australia, approximate outcrop extent of Neoproterozoic glaciogenic units and location of glaciogenic unit type sections/ areas (general map after Walter et al. 1995). (a) Officer Basin (Grey et al. 1999, 2005; Morton 1997): type sections, 1, Boondawari Fm. (Williams & Tyler 1991; Williams 1992; Grey et al. 2005); 2, Wahlgu Fm. (Grey et al. 2005); 3, Pirrilyungka Fm. (Haines et al. 2008); 4, Lupton Fm. (Jackson & van de Graaff 1981; Grey et al. 1999, 2005); 5, Turkey Hill Fm. (Jackson & van de Graaff 1981; Grey et al. 2005); 6, Chambers Bluff Tillite (Preiss 1987, 1993; Morton 1997; Preiss et al. 2011). (b) Amadeus and Ngalia Basins (Wells 1981; Lindsay 1989): type sections, 1, Areyonga Fm. (Prichard & Quinlan 1962; Preiss et al. 1978); 2, Pioneer Sandstone (Preiss et al. 1978); 3, Olympic Fm. (Wells et al. 1967; Preiss et al. 1978); 4, Inindia beds (Ranford et al. 1965; Wells 1981); 5, Boord Fm. (Wells et al. 1964; Wells 1981); 6, Naburula Fm. (Preiss et al. 1978); 7, Mount Davenport Diamictite Member, Mount Doreen Fm. (Wells et al. 1972; Preiss et al. 1978). (c) Southern Georgina Basin (Walter 1981; Dunster et al. 2007): type sections, 1, Boko Fm. (Haines et al. 1991, 2007); 2, Mount Cornish Fm. (Walter 1980); 3, Oorabra Arkose (Walter 1980; Haines et al. 2007); 4, Yardida Tillite (Walter 1980); 5, Sun Hill Arkose (Smith 1972); 6, Black Stump Arkose (Walter 1980); 7, Little Burke Tillite (de Keyser 1972).

shallow platform on the southern margin and deeper, troughconnected sub-basins along the north or northeastern margins (Fig. 67.1). Up to 15 km of Neoproterozoic sedimentary rocks are present in the Amadeus Basin (Lindsay & Korsch 1991), c. 10 km in the Officer Basin (Jackson & van de Graaff 1981; Lindsay & Leven 1996), c. 5 km in the Ngalia Basin (Wells & Moss 1983; Deckleman 1995), and c. 2 km in the southwestern Georgina Basin (Dunster et al. 2007). Basin structure is further complicated by a history of compression and extension and by halotectonics (Lindsay & Korsch 1991; Lindsay & Leven 1996; Hand & Sandiford 1999; Lindsay 2002; Simeonova & Iasky 2005; Young & Ambrose 2007; Dyson & Marshall 2007; Dunster et al. 2007). Outcrop is largely confined to narrow basin-marginal zones near major thrust faults (Fig. 67.1); hence drill holes and geophysical interpretation play a major role in basin analysis.

Stratigraphy A summary of Neoproterozoic stratigraphy and correlations of non-glacial and glacial formations in central Australia is shown in Figure 67.2. For stratigraphic details of individual basins see the references cited in the text.

Glaciogenic deposits and associated strata The order of formations in this section and in Table 67.1 is based firstly on subregional grouping (e.g. northwestern Officer Basin, northeastern Officer Basin, etc.), and then by the age of the stratigraphic units with the older glacial units described first.

AUSTRALIAN NEOPROTEROZOIC GLACIAL DEPOSITS

683

Fig. 67.2. Australian Neoproterozoic stratigraphy and correlations assuming a c. 635 Ma age (Hoffmann et al. 2004; Condon et al. 2005) for the Elatina Fm. and equivalents (see Williams et al. 2008 for an extended discussion of age constraints). The length of time breaks at major unconformities and sequence boundaries are estimates only. This figure is adapted from Preiss (1987, 2000), Walter et al. (1995), Morton (1997), Grey et al. (2005) and Dunster et al. (2007).SS, Supersequence number, Centralian Superbasin (Walter et al. 1995); CDE, Coominaree Dolomite equivalent; CVE, Cadlareena Volcanics equivalent; MCF, Mount Cornish Fm.; MDDM, Mount Davenport Diamictite Member; ?UGE, possible Umberatana Group equivalents; WDM, Wanapi Dolomite Member; WF/THF/LF, Wahlgu Fm./Turkey Hill Fm./Lupton Fm.; WV, Wantapella Volcanics; YCE, Younghusband Conglomerate equivalent.Geochronology: 1, 827 + 6 Ma (Wingate et al. 1998) and 824 + 4 Ma (Glikson et al. 1996), Gairdner Dyke Swarm; 2, 802 + 10 Ma, Rook Tuff (Fanning et al. 1986); 3, 797 + 5 Ma, Skillogalee Dolomite (Drexel 2008); 4, 725 + 11 Ma maximum depositional age (detrital zircon), uppermost Kanpa Fm. (Nelson 2002); 5, 659 + 6 Ma, Wilyerpa Fm. (Sturt Tillite correlative) (Fanning & Link 2008); 6, 657.2 + 5.4 Ma, Aralka Fm. (Kendall et al. 2006); 7, 647.2 + 10 Ma and 645.1 + 4.8 Ma, Tindelpina Shale Member (Kendall et al. 2006); 8, 657 + 17 Ma maximum depositional age (detrital zircon), Marino Arkose Member (Ireland et al. 1998); 9, 556 + 24 Ma maximum depositional age (detrital zircon), Bonney Sandstone (Ireland et al. 1998).

684

A. C. HILL ET AL.

Western Officer Basin Boondawari Fm. (NW). The Boondawari Fm. is a thick, poorly exposed unit in the northwestern Officer Basin (Fig. 67.1, Table 67.1). The lower Boondawari Fm. is a lateral equivalent of the Wahlgu Fm. (Williams 1992, 1994; Walter et al. 1994; Grey et al. 1999, 2005) and is dominated by diamictite, which consists of a dark grey to purple mudstone containing numerous polished, striated and faceted clasts that range in size from pebble to boulder, some up to 3.5 m in diameter. Clasts comprise over 20 lithological types including metamorphic, igneous and a wide variety of sedimentary rocks. There are lenticular interbeds of feldspathic sandstone and very poorly sorted conglomerate, up to 5 m thick. The diamictite grades laterally into pebbly feldspathic sandstone and conglomerate. In places, rhythmites above the diamictite resemble those from the Elatina Fm. (Williams 1985, 1989, 1991). Overlying the diamictite is a thin-bedded pink dolomite unit that may be equivalent to dolomites capping other Elatina glacial units. Only the lower part of the unit is considered to be glacial in origin. The two upper units of the Boondawari Fm. are lithologically similar to the Pertatataka Fm. and Wilpena Group and a carbonate at the top of the formation is correlated with the Julie Fm. and basal Bonney Sandstone (Walter et al. 1994; Grey et al. 2005).

Wahlgu and Boondawari formations. However, there are differences in lithology and its isolation prevents direct correlation. In the type section (Lupton Hills), the unit comprises up to 250 m of diamictite, conglomerate, sandstone and mudstone, with rare striated and faceted clasts. Turkey Hill Fm. (SW). The Turkey Hill Fm. (Jackson & van de Graaff 1981; Grey et al. 2005) is exposed in the southwestern part of the western Officer Basin and is at least several hundred metres thick. Outcrops in the type area include sandstone, siltstone, mudstone and diamictite, but it is unclear if the diamictite is part of the Turkey Hill Fm. or overlying Permian glaciogenic Paterson Fm. Jackson & van de Graaff (1981) reported that the diamictite is structurally related to the moderately to steeply dipping sandstone, siltstone and mudstone beds, and thus grouped all facies in the Turkey Hill Fm. They also proposed a correlation of the Turkey Hill Fm. with the Lupton Fm. However, Grey et al. (2005) ‘did not recognise diamictite that was unequivocally within the dipping succession’, and concluded that the diamictite could be part of the overlying Permian Paterson Fm.

Eastern Officer Basin

Wahlgu Fm. (central and NE). The Wahlgu Fm. is known only from

Chambers Bluff Tillite (NE). The type area of the Chambers Bluff

drill holes (Eyles & Eyles 1998; Grey et al. 1999, 2005; Apak & Moors 2001; Haines et al. 2004, 2008). It varies in thickness between 122 m in Lancer 1 and 478.5 m in Vines 1, and is dominated by red-brown to brownish grey sandy diamictite containing a wide variety of clast types including diverse sedimentary rocks, red jasper, reddish volcanic rocks and granite. Mafic igneous clasts are common in Lancer 1. Minor lithologies include pebbly and sandy mudstone, moderately sorted sandstone, mudstone and conglomerate, and well-sorted and normally graded sandstone to mudstone. Dropstone fabrics can be discerned in places. The interval 317.1– 318.7 m in Empress 1 and 1A is interpreted by Grey et al. (1999) as a possible ‘cap carbonate’ unit at the top of the Wahlgu Fm. It consists of interlaminated and interbedded fineto coarse-grained sandstone, mudstone and a 15-cm-thick dolomitic horizon that resembles the ‘upper marker cap dolomite’ of the Pioneer Sandstone in the Amadeus Basin, particularly that at Ellery Creek described by Preiss et al. (1978).

Tillite is in the far northeastern corner of the Officer Basin where it is up to 610 m thick (Preiss 1987, 1993; Morton 1997), and a reference section has been described in drill hole Nicholson 2 (Morton 1997; Eyles et al. 2007) (Fig. 67.1). The dominant lithologies are diamictite, siltstone and sandstone. Diamictite clasts consist of a large variety of lithologies and are often striated. In the type area a 1.2-m-thick dolomite bed occurs near the top of the formation. The Chambers Bluff Tillite is discussed in detail in a separate chapter (Preiss et al. 2011).

Pirrilyungka Fm. (NE). The Pirrilyungka Fm. (1238 m thick) is

known only in drill hole Vines 1 (Fig. 67.1, Table 67.1), where it underlies the Wahlgu Fm. (Haines et al. 2008). The formation comprises an upper unit (821 m thick) dominated by diamictite and conglomerate with lesser sandstone and mudstone, including mudstone and siltstone rhythmite, and a lower unit (417 m thick) dominated by interbedded sandstone and mudstone with minor diamictite. Diamictite throughout is predominantly grey to black, with scattered brown to reddish intervals c. 2 m thick. Clasts are mostly light grey carbonate, fine clastic rocks and chert, and a few volcanic or granitic clasts. Rare striated or faceted clasts, most commonly of fine-grained carbonate, are present throughout the diamictite intervals, as are rare clasts of stromatolitic dolostone and probable magnesite, resembling upper Buldya Group (Fig. 67.2) lithologies. Intervals of laminated chocolate-brown mudstone and siltstone (rhythmite) several metres thick are present in the upper unit. The lower unit has massive sandstone intervals (c. 2– 2.5 m thick) commonly with dish-like dewatering structures, normal and reverse grading, abundant slumping, shale intraclasts, sandstone –shale alternations, and massive diamictite units with a shaly matrix. Lupton Fm. (NE). The Lupton Fm. (Jackson & van de Graaff 1981; Grey et al. 1999, 2005) crops out in the western Officer Basin south of the Musgrave Province and is possibly a lateral equivalent of the

Elatina Fm. equivalent(?) (NE). In the type area of the Chambers Bluff Tillite there is up to 30 m of red-brown, granule-bearing sandstone that has been tentatively correlated with the Elatina Fm. (Preiss 1993; Preiss et al. 2011). Nuccaleena Fm. equivalent (?) (NE). Overlying the possible Elatina Fm. equivalent in the Chambers Bluff Tillite type area is a thin stromatolitic dolomite unit that has been very tentatively correlated with the Nuccaleena Fm. (Preiss 1993; Preiss et al. 2011).

Amadeus Basin Areyonga Fm. (north and NE). The Areyonga Fm., the most wide-

spread of the Sturt glacial units in the Amadeus Basin (Fig. 67.1), consists largely of diamictite of variable composition and texture with thinner interbeds of sandstone, conglomerate and carbonate (Preiss et al. 1978; Lindsay 1989; Walter et al. 1994, 1995). Diamictite units range in thickness from ,1 m to .40 m, with thicker units generally nearer the base of the formation (Lindsay 1989). Clast types are variable and derived from both intrabasinal (carbonate, chert, quartzite and conglomerate) and extrabasinal (granite, gneiss, porphyry, volcanic rocks and schist) sources (Lindsay 1989). Clast size and shape varies considerably throughout the formation, from 10 cm to 2 m in diameter, but there is a trend for clast size and abundance to decrease upwards and for clasts to be more rounded where there is a greater abundance of chert clasts (Lindsay 1989). Striated clasts are rare (,1%) but appear throughout the formation, and their distribution seems to be restricted to fine-grained silty sandstones (Lindsay 1989). There are rare dropstones and minor siltstone. The matrix is more or less uniform in composition, being massive, poorly sorted and generally grey-green in outcrop or dark grey to black in drill holes. The diamictite is capped by a

AUSTRALIAN NEOPROTEROZOIC GLACIAL DEPOSITS

dark, thinly laminated silty dolomite – the ‘lower marker cap dolomite’ of Preiss et al. (1978) – which is widespread in the basin. Pioneer Sandstone (north and NE). The Pioneer Sandstone crops out mainly in the northern –northeastern Amadeus Basin (Preiss et al. 1978; Shaw & Wells 1983; Field 1991; Freeman et al. 1991; Williams et al. 2007), but a 9 m section was intersected in drill hole Wallara 1 in the central part of the basin (Indigo Oil & Sirgo Exploration 1990), so its real extent is unknown. It consists mostly of feldspathic, medium- to coarse-grained, cross-bedded sandstone and conglomerate. In the lower part of the formation, medium to coarse-grained sandstone, with decimetre amplitude low-angle foresets and grit lags, grades upwards into fine-grained sandstone, with lags and lenses of medium to coarse-grained sandstone. Lonestones are locally present (Field 1991). The upper parts of the formation have centimetre-amplitude, bimodal, tabular foresets and tidal channel-fill sands, which are upward-fining, and coarse- to fine-grained, with imbricated bladed clasts of carbonate and sandstone in channel bases (Field 1991). The uppermost Pioneer Sandstone becomes increasingly dolomitic, and in places there is a discrete dolostone – the ‘upper marker cap dolomite’ (Preiss et al. 1978) – which is 1 –5 m thick, massively to weakly laminated, chert-rich and stromatolitic (Preiss et al. 1978; Walter & Bauld 1983; Field 1991; Kennedy 1996; Skotnicki et al. 2008). Erosional scours with a relief of up to 20 cm in the uppermost sandstone are filled with 10-cm-thick pink dolomite lenses containing the conspicuous columnar stromatolite Anabaria juvensis ¼ Elleria minuta that grew sometime between deposition of the Pioneer Sandstone and overlying Pertatataka Fm. (Cloud & Semikhatov 1969; Walter et al. 1979; Williams et al. 2007; Skotnicki et al. 2008). Chert occurs as either angular nodules between 1 and 10 cm across or beds up to 10 cm thick, and is often shattered and brecciated (Skotnicki et al. 2008). Olympic Fm. (NE). The Olympic Fm., restricted to the northeastern

Amadeus Basin, consists of red and green mudstone and siltstone, diamictite, sandstone, conglomerate and dolomite (Preiss et al. 1978; Lindsay 1989; Field 1991). The most distinctive lithology is a reddish diamictite that contains striated and faceted clasts and dropstones. Clasts vary in composition and include metamorphic, igneous and sedimentary lithologies. In some places, the diamictite contains boulders up to 2 m in diameter (Lindsay 1989). Red and green siltstone and mudstone with intercalated sandstone are the predominant lithology. Siltstone and mudstone are very fine-grained, laminated, normal- to reverse-graded, or massive (Field 1991). Sandstone interbeds contain trough and tabular foresets, low-angle swash stratification and symmetrical wave ripples, and have a high dolomite content. Poorly sorted decimetre-scale beds of sandy pebble conglomerate sometimes occur as channel fill and vary from clast- to matrix-supported. The clasts are mainly fine to coarse pebbles but range up to boulder size in the upper parts of the formation. Clasts in the conglomerate are mostly rounded and consist of quartzite, granite, gneiss, carbonate and basalt; some are faceted. Where the dominant clast lithology is carbonate, there are rare lonestones. Dolomite generally overlies the conglomerate facies, but can be interbedded, and is diagenetically altered and silicified at most localities. It comprises millimetre-bedded reddish grey and grey silty dolomite, or a coarse sandy dolomite with platy rip-up clasts of dolomite, coarse quartz stringers and chert nodules, and contains domical and columnar stromatolites. Overlying the Olympic Fm. in the type area is a 0 –30-m-thick, laminated to thinly bedded pure dolomite unit: the ‘upper marker cap dolomite’ (Preiss et al. 1978; Kennedy 1996). In the type section, where it is c. 4 m thick, the unit fines upward from siltyparted dolomicrite to pure dolomicrite (Kennedy 1996). In the

685

uppermost part of the unit, overlying an erosional surface, is a distinctive bed of discrete columnar and laterally linked columnar stromatolites. Individual columns are 2–20 mm in diameter and the stromatolite bed is about 10 cm thick (Kennedy 1996). Barite is commonly associated with the stromatolites, either as isolated bladed crystals or as upward-oriented rosettes of bladed crystals up to 4 cm high. Inindia beds (south). The Inindia beds crop out in the southern

Amadeus Basin and consists mostly of sandstone (beds up to 300 m thick) with major siltstone interbeds (up to 55 m thick) (Wells et al. 1966, 1970; Wells 1981; Lindsay 1989). There is also diamictite, and minor carbonate and chert beds in the lower part of the formation. Sandstone beds are mainly laminated to thinly bedded and fine- to medium-grained; the upper sandstone units are cross-bedded and kaolinite, and glauconite and phosphate are found in the thicker units. Siltstone beds are laminated to thinly bedded and are often interbedded within sandstone units. Diamictite is present in the lower and uppermost parts of the Inindia beds where it is associated with cross-bedded sandstone. The diamictite is yellow-brown, non-bedded, coarse- to finegrained and contains a mixed assemblage of clasts up to 12 mm in diameter. The clasts consist of mostly intrabasinal lithologies (quartzite, chert, jasper, siltstone, dolomite and several varieties of metamorphic rocks), and some are striated. There appears to be an abrupt facies change (possible sequence boundary) beneath the thick, cross-bedded, sandstone and interbedded diamictite units in the upper Inindia beds (Lindsay 1989). This raises the possibility of separate glaciogenic successions; the younger may even represent the Elatina glaciation. Boord Fm. (west). The Boord Fm. is a poorly exposed unit in the northwestern Amadeus Basin, which consists of calcilutite, calcarenite, dolomitic limestone, diamictite, sandstone and siltstone and conglomerate (Fig. 67.1) (Wells et al. 1964, 1970; Wells 1981; Lindsay 1989; Walter et al. 1995). The basal part of the unit consists of poorly sorted breccia and conglomerate. Clasts range in size up to 2.5 m and consist mainly of lithologies from the underlying Bitter Springs Fm. Above the basal unit, diamictite contains clasts of sedimentary and metamorphic origin in a medium-grained sandstone matrix. The clasts are rounded to angular, many are faceted and some striated. In one section, pebble and boulder size clasts of quartz, sandstone, porphyry and schist are rounded, while boulders of carbonate are angular (Wells et al. 1964). The upper part of the formation is partly oolitic, with grey and pink stromatolitic calcilutite and calcarenite, interbedded with siltstone and shale.

Ngalia Basin Naburula Fm. (north). The Naburula Fm. is a thin, poorly outcrop-

ping unit that comprises a basal diamictite, overlain by interbedded shale-siltstone and dolomite (Preiss et al. 1978; Wells 1981; Preiss & Forbes 1981; Wells & Moss 1983; Young et al. 1995). The diamictite (2 –3 m thick) has a poorly sorted mudstone – siltstone matrix containing angular feldspar and quartz grains. The matrix varies between green-brown and red (ferruginized). Diamictite clasts are 0.3– 1 m in diameter, and consist of several varieties of quartz (including vein quartz with tourmaline), quartzite, grey dolomite, glauconitic sandstone, silicified yellowgrey siltstone and a variety of basement igneous (including granite and feldspar porphyry) and metamorphic rocks (including quartz-mica schist and spotted blue-grey hornfels). The clasts are commonly striated and faceted. The overlying shale and siltstone beds are dark grey to black and well-bedded. The dolomite beds (c. 20 cm thick) are thinly bedded and fine-grained, deeply weathered, and iron stained.

686

A. C. HILL ET AL.

Mount Davenport Diamictite Member, Mount Doreen Fm. (north). The

Mount Davenport Diamictite Member is a 77-m-thick diamictite unit that crops out in the northern part of the basin, and contains pebble, cobble, and rare boulder-sized (up to 4 m in diameter) erratics, which are sub-rounded, striated and faceted (Wells et al. 1972; Preiss et al. 1978; Wells & Moss 1983; Young et al. 1995). Many erratics have weathered out of the diamictite matrix and occur as surface float. Clasts show a wide variety of lithologies and include, in decreasing abundance, granite, granite gneiss, quartzite, schist, metamorphosed sandstone, dolomite, hornfels, metamorphosed mafic igneous rocks, and rare pink stromatolitic dolomite partially replaced by jasper (Wells & Moss 1983; Young et al. 1995). The smaller erratics are schist or weathered gneiss and the larger ones are predominantly granite and quartzite (Young et al. 1995). The matrix consists of blue-green and some red-brown, poorly sorted siltstone containing angular quartz granules. It is dolomitic in places, especially where dolomite clasts are abundant. Thin lenses of poorly sorted pebbly dolomitic sandstone and angular feldspathic sandstone, occur near the base of the diamictite and some bedding planes show soft sediment deformation structures (Young et al. 1995). Wanapi Dolomite Member, Mount Doreen Fm. (north). The Wanapi Dolomite Member is exposed in the northern Ngalia Basin and small exposures are present along the northwestern margin of the basin where it is 3–4.3 m thick. It consists predominantly of finegrained, pink, laminated dolomite containing small grains and dendrites of manganese oxide and pyrite pseudomorphs (Wells & Moss 1983). Other sedimentary features include stromatolites, breccias and barite pseudomorphs after gypsum and anhydrite (Walter & Bauld 1983; Wells & Moss 1983). Barite also occurs as 5 mm crystal fans within Fe-rich stromatolite domes at the top of the member overlying an erosional surface (Kennedy 1996), and there are some bands of scattered sand- and granule-sized quartz grains and fragments of banded chert (Young et al. 1995). Locally there are slump breccias consisting of angular fragments of laminated dololutite and rare quartzite pebbles and hematitic chert nodules (Young et al. 1995).

Southern Georgina Basin Boko Fm. (west). The Boko Fm. is a massive diamictite, with cobble and boulder clasts up to 2 m in diameter in a redbrown mudstone matrix (Haines et al. 1991, 2007). The formation is lenticular, only locally preserved, and reaches an estimated 30 m in thickness. Clasts comprise quartzite, sandstone, conglomerate, granite, various metamorphic rocks and rare volcanic rocks. They commonly display facets and striations, which distinguishes the formation from conglomeratic units within the Oorabra Arkose. Diamictite units previously mapped as basal ‘Central Mount Stuart beds’ (Shaw & Warren 1975), and subsequently recognized as a distinct, unnamed unit (Walter 1980), have been defined as the Boko Fm. (Haines et al. 2007; Dunster et al. 2007). Mount Cornish Fm. (central-west). The Mount Cornish Fm., a lateral equivalent of the Yardida Tillite in the central-western area of the southern Georgina Basin, consists of poorly bedded blue-green diamictite with interbeds of green varve-like siltstone, minor sandstone, arkose and dolostone (Walter 1980, 1981; Freeman 1986; Dunster et al. 2007). Clasts are up to 1 m in diameter and lithologies consist of gneiss, pegmatite, granite, dolerite, orthoquartzite and dolomite. Some clasts are faceted and striated. As in the Yardida Tillite, the siltstone laminae are graded and have small linear ripples. Yardida Tillite (central). The Yardida Tillite, which crops out in the

central-southern Georgina Basin, consists of green-grey diamictite and laminated siltstone with minor fine to very coarse (pebbly)

grained quartz sandstone and arkose (Walter 1980, 1981; Dunster et al. 2007). Locally the diamictite is capped by grey, laminated dolomitic shale and lenticular dolomite. Laminated siltstone beds within the formation are graded and many have minute linear ripple marks. Coarse clasts up to 1 m in diameter are common, rounded to well-rounded and frequently faceted and striated (Walter 1981). Clast lithologies include dolomite, granite, gneiss, schist, metaquartzite, quartz-feldspar porphyry and conglomerate. Thin interbeds of cross-laminated pebbly arkose seem to be lenticular (Walter 1981). Oorabra (central-west), Sun Hill (central) and Black Stump Arkoses (central). In the southern Georgina Basin, the Sun Hill (eastern

area), Black Stump (eastern to central areas) and Oorabra (central to western areas) arkoses are probable equivalents, and consist of poorly sorted, medium- to coarse-grained and pebbly arkose, with interbedded sandstone, laminated micaceous siltstone and shale and conglomerate (Walter 1980, 1981; Shergold 1985; Freeman 1986; Dunster et al. 2007; Haines et al. 2007). The Black Stump Arkose contains abundant cross-stratified and rippled sands, and the sediments are immature, with abundant detrital mica and angular feldspar grains (Walter 1980; Dunster et al. 2007). The Oorabra Arkose has a basal conglomerate with faceted and striated clasts (Freeman 1986; Haines et al. 2007). The clasts are dominated by quartzite, but also include granite and vein quartz. Wonnadinna Dolostone (central). The Wonnadinna Dolostone overlies the Black Stump Arkose, and consists of purple-red to yellowbrown and green-grey dolomite that is sandy in parts (Walter 1980, 1981; Dunster et al. 2007). Sedimentary structures include oncoids, fenestrae and possible columnar stromatolites. The dolomite is locally interbedded with red-brown, purple and green-grey arkose and siltstone, some of which is dolomitic. Little Burke Tillite (east). The Little Burke Tillite is exposed in the far southeastern Georgina Basin and consists of up to 34 m of massive diamictite (de Keyser 1972; Dunster et al. 2007). Clasts are mostly rounded to sub-rounded, faceted, striated and polished, and consist of a large variety of lithologies sourced from basement. Clast sizes range from sand size up to boulders 1.5 m in diameter. The matrix is purple-brown in colour. The Little Burke Tillite is also overlain by a dolomite.

Boundary relationships with overlying and underlying non-glacial units The boundary relationships of glacial and associated non-glacial formations are summarized in Table 67.1.

Chemostratigraphy For Sturt glacial units there are d13Ccarb data only from the Areyonga Fm. (Walter et al. 2000). d13Ccarb values range between 0.8 and 2.7‰, except in the ‘lower marker cap dolomite’ (Preiss et al. 1978) immediately below the contact with the Aralka Fm. where there is one d13C value of –4.9‰. In this same section, in the Limbla Syncline, 13C-depleted values ( –5.9 to –1.5‰) continue for another 60 m up-section into the Aralka Fm. For Elatina-equivalent glacial units there are data from dolomite beds within the Olympic Fm. (d13C, 0 –3.5‰; Kennedy et al. 2001; Skotnicki et al. 2008) and Pioneer Sandstone (d13C, – 2 to 2.8‰; Calver 1995; Skotnicki et al. 2008) where there is an upward increasing trend. For overlying cap dolomites there are data from the Wahlgu Fm. (d13C, –0.8 to – 0.9‰; Walter & Hill 1999), the ‘upper marker cap dolomite’ in the Olympic Fm. –Pioneer Sandstone (d13C, þ1.0 to – 4.0‰; Calver 1995; Kennedy 1996; Kennedy et al. 2001; Skotnicki et al. 2008), and the Wanapi

AUSTRALIAN NEOPROTEROZOIC GLACIAL DEPOSITS

Dolomite Member of the Mount Doreen Fm. (d13C, – 2.5 to –5‰; Kennedy 1996). These d13C values are similar to those from the Adelaide Rift Complex (Calver 1995, 2000; Kennedy 1996; McKirdy et al. 2001) and Kimberley Region (Kennedy 1996; Corkeron 2007). 87Sr/86Sr ratios from cap dolomites in the Amadeus and Ngalia Basins are .0.7090 (Calver 1995; Kennedy 1996), and thus not considered to be primary seawater signatures.

Other characteristics A minor gas show was reported from the Pirrilyungka Fm. (western Officer Basin) during drilling (Apak et al. 2002; Haines et al. 2008). Poorly preserved fragments of the acritarch Cerebrosphaera buickii were recovered throughout the Pirrilyungka Fm. in Vines 1 (Haines et al. (2008) and in the Chambers Bluff Tillite in Nicholson 1 (Eyles et al. 2007). Badly fragmented specimens in diamictite are probably reworked, but slightly better preserved and more complete specimens in the lower 200 m of Vines 1 could be in situ. Elsewhere, Cerebrosphaera buickii is not present (except as reworked fragments) in the Sturt glacial units or in the successions overlying them (Hill et al. 2000; Grey et al. 2005; Haines et al. 2008). The columnar stromatolite Anabaria ¼ Kotuikania juvensis, and its probable synonym Elleria minuta, is widespread in the cap carbonate of the Pioneer Sandstone of the northeastern Amadeus Basin, where it tends to grow on local erosion surfaces (Walter et al. 1979; Williams et al. 2007; Skotnicki et al. 2008). The same stromatolite appears to be present in the carbonate above the Wahlgu Fm. diamictite (Grey et al. 2005), so could be a useful stratigraphic marker in central Australian basins.

Palaeolatitude and palaeogeography Low palaeolatitudes (,108) were recorded in the Elatina Fm. of the Adelaide Rift Complex (McWilliams & McElhinny 1980; Embleton & Williams 1986) and later confirmed by Schmidt & Williams (1995), Sohl et al. (1999) and Williams (2008). Similar palaeolatitudes were reported from the Centralian Superbasin, both in glaciogenic units and the underlying succession in the western Officer Basin (Pisarevsky 2001; Pisarevsky et al. 2001, 2007). At the beginning of the Neoproterozoic, Australia was part of the supercontinent Rodinia and at some time during the Neoproterozoic it separated from North America, but its position relative to the other various cratons and the break-up model is still debated (see Li et al. 2008, for a synthesis of all alternative Rodinia break-up models and palaeogeographies). There is also no consensus on when Australia separated from Rodinia, with evidence for a 780 Ma break-up (Wingate et al. 1998; Wingate & Giddings 2000) or a 600– 580 Ma break-up or second rifting event (Crawford et al. 1997; Veevers et al. 1997; Foden et al. 2001).

Geochronological constraints There are only two geochronological constraints for glacial units in the central Australian basins. One is a detrital zircon U –Pb SHRIMP maximum depositional age of 725 + 11 Ma from the upper Kanpa Fm., c. 75 m below the karstified and truncated top of the Buldya Group in Empress 1A (Nelson 2002). The other is a Re –Os date of 657.2 + 5.4 Ma for organic-rich (TOC ¼ 0.5– 1.0%) shales of the lower Aralka Fm. in drill hole Wallara 1 (Kendall et al. 2006). This date is significantly older than another Aralka Fm. Re –Os date of 592 + 14 Ma obtained by Schaefer & Burgess (2003), which Kendall et al. (2006) claimed was an artefact of the different analytical protocol employed. The Aralka Fm. age is consistent with a U –Pb zircon date of 659 + 6 Ma on an ash bed within the Wilyerpa Fm., a Sturt

687

glacial unit in the Adelaide Rift Complex (Fanning & Link 2008). Preiss et al. (2011) consider that the dated volcaniclastic bed was deposited during the waning phase of the Sturt glaciation. This date is consistent not only with the Aralka Fm. date but also Re –Os dates from post-Sturt glacial shale units that paraconformably overlie diamictites in the Adelaide Rift Complex (643 + 2.4 Ma; Kendall et al. 2006) and NW Tasmania (640.7 + 4.7 Ma; Kendall et al. 2007). All these radiometric dates are consistent with ages from early Neoproterozoic intrusive and volcanic units in Australia (see references in Fig. 67.2 caption).

Discussion A glacial origin for diamictites of the Officer, Amadeus, Ngalia and Georgina Basins is confirmed by the lithological diversity of clasts, some of which are striated, faceted and polished, and the presence of dropstones. Diamictites and interbedded siltstones of the Sturt glaciation are usually, but not universally grey-green (outcrop) and dark-grey to black (drill holes) in colour, in contrast to the typically oxidized, red-brown colour of Elatina glacial units in all but deeper basinal sections. In the Officer (Wahlgu Fm., Walter & Hill 1999), Amadeus (‘upper marker cap dolomite’ of the Pioneer Sandstone and Olympic Fm., Calver 1995, 2000; Kennedy 1996; Kennedy et al. 2001; Skotnicki et al. 2008), Ngalia (Wanapi Dolomite Member, Kennedy 1996) and Georgina basins (cap dolomite unit overlying the Little Burke Tillite) there are units of the Elatina glaciation paraconformably overlain by cap dolomites that have similarly 13C-depleted d13C ratios, which supports lithostratigraphic correlation. Sturt glacial units in the Amadeus (Areyonga Fm.), Ngalia (Naburula Fm.) and southern Georgina Basins (Yardida Tillite and Mount Cornish Fm.) are ‘capped’ by thin, silty dolomite beds – the ‘lower marker cap dolomite’ (Preiss et al. 1978) – that are sedimentologically distinct from the Elatina cap dolomite units. The ‘lower marker cap dolomite’ of the Areyonga Fm. has 13C-depleted d13C ratios (Walter et al. 2000), similar to other ‘lower marker cap dolomite’ units in the Adelaide Rift Complex (McKirdy et al. 2001), Stuart Shelf (Walter et al. 2000; McKirdy et al. 2001) and Tasmania (Calver 1998). Only the Elatina glaciation is known from the northwestern (Boondawari Fm.), central (Wahlgu Fm.) and southwestern (Turkey Hill Fm.) parts of the western Officer Basin. Diamictite of the lower Boondawari Fm. is interpreted to have been deposited in a shallow marine environment, distant from the ice source (Williams 1987, 1992). The Wahlgu Fm., in this part of the basin, has been interpreted as glaciomarine (Eyles & Eyles 1998), but in the northeastern part of the basin it has been interpreted as fluvioglacial because of the presence of conglomerates and the sandy character of the diamictite (Haines et al. 2008). Eyles & Eyles (1998) concluded that the clastic sediments were deposited rapidly by debris flows and turbidity currents in a shallow marine environment, with later channelling and reworking of the sediments to form well-sorted and stratified deposits. Diamictite of the Turkey Hill Fm. could have been deposited under floating ice shelves or deposited as mass flow deposits unrelated to glaciation, and could alternatively be of Permian age (Grey et al. 2005). In the northeastern part of the western Officer Basin, south of the Musgrave Province, there are records of both Sturt (Pirrilyungka Fm.) and Elatina (Wahlgu and Lupton formations) glacial units. In the sandstone –mudstone-dominated lower unit of the Pirrilyungka Fm. there are sedimentary facies (e.g. normal and reverse graded beds, slump structures, diamictite with shaly matrix) that are suggestive of a relatively deep-water (?marine) mass-flow to turbidity current setting (Haines et al. 2008). In the upper unit of the Pirrilyungka Fm., clast-supported pebble conglomerate beds suggest a shallow marine or continental environment of deposition (Haines et al. 2008). Rhythmites in the upper unit have been interpreted as distal turbidites (Apak et al. 2002;

688

A. C. HILL ET AL.

Stevens et al. 2002); however, the extreme regularity of the laminations and rare dropstone fabrics suggest they could be glacial varves (Haines et al. 2008). The depositional environment of the Lupton Fm. is probably glaciogenic, but there is only limited evidence for glacial activity (Grey et al. 1999) and the sediments could have been deposited in alluvial fan and fluvial environments (Grey et al. 2005). In the northeastern area of the eastern Officer Basin there are Sturt-equivalent glacial deposits (Chambers Bluff Tillite) and possible Elatina-equivalent glacial deposits, which are overlain by a dolostone unit inferred to be equivalent to the Nuccaleena Fm.. The Chambers Bluff Tillite is interpreted to have been deposited in both fluvio-lacustrine and marine environments (Preiss 1987, 1993; Morton 1997). Laminated fine sandstone and siltstone facies at the top of the Chambers Bluff Tillite in drill hole Nicholson 2 have been interpreted as tidal rhythmites (Williams 1991; Williams & Schmidt 2004). See Preiss et al. (2011) for a detailed discussion of the Chambers Bluff Tillite and the possible Elatina and Nuccaleena Fm. equivalents. In the northern and northeastern Amadeus Basin there are both Sturt- (Areyonga Fm.) and Elatina-equivalent (Pioneer Sandstone and Olympic Fm.) glacial deposits. According to Lindsay (1989) there are three major facies associations in the Areyonga Fm. A massive basal diamictite also fills gullies eroded into underlying rocks and appears to have been deposited subglacially, perhaps as lodgement till. Locally, clean quartz sandstones are present at the base of the Areyonga Fm. suggesting fluviatile deposition. In the middle and upper parts of the Areyonga Fm., diamictite units are separated by channel-like, cross-bedded sandstone and conglomerate units, which suggests a fluvial depositional environment more distant from the ice source. In the uppermost Areyonga Fm. diamictites are finer grained and grade upwards into fissile siltstone with interbeds of stromatolitic dolomite, which suggests a transition from fluviatile to marine sedimentation. The Pioneer Sandstone is interpreted to have been deposited in a shallow-marine tidal setting (Lindsay 1989) as a glacial outwash fan that is a lateral equivalent to the diamictite-bearing Olympic Fm. (Preiss et al. 1978; Shaw & Wells 1983; Walter et al. 1995), which is interpreted to have been deposited as a result of fluvial and mass-flow processes (Field 1991). This correlation is supported by d13C values of discrete dolomite beds within the glacial units and overlying ‘upper marker cap dolomite’ units (Calver 1995; Kennedy 1996; Skotnicki et al. 2008). In the southern Amadeus Basin, the Inindia beds contain lower and upper diamictite units that are separated by a sequence boundary so probably represent both Sturt and Elatina glacial units (Lindsay 1989; Walter et al. 1995), although Lindsay (1989) speculated that the diamictite in the upper Inindia beds could be reworked from the Areyonga Fm. Facies associations suggest the lower Inindia beds were deposited in a marine environment, whereas the upper Inindia beds were deposited in a fluvial environment (Lindsay 1989). In the western Amadeus Basin, the Boord Fm. was deposited in fluviatile to shallow marine environments (Wells et al. 1964, 1970). Some authors have correlated the Boord Fm. with the Areyonga Fm. (Wells et al. 1970; Wells 1981; Preiss & Forbes 1981), whereas others have correlated it with the Olympic Fm. –Pioneer Sandstone sequence (Walter et al. 1995, 2000; Grey et al. 1999). In the Ngalia Basin there are both Sturt (Naburula Fm.) and Elatina (Mount Davenport Diamictite Member, Mount Doreen Fm.) glacial units, including a cap dolomite over the latter (Wanapi Dolomite Member, Mount Doreen Fm.). Dark grey to black siltstones and thin-bedded dolomite suggests that the Naburula Fm. was probably deposited in a shallow marine environment proximal to an ice shelf (Preiss et al. 1978; Wells & Moss 1983). Beds of pebbly quartz sandstone and medium-grained sandstone suggest a fluvioglacial environment of deposition for the Mount Davenport Diamictite Member (Young et al. 1995). In the western area of the southern Georgina Basin, the Boko Fm. is interpreted to have been deposited as a non-marine

moraine (till) (Haines et al. 2007) and to be an Elatina glacial deposit (Haines et al. 1991). In the central –western and central areas of the southern Georgina Basin, there are both Sturt (Mount Cornish Fm. and Yardida Tillite) and Elatina (Oorabra, Sun Hill and Black Stump arkoses) glacial deposits, including a cap dolomite unit (Wonnadinna Dolostone). There are no new data for the Mount Cornish Fm. or Yardida Tillite and the authors accept the conclusions of Walter (1980, 1981) that they contain evidence for deposition in glacial and periglacial environments. Diamictite in the Mount Cornish Fm. is capped by a very thin, dark grey laminated dolomite, possibly equivalent to the ‘lower marker cap dolomite’ in the Areyonga Fm., Amadeus Basin (Preiss et al. 1978). The Oorabra, Sun Hill and Black Stump Arkoses were probably deposited as proximal glacial outwash; they show widespread evidence of minimal chemical weathering (Walter 1980, 1981; Shergold 1985; Freeman 1986; Dunster et al. 2007; Haines et al. 2007), and there are faceted and striated cobbles in the Oorabra Arkose (Freeman 1986; Haines et al. 2007). Haines et al. (2007) further concluded that the Oorabra Arkose was deposited in a fluvial environment during the waning phase of the Elatina glaciation because of evidence that it post-dates the Boko Fm. in the southwestern Georgina Basin. The Wonnadinna Dolostone has been correlated with other Elatina glaciation cap dolomite units (Preiss et al. 1978). In the eastern area of the southern Georgina Basin there is an Elatina glacial unit, the Little Burke Tillite, which is interpreted as a shallow marine till (Plumb 1981), and a cap dolomite unit that can be correlated Australia-wide. Thanks go to C. Calver, R. Hocking, J. Gehling and W. Preiss for helpful discussions. Our two reviewers, W. Preiss and G. Halverson, greatly improved the manuscript. K. Grey and P. Haines published with the permission of the Executive Director of the Geological Survey of Western Australia. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Apak, S. N. & Moors, H. T. 2001. Basin development and petroleum exploration potential of the Lennis area, Officer Basin, Western Australia. Western Australia Geological Survey Report, 77, 42. Apak, S. N., Moors, H. T. & Stevens, M. K. 2002. Vines 1 Well Completion Report, Waigen area, Officer Basin, Western Australia. Geological Survey of Western Australia Record, 2001/18. Calver, C. R. 1995. Ediacarian isotope stratigraphy of Australia. PhD thesis, Macquarie University, Sydney, Australia. Calver, C. R. 1998. Isotope stratigraphy of the Neoproterozoic Togari Group, Tasmania. Australian Journal of Earth Sciences, 45, 865– 874. Calver, C. R. 2000. Isotope stratigraphy of the Ediacarian (Neoproterozoic III) of the Adelaide Rift Complex, South Australia, and the overprint of water column stratification. Precambrian Research, 100, 121– 150. Calver, C. R. & Lindsay, J. F. 1998. Ediacarian sequence and isotope stratigraphy of the Officer Basin, South Australia. Australian Journal of Earth Sciences, 45, 513–532. Cloud, P. E. & Semikhatov, M. A. 1969. Proterozoic stromatolite zonation. American Journal of Science, 267, 1017– 1061. Coats, R. P. & Preiss, W. V. 1980. Stratigraphic and geochronological reinterpretation of Late Proterozoic glaciogenic sequences in the Kimberley Region, Western Australia. Precambrian Research, 13, 181– 208. Condon, D., Zhu, M., Bowring, S., Wang, W., Yang, A. & Jin, Y. 2005. U –Pb ages from the Doushantuo Formation, China. Science, 308, 95 – 98. Corkeron, M. 2007. ‘Cap carbonates’ and Neoproterozoic glacigenic successions from the Kimberley region, north-west Australia. Sedimentology, 54, 871– 903. Crawford, A. J., Stevens, B. P. J. & Fanning, M. 1997. Geochemistry and tectonic setting of some Neoproterozoic and Early Cambrian

AUSTRALIAN NEOPROTEROZOIC GLACIAL DEPOSITS

volcanics in western New South Wales. Australian Journal of Earth Sciences, 44, 831– 852. Deckleman, J. A. 1995. Central Australia’s Ngalia Basin has geologic, marketing obstacles. Oil and Gas Journal, 93, 74 –81. De Keyser, F. 1972. Proterozoic Tillite at Duchess, northwestern Queensland. Bureau of Mineral Resources Australia Bulletin, 125, 1– 6. Drexel, J. F. 2008. Review of the Burra Mine Project, 1980–2008 – a progress report. South Australian Department of Primary Industries & Resources Report Book 2008/16, 76. Dunn, P. R., Thomson, B. P. & Rankama, K. 1971. Late Precambrian glaciation in Australia as a stratigraphic boundary. Nature, 231, 498– 502. Dunster, J. N., Kruse, P. D., Duffett, M. L. & Ambrose, G. J. 2007. Geology and resource potential of the southern Georgina Basin. Northern Territory Geological Survey, Digital Information Package, DIP007. Dyson, I. A. & Marshall, T. R. 2007. Neoproterozoic salt nappe complexes and salt-withdrawal mini-basins in the Amadeus Basin. In: Munson, T. J. & Ambrose, G. J. (eds) Proceedings of the Central Australian Basins Symposium, Alice Springs, 16 –18 August, 2005. Northern Territory Geological Survey Special Publication, 2. Available online: http://conferences.minerals.nt.gov.au/ cabsproceedings/ Embleton, B. J. J. & Williams, G. E. 1986. Low palaeolatitude of deposition for the late Precambrian periglacial varvites in South Australia: implications for palaeo-climatology. Earth & Planetary Science Letters, 79, 419– 430. Eyles, N. & Eyles, C. H. 1998. Summary of Late Proterozoic glaciogenic succession in Empress 1, Officer Basin. Western Australia Geological Survey, Statutory Petroleum Exploration report, S20424 A6 (unpublished). Eyles, C. H., Eyles, N. & Grey, K. 2007. Palaeoclimate implications from deep drilling of Neoproterozoic strata in the Officer Basin and Adelaide Rift Complex of Australia: a marine record of wet-based glaciers. Palaeogeography, Palaeoclimatology, Palaeoecology, 248, 291– 312. Fanning, C. M. & Link, P. K. 2008. Age constraints for the Sturtian Glaciation; data from the Adelaide Geosyncline, South Australia and Pocatello Formation, Idaho, USA. In: Gallagher, S. J. & Wallace, M. W. (eds) Selwyn Symposium 2008: Neoproterozoic Extreme Climates and the Origin of Early Metazoan Life. Geological Society of Australia, Extended Abstracts No. 91, 57 –62. Fanning, C. M., Ludwig, K. R., Forbes, B. G. & Preiss, W. V. 1986. Single and multiple grain U– Pb zircon analyses for the early Adelaidean Rook Tuff, Willouran Ranges, South Australia. Geological Society of Australia Abstracts, 15, 255–304. Field, B. D. 1991. Paralic and periglacial facies and contemporaneous deformation of the Late Proterozoic Olympic Formation, Pioneer Sandstone and Gaylad Sandstone, Amadeus Basin, central Australia. Australian Bureau of Mineral Resources Bulletin, 236, 127– 136. Foden, J., Barovich, K., Jane, M. & O’Halloran, G. 2001. Sr-isotopic evidence for Late Neoproterozoic rifting in the Adelaide Geosyncline at 586 Ma: implications for a Cu ore forming fluid flux. Precambrian Research, 106, 291– 308. Freeman, M. J. 1986. Huckitta, Northern Territory (Second Edition). 1:250 000 geological series explanatory notes, SF 5311. Northern Territory Geological Survey, Darwin. Freeman, M. J., Oaks, R. Q., Jr. & Shaw, R. D. 1991. Stratigraphy of the late Proterozoic Gaylad Sandstone, northeastern Amadeus Basin, and recognition of an underlying regional unconformity. Bureau of Mineral Resources Geological and Geophysical Bulletin, 236, 137– 154. Glikson, A. Y., Stewart, A. J., Ballhaus, C. G., Clarke, G. L., Feeken, E. H. J., Sheraton, J. W. & Sun, S-S. 1996. Geology of the western Musgrave Block, central Australia, with particular reference to the mafic–ultramafic Giles Complex. Australian Geological Survey Organisation Bulletin, 239, 41 –68. Gorjan, P., Veevers, J. J. & Walter, M. R. 2000. Neoproterozoic sulfurisotope variation in Australia and global implications. Precambrian Research, 100, 151– 179. Grey, K. 2005. Ediacaran Palynology of Australia. Memoir of the Association of Australian Paleontologists, 31, 1 –439.

689

Grey, K., Apak, S. N., Eyles, N., Stevens, M. K. & Carlsen, G. M. 1999. Neoproterozoic glacigene successions, western Officer Basin, Western Australia. Geological Survey of Western Australia Annual Review, 74– 80. Grey, K., Hocking, R. M. et al. 2005. Lithostratigraphic nomenclature of the Officer Basin and correlative parts of the Paterson Orogen, Western Australia. Western Australia Geological Survey Report, 93, 89. Haines, P. W., Bagas, L., Wyche, S., Simons, B. & Morris, D. G. 1991. Barrow Creek, Northern Territory (Second Edition). 1:250 000 geological series explanatory notes, SF 5306. Northern Territory Geological Survey, Darwin. Haines, P. W., Hand, M. & Sandiford, M. 2001. Palaeozoic synorogenic sedimentation in central and northern Australia: a review of distribution and timing with implications for the evolution of intracratonic orogens. Australian Journal of Earth Sciences, 48, 911– 928. Haines, P. W., Mory, A. J., Stevens, M. K. & Ghori, K. A. R. 2004. GSWA Lancer 1 well completion report (basic data), Officer and Gunbarrel Basins, Western Australia. Geological Survey of Western Australia Record, 2004/10, 39. Haines, P. W., Scrimgeour, I. R. & Duffett, M. L. 2007. Woodgreen, Northern Territory. 1:100 00 geological map explanatory notes, 5753. Northern Territory Geological Survey, Darwin. Haines, P. W., Hocking, R. M., Grey, K. & Stevens, M. K. 2008. Vines 1 revisited: are older Neoproterozoic glacial deposits preserved in Western Australia? Australian Journal of Earth Sciences, 55, 397– 406. Hand, M. & Sandiford, M. 1999. Intraplate deformation in central Australia, the link between subsidence and fault reactivation. Tectonophysics, 305, 121– 140. Hill, A. C. 2005. Stable isotope stratigraphy, GSWA Lancer 1, Officer Basin, Western Australia. In: Mory, A. J. & Haines, P. W. (eds) GSWA Lancer 1 Well Completion Report (Interpretive Papers), Officer and Gunbarrel Basins, Western Australia. Geological Survey of Western Australia Record, 2005/4, 1– 11. Hill, A. C. & Walter, M. R. 2000. Mid-Neoproterozoic (830– 750 Ma) isotope stratigraphy of Australia and global correlation. Precambrian Research, 100, 181–211. Hill, A. C., Cotter, K. L. & Grey, K. 2000. Mid-Neoproterozoic biostratigraphy and isotope stratigraphy in Australia. Precambrian Research, 100, 281–298. Hoffmann, K.-H, Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. A U– Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817– 820. Indigo Oil & Sirgo Exploration. 1990. Wallara-Well Completion Report. Indigo Oil Pty Ltd and Sirgo Exploration Inc. Northern Territory Geological Survey, Open File Petroleum Report, PR 1990-0101B. Ireland, T. R., Flo¨ttmann, T., Fanning, C. M., Gibson, G. M. & Preiss, W. V. 1998. Development of the Early Paleozoic Pacific Margin of Gondwana from detrital zircon ages across the Delamerian Orogen. Geology, 26, 243–246. Jackson, M. J. & van de Graaff, W. J. E. 1981. Geology of the Officer Basin. Bureau of Mineral Resources Bulletin, 206, 1– 102. Kendall, B., Creaser, R. A. & Selby, D. 2006. Re– Os geochronology of postglacial black shales in Australia: consequences for timing of the Sturtian glaciation. Geology, 34, 729– 732. Kendall, B., Creaser, R. A., Calver, C. R., Raub, T. D. & Evans, D. A. D. 2007. Neoproterozoic paleogeography, Rodinia breakup, and Sturtian glaciation: constraints from Re– Os black shale ages from southern Australia and northwestern Tasmania. Geological Society of America Abstracts with Programs, 39, 335. Kennedy, M. J. 1996. Stratigraphy, sedimentology, and isotope geochemistry of Australian Neoproterozoic postglacial cap dolostones: deglaciation, d13C excursions, and carbonate precipitation. Journal of Sedimentary Research, 66, 1050– 1064. Kennedy, M. J., Christie-Blick, N. & Prave, A. R. 2001. Carbon isotopic composition of Neoproterozoic glacial carbonates as a test of paleoceanographic models for snowball Earth phenomena. Geology, 29, 1135– 1138.

690

A. C. HILL ET AL.

Li, Z. X., Bogdanova, S. V. et al. 2008. Assembly, configuration, and break-up history of Rodinia: a synthesis. Precambrian Research, 160, 179– 210. Lindsay, J. F. 1989. Depositional controls on glacial facies associations in a basinal setting, Late Proterozoic, Amadeus Basin, central Australia. Palaeogeography, Palaeoclimatology, Palaeoecology, 73, 205– 232. Lindsay, J. F. 2002. Supersequences, superbasins, supercontinents – evidence from the Neoproterozoic – Early Palaeozoic basins of central Australia. Basin Research, 14, 207–223. Lindsay, J. F. & Korsch, R. J. 1991. The evolution of the Amadeus Basin, central Australia. Australian Bureau of Mineral Resources Bulletin, 236, 7 –32. Lindsay, J. F. & Leven, J. F. 1996. Evolution of a Neoproterozoic to Palaeozoic intracratonic setting. Basin Research, 8, 403–424. Mawson, D. 1949. The Elatina glaciation. Royal Society of South Australia, Transactions, 73, 117–121. McKirdy, D. M., Burgess, J. M. et al. 2001. A chemostratigraphic overview of the late Cryogenian interglacial sequence in the Adelaide Fold– Thrust Belt, South Australia. Precambrian Research, 106, 149– 186. McWilliams, M. O. & McElhinny, M. W. 1980. Late Precambrian paleomagnetism of Australia: the Adelaide Geosyncline. Journal of Geology, 88, 1– 26. Morton, J. G. G. 1997. Lithostratigraphy and environments of deposition. In: Morton, J. G. G. & Drexel, J. F. (eds) Petroleum Geology of South Australia, Volume 3: Officer Basin, South Australian Department of Mines and Energy Resources Report Book, 97/19, 47 –86. Nelson, D. R. 2002. Compilation of geochronology data, 2001. Western Australia Geological Survey Record, 2002/2, 282. Pisarevsky, S. A. 2001. Paleomagnetic study of Vines 1. Geological Survey of Western Australia Record, 2001/18, Appendix 4. Pisarevsky, S. A., Li, Z. X., Grey, K. & Stevens, M. K. 2001. A palaeomagnetic study of Empress 1A, a stratigraphic drill hole in the Officer Basin: evidence for a low-latitude position of Australia in the Neoproterozoic. Precambrian Research, 110, 93 –108. Pisarevsky, S. A., Wingate, M. T. D., Stevens, M. K. & Haines, P. W. 2007. Paleomagnetic results from the Lancer 1 stratigraphic drill hole, Officer Basin, Western Australia, and implications for Rodinia reconstructions. Australian Journal of Earth Sciences, 54, 561– 572. Plumb, K. A. 1981. Late Proterozoic (Adelaidean) tillite of the Duchess area, northwestern Queensland. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 528– 530. Preiss, W. V. 1987. The Adelaide Geosyncline –late Proterozoic stratigraphy, sedimentation, paleontology and tectonics. South Australian Geological Survey Bulletin, 53, 438. Preiss, W. V. 1993. Neoproterozoic. In: Drexel, J. F., Preiss, W. V. & Parker, A. J. (eds) The Geology of South Australia, Volume 1, The Precambrian. South Australian Geological Survey Bulletin, 54, 171– 203. Preiss, W. V. 2000. The Adelaide geosyncline of South Australia and its significance in Neoproterozoic continental reconstruction. Precambrian Research, 100, 21 –63. Preiss, W. V. & Forbes, B. G. 1981. Stratigraphy, correlation and sedimentary history of Adelaidean (late Proterozoic) basins in Australia. Precambrian Research, 15, 255–304. Preiss, W. V., Walter, M. R., Coats, R. P. & Wells, A. T. 1978. Lithological correlations of Adelaidean glaciogenic rocks in parts of the Amadeus, Ngalia and Georgina Basins. Bureau of Mineral Resources Geology & Geophysics Australia Journal, 3, 43 – 53. Preiss, W. V., Gostin, V. A, McKirdy, D. M., Ashley, P. M., Williams, G. E. & Schmidt, P. W. 2011. The glacial succession of Sturtian age in South Australia – the Yudnamutana Subgroup. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 701– 712. Prichard, C. E. & Quinlan, T. 1962. The geology of the southern half of the Hermannsburg 1:250 000 sheet. Bureau of Mineral Resources Australia Report, 61. Ranford, L. C., Cook, P. J. & Wells, A. T. 1965. The geology of the central part of the Amadeus Basin, Northern Territory. Bureau of Mineral Resources Australia Report, 86.

Schaefer, B. F. & Burgess, J. M. 2003. Re– Os isotopic age constraints on deposition in the Neoproterozoic Amadeus Basin: implications for the ‘snowball Earth’. Geological Society of London Journal, 160, 825– 828. Schmidt, P. W. & Williams, G. E. 1995. The Neoproterozoic climatic paradox: equatorial palaeolatitude for Marinoan glaciation near sea level in South Australia. Earth & Planetary Science Letters, 134, 107– 124. Shaw, R. D. & Warren, R. G. 1975. Alcoota, Northern Territory, 1:250 000 geological series, explanatory notes. Bureau of Mineral Resources Australia, SF/5310. Shaw, R. D. & Wells, A. T. 1983. Alice Springs (second edition), Northern Territory, 1:250000 geological series, explanatory notes. Bureau of Mineral Resources Australia, SF/53-14. Shergold, J. H. 1985. Notes to accompany the Hay River-Mount Whelan Special 1: 250 000 Geological Sheet, southern Georgina Basin. Bureau of Mineral Resources, Australia, Report, 251. Simeonova, A. P. & Iasky, R. P. 2005. Seismic mapping, salt deformation and hydrocarbon potential of the central western Officer Basin, Western Australia Report 98. Skotnicki, S. J., Hill, A. C, Walter, M. R. & Jenkins, R. 2008. Stratigraphic relationships of Cryogenian strata disconformably overlying the Bitter Springs Formation, northeastern Amadeus Basin, Central Australia. Precambrian Research, 165, 243– 259. Smith, K. G. 1963a. Hay River, Northern Territory, 1:250 000 geological series, explanatory notes. Bureau of Mineral Resources Australia, SF/53-16. Smith, K. G. 1963b. Huckitta, Northern Territory, 1:250 000 geological series, explanatory notes. Bureau of Mineral Resources Australia, SF/53-11. Smith, K. G. 1972. Stratigraphy of the Georgina Basin. Bureau of Mineral Resources Australia Bulletin, 111. Sohl, L. E., Christie-Blick, N. & Kent, D. V. 1999. Paleomagnetic polarity reversals in Marinoan (ca. 600 Ma) glacial deposits of Australia: implications for the duration of low-latitude glaciation in Neoproterozoic time. Geological Society of America Bulletin, 111, 1120– 1139. Stevens, M. K., Apak, S. N. & Moors, H. T. (compilers) 2002. GSWA Vines 1 well completion report, Waigen area, Officer Basin, Western Australia. Western Australia Geological Survey Record, 2001/18, 32. Veevers, J. J., Walter, M. R. & Scheibner, E. 1997. Neoproterozoic tectonics of Australia –Antarctica and Laurentia and the 560 Ma birth of the Pacific Ocean reflect the 400 million year Pangean supercycle. Journal of Geology, 105, 225– 242. Wade, B. P., Hand, M. & Barovich, K. M. 2005. Nd isotopic and geochemical constraints on provenance of sedimentary rocks in the eastern Officer Basin, Australia: implications for the corpduration of the intracratonic Petermann Orogeny. Journal of the Geological Society of London, 162, 513– 530. Walter, M. R. 1980. Adelaidean and Early Cambrian stratigraphy of the southwestern Georgina Basin: correlation chart and explanatory notes. Bureau of Mineral Resources Australia Report, 214, 21. Walter, M. R. 1981. Late Proterozoic tillites of the southwestern Georgina Basin, Australia. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 525– 527. Walter, M. R. & Bauld, J. 1983. The association of sulphate evaporites, stromatolitic carbonates and glacial sediments: examples from the Proterozoic of Australia and the Cainozoic of Antarctica. Precambrian Research, 21, 129–148. Walter, M. R. & Hill, A. C. 1999. Carbon isotope stratigraphy of Empress 1/1A, Supersequence 1, western Officer Basin. Western Australia Geological Survey, Statutory Petroleum Exploration Report, S20424 A8 (unpublished). Walter, M. R., Krylov, I. N. & Preiss, W. V. 1979. Stromatolites from Adelaidean (Late Proterozoic) sequences in central and South Australia. Alcheringa, 3, 287– 305. Walter, M. R., Grey, K., Williams, I. R. & Calver, C. R. 1994. Stratigraphy of the Neoproterozoic to early Palaeozoic Savory Basin, Western Australia, and correlation with the Amadeus and Officer Basins. Australian Journal of Earth Sciences, 41, 533–546.

AUSTRALIAN NEOPROTEROZOIC GLACIAL DEPOSITS

Walter, M. R., Veevers, J. J., Calver, C. R. & Grey, K. 1995. Neoproterozoic stratigraphy of the Centralian Superbasin, Australia. Precambrian Research, 73, 173– 195. Walter, M. R., Veevers, J. J., Calver, C. R., Gorjan, P. & Hill, A. C. 2000. Dating the 840– 544 Ma Neoproterozoic interval by isotopes of strontium, carbon, and sulfur in seawater, and some interpretative models. Precambrian Research, 100, 371–433. Wells, A. T. 1972. Mount Doreen, Northern Territory, 1:250 000 geological series, explanatory notes. Bureau of Mineral Resources Australia, SF/53-12. Wells, A. T. 1981. Late Proterozoic diamictites of the Amadeus and Ngalia Basins, central Australia. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. University of Cambridge Press, Cambridge, 515–524. Wells, A. T. & Moss, J. F. 1983. The Ngalia Basin, Northern Territory: stratigraphy and structure. Australian Bureau of Mineral Resources (now Geoscience Australia) Bulletin, 212, 88. Wells, A. T., Forman, D. J. & Ranford, L. C. 1964. Geological reconnaissance of the Rawlinson and MacDonald 1:250 000 Sheet areas, Western Australia. Bureau of Mineral Resources Australia Report, 65. Wells, A. T., Cook, P. J., Ranford, L. C., Shaw, R. D. & Stewart, A. J. 1965. The geology of the north-eastern part of the Amadeus Basin, Northern Territory. Bureau of Mineral Resources Australia Record, 108. Wells, A. T., Stewart, A. J. & Skwarko, S. K. 1966. Geology of the south-eastern part of the Amadeus Basin, Northern Territory. Bureau of Mineral Resources Australia Report, 88. Wells, A. T., Ranford, L. C., Stewart, A. J., Cook, P. J. & Shaw, R. I. 1967. Geology of the North-eastern part of the Amadeus Basin, Northern Territory. Bureau of Mineral Resources Australia Report, 88. Wells, A. T., Forman, D. J., Ranford, L. C. & Cook, P. J. 1970. Geology of the Amadeus Basin, central Australia. Bureau of Mineral Resources Australia Report, 100. Wells, A. T., Moss, F. J. & Sabitay, A. 1972. The Ngalia Basin, Northern Territory – recent geological and geophysical information upgrades petroleum prospects. Australian Petroleum Exploration Association Journal, 12, 144– 151. Williams, G. E. 1985. Solar affinity of sedimentary cycles in the late Precambrian Elatina Formation. Australian Journal of Physics, 38, 1027– 1043. Williams, I. R. 1987. Late Proterozoic glacigene deposits in the Little Sandy Desert, Western Australia. Australian Journal of Earth Sciences, 34, 153– 155. Williams, G. E. 1989. Late Precambrian tidal rhythmites in South Australia and the history of Earth’s rotation. Geological Society of London Journal, 146, 97– 111. Williams, G. E. 1991. Upper Proterozoic tidal rhythmites, South Australia: sedimentary features, deposition, and implications for the Earth’s paleorotation. In: Smith, D. G., Reinson, G. E., Zaitlin, B. A. & Rahmani, R. A. (eds) Clastic Tidal Sedimentology. Canadian Society of Petroleum Geologists Memoirs, 16, 161– 177.

691

Williams, I. R. 1992. Geology of the Savory Basin, Western Australia. Western Australia Geological Survey Bulletin, 141, 115. Williams, I. R. 1994. The Neoproterozoic Savory Basin, Western Australia. In: Purcell, P. G. & Purcell, R. R. (eds) The Sedimentary Basins of Western Australia. Petroleum Exploration Society of Australia, West Australian Basins Symposium, Perth, WA, 1994, Proceedings, 841– 850. Williams, G. E. 2008. Proterozoic (pre-Ediacaran) glaciation and the high obliquity, low-latitude ice, strong seasonality (HOLIST) hypothesis: Principles and tests. Earth-Science Reviews, 87, 61– 93. Williams, I. R. & Tyler, I. M. 1991. Robertson, Western Australia (Second edition). Western Australia Geological Survey 1:250 000 Geological Series Explanatory Notes, 36. Williams, G. E. & Schmidt, P. W. 2004. Neoproterozoic glaciation: reconciling low paleolatitudes and the geologic record. In: Jenkins, G. S., McMenamin, M. A. S., McKay, C. P. & Sohl, L. (eds) The Extreme Proterozoic: Geology, Geochemistry and Climate. American Geophysical Union Geophysical Monograph, 146, 145–159. Williams, G. E., Jenkins, R. J. F. & Walter, M. R. 2007. No heliotropism in Neoproterozoic columnar stromatolite growth, Amadeus Basin, central Australia: geophysical implications. Palaeogeography, Palaeoclimatology, Palaeoecology, 249, 80– 89. Williams, G. E., Gostin, V. A., McKirdy, D. M. & Preiss, W. V. 2008. The Elatina glaciation, late Cryogenian (Marinoan Epoch), South Australia: sedimentary facies and palaeoenvironments. Precambrian Research, 163, 307–331. Williams, G. E., Gostin, V. A., McKirdy, D. M., Preiss, W. V. & Schmidt, P. W. 2011. The Elatina glaciation (late Cryogenian), South Australia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 713–721. Wingate, M. T. D. & Giddings, J. W. 2000. Age and paleomagnetism of the Mundine Well dyke swarm, Western Australia: implications for an Australia– Laurentia connection at 755 Ma. Precambrian Research, 100, 335–357. Wingate, M. T. D., Campbell, I. H., Compston, W. & Gibson, G. M. 1998. Ion microprobe U– Pb ages for Neoproterozoic basaltic magmatism in south-central Australia and implications for the breakup of Rodinia. Precambrian Research, 87, 135–159. Young, I. F. & Ambrose, G. J. 2007. Petroleum geology of the southeastern Amadeus Basin: the search for sub-salt hydrocarbons. In: Munson, T. J. & Ambrose, G. J. (eds) Proceedings of the Central Australian Basins Symposium, Alice Springs, 16 – 18 August, 2005. Northern Territory Geological Survey Special Publication, 2. Available online at http://conferences.minerals.nt.gov. au/cabsproceedings/ Young, D. N., Edgoose, C. J., Blake, D. H. & Shaw, R. D. 1995. Explanatory Notes, Mount Doreen 1:250 000 Geological Map Series, Northern Territory Geological Survey, 26 – 31. Zhao, J.-X., McCulloch, M. T. & Korsch, R. J. 1994. Characterization of a plume-related 800 Ma magmatic event and its implications for basin formation in central-southern Australia. Earth & Planetary Science Letters, 121, 349–367.

Chapter 68 Billy Springs glaciation, South Australia RICHARD J. F. JENKINS South Australian Museum, North Terrace, Adelaide, South Australia 5000 (e-mail: [email protected]) Abstract: An interval several hundred metres thick in the lower part of the late Ediacaran Billy Springs Formation (Fm.) of the NE Flinders Ranges, South Australia, includes both diamictic levels, ‘dropstones’ and isolated ‘stone-clusters’ in a thin-bedded, silty or fine sandy matrix, which is commonly laminated. Slumping at a variety of scales is prevalent, but extensive panels of ‘right-way-up’, moderately dipping beds host dropstones, which disturbed the laminae. It is difficult to explain laminated sediments peppered with dropstones other than by ice-rafting, and the stone-clusters comply with sinking pieces of ice loaded with debris. Although common reworking occurred in channels, larger erratics in this material are out of hydrodynamic equilibrium. Deposition occurred offshore on an unstable slope. Isotopic d13Ccarb measurements show a strong negative excursion through the preceding Wonoka Fm., and several further similar negative excursions in the Billy Springs Fm. This record shows similarities to that of Ediacaran carbonates on the Yangtze platform, south China. Compilation based on a survey of palaeomagnetic data for Gondwana continents, Baltica and Laurasia permits a possible palaeogeography indicating a relatively high palaeolatitude at the time of deposition. Indications of age are imprecise but may be comparable with, or younger(?) than the c. 580 Ma Gaskiers Fm. of Newfoundland.

Defined originally as the Billy Springs Beds by Coats (in Coats & Blisset 1971), a thick late Neoproterozoic unit of siltstones, sandstones and less common carbonates crops out in the far NE Flinders Ranges in several large synformal structures referred to as the Umberatana and Mount Freeling –Mount Gardiner synclines (Fig. 68.1). It was accorded formation status by Forbes & Preiss (1987). Coats proposed a type section in the Mount Freeling syncline in the area between Tardlapinna Well and Village Well from lat. 298560 3500 S, long. 1398200 4100 E to lat. 298520 2000 S, long.1398100 2700 E, where the Billy Springs Fm. was considered to be at least 4800 m thick. It is commonly considered as correlating with the Pound Subgroup (Forbes & Preiss 1987), which comprises the upper part of the stratotypic Ediacaran System in the western central Flinders Ranges (Preiss 2005; Knoll et al. 2006). Coats (in Coats & Blisset 1971) recorded that in the Umberatana syncline a ‘breccia’ near the base of the formation included blocks of dolomite and rare boulders of granite. Similarly Von der Borch & Grady (1982) described a 50-m-thick interval of slump-folded diamictites forming a basal part of the formation in the southern flank of the Umberatana syncline. DiBona (1991) first advanced the notion that diamictites in this area were generated as a result of ice-rafting and this was extended by Jenkins (1993, p.12, Jenkins et al. 1998). However, the notion of ice-rafting was disputed by Christie-Blick (1993), who considered gravity flow on a palaeoslope as an adequate mechanism for emplacement of the diamictite and opined ‘there is absolutely nothing else in the regional stratigraphy that remotely implies a glacial or glacial-marine environment at this horizon’. The present chapter disputes this assertion.

movements directed towards the NW, with a large local thrust system developing (Jenkins 1992; Jenkins & Sandiford 1992). A second compressional movement in the northern part of the complex was directed towards the south. This resulted in crossfolded structural elements, such that the dominant resistive unit represented by the sandstones and quartzites of the Pound Subgroup forms sub-circular, saucer-shaped structures or ‘pounds’, well represented by Wilpena Pound and the Gammon Ranges. The dominant fold axis in the Gammon Ranges is roughly east – west and this same axis predominates in the more northerly sited Umberatana and Mount Freeling-Gardiner synclines which host the Billy Springs Fm. The region experienced greenschist facies metamorphism with attendant cleavage development during the Delamerian deformation (McKirdy et al. 1975) with local emplacement of uraniferous granites in the adjacent Mount Painter complex. Deformation becomes generally more intense in a northeasterly direction. The Umberatana syncline is broad and relatively open, with dips of c. 158 NE in the Billy Springs Fm. on its southwestern flank. In contradistinction, numerous tight approximately upright folds run east –west through the Mount Freeling syncline, which also shows various shears, and it is unlikely the Billy Springs Fm. is as thick as claimed. Pell (1989) and Reid (1992) measured a section of c. 2900 m through the stratotype (not including the interval presently assigned to the Rawnsley Quartzite, Fig. 68.2). The thick sequences in this NE part of the Flinders Ranges are probably linked to local late Ediacaran extension forming a restricted depocentre passing into a slope setting towards the north (Preiss 1987, pp. 404– 407).

Structural framework

Stratigraphy

The Billy Springs Fm. lies within the NE sector of the Adelaide Rift Complex, a large Neoproterozoic to mid-Cambrian(?) basinal complex extending some 700 km north –south with an WSW –ENE extension through Kangaroo Island and a similarly trending side fork, the Nackara Arc, curving towards Broken Hill (Fig. 68.1). At least four major phases of extension are represented, beginning somewhat earlier than 800 Ma, and culminating within the Early Cambrian (Jenkins 1992; Walter & Veevers 1997). In the Adelaide region, a major compressional event known as the Delamerian Orogeny commenced a little later than 522 Ma (Jenkins et al. 2002; Foden et al. 2006). Southern and mid parts of the Rift Complex experienced compressional

As noted above, the Billy Springs Fm. is generally correlated with the upper part of the stratotypic Ediacaran Wilpena Group of the central western Flinders Ranges. In both the Umberatana and Mount Freeling synclines, the Billy Springs Fm. is conformable above the Wonoka Fm. With reference to the type area, Coats (in Coats & Blisset 1971) divided the formation informally into a lower member of grey-green fine sandstone, quartzite, marble and dolomite, and an upper member forming most of the section west of Mount Freeling and comprising grey-green slate, quartzite and minor lenticular marbles. Coats (in Coats & Blisset 1971) referred to the ‘breccia’ in the older part of the lower member in the Umberatana syncline.

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 693– 699. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.68

694

R. J. F. JENKINS

several kilometres SW of Mount Freeling, and Reid (1999, pers. comm.) located sparse small discs and a small frondose structure presumably representing an outlier of the Ediacara assemblage at an undisclosed nearby site. This is an important find because it supports an occurrence of the hosting unit of the type Ediacara assemblage, the Ediacara Member of the Rawnsley Quartzite, which crops out widely through much of the remainder of the Flinders Ranges. Von der Borch & Grady (1982) described a section through the upper Wonoka Formation and Billy Springs Fm. NW of the ‘Umberatana’ pastoral homestead. They recorded the Billy Springs section as comprising a lower c. 60 m of fine sandstone, a further 50-m-thick interval of slump-folded diamictites, followed by 230 m of rhythmically bedded fine calcareous sandstone, and a succeeding 520 m of grey-green siltstone and fine sandstone grading up into quartzite bands interfingering with mudstone. DiBona (1991, fig.1) researched the same area as Von der Borch & Grady (1982) and extended mapping around the southwestern closure of the Umberatana syncline. He reported on the ‘succession of dropstones, rhythmites, thin-bedded sandstone units, sandsilt lenses and a diamictite bed or horizon’ comprising principally a 50 m thickness of the lower Billy Springs Fm., but recognized dropstone facies reaching a thickness of some 80 m, and demonstrated marked lithological changes along strike.

Glaciogenic deposits and associated strata Southwestern Umberatana syncline

Fig. 68.1. Locality maps of the Adelaide Rift Complex and placement of Billy Springs glacials.

Forbes (1966) reported the highest part of the succession SW of Village Well as quartzite and red-brown silstones overlying carbonates. This interval has been considered by N. Lemon (1997, pers. comm.) as equivalent to a later part of the local Early Cambrian, but field investigations imply all contacts as highly sheared and any stratigraphic contiguity is uncertain. The author has observed rare trace fossils in flaggy quartzites

The present investigation concentrated on relatively shallowly dipping strata exposed in the incised channel of Old Station Creek (Fig. 68.1). A measured section documents an interval of some 220 m thickness including diamictite associations (Fig. 68.2) and side traverses confirm lateral variability of the sediments. The older c. 40-m-thick ‘slumped deposits’ consist of fine silty sandstone with large (several metres) slump folds tending to resemble ‘flat’ isoclinal tectonic folds, but shows rapid changes in the thickness of bedding packages. Thin internal lamination is common and pebble- to boulder-sized clasts are randomly dispersed rarely to commonly through this lithology. The abundance of clasts varies considerably along strike, with boulder-rich slump folds tending to show a clear association over a strike distance of several hundred metres. The greater number of clasts comprise white ‘vein-quartz’ and are sharply angular to subangular, without obvious surface abrasion. Angular to sub-rounded carbonate lithologies are common with blocks up to several tens of centimetres; black chert occurs mainly as pebbles, and small fragments of granite are extremely rare. A large clast of amygdaloidal basalt has been located. The carbonates do not resemble either carbonates known elsewhere in the Billy Springs Fm. or in the underlying Wonoka Fm. Medium to coarse rounded quartz grains form c. 5% of the matrix (Di Bona 1991). Immediately up-section, clasts become rarer, but the largest ‘vein-quartz’ block located is 1.2 m in its longest dimension and the thin lamination of its enclosing matrix is tilted up strongly on several sides consistent with soft sediment cratering as this block was emplaced (Fig. 68.2). The succeeding c. 40 m of thin-bedded sandstones include several beds of platy, calcareous sandstones. The ensuing 15 m of platy or laminated siltstone include increasingly common lonestones, up-section, with angular sub-rounded yellow dolomite and grey limestone fragments common along with rarer ‘vein-quartz’. The fine lamination of these beds indicates large extensive panels representing intact stratigraphy, and ‘way-up’ can be determined by soft-sediment load-casting and flame structures. Lonestones deform underlying laminae and also show overlying sediment drape. Pebble clusters with associated grit patches are common. This and the underlying level also include lenses or pods of

BILLY SPRINGS GLACIATION, SOUTH AUSTRALIA

695

Fig. 68.2. (a)–(c) Stratigraphic sections through intervals incorporating glacial sediments and correlative part of central northern Flinders Ranges. Section (c) has the same scale as (b). Carbonate C-isotope data: designation of negative excursions indicates suggested correlation with notation of McFadden et al. (2008). *, Urlwin (in Urlwin et al. 1993); , Pell (1989); †, Reid (1992).

black, carbonate-rich, pebbly diamictite, which weather out of their hosting siltstones in the harsh, arid climate. Traced a kilometre to the NW from the channel of Old Station Creek into a cliff section where the diamictic pods can be seen in situ, they occur as flat-topped lenses in laminated siltstones and fine sandstones, and are some 50 cm to 1 m thick, occasionally with crude, large-scale internal crossbedding formed by an alignment of pebble-sized clasts. However, much larger clasts outside of the limits of hydrodynamic equilibrium stud the diamictite and commonly form associated clusters. Cut sections of the black diamictites reveal all manner of wisps and blobs of unsorted grit, and similar poorly sorted ‘blobs’ of such grits also occur in the thin-bedded matrix siltstones as ‘clasts’. A single example of a facetted quartzite clast or ‘dreikanter’ was found loose at about this level; no surface striations are evident. Near the top of the ‘15-m-thick’ interval, an alternation of thin siltstone and graded fine sandstone layers about 1 cm thick individually, and 25 cm thick overall, is peppered with ‘dropstones’ and other finer angular gravel lithoclasts of carbonates, quartz and chert. The ensuing c. 65 m of platy siltstone grades up into medium bedded sandstone with scattered lonestones of clastic lithologies and rarer quartzite and carbonate clasts. Clusters of dirty sandstone clasts and granules may occur closely associated with one another. This level appears to mark the top of glaciogenic influence, and sandstones and laminated siltstones continue upwards.

Southern Mount Freeling syncline The lower ‘breccia’ is traced in cliff sections along Tardlapinna Creek, and 200 m SW of the Tardlapinna windmill. The basal Billy Springs Fm. is a highly cleaved, medium-bedded greenish grey siltstone dipping 168 to 258 SW. About 20 m above the base, clasts greatly elongated in the plane of the bedding are common, and include a quartoze, chlorite augen gneiss and stretched limestone boulders. At a location 30 m higher, the siltstones show highly convoluted but parallel bedding with decimetre-sized holes where sub-rounded carbonate clasts have dissolved out. Along strike, clearly bedded siltstones and sandstones include further holes representing dissolved carbonate clasts and lone siliceous pebbles. Strange, angular zig-zag folds in a banded sandstone may indicate iceberg scour.

Central northern and southern Flinders Ranges In the central-northern Flinders Ranges the direct equivalents of the Billy Springs glacials are not clearly apparent. Further west, clast-supported pebble bands (clasts 1– 3 cm, rarely up to 8– 9 cm) marking the base of the Bonney Sandstone at Beltana are attributed to shedding from an active uplifting diapir (Leeson 1970; Forbes & Preiss 1987, p. 242). Located 30 km ENE of

696

R. J. F. JENKINS

Beltana, in the northern limb of the Mount Goddard syncline (Fig. 68.1), presumed marine limey-shales and grey limestones of the Wonoka Fm. are succeeded by a few metres of red-brown siltstone, shales and sandstones with limestones and some conglomerate, a thick (c. 70 m) bleached sandstone and further upward platy sandstones and carbonates (c. 60 m). This overall interval (post-dating the Wonoka limestones) is probably the downward extension of the Bonney Sandstone represented in its Patsy Hill Member (Reid & Preiss 1999; Preiss 1999). Overlying repeated cycles of red non-marine silty-sandstones, possible fluvial channel deposits, and granule beds at the base of flaggy sandstones with ripple-marked tops, indicate a deltaic setting with episodic marine incursions (e.g. Gehling 1982). Further west, clast-supported pebble bands (clasts 1–3 cm, rarely up to 8 –9 cm) marking the base of the Bonney Sandstone at Beltana are attributed to shedding from an active uplifting diapir (Leeson 1970; Forbes & Preiss 1987, p. 242). The equivalent of this latter interval in the Stirrup Iron Range shows stacked, thick, lensoidal sand bodies dispersed in the red maroon silty-sandstone facies. The sandstones show impressive sigmoidal crossbedding, with rapid dumping leading to irregular internal layers. A riverine braid-plain is interpreted with overbank flood-deposited siltstones. Sandy, matrix-supported pebbly beds a few tens of metres thick form part of a fluviatile phase a narrow distance stratigraphically above the Patsy Hill Member at the Devil’s Peak, southern Flinders Ranges.

Boundary relationships with underlying and overlying non-glacial units In the Umberatana syncline, the base of the Billy Springs Fm. is at the indistinct transition between the blue-grey calcareous siltstones and light bluish-grey weathering, internally grey limestones of the Wonoka Fm., and a succeeding c. 60 m of finely laminated greygreen fine sandstones. Upwards, these make an abrupt transition into rather massive similarly coloured fine impure sandstones, which commonly include exotic stones and show pervasive evidence of slumping. However, nested stones in a laminated matrix suggest the diamictite association pre-dates the slumping. At the ‘top’ of the stone-bearing section, isolated clasts become increasingly rare, and the fine sandstones are conformably succeeded by coarser, flat- bedded sandstones. There is no ‘cap carbonate’. In the Mount Freeling syncline the one lower contact of the stone-bearing beds examined at Tardlapinna Springs apparently shows a sharp contact over intercalated siltstones and carbonate beds of the Wonoka Fm., but is intensely sheared such that structural rodding of the clasts has occurred. Outcrop of the stonebearing interval is limited upwards.

within the upper member. d34S values on pyrite from the Bunyeroo Fm. at Bunyeroo Gorge vary between þ4.4 and þ30 to þ 35‰ Canyon Diablo meteorite troilite standard (CDT) (McKirdy & Jenkins, unpublished data).

Palaeolatitude and palaeogeography The palaeomagnetic record of the Adelaide Rift Complex is comparatively well studied and is reviewed by Wingate & Giddings (2000). No palaeomagnetic measurements are known from the Billy Springs Fm., but data are available for the correlative Bonney Sandstone (Embleton & Giddings 1974; McWilliams & McElhinny 1980), and analyses have been made on the upper Pertatatataka Fm./Arumbera Sandstone and Todd River Dolomite of the Amadeus Basin central Australia (Embleton 1972; Kirschvink 1978). Klootwijk (1980) drew attention to the likelihood that many Neoproterozoic palaeomagnetic records for the northern Adelaide Rift Complex made up to that time may be unreliable due to overprints related to subsequent tectonism (Wingate & Embleton 2000). The same difficulty was highlighted by Li (2000), who obtained new results from a relatively undisturbed Neoproterozoic glacial sequence in the Kimberley block of Western Australia, and recorded the lack of correspondence with palaeomagnetic measurements of supposedly correlative levels in the Flinders Ranges. Klootwijk (1980) was able to confirm that his palaeomagnetic measurements through the Cambrian to oldest Ordovician of the Amadeus basin agreed closely with results previously made elsewhere in Australia and several parts of the Adelaide Rift Complex. The Australian Cambro –Ordovician apparent polar wander path can be compared with data from Africa (Piper 1976; McClausland et al. 2007) with rotation of the Australian Shield, fragmentary information for Antarctica (tabled in McElhinny & Embleton 1976) and limited data from India (Powel & Li 1993; Torsvik et al. 2001) incorporated in a plate reconstruction (Fig. 68.3). The relatively extensive North America and Baltica measurements tabled in Torsvik et al. (1996, 2001) and McClausland et al. (2007) may also be referenced relative to African data. Striking geological and palaeontological similarities between the Adelaide Rift Complex and northern Baltica during the Ediacaran are supportive of these regions then lying adjacent (Jenkins 2007; Jenkins & Nedin 2007). Based on the above reconstruction, during the latest Ediacaran (c. Pound Subgroup time) to earliest Cambrian the northern Adelaide Rift Complex was located approximately 20 –308 from the adjacent palaeopole, and assuming this corresponded to the Earth’s pole of rotation, was located at high palaeolatitude of c. 70– 808 during deposition of the Billy Springs Fm. (Fig. 68.3).

Chemostratigraphy Geochronological constraints and biostratigraphy An initial study of the C-isotope stratigraphy of the local Ediacaran included results for the Wonoka Fm. and parts of the Billy Springs Fm. (Pell et al.1993). Further analyses made for the Billy Springs Fm. in the Mount Freeling syncline are given by Reid (1992). The prevalent sandstone and siltstone lithologies in the lower parts of the Billy Springs Fm. largely preclude carbonate analyses. Nor are results available for the Umberatana syncline. Measurements through the Wonoka Fm. and Patsy Hill Member of the Bonney Sandstone made in the mid-western and middle northern Flinders Ranges are summarized in Urlwin et al. (1993). A d13Ccarb-isotope curve derived from these data sets is presented in Figure 68.2. Within the Billy Springs Fm., the dominant features are a sharp rise from highly negative (, –6‰) to positive (.6‰) values straddling the boundary between the lower and upper members, followed by a drop back to negative values (as low as – 10‰)

Subsequent to a recent U –Pb zincon age of 794 + 4 Ma indicating a minimum time of deposition for the Skillogalee Dolomite of the Burra Group (Preiss et al. 2009), the dating of the Neoproterozoic section of the Adelaide Rift Complex is uncertain (e.g. Preiss 2000). The Bonney Sandstone includes one detrital zircon indicating an age of 556 + 24 Ma (Ireland et al.1998), representing an imprecise maximum for deposition. The older Arumbera Sandstone of central Australia is similar in lithology and occupies the same homotaxial placement as the Bonney Sandstone. The former has been given an interpreted stratigraphic age of c. 600 Ma, in agreement with the younger (c. 580–550 Ma) appropriate intersections on the 1Nd v. time evolutionary trajectory of the Musgrave (crustal) Block from which the sediments were evidently derived (Zhao et al. 1992). The Croles Hill Diamictite of

BILLY SPRINGS GLACIATION, SOUTH AUSTRALIA

697

Fig. 68.3. Mid Ediacaran palaeogeography of ‘eastern’ hemisphere permitted by palaeomagnetic measurements, with Africa positioned as on the modern globe (Africa and the conjoined South America are sited on the opposite hemisphere and hence not seen). Africa, Laurentia and Baltica are placed by a reference palaeomagnetic pole of common age, dated 613 + 3 Ma for the Adama Diorite (Mali), 615 + 2 Ma for a Long Range dyke (Labrador), and 616 + 3 Ma for the Egersund Dykes, Norway (see data of Torvisk et al. 1996; McClusland et al. 2007). South Island of New Zealand, NZ; Tasmania, T.

the Kununnah Subgroup, NW Tasmania, dated as younger than 582 + 4 Ma, and the Cottons Breccia of the Grassy Group, King Island (Tasmania), likely to be older than 575 + 3 Ma (Calver et al. 2004) are other possible correlatives of the Billy Springs glaciation. The latter is also potentially coeval with the glaciogenic Egan Formation of the Kimberley region, NW Australia, reflecting an event of regional extent (Grey & Corkeron 1998). Two detrital zircons from the older part of the Marino Arkose south of Adelaide give closely comparable 206Pb– 238U SHRIMP ages 655 + 17 and 649 + 17 Ma, providing a combined mean of 652 + 12 Ma (Ireland et al. 1998; Preiss 2000). This level is c. 200 m stratigraphically below a local equivalent of the terminal Cryogenian, glaciogenic Elatina Fm., and the mean age is effectively a maximum for the stratotypic Ediacaran System. The refrigeration corresponding to Bunyeroo Fm. time (Gostin et al. 2011) occupies a homotaxial placement identical to the Mortensnes Tillite of Norway (Jenkins 2007). Furthermore, Zhou et al. (2007) consider that there are sufficient distinctive genera of mid-Ediacaran large acanthomorphic acritarchs in common between south China and Australia (including Gyalosphaeridium Grey ¼ in part to the Chinese ‘Meghystrichospaeridium’) as to indicate that the Upper Member of the Doushantuo Fm. is not older than the upper Dey-Dey Mudstone and Karlaya Limestone of the Officer basin (central southern Australia), implying that the lesser numbers of such taxa in the Lower Member of the Doushantuo Fm. are correlatives of the similar trickle of taxa Grey (2005) documents in the mid Dey-Dey Mudstone. Homotaxial correlation with the stratotypic Ediacaran of the Flinders Ranges places the Dey-Dey Mudstone as equivalent to the Bunyeroo Fm. and the Kalaya Limestone as overlapping the basal Wearing Dolomite Member of the Wonoka Fm. The ‘Acraman impact ejecta layer’ is only a few metres below the ‘top’ of the stone-bearing interval representing the Bunyeroo refrigeration, which is thus likely to be a direct correlative of the (upper) Nantuo Tillite, for which the succeeding cap carbonate

has been given an age of 635 +1.5 Ma (Condon et al. 2005). The true lower boundary of the statotypic Ediacaran lies at the base of the Nuccaleena Fm., c. 1.4 km lower in the succession than the Bunyeroo Fm. (Jenkins 2007). Hence if the 635 + 1.5 Ma for the close of the Nantuo refrigeration is correlative with part of the Bunyeroo Fm., it is unlikely that it dates the beginning of the Ediacaran as widely promulgated. Evidently, the true age for the base of the stratotypic Ediacaran is older, but on present evidence not greater than c. 650 Ma. Fanning & Link (2006) give an indication of a U –Pb zircon SHRIMP age of c. 658 Ma for a tuff in the upper Merinjina Tillite representing the later phase of the ‘Sturtian’ glacials in the northern Flinders Ranges, but the formal evaluation of this work has yet to be published. In terms of known fossil elements of the world-wide Ediacaran soft-bodied biotas (Narbonne 2005), the Billy Springs glaciation pre-dates the local appearance of diverse ‘animalian’ remains (e.g. Jenkins 1995) just as does the 580 + 1 Ma Gaskiers glaciation in Newfoundland (Narbonne & Gehling 2003; Bowring et al. 2003; Condon et al. 2005). It is therefore tempting to regard these glaciations as related in time. With reference to Figure 68.3, existing palaeomagnetic data support both of these ice events as occurring at high or relatively high latitudes, but clearly at a time when the Laurentian and ‘East Gondwana’ apparent polar wander paths had diverged. Thus there is no a priori reason for these late Ediacaran ice ages to be synchronous, and the imprecise dating for the Bonney Sandstone allows for the Billy Springs glaciation to be younger.

Discussion and conclusions The crux of the argument that parts of the Billy Springs Fm. represent a glacial cycle is that a variety of stones of differing lithology occur in finely layered sediment. On various scales, layers

698

R. J. F. JENKINS

were depressed by apparently impacting stones, which rarely reached over a metre in dimensions. Thin carbonate-rich diamictites contain till pellets. Deposition was presumably marine, and clearly on an unstable slope with abundant slumping. Distributary channels evidently concentrated coarse detritus through currents, but this material is also studded with boulders, out of transport equilibrium. While ice blocks overloaded with detritus sank to the bottom, locally, icebergs may also have touched the substrate. The angularity of the common vein quartz erratics suggests they rode as top moraines, implying mountain peaks shedding detritus through frost-wedging above a valley glacier. More rounded carbonate clasts may represent glacial bed load eroded from the local Callanna and Burra Groups. The rarity of basement clasts is an indication the ice was not part of a continental glaciation. A local source is suspected, perhaps uplifted parts of the Mt. Painter and Mt Babbage inliers (Fig. 68.1). Research Associates V.A. Gostin and D. McKirdy, Adelaide University, are thanked for encouragement and helpful discussion. M. Fuller gave technical assistance in preparing the figures. This represents a contribution of the IUGSand UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Bowring, S. A., Myrow, P., Landing, E. & Ramenzani, J. 2003. Geochronological constraints on terminal Neoproterozoic events and the rise of metazoans: Geophysical Research Abstracts, 5, 219. Calver, C. R., Black, L. P., Everard, J. L. & Seymour, D. B. 2004. U –Pb zircon age constraints on late Neoproterozoic glaciation in Tasmania. Geology, 32, 893–896. Coats, R. P. & Blissett, A. H. 1971. Regional and economic geology of the Mount Painter Province. Geological Survey of South Australia Bulletin, 43. Christie-Blick, N. 1993. Billy Springs Formation – possible glacials (?). In: Jenkins, R. J. F., Lindsay, J. F. & Walter, M. R. (eds) Field Guide to the Adelaide Geosyncline and Anadeus Basin. Australian Geological Survey Organization Record 1993/35, 29 – 29. Condon, D., Zhu, M., Bowring, S., Wang, W., Yang, A. & Jin, Y. 2005. U –Pb ages from the Neoproterozoic Doushantuo Formation, China. Science, 1126, 1 –10. DiBona, P. A. 1991. A previously unrecognized late Proterozoic succession: upper Wilpena Group, northern Flinders Ranges, South Australia. South Australian Geological Survey Quarterly Geological Notes, 117, 2 –9. Embleton, B. J. J. 1972. The palaeomagnetism of some Proterozoic – Cambrian sediments from the Amadeus Basin. Earth and Planetary Science Letters, 17, 217– 226. Embleton, B. J. J. & Giddings, J. W. G. 1974. Late Precambrian and Lower Palaeozopic palaeomagnetic results from South Australia and Western Australia. Earth and Planetary Science Letters, 22, 355– 356. Fanning, C. M. & Link, P. K. 2006. Constraints on the timing of the Sturtian glaciation from southern Australia: that is for the true Sturtian. Geological Society of America Abstracts with Programs, 38, 115. Foden, J., Elburg, M. A., Dougherty-Page, J. & Burtt, A. 2006. The timing and duration of the Delamerian Orogeny: correlation with the Ross Orogen and implications for Gondwana assembly. Journal of Geology, 114, 189–210. Forbes, B. G. 1966. The geology of the Maree 1:250 000 map area. Reports of Investigations, Geological Survey of South Australia, 29. Forbes, B. G. & Preiss, W. V. 1987. Stratigraphy of the Wilpena Group. In: Preiss, W. V. (compiler) The Adelaide Geosyncline– Late Proterozoic Stratigraphy, Palaeontology and Tectonics. Geological Survey of South Australia Bulletin, 53, 211–248. Gehling, J. G. 1982. The sedimentology and stratigraphy of the late Precambrian Pound Subgroup, central Flinders Ranges, South Australia. MSc thesis, University of Adelaide, Department of Geology and Geophysics (unpublished). Gostin, V. A., McKirdy, D. M., Webster, L. & Williams, G. E. 2011. Mid-Ediacaran ice-rafting in the Adelaide Geosyncline and Officer

Basin, South Australia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 673–676. Grey, K. 2005. Ediacaran palynology of Australia. Association of Australasian Palaeontologists Memoir, 31, 1 –439. Grey, K. & Corkeron, M. 1998. Late Neoproterozoic stromatolites in glaciogenic successions of the Kimberley region, Western Australia: evidence of a younger Marinoan glaciation. Precambrian Research, 92, 65 – 87. Ireland, T. R., Flo¨ttman, T., Fanning, C. M., Gibson, G. M. & Preiss, W. V. 1998. Development of Early Paleozoic Pacific Margin of Gondwana from detrital zircon ages across the Delamerian Orogen. Geology, 26, 243–246. Jenkins, R. J. F. 1992. The Adelaide Fold Belt: tectonic reappraisal. In: Jago, J. B. & Moore, P. S. (eds) The Evolution of a Late Precambrian – Early Palaeozoic Rift Complex: The Adelaide Geosyncline. Geological Society of Australia, Special Publication, 16, 177– 198. Jenkins, R. J. F. 1993. Concepts of Ediacaran and Ediacarian Systems. In: Jenkins, R. J. F., Lindsay, J. F. & Walter, M. R. (eds) Field Guide to the Adelaide Geosyncline and Anadeus Basin.Australian. Geological Survey Organization Record 1993/35, 7 –14. Jenkins, R. J. F. 1995. The problem of using animal fossils and trace fossils in terminal Proterozoic stratigraphy. Precambrian Research, 73, 51 – 69. Jenkins, R. J. F. 2007. ‘Ediacaran’ as a name for the newly designated terminal Proterozoic period. In: Vickers-Rich, P. & Komarower, P. (eds) The Rise and Fall of the Ediacara Biota. Geological Society of London, Special Publications, 286, 137– 142. Jenkins, R. J. F. & Sandiford, M. 1992. Observations on the tectonic evolution of the southern Adelaide Fold Belt. Tectonophysics, 214, 27 – 36. Jenkins, R. J. F. & Nedin, C. 2007. The provenance and palaeobiology of a new multi-vaned, chambered frondose organism from the Ediacaran (later Neoproterozoic) of South Australia. In: Vickers-Rich, P. & Komarower, P. (eds) The Rise and Fall of the Ediacara Biota. Geological Society of London, Special Publications, 286, 195–222. Jenkins, R. J. F., McKirdy, D. M. & Nedin, C. (compilers) 1998. The Ediacaran in South Australia: proposal and field guide supporting GSSP position ‘C’ at Wearing Dolomite, Flinders Ranges. IUGS Terminal Proterozoic Period Working Group excursion, 16 –22 June 1998, University of Adelaide. Jenkins, R. J. F., Cooper, J. A. & Compston, W. 2002. Age and biostratigraphy of Early Cambrian tuffs from SE Australia and southern China. Journal of the Geological Society of London, 159, 645– 658. Jiang, G., Kaufman, A. J., Christie-Blick, N., Zhang, S. & Wu, H. 2007. Carbon isotope variability across the Ediacaran Yangtze platform in South China: Implications for a large surface to-deep ocean d13C gradient. Earth and Planetary Science Letters, 261, 303– 320. Kirschvink, J. L. 1978. The Precambrian –Cambrian boundary problem: primary, secondary, and transitional paleomagnetic directions from the Amadeus Basin, central Australia. Geological Magazine, 115, 139– 150. Klootwijk, C. T. 1980. Early Palaeozoic palaeomagnetism in Australia. Tectonophysics, 64, 249–332. Knoll, A. H., Walter, M. R., Narbonne, G. M. & Christie-Blick, N. 2006. The Ediacaran Period: a new addition to the geological time scale. Lethaia, 39, 13 – 30. Leeson, B. 1970. Geology of the Beltana 1: 630 000 map area. Reports of Investigations, Geological Survey of South Australia, 35. Li, Z. X. 2000. New palaeomagnetic results from the ‘capdolomite’ of the Neoproterozoic Walsh Tillite, northwestern Australia. Precambrian Research, 100, 59– 370. McCausland, P. J. A., Van der Voo, R. & Hall, C. M. 2007. Circum-Iapetus paleogeography of the Precambrian-Cambrian transition with a new paleomagnetic constraint from Laurentia. Precambrian Research, 156, 125– 152. McElhinny, M. W. & Embleton, B. J. J. 1976. Precambrian and early Palaeozoic palaeomagnetism in Australia. Philosophical Transactions of the Royal Society of London, Series A, 280, 417–431. McFadden, K. A., Huang, J. et al. 2008. Pulsed oxidation and biological evolution in the Ediacaran Doushantuo Formation. Proceedings of the National Academy of Sciences of the USA, 105, 3197– 3202.

BILLY SPRINGS GLACIATION, SOUTH AUSTRALIA

McKirdy, D. M., Sumartojo, J., Tucker, D. H. & Gostin, V. 1975. Organic, mineralogic and magnetic indications of metamorphism in the Tapley Hill Formation, Adelaide Geosyncline. Precambrian Research, 2, 345– 373. McWilliams, M. O. & McElhinny, M. W. 1980. Late Precambrian palaeomagnetism in Australia: the Adelaide Geosyncline. Journal of Geology, 88, 1– 26. Narbonne, G. M. 2005. The Ediacara biota: Neoproterozoic origin of animals and their ecosystems. Annual Review of Earth and Planetary Sciences, 33, 421– 442. Narbonne, G. M. & Gehling, J. G. 2003. Life after snowball: the oldest complex Ediacaran fossils. Geology, 31, 27– 30. Pell, S. D. 1989. Stable isotope composition of organic matter and co-existing carbonate in the late Precabrian of the officer basin:stratigraphic relationships with neighbouring basins and environmental significance. Honours thesis, Adelaide University, Department of Geology and Geophysics (unpublished). Pell, S. D., McKirdy, D. M., Jansyn, J. & Jenkins, R. J. F. 1993. Ediacaran carbon isotope stratigraphy of South Australia-an initial study. Transactions of the Royal Society of South Australia, 117, 153– 161. Piper, J. D. A. 1976. Paleomagnetic evidence for a Proterozoic supercontinent. Philosophical Transactions of the Royal Society of London, Series A, 280, 469–490. Powell, C. Mc. A., Li, Z. X., McElhinny, M. W., Meert, J. G. & Park, J. K. 1993. Palaeomagnetic constraints on timing of the Neoproterozoic breakup of Rodinia and the Cambrian formation of Gondwana. Geology, 21, 889–892. Preiss, W. V. 1987. A synthesis of palaeogeographic evolution of the Adelaide Geosyncline. In: Preiss, W. V. (compiler) The Adelaide Geosyncline – Late Proterozoic Stratigraphy, Palaeontology and Tectonics. Geological Survey of South Australia Bulletin, 53, 315– 409. Preiss, W. V. 1999. PARACHILNA South Australia 1:250 000 Geological Series – Explanatary notes. Primary Industries and Resources South Australia, Adelaide. Preiss, W. V. 2000. The Adelaide Geosyncline of South Australia and its significance in Neoproterozoic continental reconstruction. Precambrian Research, 100, 21 –63. Preiss, W. V. 2005. Global stratotype for the Ediacaran system and Period – the golden spike has been placed in South Australia. MESA Journal, 37, 20– 25. Preiss, W. V., Drexel, J. F. & Reid, A. J. 2009. Definition and age of the Kooringa Member of the Skillogalee Dolomite: host for

699

Neoproterozoic (c. 790 Ma) porphyry-related copper mineralisation at Burra. MESA Journal, 55, 19– 33. Reid, P. W. 1992. Ediacaran (latest Proterozoic) stratigraphic, isotopic and palaeobiological studies in the Flinders Ranges: stratigraphy, structure and stable isotope analysis of the Billy Springs Formation, Mt Freeling syncline, S.A. Preservation and palaeobiology of the Ediacara fauna, central Flinders Ranges, S.A. Honours thesis, Adelaide University, Department of Geology and Geophysics (unpublished). Reid, P. & Preiss, W. V. 1999. PARACHILNA map sheet. South Australia geological survey. Geological Atlas 1; 25000 Series, Sheet SH 54-13. Torsvik, T. H., Smethurst, M. A. et al. 1996. Continental break-up and collision in the Neoproterozoic and Palaeozoic: a tale of Baltica and Lurentia. Earth Science Reviews, 40, 229– 258. Torsvik, T. H., Carter, L. M., Ashwal, L. D., Bhushan, S. K., Pandit, M. K. & Jamviet, B. 2001. Rodinia refined or obscured: palaeomagnetism of the Malani Igneous Suite (N.W. India). Precambrian Research, 108, 319–333. Urlwin, B., Ayliffe, D. J., Jansyn, J., McKirdy, D. M., Jenkins, R. J. F. & Gostin, V. A. 1993. Appendix C: a d13C survey of carbonates in the early Ediacaran Wonoka Formation. In: Jenkins, R. J. F., Lindsay, J. F. & Walter, M. R. (eds) Field Guide to the Adelaide Geosyncline and Amadeus Basin. Australian Geological Survey Organization Record 1993/35, 92 –96. Von der Borch, C. C. & Grady, A. E. 1982. Wonoka Formation and Billy Springs Beds: reconnaissance interpretation. Transactions of the Royal Society of South Australia, 106, 217– 219. Walter, M. R. & Veevers, J. J. 1997. Australian Neoproterozoic palaeogeography tectonics and supercontinent connections. AGSO Journal of Australian Geology and Geophysics, 17, 73 –92. Wingate, M. T. D. & Giddings, J. W. 2000. Age and palaeomagnetism of the Mundine Well dyke swarm, Western Australia: implications for an Australia –Laurentia connection at 755 Ma. Precambrian Research, 100, 335–357. Zhao, J. X., McCulloch, M. T. & Bennett, V. C. 1992. Sm–Nd and U–Pb zircon isotopic constraints on the provenance of sediments from the Amadeus basin, central Australia: evidence for REE fractionation. Geochimica et Cosmochimica Acta, 56, 921– 940. Zhou, C., Xie, G., Mcfadden, K., Xiao, S. & Yuan, X , 2007. The diversification of Doushantuo-Pertatataka acritarchs in south China: causes and biostratigraphic significance. Geological Journal, 42, 229– 262.

Chapter 69 The glacial succession of Sturtian age in South Australia: the Yudnamutana Subgroup WOLFGANG V. PREISS1*, VICTOR A. GOSTIN2, DAVID M. MCKIRDY2, PAUL M. ASHLEY3, GEORGE E. WILLIAMS2 & PHILIP W. SCHMIDT4 1

Geological Survey Branch, Primary Industries and Resources South Australia, GPO Box 1671 Adelaide, South Australia 5001, Australia 2

School of Earth and Environmental Sciences, University of Adelaide, South Australia 5005, Australia 3

School of Earth Sciences, University of New England, Armidale, New South Wales 2351, Australia 4

CSIRO Exploration and Mining, PO Box 136, North Ryde, New South Wales 1670, Australia *Corresponding author (e-mail: [email protected])

Abstract: The record of two Neoproterozoic glaciations in South Australia has been known for about a century. The earlier glaciation, of Sturtian age, is represented by the Yudnamutana Subgroup and is characterized by widespread diamictites with both intrabasinal and extrabasinal clasts, some locally faceted and striated. Associated facies include shallow-water sandstone, bedded and laminated siltstone with lonestones and dropstones, and sedimentary ironstones (mainly ferruginous siltstone and diamictite). Proximal settings adjacent to the Curnamona Province display massive basement-derived conglomerate and gigantic basement megaclasts (up to hundreds of metres across). Sturtian glaciogenic sediments of the Yudnamutana Subgroup unconformably overlie a variety of older rock units, including crystalline basement near basin margins and uppermost Burra Group sediments in the depocentre, and were deposited both in shallow marine shelf environments and in tectonically active rift basins encircling the Curnamona Province, with corresponding increases in total thickness from 100– 300 m to more than 5 km. Recent U–Pb zircon SHRIMP dating of a thin volcaniclastic layer indicates that the waning stages of the Sturtian glaciation occurred at c. 660 Ma. Unlike the deposits of the younger Elatina glaciation, the Yudnamutana Subgroup has so far not yielded reliable palaeomagnetic data.

The deposits of ancient glaciation in South Australia, now known to be of Neoproterozoic age, were discovered near Adelaide by Howchin (1901) and quickly traced during the early 20th century throughout the Mount Lofty, Flinders and Olary ranges (Fig. 69.1). Poorly sorted and poorly bedded diamictite is the most characteristic facies of these deposits, which have mostly been interpreted as glaciomarine sediments on the basis of striated and faceted extrabasinal as well as intrabasinal clasts, and dropstones in associated laminated fine-grained sediments. The genetic term ‘tillite has historically been applied not only to individual diamictite facies, but also as formation names for lithostratigraphic units comprising a variety of rock types beside diamictite, including sandstone, siltstone, grit, conglomerate, ironstone, dolomite and limestone. In this chapter, the non-genetic term ‘diamictite’ is used in the characterization of sedimentary facies to describe unsorted, poorly bedded sediments ranging in grain size from mud to boulders. For formal lithostratigraphic nomenclature, the historical usage of ‘tillite’ is maintained, but with the understanding that formation names using this as the lithological component of the name were deposited under general glacial conditions but are not necessarily exclusively or even dominantly composed of diamictite. To some extent, this lithological variability is accommodated with the use of members. The Neoproterozoic glaciogenic sediments in South Australia (Fig. 69.1) are found mainly among the deposits of an extensive and deeply subsident, mostly marine basin, historically known as the ‘Adelaide Geosyncline’ (Sprigg 1952). Glaciogenic deposits of similar age are also found along the northern margin of the intracratonic Officer Basin, part of the Centralian Superbasin. Deposits of at least two separate Neoproterozoic glacial episodes are widespread in South Australia. The older of these, of mid-Sturtian age (see Table 69.1 and below), is described in this chapter and is termed the ‘Sturt glaciation’ (after the archetypal Sturt Tillite).

The younger Elatina glaciation, of early Marinoan age, is the subject of a companion paper (Williams et al. 2011). In addition, local evidence of glaciation in the mid-Marinoan Bunyeroo Formation is documented by Gostin et al. (2011), while the origin of another, local development of diamictite in the youngest Neoproterozoic deposits of the northern Flinders Ranges (Dibona 1991) has been controversial but a glacial origin is advocated by Jenkins (2011). In some earlier publications, it has been claimed that there were two separate glacial events during Sturtian times, the deposits of which were thought to be separated by a major unconformity (Coats 1973, 1981; Coats & Preiss 1987). However, the present authors accept the arguments by Murrell et al. (1977) of a record of only one glaciation, with advances and retreats, represented by massive diamictite units interfingering with other glacialrelated facies of the Yudnamutana Subgroup.

Stratigraphy The Adelaide(an) System of the Adelaide Geosyncline (Table 69.1) is divided into four chronostratigraphic units, from oldest to youngest: the Willouran, Torrensian, Sturtian and Marinoan Series (Mawson & Sprigg 1950; Sprigg 1952). In terms of the chronometric time scale, the Adelaidean sediments fall within the Cryogenian Period of the Neoproterozoic Era (Table 69.1). Formal lithostratigraphic subdivision of the complete Neoproterozoic succession (Thomson et al. 1964), and subsequent refinements to it, have been summarized by Preiss & Cowley (1999) Preiss et al. (1999) and Preiss (2000). The adjustments to the original definitions of Thomson et al. (1964) described in these papers resulted from sequence stratigraphic analysis, which also confirms the broad temporal significance of the Willouran, Torrensian, Sturtian and Marinoan Series as conceived by the original authors.

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 701– 712. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.69

702

W. V. PREISS ET AL.

Fig. 69.1. Outcropping deposits of the Sturt glaciation in South Australia.

Group and subgroup boundaries now correspond to major sequence boundaries, but further genetic subdivision may be possible with more detailed study. The Yerelina Subgroup contains the glaciogenic deposits of the Elatina glaciation, while the Yudnamutana Subgroup encompasses all the facies deposited during the Sturt glaciation. The latter is the subject of this chapter. The Sturtian Series comprises more than the Sturt glaciation (Table 69.1). The Belair Subgroup, although pre-dating the major regional unconformity at the base of the Yudnamutana Subgroup and lithostratigraphically part of the Burra Group, is by definition of early Sturtian age, though its sedimentary style resembles the remainder of the Burra Group, which is of Torrensian age. However, Mawson and Sprigg (1950) considered the Belair Subgroup to be related to the glaciogenic deposits on the basis of varve-like lamination and fresh feldspar in its basal arkosic quartzite. Although these criteria are not definitive, the presence of extremely rare lonestones in the Mintaro Shale of the Belair Subgroup, 150 km north of the type area, does allow the possibility of

local sea ice (Forbes & Preiss 1987) long before the Sturt glaciation. In addition, the interglacial Nepouie Subgroup was deposited during late Sturtian time (Table 69.1).

Structural framework Adelaide Geosyncline Tectonic subdivision. The Adelaide Geosyncline and its basement

were variably deformed and metamorphosed in the c. 0.5 Ga Delamerian Orogeny. The Delamerian Orogen is divided into a number of domains on the basis of tectonic characteristics (Fig. 69.2), but these, to some extent, reflect original thickness and facies variations in the sedimentary basin: † Stuart Shelf – the thin, relatively undeformed platform cover of the eastern Gawler Craton;

YUDNAMUTANA SUBGROUP, SOUTH AUSTRALIA

703

Table 69.1. Relationship of major lithostratigraphic units of the Adelaide Geosyncline to chronometric and chronostratigraphic timescales Chronometric time scale

Chronostratigraphic units

Lithostratigraphy Supergroup

Early-Mid Cambrian

Moralana

Early Cambrian

Group

Subgroup

Lake Frome Unnamed Hawker

Neoproterozoic

Adelaidean System

Marinoan Series

Ediacaran

Heysen

Wilpena

Unnamed Sandison

Cryogenian

Umberatana

Yerelina Upalinna

Sturtian Series

Nepouie c. 650 Ma Yudnamutana c. 660 Ma Warrina

Burra

Torrensian Series

Belair Bungarider Mundallio c. 790 Ma Emeroo

Willouran Series

Callanna

Curdimurka c. 800 Ma Arkaroola c. 830 Ma

The approximate age estimates in Ma listed for some subgroups are based on various but limited geochronological data from the Adelaide Geosyncline. Estimates based on correlation with dated successions elsewhere are deliberately excluded.

† Torrens Hinge Zone – a mildly deformed zone of transition between the Stuart Shelf and the thick, folded succession of the Adelaide Geosyncline; † Fleurieu Arc – a zone of arcuate NW-verging folds and thrusts adjacent to a southeastern promontory of the Gawler Craton; † Nackara Arc – a southern arcuate belt of relatively upright folds; † North Flinders Zone – a northern arcuate belt of relatively upright folds suggesting dominantly north –south contractional deformation; † Central Flinders Zone – a zone of open dome and basin folds between the Nackara Arc, and North Flinders Zone. As a result of the break-up of Rodinia, an eastern passive continental margin developed during the late Neoproterozoic, although the exact timing of continental separation is still debated. The Fleurieu and Nackara arcs probably lay adjacent to but inboard from this margin, while the Central Flinders and North Flinders Zones are largely intracratonic between the Gawler Craton to the west and the Curnamona Province and the buried Muloorina Ridge to the east and north respectively. Basement geology. Igneous, metamorphic and sedimentary rocks

ranging in age from latest Archaean to Mesoproterozoic make up the complex array of orogenic belts and sedimentary basins preserved in the basement provinces of the Gawler Craton and Curnamona Province. Very little is known of the basement directly underlying the deepest parts of the Neoproterozoic succession of the Adelaide Geosyncline, apart from c. 1.7 Ga psammitic, psammopelitic, pelitic and quartzo-feldspathic metasediments and orthogneisses that form basement inliers in the Mount Lofty Ranges, and the 1720– 1640 Ma Willyama Supergroup of similar lithologies in the Curnamona Province. The latter also contains 1590– 1560 Ma granites, felsic volcanics and clastic metasediments in its northern regions (Benagerie Ridge and Mount Painter Inlier).

Early rifting. The early history of the Adelaide Geosyncline is

characterized by rifting on a large scale. Major rifts of NW – SE (present-day orientation) to north –south orientation are broadly spatially coincident with the neotectonic feature of the Flinders and Mount Lofty Ranges. After initial early Willouran epicratonic clastic and carbonate deposition, the oldest evidence of extension is in the NW-trending Gairdner Dolerite dyke swarm, dated at 827 + 6 Ma (Wingate et al. 1998) and associated mafic lavas, followed by deposition of evaporitic, mixed clastic and carbonate successions .6 km thick in NW-trending late Willouran grabens. The latter sediments were later brecciated and intruded as diapirs in the Flinders Ranges region during deposition of younger strata, and greatly influenced thickness and facies variations in these sediments (Dalgarno & Johnson 1968; Lemon 1985; Dyson 1996). Torrensian-age rifting established the meridional Torrens Hinge Zone as a western rift boundary. Basal arkosic and pebbly clastic fluvial sediments pass up into a series of marine transgressive –regressive cycles. Sturtian rifting. The locus of rifting shifted to the east and north in

Sturtian times, with the development of the Baratta and Yudnamutana Troughs, in which 3 to over 5 km of glacial, glaciomarine and fluvioglacial sediments accumulated, while only a few hundred metres of equivalent sediments were deposited in the shelf regions to the west and south. Subsequent post-glacial transgression spread far westward onto the Gawler Craton, and records a rift-to-sag phase transition. Younger parts of the Neoproterozoic succession, including deposits of the Elatina glaciation, do not show evidence of rifting on a large scale. Young & Gostin (1988) proposed that deposition during the Sturt glaciation took place in two distinct basins corresponding approximately to the North Flinders Zone and Nackara Arc and separated by a depositional high in the Central Flinders Zone. This concept was based on the palaeogeography of the northern basin and the scarce outcrop of Yudnamutana Subgroup in the Central Flinders Zone. However, subsequent drilling of the

704

W. V. PREISS ET AL.

Fig. 69.2. Subdivision of the Delamerian Orogen in South Australia.

Blinman-2 drill hole has shown that a relatively thick and complete Sturtian glacial succession is present at depth in this region.

margin of the Musgrave Province and have not been intersected in drill holes further south toward the depocentre.

Officer Basin

Glaciogenic deposits and associated strata of the Yudnamutana Subgroup

The Officer Basin is part of the Centralian Superbasin (Walter et al. 1995) that underwent large-scale subsidence during the Neoproterozoic. Unlike the Adelaide Geosyncline, there is little evidence for rifting during deposition in most of the Centralian Superbasin, and sedimentation was punctuated by numerous unconformities. Sturtian glaciogenic deposits (Preiss et al. 1978) are known as erosional remnants in the Amadeus, Ngalia and Officer Basins (Areyonga Fm., Naburula Fm. and Chambers Bluff Tillite, respectively), and only in the southeastern part of the Georgina Basin is there evidence of major Sturtian fault-bounded troughs (Walter 1980). In the Officer Basin, the Chambers Bluff Tillite and equivalents in Western Australia are restricted to the southern

The Umberatana Group consists of glacial-related deposits of the Yudnamutana Subgroup (Sturtian age) at the bottom and Yerelina Subgroup (Marinoan age) at the top, separated by the interglacial Nepouie and Upalinna Subgroups.

Description of lithofacies A number of distinct lithofacies make up the formations of the Yudnamutana Subgroup. Most characteristic are the diamictites, but in many sections these are not volumetrically dominant.

YUDNAMUTANA SUBGROUP, SOUTH AUSTRALIA

Diamictite. Diamictite in the Yudnamutana Subgroup is character-

ized by extremely poor to absent sorting and consequent great range in grain size from clay to boulders. The matrix varies from mudstone through silty and sandy mudstone to muddy and silty sandstone, and carbonate (calcite or dolomite) is a major constituent in some areas. Clasts vary from angular to well-rounded, reflecting different pre-glacial transport histories, that is freshly eroded joint fragments or water-worn clasts of fluvial origin. Glacial faceting and striation of clasts are common in some sections, especially in the northeastern Flinders Ranges. Diamictite generally shows little or no internal stratification and individual diamictite beds vary from one metre to tens of metres in thickness. Commonly, bedding can be discerned only at a gross scale. Clast-rich diamictite is characteristic of areas near glaciated source regions, and in places there is a gradation into clastsupported conglomerate facies in the most proximal sections. Clast-poor diamictite is common in distal regions in the basin centre, where clasts tend to be small and widely separated. In extreme cases, clasts are very rare to absent, and massive siltstone is the dominant facies. This type of diamictite matrix is distinguished from other silty facies by the presence of occasional outsized quartz or feldspar sand grains and the general absence of bedding. Siltstone. Siltstone is the most abundant lithotype associated with

diamictite in the Yudnamutana Subgroup. Planar lamination is the most common sedimentary structure and generally reflects subtle variations in grain size and mineralogy. These fine-grained sediments are commonly medium to dark grey or grey-green in colour, though very dark grey, carbonaceous siltstone occurs in distal basin depocentres. Current ripples and small-scale crossbedding occur in some sections. Laminated siltstone with lonestones is a common lithofacies in all formations of the Yudnamutana Subgroup. In special cases, the clasts can be demonstrated to be dropstones, by deforming the underlying laminations and being onlapped by overlying laminae. It is more common, however, to see laminations deflected both above and below clasts by the effects of compaction. One variant of laminated siltstone shows well-defined planar laminae 1–5 mm thick marked by pronounced grain size variation from very fine silt or mud to coarse silt or very fine sand. In places the coarser laminae are graded. One possible interpretation of such laminae is as seasonal varves. Sandstone. Various types of sandstone are also abundant lithofacies associated with diamictite, and commonly show interbedded and interfingering relationships. Quartzose sandstone is mostly well-sorted and dominated by well-rounded quartz of medium sand size, but fine-grained and coarse-grained sandstones are also common. Planar and trough cross-stratification are the dominant sedimentary structures; small-scale trough cross-bedding, in places outlined by fine-grained heavy mineral lamination, is common in many sections. In outcrop, sandstones vary from relatively friable where feldspar is weathered, to dense quartzite where intensely silicified, and some sandstones have a carbonate cement. In some sections, sandstones are highly feldspathic, and may grade into gritty arkose. Feldspathic sandstone tends to be coarsergrained. Graded sandstone beds are much less common than crossbedded units, but do occur where interbedded with siltstone, in particular in the Wilyerpa Fm. where deposited as event beds in the major depocentre of the Baratta Trough. Conglomerate. Conglomerate is a minor lithofacies in the

Yudnamutana Subgroup, and substantial thicknesses are restricted to the most proximal regions. Rarely, thin lenses of sorted conglomerate mark the unconformable base of the Yudnamutana Subgroup, but diamictite or sandstone are more common basal facies.

705

Granite-boulder conglomerate of the Old Boolcoomata Conglomerate Member is a unique occurrence immediately adjacent to the glaciated basement terrain of the Curnamona Province. It interfingers at all scales with fine-grained sediments (Preiss 2006). Arkosic grit, commonly carbonate-cemented, interfingers with both conglomerate and finer grained sediments. Ironstone. Sedimentary ironstone in the Yudnamutana Subgroup is an iron-rich clastic facies restricted to the Baratta Trough. In the least metamorphosed areas in the Central Flinders Zone, the iron is in the form of very finely divided hematite deposited along with clastic silt-sized quartz and clay minerals. Locally there are very thin (,10 mm) red chert bands within the clastic ferruginous sediments. In more highly metamorphosed sections (informally referred to as ‘Braemar ironstone facies’ of the Benda Siltstone), hematite has been recrystallized to metamorphic magnetite; the resulting magnetic susceptibility of the ironstone varies from about 1000 to more than 50 000  1025 S.I. units. All ironstones display the normal sedimentary structures of fine-grained clastic sediments, that is planar lamination, ripple marks, crosslamination and rare graded bedding. Ferruginous diamictite is an important variant, and is texturally similar to non-ferruginous varieties. In places, silty and/or iron-rich limestone and dolomite interfinger with ferruginous siltstone. The geochemistry and mineralogy of the ‘Braemar ironstone facies’ have been studied in the region south of Yunta (Braemar, Razorback Ridge) and north of Olary (Bimbowrie, Outalpa) (Lottermoser & Ashley 2000). The ironstones commonly occur as thin units (up to several metres thick) intercalated with dolomitic and calcareous siltstone (locally manganiferous), diamictite, sandstone, quartzite and dolomite. Two ironstone facies are recognized, one thinly laminated on a millimetric scale, the other diamictic and massive. These are substantially different in macroscopic appearance, but compositionally identical apart from the clasts in the latter. Mineralogically, the ironstones are relatively simple, being commonly composed of dominant fine-grained magnetite, hematite, quartz, carbonate (ferroan dolomite, ferroan calcite), plagioclase, muscovite, chlorite and biotite. There is an antithetic relationship between iron oxides and silicates/carbonate, with a gradation into siltstone and dolomite. Textures imply metamorphic growth of magnetite and hematite, with no preserved evidence for the presence of detrital iron oxides. Mineral assemblages in the ironstones are consistent with the rocks having been metamorphosed to greenschist facies (chlorite to biotite grade). Geochemically, the ironstones display a wide range of iron content (c. 20–80% Fe as Fe2O3), with a corresponding antithetic range of elements that are present as clastic and chemical precipitate components (e.g. Si, Al, Mg, Ca, Na, K, Ti, Zr, Nb, Rb, Sr, Ba). Higher Ca and Mg contents are largely accommodated in carbonates, Na in plagioclase and K in micas. Although the correlations are not strong, rocks with higher iron contents commonly display higher P, As, Cu, Zn and V than the enclosing sedimentary rocks, implying possible co-precipitation of these elements with iron, for example from exhalative sources. Locally, iron-rich siltstone associated with ironstone has high Mn contents, with Mn accommodated in metamorphic garnet (spessartine) and carbonate. Rare earth element (REE) patterns in the ironstones fall into two groups. In lower-iron samples, the patterns are consistent with the REE being derived from clastic (detrital) components, whereas in higheriron samples, they indicate chemical precipitation from seawater. C-isotopic signatures of carbonate minerals from the ironstones are generally negative (d13C values of –5.5 to 0.9‰, mean value –2.6‰: Lottermoser & Ashley 2000) and similar to those of dolomites elsewhere in Neoproterozoic sedimentary successions (see ‘Isotope chemostratigraphy’ below). In summary, the geochemical compositions of the ‘Braemar ironstone facies’ appear to reflect a mixing of clastic detritus with materials precipitated from exhalative sources and coastal seawater.

706

W. V. PREISS ET AL.

Lithostratigraphy

Bungarider Subgroup east of Terowie. Highly variable thicknesses of clean and feldspathic sandstone, commonly with heavy mineral cross-lamination, and laminated siltstone, interfinger with massive diamictite in many sections. Sedimentary clasts derived from the Burra Group are very abundant, but granitic and felsic volcanic clasts are also common. The depositional environment is interpreted as shallow glaciomarine. The previously defined Hansborough Tillite of the Eudunda area corresponds exactly in both lithological association and stratigraphic position to the Appila Tillite, and the term is no longer needed. A thin, local lens of basaltic tuff and agglomerate occurs 7 m below the top of the Appila Tillite south of Depot Creek near Port Augusta (Hopton 1983), but has no known equivalents elsewhere. Attempts to extract zircons for dating have not been successful. The overlying Wilyerpa Fm. is dominated by grey to grey-green laminated siltstone with lonestones and dropstones. Cross-bedded sandstone interbeds are common in the Nackara Arc. A pebbly dolomite (Warcowie Dolomite Member) commonly occurs at or near the base, which appears transitional in the southern part of the Nackara Arc. The Wilyerpa Fm. is interpreted to record Sturtian deglaciation, with increasing water depth and persistence of floating ice. The siltstone-sandstone-diamictite succession originally mapped as ‘Appila Tillite’ in the Eudunda region is now recognized as the Wilyerpa Fm., while the ‘Hansborough Tillite’ is actually Appila.

The lithostratigraphy of the Yudnamutana Subgroup comprises formations and members as defined in the following regions (Table 69.2). Mount Lofty Ranges. Sturt Tillite, the first recorded occurrence of glaciogenic sediments – Howchin (1901). In its type area in Sturt Gorge, the Sturt Tillite is c. 360 m thick and disconformably overlies sediments of the Belair Subgroup. It is dominated by silty-sandy-matrix diamictite showing strong east-dipping cleavage. Clasts are mostly basement derived, possibly from the Gawler Craton, though few clast lithologies have been positively identified as to source. Quartzite, chert and dolomite clasts are probably derived from erosion of lithified Burra Group sediments. Thin laminated siltstone and feldspathic sandstone units are interbedded in the lower part, whereas arkosic grit lenses at the top may occupy fluvioglacial channels. These lenses are disconformably overlain by laminated carbonaceous siltstone and dolomite of the basal Tindelpina Shale Member of the Tapley Hill Fm. The Sturt Tillite is interpreted as glaciomarine. Nackara Arc. Appila Tillite, conformably overlain by Wilyerpa Fm. With a type section in Appila Gorge, the Appila Tillite increases in thickness from 23 m near Bute on the eastern Gawler Craton to c. 1500 m east of Terowie. It disconformably overlies a variety of older formations of the Burra Group: Rhynie Sandstone near Bute and Mount Remarkable, Woolshed Flat Shale and Skillogalee Dolomite at Depot Creek, Belair Subgroup in the Kapunda-Clare-Spalding-Jamestown region, and

Baratta Trough. Pualco Tillite, Benda Siltstone (including Old Boolcoomata Conglomerate Member), Braemar ironstone facies and Holowilena Ironstone, unconformably overlain by Wilyerpa Fm. The great thickness of Yudnamutana Subgroup in the

Table 69.2. Lithostratigraphy of the Yudnamutana Subgroup and relationship with overlying and underlying rocks

NEPOUIE SUBGROUP

Mt Lofty Ranges

Nackara Arc

Tapley Hill Fm.

Tapley Hill Fm.

Tindelpina Shale Member

Tindelpina Shale Member

North Flinders Zone

Yudna-mutana Trough

Peake and Denison Ranges

Officer Basin

Tapley Hill Fm.

Tapley Hill Fm.

Tapley Hill Fm.

(correlatives uncertain)

(correlatives uncertain)

Tindelpina Shale Member

Tindelpina Shale Member

Tindelpina Shale Member

Calthorinna Tillite

Chambers Bluff Tillite

Burra Group

?Burra Group

Baratta Trough

Serle Conglomerate Regional disconformity YUDNAMUTANA SUBGROUP

Sturt Tillite

Wilyerpa Fm.

Wilyerpa Fm.

Warcowie Dolomite Member

Warcowie Dolomite Member

Appila Tillite

Benda Siltstone

Wilyerpa Fm.

Lyndhurst Fm.

Merinjina Tillite

Bolla Bollana Tillite

Local unconformity Holo-wilena Iron-stone

Braemar Ironstone facies Pualco Tillite Fitton Fm. Hamilton Creek Member Regional unconformity Belair Subgroup

Burra Group

Burra Group or basement

Burra Group

Burra Group or basement

YUDNAMUTANA SUBGROUP, SOUTH AUSTRALIA

Baratta Trough, and the locally angular unconformity between the Wilyerpa Fm. and underlying units is attributed to extensional growth faulting. Such basinward down-stepping faults have been mapped at Worumba (Preiss 1985). The Pualco Tillite is similar to the Appila but generally thicker, and contains ferruginous facies variants (Forbes & Cooper 1976) not seen in the southern part of the Nackara Arc. Near Olary, major clean and pebbly sandstone units are interbedded, probably derived from erosion of nearby Curnamona Province basement. The Pualco passes gradationally up into siltstone and fine-grained sandstone of the Benda Siltstone, which also has shallow-water sandstone interdigitations. Unusual lip-shaped quartz-filled voids and/or concretions in bedded siltstone are possibly due to replacement of diagenetic ?evaporite mineral growth, but the lack of distinct crystal form precludes identification of the original mineral. Ferruginous siltstone of the Benda is referred to informally as Braemar ironstone facies. The Wilyerpa Fm. unconformably overlies Benda Siltstone in the Baratta Trough, with basal pebbly Warcowie Dolomite Member. The Wilyerpa Fm. is over 4 km thick in the depocentre (e.g. Bibliando Dome area) and comprises graded sandstone and siltstone event beds deposited in relatively deep water. In the MacDonald Corridor, adjacent to Curnamona Province basement, very thick Wilerpa Fm. overlies the Pualco-Benda succession, c. 4 km thick, with high-angle unconformity (Forbes 1991). North Flinders Zone and Yudnamutana Trough. Fitton Fm., Hamilton Creek Member, Bolla Bollana Tillite (and equivalent Merinjina Tillite), Lyndhurst Fm. The stratigraphy of the whole glaciogenic sequence is generally similar across the northern Flinders Ranges, but shows some differences from the southern regions, for example there are no associated ironstones. In some 50 measured sections, Young & Gostin (1991) identified two major diamictite units, each followed by a mudstone-dominated heterogeneous facies, and interpreted these as resulting from two glacial advance –retreat cycles, but there is uncertainty as to how these correlate between the shelf region to the west and the rifted Yudnamutana Trough. The whole succession unconformably overlies older Adelaidean sedimentary and volcanic rocks or Mesoproterozoic basement granites of the Mount Painter Inlier. Basal diamictites in the North Flinders Zone show the greatest variability: they are discontinuous and contain sandstone and conglomerate lenses. In the western area near Copley (Link & Gostin 1981), the basal unit includes diamictite beds ranging in thickness from a few centimetres to 2 m that may have been deposited directly from ice, while dropstones in laminated siltstone indicate floating ice (Young & Gostin 1988). Further east there is abundant evidence of sediment gravity flows, suggesting steeper depositional slopes. In a section near Mulga Well, on the north limb of the Yankannina Anticline, the basal diamictite is overlain by laminated mudstone containing a diamictite unit up to 25 m thick with abundant granite and pegmatite boulders (Young & Gostin 1990, 1991). Generally, this unit is more homogeneous than the basal diamictite, and the presence of stratified, graded and convolute-bedded diamictites suggests a relatively deep-water, glacial marine environment during a general recession of the glacial ice. In the Yudnamutana Trough, the Hamilton Creek Member of the Fitton Fm. was considered by Young & Gostin to be equivalent to the basal diamictite near Copley, although Coats & Blissett (1971) had mapped the whole Fitton Fm. as confined to the trough. The latter mapping further suggests that the lowest beds (the Hamilton Creek Member of Young & Gostin 1991), occur only in a basal sub-basin within the trough, the origin of which is unclear but may be due to earliest extensional faulting, or earliest glacial scouring, or a combination of both. The total thickness of the Fitton Fm. is 1560 m. The basal Hamilton Creek Member, c. 350 m thick, non-conformably overlies the crystalline basement of the Mount Painter Inlier and consists predominantly of

707

laminated silty mudstone interbedded with granule to boulder conglomerate and some graded sandstone beds. Some of the mudstones are slumped. The conglomerates range from a few centimetres to 20 m in thickness, occurring mainly in the basal 80 m. Most are massive to graded orthoconglomerates but diamictite beds are also present. These carry abundant locally derived granitic clasts, some up to 50 cm in diameter, quartz pebbles and clasts of mafic rock. Isolated clasts, possibly dropstones, occur in the uppermost part of the unit. The Hamilton Creek Member was probably formed as resedimented material derived from local meltwater deposits near the ice front (Young & Gostin 1989a). The laminated mudstones (some slumped) were deposited in a deep tectonic basin adjacent to a growth fault, into which sandstones and conglomerates were brought by episodic sediment gravity flows. The coarse, unweathered granitic debris is reminiscent of locally derived glacial outwash, and the rare dropstones support the presence of floating ice. The remainder of the Fitton Fm. consists mostly of grey-green laminated and cross-laminated mudstone with occasional dropstones, but includes sandstone, calc-silicate, massive to stratified and graded diamictite, pebbly orthoconglomerate and mud-chip conglomerate. The Bolla Bollana Tillite, with its type section also in the Yudnamutana Trough, represents the second glacial advance and consists mainly of stratified diamictite with minor interbeds of mudstone, sandstone, quartzite, dololutite, conglomerate and dolomitic conglomerate that show features of traction and turbidity currents, and other sediment gravity-flow mechanisms. Belperio (1973) mapped intertonguing relationships between these various facies. The thickness of the Bolla Bollana Tillite increases abruptly from south to north across a series of down-stepping growth faults in the Yudnamutana area from c. 200 m on the rift shoulder to c. 2000 m in the rifted Yudnamutana Trough. Dropstones are common in the stratified interbeds. Basementderived clasts such as granite and gneiss are more abundant than in the lower diamictites. The Bolla Bollana Tillite is interpreted to have been deposited mainly by rain-out from floating sediment-laden glacial ice, with some slumped and graded layers indicating submarine mass movement (Young & Gostin 1991). It is uncertain which diamictite units outside the Yudnamutana Trough correlate strictly with the type Bolla Bollana Tillite. Young & Gostin (1991, p. 214) correlated the basal diamictites of the thinner glaciogenic successions in the western shelf region with the basal Hamilton Creek Member in the Yudnamutana Trough. Alternatively, if the Fitton Fm. is confined to the trough as implied by the mapping of Coats & Blissett (1971), then the western basal diamictites may correlate with a lower part of the Bolla Bollana Tillite (Fig. 69.3b). The Lyndhurst Fm. is an upper heterogeneous mudstone facies in the Yudnamutana Trough and represents the waning stages of the Sturtian glaciation. It is absent in some areas but reaches 1200 m in the Yudnamutana Trough. Most of this unit is finely bedded and laminated mudstone, with minor pebbly diamictite, ripple cross-laminated sandstone and conglomerate interbeds. Dropstones are scattered throughout, indicating sporadic iceberg rafting. The Lyndhurst Fm. resembles the Wilyerpa Fm. in the Nackara Arc and Baratta Trough, similarly recording deglaciation, but there is no sign of angular unconformity at its base, such as is evident at the base of the Wilyerpa Fm. near Olary. The thickness of preserved Lyndhurst Fm. varies greatly because of differential erosion at the base of the overlying Tapley Hill Fm. The Yudnamutana Subgroup in the remainder of the Northern Flinders Ranges and Willouran Ranges is only moderately thick, attaining a maximum thickness of c. 1500 m in the Wooltana area. Here the Merinjina Tillite (Coats & Preiss 1987) contains numerous striated and faceted quartzite and felsic volcanic clasts, the latter derived from the c. 1580 Ma Benagerie Volcanics, known from the subsurface Benagerie Ridge. The lower part of the

708

W. V. PREISS ET AL.

Fig. 69.3. Diagrammatic representation of stratigraphic and tectonic relationships of Sturtian lithostratigraphic units along transects in the Nackara Arc (a) and North Flinders Zone (b).

Merinjina Tillite contains green fine-grained mudstone interbeds with dropstones. At Merinjina Well, these rest unconformably on early Willouran Wooltana Volcanics. A small area of a glaciated pavement is preserved here as a cast on the sole of the overlying mudstone (Mirams 1964; Coats & Preiss 1987); this is the only recorded pavement associated with Neoproterozoic glacials in South Australia. The Serle Conglomerate, unconformably overlying the Yudnamutana Subgroup north of Mount Painter, was once considered (Coats 1981; Coats & Preiss 1987) to be equivalent to the Merinjina Tillite, but is better assigned to the Nepouie Subgroup (Dyson 2004) and interpreted as a conglomerate generated by submarine mass flow immediately after the Sturtian glaciation (Young & Gostin 1989b). In the North Flinders Zone and elsewhere in the Adelaide Geosyncline, the Sturtian glacial succession is sharply overlain by widespread transgressive, thick, laminated siltstone of the Tapley Hill Fm. The glaciated shelf appears to have been drowned by a post-glacial eustatic rise in sea level. The lower part of the Tapley Hill Fm., the Tindelpina Shale Member, is characteristically very fine-grained, extremely thinly laminated, pyritic and carbonaceous with interbeds of carbonaceous, planar laminated silty dolomite. Peake and Denison Ranges. Calthorinna Tillite. The Calthorinna Tillite (Ambrose et al. 1981) resembles other occurrences of Sturtian glacials in having massive diamictite units interbedded with siltstone and sandstone and disconformably overlying the Burra Group. A thick feldspathic sandstone unit occurs at its top.

Officer Basin equivalents. Chambers Bluff Tillite. At the northern

margin of the Officer Basin, the Chambers Bluff Tillite disconformably overlies a quartzite-siltstone succession forming an uncertain part of the Burra Group, which in turn unconformably overlies metamorphic and igneous rocks of the Musgrave Province. The type section in the Indulkana Range is c. 500 m thick and comprises massive diamictite with interbedded sandstone and siltstone. Like the Calthorinna Tillite, it has a thick sandstone package at the top. Preiss et al. (1993) raised the possibility that an upper part of this sandstone might be of Marinoan age, separated from the lower by a possible Tindelpina Shale Member equivalent and overlain by a potential cap dolomite. This alternative is, however, far from certain, as no Tapley Hill Fm. is known from elsewhere in the Officer Basin. The mafic Wantapella Volcanics overlie the arenaceous interval, and may be of either Sturtian or Marinoan age. They have no known equivalent in other basins and they are not associated with any glacial facies. Chambers Bluff Tillite intersected in the Nicholson-2 drill hole in the eastern Officer Basin displays cyclic tidal rhythmites in laminated siltstone and fine-grained sandstone interbedded with diamictite (Comalco 1983; Williams 1991). The rhythmites record semidiurnal, diurnal and fortnightly periods. They also display a longer period interpreted as recording the non-tidal annual oscillation of sea level, which is a response to seasonal changes in sea surface temperature and winds and implies open seas (Williams & Schmidt 2004). Deposition of tidal rhythmites is favoured by the drowning of coastal rivers and valleys through a rise of relative sea level (Williams 1991, 2000), and occurs in modern glaciomarine settings (Smith et al. 1990).

YUDNAMUTANA SUBGROUP, SOUTH AUSTRALIA

Facies architecture and depositional environments of the Yudnamutana Subgroup Complex facies relationships characterize the Yudnamutana Subgroup. The lithostratigraphic nomenclature recorded in the literature is largely of historical origin, where different names have been applied to glaciogenic formations in various regions. These represent assemblages of lithofacies, and mapping has generally been insufficiently detailed to permit formal subdivision into genetically meaningful units. Hence the existing formal nomenclature only partly reflects the complexity of facies relationships. Glaciogenic sediments lap out westward near the western margin of the Torrens Hinge Zone, and gradually thicken eastward across a wide shelf region preserved in the Nackara and Fleurieu Arcs, attaining a maximum of c. 1500 m in the eastern Nackara Arc. The Baratta Trough developed by NE–SW extension between this shelf region and the Curnamona Province; this rift event defined the present SW margin of the latter basement province. The southwestern bounding faults of the Baratta Trough are exposed near Worumba, where the thickness of the Wilyerpa Fm. increases dramatically across a series of down-stepping growth faults (Preiss 1985). In the North Flinders Zone, two alternative stratigraphic correlations for the lower units have been discussed above. Current direction measurements by Young & Gostin (1991) suggest a central depositional high within the North Flinders Zone, possibly caused by syn-depositional diapirism, while the overall palaeoslope was to the north. Interpreted stratigraphic and tectonic relationships between Sturtian lithostratigraphic units in southern and northern transects are illustrated diagrammatically in Figure 69.3, showing the effects of syn-depositional faulting.

Isotope chemostratigraphy A limited number of C, O, S and Sr isotopic measurements on the Sturtian glacial succession and overlying carbonate form the basis of a fragmentary chemostratigraphy. Dolostone clasts in the Appila Tillite of the Emeroo Range retain the d13C signatures of their parent carbonate units in the underlying Burra Group (þ2 to þ6‰), while its matrix dolomite has an isotopic composition (mean d13Cdol ¼ –1‰, d18Odol ¼ – 7‰: Crossing & Gostin 1994) suggestive of early diagenetic interaction of rock flour with isotopically light meltwater. The dolomicrite of the lower Wilyerpa Formation in the Blinman-2 drill hole is likewise depleted in 13C, although the unit as a whole records a positive Cisotopic excursion (d13Cdol ¼ –5 to þ2‰: McKirdy et al. 2001). This pattern is at odds with the negative excursions documented by Walter et al. (2000) in organic C from drill hole sections of Sturtian-age tillites on the Stuart Shelf (Amoco SCYW-1a: d13Corg ¼ – 25 to – 30‰) and further south in the Torrens Hinge Zone (SADME Wokurna-2: d13Corg ¼ – 29.5 to –31‰). Dark grey laminated dolomite of the Tindelpina Shale Member at the base of the post-glacial Tapley Hill Fm. at Depot Creek has d13Cdol and d18Odol values of þ1.8‰ and –5.5‰, respectively (Veizer & Hoefs 1976). In the Blinman-2 drill core, dolomitic siltstone in the lower part of the Tapley Hill Fm. hosts a sharp initial negative excursion (d13Cdol ¼ –1.7 to –5‰) followed by a steady climb to þ1.5‰ at the top of the unit (McKirdy et al. 2001, 2005). Here the d13Cdol profile is paralleled by a similar trend in the co-existing organic carbon (d13Corg ¼ –29 to – 26‰ over 79 m; cf. – 34 to –31‰ over 13 m at SCYW-1A on the Stuart Shelf ). At Myall Creek, further south on the Stuart Shelf, the kerogen is even more depleted in 13C (d13Corg ¼ – 36‰). Here Lambert et al. (1984) reported a narrow spread of d13Cdol ( –4 to – 2‰) through the basal 15 m of the Tapley Hill Fm., but without any stratigraphic trend, and isotopically heavy sulphides (d34S ¼ –3 to þ45‰). The latter sulphur isotopic values are typical of the

709

lower Tapley Hill Fm. throughout the Adelaide Geosyncline (see also Lambert et al. 1987; Walter et al. 2000), although in Blinman-2 there is an upward increase in mean sulphide d34S from þ9‰ in the Wilyerpa Fm., through þ19‰ in the lower Tapley Hill Fm., to þ24‰ higher up in the same unit (McKirdy et al. 2001). Carbonates from the Tapley Hill Fm. have 87Sr/86Sr ratios of 0.7105 –0.7400 (Veizer & Compston 1976; Lambert et al. 1984; Foden et al. 2001), significantly higher than the currently accepted Sr-isotopic signature of post-Sturtian glacial seawater (0.7066 – 0.7068: Shields et al. 1997; Jacobsen & Kaufman 1999; Halverson et al. 2007). Across the Flinders Ranges kerogen in the Tindelpina Shale Member of the Tapley Hill Formation displays a west-to-east decrease of its atomic H/C ratio (from 0.25 to 0.01), measuring a regional increase of thermal maturity (McKirdy et al. 1975). Within this data set, d13Corg correspondingly increases from –24 to –15‰. Higher kerogen H/C values (0.5 –0.6) have subsequently been reported for the Tapley Hill Fm. on the Stuart Shelf (Lambert et al. 1984) and adjacent to the Blinman and Worumba Diapirs (McKirdy et al. 2001). The strong, but geographically variable, negative correlation between total organic carbon content (up to 1.1%) and d13Corg evident throughout the Tapley Hill Fm. highlights the influence of post-glacial palaeobathymetry on precursor marine biota (benthic v. planktonic) and hence the C-isotopic signature of the resultant kerogen (McKirdy et al. 2001).

Palaeolatitude and palaeogeography The Sturtian glacial deposits typically display drab colours, indicating a lack of hematitic pigment. In contrast to red beds from the Marinoan Elatina Formation (Williams et al. 2011), most Sturtian rocks are unsuitable for palaeomagnetic study and thus far have not provided reliable palaeomagnetic data. The palaeomagnetism of 82 core and block samples from 18 sites in the Sturtian Holowilena Ironstone in the Worumba and Holowilena South areas and 24 block samples from three sites in the correlative Braemar ironstone facies in the Braemar area has been studied (P. W. Schmidt & G. E. Williams, unpublished data 1996). Natural remanent magnetization (NRM) intensities showed a large variation, probably resulting from lightning strikes. Unblocking temperatures close to the Curie temperature of hematite (680 8C) indicated that hematite is the predominant magnetic phase, although the high intensities after lightning strikes imply that at least some magnetite is present in those samples. Directions of magnetic components identified using principal component analysis are scattered both in geographical coordinates and, after correcting for dip, in stratigraphic coordinates. This indicates that the remanence is most probably not primary, that is it does not date from near the time of deposition but rather is secondary, perhaps related to the time of deformation. Many iron formations and some red beds possess immature iron-bearing mineral assemblages and appear vulnerable to remagnetization in the presence of oxidizing fluids (Turner 1980). The iron mineralogy of the more highly metamorphosed Braemar ironstone facies of the southeastern Nackara Arc is entirely metamorphic (mainly magnetite, lesser hematite; Lottermoser & Ashley 2000), so that remanence at best would date from the time of the metamorphism (i.e. the Delamerian Orogeny).

Geochronology Owing to the dearth of syn-depositional magmatism, the geochronological controls on the ages and duration of Neoproterozoic sedimentation in the Adelaide Geosyncline are extremely

710

W. V. PREISS ET AL.

sparse. Until recently there have been only very broad constraints on the age of the Sturt glaciation, namely the age of magmatism related to the earliest phases of rifting, the age of the base of the Cambrian, and some Rb –Sr shale dates with very large errors. Relevant Rb –Sr isochron data include the age estimate by Compston et al. (1966) of c. 830 Ma (recalculated with new decay constant) on the Willouran Wooltana Volcanics, which are unconformably overlain by Sturtian glacials at Merinjina Well, and the 750 + 53 Ma shale date on the overlying Tapley Hill Fm., sampled from drilling on the Stuart Shelf (Webb & Coats 1980). Given the presence of Rb-bearing detrital minerals, such an age is likely to be an overestimate. The Rb –Sr isochron on the Wooltana Volcanics is consistent with the widely accepted correlation with the Beda Volcanics and the 827 Ma Gairdner Dolerite dykes. Re –Os analysis of organic matter in the Tapley Hill Fm. produced a much younger age of 643.0 + 2.4 Ma (Kendall et al. 2006). This estimate is consistent with the latest U –Pb data on a newly discovered 3-cm-thick volcaniclastic layer in the Yudnamutana Subgroup near Copley (Fanning & Link 2006), giving a zircon SHRIMP age of c. 658 Ma. This is the first direct age determination on the waning stages of the Sturtian glaciation, as the volcaniclastic layer occurs in the upper bedded siltstone unit (possibly Wilyerpa Fm. equivalent).

Discussion The Yudnamutana Subgroup of South Australia contains a wide variety of lithofacies with complex interfingering relationships deposited in different tectonic settings during the very extensive Sturt glaciation of mid-Sturtian age. Extremely rare lonestones in the early Sturtian Belair Subgroup could be a distant echo of earlier glaciation elsewhere. The Yudnamutana Subgroup represents a palaeogeography in which the stable, low-lying Gawler Craton in the west, with no preserved evidence of glacial features, is onlapped by thin glaciomarine deposits at its eastern margin. These deposits thicken gradually eastward, reflecting increased subsidence, across wide shelf regions in the Nackara Arc and North Flinders Zone. The Torrens Hinge Zone at the eastern margin of the Gawler Craton, which had been the site of major rifting during Torrensian times, displays only gradual increases in thickness of Sturtian deposits across it. Active rifting in the Sturtian was confined to the eastern regions – Baratta Trough in the SE, where very thick glaciomarine deposits include ironstone, and Yudnamutana Trough in the NE, where ironstone is absent but there is evidence of steep slopes and resedimentation of glacial deposits. These troughs, together with NNW-trending grabens in western New South Wales, define the margins of the Curnamona Province, which is likely to have been a major source of glacial debris. An analogue to the Sturtian glaciogenic succession may be the facies exposed in the Miocene/Pliocene Yakataga Fm. of coastal Alaska (Eyles 1987). However, the tectonic setting of the Sturtian was in shelf and rifted basin environments rather than a collisional zone. The abundant, crudely stratified diamictites of the Bolla Bollana Tillite and its equivalents, interpreted as glacial-marine outwash deposits from sediment-laden floating ice, may have been concentrated in restricted fault-controlled basins (Young & Gostin 1991). There is great potential for further studies on provenance, both by detrital zircon dating and by petrological comparisons of igneous and metamorphic clasts with exposed basement rocks, to determine likely sources for the glacial deposits and hence to refine the palaeogeographical interpretation. Correlatives of the Sturt glaciation are known in the Centralian Superbasin (Chambers Bluff Tillite, Areyonga Fm., Naburula Fm.), but the presence of glacials of the same age in the Kimberley Region of Western Australia is controversial (contrast the views of Coats & Preiss 1980 and Grey & Corkeron 1998).

Whether exact correlatives exist on other continents remains to be tested by further geochronology. Until then, it should not be assumed that all other occurrences of a lower of two stratigraphically separated Cryogenian glacial successions correlates exactly with the Sturt glaciation. One instance where this has been tested is in the Pocatello Fm. of Idaho, where recent geochronology on volcaniclastics younger than the glacials has yielded an age slightly older than the date obtained for the waning stage of the Sturt glaciation in the Flinders Ranges (Fanning & Link 2003). This chapter summarizes the current understanding of the Sturt glaciation resulting from the work of many generations of investigators, ranging from the pioneer geologists of South Australia, to the regional mappers of the Geological Survey, and to the academic specialists who have elucidated the sedimentology, geochemistry and palaeomagnetic properties of the glaciogenic succession. It includes data derived from unpublished university theses that supplement the authors’ field observations. G. Young, ably assisted in the field by M. Young, is especially thanked for his extensive contribution and collaboration that resulted in a number of published papers on the Sturtian glaciogenic deposits of the northern Flinders Ranges. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512. Preiss publishes with permission of the Deputy Chief Executive, Resources and Infrastructure, PIRSA.

References Ambrose, G. J., Flint, R. B. & Webb, A. W. 1981. Precambrian and Palaeozoic Geology of the Peake and Denison Ranges. Bulletin of the Geological Survey of South Australia, 50. Belperio, A. P. 1973. The stratigraphy and facies of the late Precambrian lower glacial sequence, Mount Painter, South Australia. BSc (Hons) thesis, University of Adelaide (unpublished). Coats, R. P. 1973, COPLEY, South Australia. Explanatory Notes, 1:250 000 geological series. Sheet SH/54-9. Geological Survey of South Australia. Coats, R. P. 1981. Late Proterozoic (Adelaidean) tillites of the Adelaide Geosyncline. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. International Geological Correlation Programme, Project 38: Pre-Pleistocene tillites. Cambridge University Press, Cambridge, 537–548. Coats, R. P. & Blissett, A. H. 1971. Regional and economic geology of the Mount Painter Province. Bulletin of the Geological Survey of South Australia, 43. Coats, R. P. & Preiss, W. V. 1980. Stratigraphic and geochronological reinterpretation of late Proterozoic glaciogenic sequences in the Kimberley region, Western Australia. Precambrian Research, 13, 181– 208. Coats, R. P. & Preiss, W. V. 1987. Stratigraphy of the Umberatana Group. In: Preiss, W. V. (Compiler). The Adelaide Geosyncline. Late Proterozoic Stratigraphy, Sedimentation, Palaeontology and Tectonics. Bulletin of the Geological Survey of South Australia, 53, 125–209. Comalco Aluminium Ltd. 1983. Well-completion report, Nicholson no. 2. South Australian Department of Mines and Energy, Adelaide, Open-file Report 3938. Compston, W., Crawford, A. R. & Bofinger, V. M. 1966. A radiometric estimate of the duration of sedimentation in the Adelaide Geosyncline, South Australia. Journal of the Geological Society of Australia, 13, 229– 276. Crossing, A. R. & Gostin, V. A. 1994. Isotopic signatures of carbonates associated with Sturtian (Neoproterozoic) glacial facies, central Flinders Ranges, South Australia. In: Deynoux, M., Miller, J. M. G., Domack, E. W., Eyles, N., Fairchild, I. J. & Young, G. M. (eds) Earth’s Glacial Record. Cambridge University Press, Cambridge, 165– 175. Dalgarno, C. R. & Johnson, J. E. 1968. Diapiric structures and late Precambrian-early Cambrian sedimentation in Flinders Ranges, South Australia. American Association of Petroleum Geologists, Memoir, 8, 301– 314. Dibona, P. A. 1991. A previously unrecognised Late Proterozoic succession: Upper Wilpena Group, northern Flinders Ranges, South Australia. Quarterly Geological Notes, Geological Survey of South Australia, 117, 2– 9.

YUDNAMUTANA SUBGROUP, SOUTH AUSTRALIA

Dyson, I. A. 1996. A new model for diapirism in the Adelaide Geosyncline. MESA Journal, 3, 41 –48. Dyson, I. A. 2004. Geology of the eastern Willouran Ranges – evidence for earliest onset of salt tectonics in the Adelaide Geosyncline. MESA Journal, 35, 48– 56. Eyles, C. H. 1987. Glacially influenced submarine-channel sedimentation in the Yakataga Formation, Middleton Island, Alaska. Journal of Sedimentary Petrology, 57, 1004–1017. Fanning, C. M. & Link, P. 2003. 700 Ma U– Pb SHRIMP age for Sturtian (!) diamictites of the Pocatello Formation, southeastern Idaho. Geological Society of America Abstracts with Programs, 35, 389. Fanning, C. M. & Link, P. 2006. Constraints on the timing of the Sturtian glaciation from southern Australia; that is for the true Sturtian. Geological Society of America Abstracts with Programs, 38, 115. Foden, J., Barovich, K., Jane, M. & O’Halloran, G. 2001. Sr-isotopic evidence for Late Neoproterozoic rifting in the Adelaide Geosyncline at 586 Ma: implications for a Cu ore-forming fluid flux. Precambrian Research, 106, 291– 308. Forbes, B. G. 1991. OLARY, South Australia, sheet SI542. South Australia. Geological Survey. 1:250 000 Series – Explanatory Notes. Forbes, B. G. & Cooper, R. S. 1976. The Pualco Tillite of the Olary region, South Australia. Quarterly Geological Notes, Geological Survey of South Australia, 60, 2 – 5. Forbes, B. G. & Preiss, W. V. 1987. Stratigraphy of the Burra Group. In: Preiss, W. V. (Compiler). The Adelaide Geosyncline. Late Proterozoic Stratigraphy, Sedimentation, Palaeontology and Tectonics. Bulletin of the Geological Survey of South Australia, 53, 73 –123. Gostin, V. A., McKirdy, D. M., Webster, L. J. & Williams, G. E. 2011. Mid-Ediacaran ice-rafting in the Adelaide Geosyncline and Officer Basin, South Australia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 673 – 676. Grey, K. & Corkeron, M. 1998. Late Neoproterozoic stromatolites in glaciogenic successions of the Kimberley region, Western Australia; evidence for a younger Marinoan glaciation. Precambrian Research, 92, 65 – 87. Halverson, G. P., Duda´s, F. O., Maloof, A. C. & Bowring, S. A. 2007. Evolution of the 87Sr/86Sr composition of Neoproterozoic seawater. Palaeogeography, Paleoclimatology, Palaeoecology, 256, 103– 129. Hopton, D. L. 1983. Environmental analysis of the late Precambrian Appila Tillite equivalent at Depot Flat, southern Flinders Ranges, South Australia. BSc (Hons) thesis, University of Adelaide (unpublished). Howchin, W. 1901. Preliminary note on the existence of glacial beds of Cambrian age in South Australia. Transactions of the Royal Society of South Australia, 25, 10 –13. Jacobsen, S. B. & Kaufman, A. J. 1999. The Sr, C and O isotopic evolution of Neoproterozoic seawater. Chemical Geology, 161, 37 – 57. Jenkins, R. J. F. 2011. Billy Springs glaciation, South Australia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 693–699. Kendall, B., Creaser, R. A. & Selby, D. 2006. Re– Os geochronology of postglacial black shales in Australia: constraints on the timing of ‘Sturtian’ glaciation. Geology, 34, 729–732. Lambert, I. B., Knutson, J., Donnelly, T. H., Etminan, H. & Mason, M. G. 1984. Genesis of copper mineralisation, Myall Creek Prospect, South Australia. Mineralium Deposita, 19, 266– 273. Lambert, I. B., Knutson, J., Donnelly, T. H. & Etminan, H. 1987. Stuart Shelf-Adelaide Geosyncline Copper Province. Economic Geology, 82, 108–123. Lemon, M. N. 1985. Physical modelling of sedimentation adjacent to diapirs and comparison with late Precambrian Oratunga breccia body in central Flinders Ranges, South Australia. Bulletin of the American Association of Petroleum Geologists, 69, 1327–1338. Link, P. K. & Gostin, V. A. 1981. Facies and palaeogeography of Sturtian glacial strata (late Precambrian), South Australia. American Journal of Science, 281, 353–374. Lottermoser, B. G. & Ashley, P. M. 2000. Geochemistry, petrology and origin of Neoproterozoic ironstones in the eastern part of the Adelaide Geosyncline, South Australia. Precambrian Research, 101, 49 –67.

711

Mawson, D. & Sprigg, R. C. 1950. Subdivision of the Adelaide System. Australian Journal of Science, 13, 69 – 72. Mirams, R. C. 1964. A Sturtian glacial pavement at Merinjina Well, near Wooltana. Quarterly Geological Notes, Geological Survey of South Australia, 11, 4– 6 McKirdy, D. M., Sumartojo, J., Tucker, D. H. & Gostin, V. A. 1975. Organic, mineralogic and magnetic indicators in the Tapley Hill Formation, Adelaide Geosyncline. Precambrian Research, 2, 345– 373. McKirdy, D. M., Burgess, J. M. et al. 2001. A chemostratigraphic overview of the late Cryogenian interglacial sequence in the Adelaide Fold-Thrust Belt, South Australia. Precambrian Research, 106, 149– 186. McKirdy, D. M., Gammon, P. R., Smith, H. D., Hayward, H. R. & Sonter, S. 2005. Biogeochemistry of Neoproterozoic cap carbonates – a speculative hypothesis. 22nd International Meeting on Organic Geochemistry, Seville, Spain, Abstracts, 2, 815– 816. Murrell, B., Link, P. K. & Gostin, V. A. 1977. Evidence for only one Sturtian glacial period in the COPLEY map area. Quarterly Geological Notes, Geological Survey of South Australia, 64, 16– 19. Preiss, W. V. 1985. Stratigraphy and tectonics of the Worumba Anticline and associated intrusive breccias. Bulletin of the Geological Survey of South Australia, 52, 85. Preiss, W. V. 2000. The Adelaide Geosyncline of South Australia and its significance in Neoproterozoic continental reconstruction. Precambrian Research, 100, 21 –63. Preiss, W. V. 2006. Old Boolcoomata Conglomerate Member of the Benda Siltstone – Neoproterozoic glacial sedimentation in terrestrial and marine environments in an active rift basin. MESA Journal, 41, 15– 23. Preiss, W. V. & Cowley, W. M. 1999. Genetic stratigraphy and revised lithostratigraphic classification of the Burra Group in the Adelaide Geosyncline. MESA Journal, 14, 30 –40. Preiss, W. V., Walter, M. R., Coats, R. P. & Wells, A. T. 1978. Lithological correlations of Adelaidean glaciogenic rocks in parts of the Amadeus, Ngalia and Georgina Basins. BMR Journal of Australian Geology and Geophysics, 3, 43– 53. Preiss, W. V., Belperio, A. P., Cowley, W. M. & Rankin, L. R. 1993. Neoproterozoic. In: Drexel, J. F., Preiss, W. V. & Parker, A. J. (eds) The Geology of South Australia. Volume 1. Geological Survey of South Australia, Bulletin, 54, 171–203. Preiss, W. V., Dyson, I. A., Reid, P. W. & Cowley, W. M. 1998. Revision of lithostratigraphic classification of the Umberatana Group. MESA Journal, 9, 36 –42. Shields, G. A., Stille, P., Brasier, M. D. & Atudorei, N.-V. 1997. Stratified oceans and oxygenation of the late Precambrian environment: a post glacial geochemical record from the Neoproterozoic of W. Mongolia. Terra Nova, 9, 218– 222. Smith, N. D., Phillips, A. C. & Powell, R. D. 1990. Tidal drawdown: a mechanism for producing cyclic sediment laminations in glaciomarine deltas. Geology, 18, 10 – 13. Sprigg, R. C. 1952. Sedimentation in the Adelaide Geosyncline and the formation of the continental terrace. In: Glaessner, M. F. & Rudd, E. A. (eds) Sir Douglas Mawson Anniversary Volume. University of Adelaide, Adelaide, 153–159. Thomson, B. P., Coats, R. P., Mirams, R. C., Forbes, B. G., Dalgarno, C. R. & Johnson, J. E. 1964. Precambrian rock groups in the Adelaide Geosyncline: a new subdivision. Geological Survey of South Australia. Quarterly Geological Notes, 9, 1 –19. Turner, P. 1980. Continental Red Beds. Elsevier, Amsterdam. Veizer, J. & Compston, W. 1976. The nature of 87Sr/86Sr in Precambrian carbonates as an index of crustal evolution. Geochimica et Cosmochimica Acta, 40, 905– 914. Veizer, J. & Hoefs, J. 1976. The nature of O18/O16 and C13/C12 secular trends in sedimentary carbonate rocks. Geochimica et Cosmochimica Acta, 40, 1387–1395. Walter, M. R. 1980. Adelaidean and Early Cambrian stratigraphy of the southwestern Georgina Basin: correlation chart and explanatory notes. Report Bureau of Mineral Resources, Geology and Geophysics, Australia.

712

W. V. PREISS ET AL.

Walter, M. R., Veevers, J. J., Calver, C. R. & Grey, K. 1995. Neoproterozoic stratigraphy of the Centralian Superbasin, Australia. Precambrian Research, 73, 173–195. Walter, M. R., Veevers, J. J., Calver, C. R., Gorjan, P. & Hill, A. C. 2000. Dating the 840–544 Ma Neoproterozoic interval by isotopes of strontium, carbon, and sulfur in seawater, and some interpretative models. Precambrian Research, 100, 371– 433. Webb, A. W. & Coats, R. P. 1980. A reassessment of the age of the Beda Volcanics on the Stuart Shelf, South Australia. South Australia. Department of Mines and Energy. Report 80/6 (unpublished). Williams, G. E. 1991. Upper Proterozoic tidal rhythmites, South Australia: sedimentary features, deposition, and implications for the earth’s paleorotation. In: Smith, D. G., Reinson, G. E., Zaitlin, B. A. & Rahmani, R. A. (eds) Clastic Tidal Sedimentology. Canadian Society of Petroleum Geologists Memoir, 16, 161– 177. Williams, G. E. 2000. Geological constraints on the Precambrian history of Earth’s rotation and the Moon’s orbit. Reviews of Geophysics, 38, 37 –59. Williams, G. E. & Schmidt, P. W. 2004. Neoproterozoic glaciation: reconciling low paleolatitudes and the geologic record. In: Jenkins, G. S., McMenamin, M., Sohl, L. E. & McKay, C. P. (eds) The Extreme Proterozoic: Geology, Geochemistry and Climate. American Geophysical Union Geophysical Monograph, 146, 145– 159. Williams, G. E., Gostin, V. A., McKirdy, D. M., Preiss, W. V. & Schmidt, P. W. 2011. The Elatina glaciation (late Cryogenian), South Australia. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G.

(eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 713–721. Wingate, M. T. D., Campbell, I. H., Compston, W. & Gibson, G. M. 1998. Ion-probe U– Pb ages for Neoproterozoic basaltic magmatism in south –central Australia and implications for the breakup of Rodinia. Precambrian Research, 87, 135–159. Young, G. M. & Gostin, V. A. 1988, Stratigraphy and sedimentology of Sturtian Glaciogenic deposits in the western part of the North Flinders Basin, South Australia. Precambrian Research, 39, 151–170. Young, G. M. & Gostin, V. A. 1989a. An exceptionally thick upper Proterozoic (Sturtian) glacial succession in the Mount Painter Area, South Australia. Geological Society of America Bulletin, 101, 834– 845. Young, G. M. & Gostin, V. A. 1989b. Depositional environment and regional stratigraphic significance of the Serle Conglomerate: a late Proterozoic submarine fan complex, South Australia. Palaeogeography, Palaeoclimatology, Palaeoecology, 71, 237– 252. Young, G. M. & Gostin, V. A. 1990. Sturtian glacial deposition in the vicinity of the Yankaninna Anticline, North Flinders Basin, South Australia. Australian Journal of Earth Sciences, 37, 447– 458. Young, G. M. & Gostin, V. A. 1991. Late Proterozoic (Sturtian) succession of the North Flinders Basin, South Australia; an example of temperate glaciation in an active rift setting. In: Anderson, J. B. & Ashley, G. M. (eds) Glacial Marine Sedimentation; Paleoclimatic Significance. Geological Society of America, Special Paper, 261, 207– 222.

Chapter 70 The Elatina glaciation (late Cryogenian), South Australia GEORGE E. WILLIAMS1*, VICTOR A. GOSTIN1, DAVID M. MCKIRDY1, WOLFGANG V. PREISS2 & PHILLIP W. SCHMIDT3 1

Discipline of Geology and Geophysics, School of Earth and Environmental Sciences, University of Adelaide, SA 5005, Australia 2

Geological Survey Branch, Primary Industries and Resources South Australia, GPO Box 1671, Adelaide, SA 5001, Australia 3

CSIRO Earth Science and Resource Engineering, PO Box 136, North Ryde, NSW 1670, Australia *Corresponding author (e-mail: [email protected])

Abstract: Deposits of the late Cryogenian Elatina glaciation constitute the Yerelina Subgroup in the Adelaide Geosyncline region, South Australia. They have a maximum thickness of c. 1500 m, cover 200 000 km2, and include the following facies: basal boulder diamictite with penetrative glaciotectonites affecting preglacial beds; widespread massive and stratified diamictites containing faceted and striated clasts, some derived from nearby emergent diapiric islands and others of extrabasinal provenance; laminated siltstone and mudstone with dropstones; tidalites and widespread glaciofluvial, deltaic to marine-shelf sandstones; a regolith of frost-shattered quartzite breccia up to 20 m thick that contains primary sand wedges 3 þ m deep and other large-scale periglacial forms; and an aeolian sand sheet covering 25 000 km2 and containing primary sand wedges near its base. These deposits mark a spectrum of settings ranging from permafrost regolith and periglacial aeolian on the cratonic platform (Stuart Shelf) in the present west, through glaciofluvial, marginal-marine and inner marine-shelf in the central parts of the Adelaide Geosyncline, to outer marine-shelf in sub-basins in the present SE and north. The Elatina glaciation has not been dated directly, and only maximum and minimum age limits of c. 640 and 580 Ma, respectively, are indicated. Palaeomagnetic data for red beds from the Elatina Formation (Fm.) and associated strata indicate deposition of the Yerelina Subgroup within 108 of the palaeoequator. The Yerelina Subgroup is unconformably to disconformably overlain by the dolomitic Nuccaleena Fm., which in most places is the lowest unit of the Wilpena Group and marks Early Ediacaran marine transgression. Supplementary material: Photographs are available at http://www.geolsoc.org.uk/SUP18481.

This chapter discusses the extensive (200 000 km2) glaciogenic facies associated with the late Cryogenian Elatina glaciation of the Marinoan Epoch in the Adelaide Geosyncline region, South Australia (Fig. 70.1), and their palaeomagnetism and palaeoenvironments. Preiss et al. (2011) discuss the preceding Cryogenian (Sturtian age) glacial succession in South Australia and the lithostratigraphy of the Adelaide Geosyncline. The Marinoan Epoch as defined in South Australia encompasses both the late Cryogenian and the Ediacaran (Preiss 1987; Williams et al. 2008). The term ‘Elatina glaciation’ was proposed by Mawson (1949) following his discovery of diamictite containing faceted and striated clasts in Elatina Creek in the central Flinders Ranges. The diamictite belongs to the Elatina Fm. (Preiss 1987; Lemon & Gostin 1990), which has its type section near Enorama Creek (Fig. 70.1). We use Mawson’s terminology for the late Cryogenian glaciation in South Australia. The Elatina glaciation is of global importance for several reasons: (i) its diverse and excellently preserved glacial and periglacial facies represent a de facto type region for late Cryogenian glaciation in general; (ii) the Elatina Fm. has yielded the most robust palaeomagnetic data for any Cryogenian glaciogenic succession; and (iii) the recently established Ediacaran System and Period (Knoll et al. 2004, 2006; Preiss 2005) has its Global Stratotype Section and Point (GSSP) placed near the base of the Nuccaleena Fm. overlying the Elatina Fm. in the central Flinders Ranges (Fig. 70.1).

Structural framework The Adelaide Geosyncline (Preiss 1987) was initiated by rifting of a Precambrian craton, the post-rifting parts of which are now represented by the Gawler Craton in the west and the Curnamona Province in the NE (Fig. 70.1). Feeder dykes for volcanic rocks near the base of the sedimentary succession have been dated at 867 + 47 and 802 + 35 Ma (Zhao & McCulloch 1993; Zhao et al. 1994) and 827 + 6 Ma (Wingate et al. 1998). The early stages of development of the Adelaide Geosyncline were marked by a succession

of major rift cycles, but rifting was less important during the later stages including the late Cryogenian (Preiss 1987, 2000). No volcanism is known in the region during the Elatina glaciation. The Neoproterozoic –early Palaeozoic succession in the Adelaide Geosyncline was deformed by the Delamerian Orogeny at 514– 490 Ma (Drexel & Preiss 1995; Foden et al. 2006). The folded strata of the Delamerian Orogen now form the Flinders Ranges in the north and Mount Lofty Ranges in the Adelaide area in the south. Preiss (2000) identified the following major subdivisions of the Adelaide Geosyncline region based on Delamerian tectonic style (Fig. 70.1): † Cratonic platforms of the Stuart Shelf and the Curnamona Province, with thin, little deformed Neoproterozoic and Cambrian cover; † Torrens Hinge Zone of gentle folding; † Central Flinders Zone of broad dome and basin structures; † North Flinders Zone of arcuate, open to tight folds; † Nackara Arc of long, arcuate, relatively upright folds; † Fleurieu Arc marked by thrusting and tight folding.

Stratigraphy The Neoproterozoic stratigraphy of the Adelaide Geosyncline region is discussed by Coats & Preiss (1987), Preiss (1993, 2000) and Preiss et al. (1998). The Yerelina Subgroup at the top of the Cryogenian Umberatana Group embraces all the glaciogenic formations of the Elatina glaciation (Preiss et al. 1998). The Elatina glaciation is recorded by a wide range of facies, including permafrost regolith and periglacial– aeolian sandstone on the Stuart Shelf, and sandstone, mudstone, siltstone and diamictite of interpreted glaciofluvial, deltaic, marginal marine and marine shelf environments to the east, north and SE (Fig. 70.1). Proposed correlations within the Yerelina Subgroup, its possible division into three sequences, and a suggested relative sea-level curve are shown in Figure 70.2. Measured sections and fence

From: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 713– 721. 0435-4052/11/$15.00 # The Geological Society of London 2011. DOI: 10.1144/M36.70

714

G. E. WILLIAMS ET AL.

Fig. 70.1. Map of the Adelaide Geosyncline and Stuart Shelf showing the distribution of the late Cryogenian Yerelina Subgroup. The periglacial –aeolian Whyalla Sandstone on the Stuart Shelf passes eastwards into glaciofluvial, deltaic and inner marine-shelf facies of the Elatina Fm. and outer marine-shelf (basinal) facies of sandstone, diamictite, and mudstone– siltstone with dropstones of the Yerelina Subgroup in the Adelaide Geosyncline. Isopachs in metres. NFZ, North Flinders Zone; CFZ, Central Flinders Zone. Emergent diapiric islands: 1, Oraparinna Diapir; 2, Enorama Diapir; 3, Blinman Diapir. The Ediacaran GSSP is located in Enorama Creek. The palaeowind rose diagram for the Whyalla Sandstone (308 class interval, radius of circle ¼ 10 observations) represents 29 observations of maximum-dip direction of aeolian cross-strata (Williams 1998). The inset shows palaeolatitudes for Australia during the Elatina glaciation based on combined palaeomagnetic data for the Elatina Fm. (see text); the arrow shows the dominant palaeowind direction determined for the Whyalla Sandstone. Modified from Preiss (1987, 1993) and Williams (1998).

diagrams showing various facies of the subgroup and their suggested correlations are provided by Coats & Preiss (1987) and Lemon & Gostin (1990). The Yerelina Subgroup is unconformably to disconformably overlain by the Ediacaran Wilpena Group.

Glaciogenic deposits and associated strata The thickest deposits and most complete successions of the Yerelina Subgroup occur in the North Flinders Zone and the Nackara Arc (Fig. 70.1).

North Flinders Zone Deposition in the North Flinders Zone commenced, possibly following an erosional break, with the 1070-m-thick Fortress Hill Fm., which comprises laminated siltstone with gritty lenses and scattered dropstones, some faceted, marking the onset of glacial deposition (Coats & Preiss 1987; Preiss et al. 1998). Clast lithologies include granite, quartzite, limestone, oolitic limestone and dolostone. The Fortress Hill Fm. is typical of the dominantly finegrained units of the Yerelina Subgroup that are interpreted by Preiss (1992) as outer marine-shelf deposits.

THE ELATINA GLACIATION, SOUTH AUSTRALIA

NORTH FLINDERS ZONE

HALLETT COVE

WILPENA GROUP

WILPENA GROUP

WILPENA GROUP

WILPENA GROUP

WILPENA GROUP

Elatina

diamictite tidal rhythmites

Ketchowla Siltstone

Ketchowla Siltstone

sandstone

Grampus Quartzite

Grampus Quartzite

Balparana Sandstone

Pepuarta Tillite

Pepuarta Tillite

Mount Curtis Tillite

Gumbowie Arkose

Gumbowie Arkose

sandstone

siltstone

YERELINA SUBGROUP

CENTRAL NACKARA ARC FLINDERS Z., W Flinders Ra. Burra region Olary region

STUART SHELF

Whyalla Sandstone

Reynella Siltstone Member

Formation

Elatina Formation diamictite

Cattle Grid Breccia ?

?

?

sandstone basal diamictite

Fortress Hill Formation ?

WILPENA GROUP

SEQUENCES & RELATIVE SEA-LEVEL < HIGH LOW > M2.3

Fortress Hill Formation

M2.2

M2.1

?

UPALINNA SUBGROUP

The Fortress Hill Fm. is sharply overlain by sandstone and conglomerate at the base of the Mount Curtis Tillite (90 m) that may record a lowering of relative sea level and mark a sequence boundary (Preiss et al. 1998). (Use of ‘Tillite’ in the formal lithostratigraphic terminology of Cryogenian diamictite units in South Australia reflects early practice and does not imply any specific glacial environment.) The Mount Curtis Tillite is a sparse diamictite with erratics of pebble to boulder size, some faceted and striated, in massive and laminated, grey-green dolomitic siltstone. Clast lithologies are mostly quartzite, limestone and dolostone, but also include granite and porphyry (Coats & Preiss 1987). Granite boulders attain 3  8 m. The Mount Curtis Tillite is overlain by the medium-grained, feldspathic Balparana Sandstone (130 m), which contains interbeds and lenses of calcareous siltstone and pebble conglomerate. The Balparana Sandstone is disconformably overlain by the Wilpena Group. The main source for the glaciogenic deposits may have been the Curnamona Province to the present east (Fig. 70.1) and possibly the now-buried Muloorina Ridge immediately north of the North Flinders Zone (Preiss 1987).

Nackara Arc The glaciogenic succession in the Nackara Arc is up to 1500 m thick and has some similarity to that in the North Flinders Zone (Coats & Preiss 1987; Preiss 1992; Preiss et al. 1998). The lowermost, laminated siltstone facies of the Fortress Hill Fm. shows progressively greater amounts of scattered, ice-rafted granules and pebbles. The shallow-water Gumbowie Arkose (45– 90 m) disconformably overlies these early deposits at a possible sequence boundary and is conformably succeeded by the Pepuarta Tillite (120 –197 m), which is a sparse diamictite with scattered clasts up to boulder size in massive and laminated, grey calcareous siltstone. Faceted and striated boulders reach 2.5 m in diameter. Clast lithologies include pink granite, granite gneiss, grey porphyry, quartz-granule conglomerate, various quartzites, and vein quartz. The siltstone facies with scattered large clasts of extrabasinal provenance implies deposition from floating ice. The widespread Grampus Quartzite (60 m) disconformably overlies the Pepuarta Tillite, possibly at a sequence boundary defining a third genetic sequence of the Yerelina Subgroup (Preiss et al. 1998). It is conformably overlain by the laminated to cross-laminated, calcareous, pale grey Ketchowla Siltstone (271 m) (Preiss 1992). The Ketchowla Siltstone contains scattered ice-rafted granules, pebbles and boulders up to 1 m across, and is ascribed by Preiss (1992) to outer marine-shelf deposition under generally waning glacial conditions. It is overlain disconformably by the Nuccaleena Fm., with any Ketchowla Siltstone deposited in the North Flinders Zone having been completely removed by erosion at this sequence boundary (Preiss 2000).

715

Fig. 70.2. Stratigraphy and suggested correlation of the Yerelina Subgroup, which encompasses all the late Cryogenian glaciogenic formations on the Stuart Shelf and in the Adelaide Geosyncline. The Yerelina Subgroup occurs at the top of the Cryogenian Umberatana Group and is unconformably to disconformably overlain by the Ediacaran Wilpena Group. Modified from Preiss et al. (1998), with Marinoan sequence sets M2.1–M2.3 and relative sea-level curves from Preiss (2000).

The outer marine-shelf successions of the Fortress Hill Fm. and Ketchowla Siltstone record the waxing and waning of glacial conditions, respectively. The Pepuarta Tillite and the correlative Mount Curtis Tillite mark the glacial maximum of the Elatina glaciation (Preiss et al. 1998).

Central Flinders Zone Deposition of the relatively thin (30– 130 m), inner marine-shelf, glaciofluvial and deltaic facies of the Elatina Fm. in the Central Flinders Zone (Lemon & Gostin 1990) commenced about the time of the glacial maximum. The Elatina Fm. attests to punctuated glacial retreat, although the entire formation accumulated under glacial conditions. Its reddish hue results from the presence of ultra-fine hematitic pigmentation, which makes the formation ideal for palaeomagnetic study. Regionally, the pre-Elatina surface has a relief of .450 m but shows little relief at outcrop apart from local karst solution features. In its type area the Elatina Fm. unconformably overlies the Trezona Fm. and the Yaltipena Fm. of the Upalinna Subgroup (Lemon & Gostin 1990; Lemon & Reid 1998). The Elatina Fm. around the emergent Enorama Diapir (Fig. 70.1) is c. 100 m thick and includes the following facies: (1)

(2)

(3)

(4)

Boulder conglomerate and basal diamictite (5 m) containing boulders of granite gneiss and other extrabasinal clasts in a muddy and sandy matrix that includes some material from the underlying beds. At Trezona Bore and Bulls Gap 5 – 12 km north of Enorama Creek, the top 1–3 m of the then poorly lithified, preglacial Yaltipena Formation are truncated and deformed (Lemon & Gostin 1990; Lemon & Reid 1998; Williams et al. 2008). Locally, boulders of granite gneiss have been forced into the top of the Yaltipena Fm. These features are classifiable as penetrative, Type A glacitectonites (Evans et al. 2006). Cross-bedded, coarse-grained sandstone (c. 5 m) interpreted as fluvial channel deposits (Lemon & Gostin 1990). This unit is correlative with trough cross-bedded, labile coarsegrained sandstone at the base of the Elatina Fm. in Pichi Richi Pass. Flaser-bedded, muddy and silty sandstone, overlain by sandstone beds 1 m thick showing large ball-and-pillow structures ascribed by Lemon & Gostin (1990) to collapse of the sand bodies into the underlying muds. White, pink to red brown, poorly sorted feldspathic sandstone (20– 40 m) that is regionally extensive and in many places rests on the basal unconformity. The sandstone has indistinct bedding with slumped trough cross-bedded sets 1 m thick and small water-escape structures, and contains thin discontinuous granule layers and rare pebbles to

716

(5)

(6)

(7)

(8)

G. E. WILLIAMS ET AL.

boulders. The bimodal nature of the sandstone, expressed as a dominance of coarse silt to fine sand and very coarse sand to granule fractions, suggests aeolian winnowing in the source area. Lemon & Gostin (1990) concluded the sandstone was rapidly deposited in a subaqueous environment with almost continuous slumping and dewatering. Eyles & Gammon (2007) identified hummocky cross-bedding in sandstone 5–10 km north of Enorama Creek, which they interpreted as indicating a storm-influenced shore-face setting. Laminated reddish grey siltstone and red mudstone with rare dropstones that pass laterally into diamictite with sparse clasts, suggesting deposition from floating ice and a return to glacial conditions. Diamictite partly reworked by shallow-marine currents as marked by the presence of lag gravel layers capped by ripple cross-laminated sandstone within otherwise massive units, in the upper third of the Elatina Fm. west of the Enorama Diapir. A diamictite (Supplementary material, Photo 3) with numerous striated clasts that underlies the Nuccaleena Fm., indicating that a glacial influence persisted to the end of deposition of the Elatina Fm. (Lemon & Gostin 1990). The Elatina Fm. east of the line of diapiric islands is c. 120 m thick and has a thin basal conglomerate overlain by massive diamictite carrying boulders of extrabasinal origin (Lemon & Gostin 1990). This is succeeded by c. 50 m of pink sandstone with indistinct bedding, followed by ripple cross-laminated fine-grained sandstone with rare lonestones. A subaqueous, possibly marine shelf, environment is suggested. Further east in the Chambers Gorge area, a cross-bedded sandstone facies up to 200 m thick may mark a delta at the western margin of the Curnamona Province (Coats & Preiss 1987). Halite casts occur in purple fine-grained sandstones of the Elatina Fm. in this area (Coats 1973).

Near the western margin of the Adelaide Geosyncline, the Elatina Fm. includes tidal rhythmites of siltstone and fine-grained sandstone that formed on a series of ebb-tidal deltas (Williams 1989, 1991, 2000). The rhythmites locally display gravity-slide fold structures (wavelength typically 30–50 cm, height 3–5 cm) and wave-generated ripple marks (wavelength 3 –5 cm, height 1.5 cm), best seen at Warren Gorge where the unit is c. 18 m thick (Williams 1996). Tidal rhythmites are well formed at a more distal setting in Pichi Richi Pass, although exposure is limited. Detailed study of cores from three diamond drill holes through the rhythmites in Pichi Richi Pass, supplemented by data for rhythmites near Hallett Cove, has provided a self-consistent palaeotidal data set spanning 60 years (continuous log 9.4 m long comprising 1580 successive fortnightly neap –spring cycles; Williams 1991) that records information on the Earth’s palaeorotation and the Moon’s orbit in the late Cryogenian: data include 13.1 + 0.1 lunar months/year, 400 + 7 solar days/year, 21.9 + 0.4 hours/solar day, and a mean Earth –Moon distance of 96.5 + 0.5% of the present distance (Williams 1989, 1991, 2000). The rhythmites also record the non-tidal, annual oscillation of sea level (Williams 2004; Williams & Schmidt 2004; Williams et al. 2008), which is a response mostly to seasonal changes in water temperature as well as variation in winds and atmospheric pressure (Roden 1963; Pattullo 1966; Wunsch 1972; Komar & Enfield 1987). The rhythmites were deposited during a high stand of sea level during temporary glacial retreat (Williams et al. 2008). Comparable tidal rhythmites are known from modern glaciomarine settings (Smith et al. 1990; Cowan et al. 1999). The extensive sandstone (facies 4) of the Elatina Fm. may equate with the Gumbowie Arkose and Grampus Quartzite (Fig. 70.2). Diamictite units near the base and at the top of the Elatina Fm. may correlate with the Pepuarta Tillite –Mount Curtis Tillite and the Ketchowla Siltstone, respectively.

Numerous clasts in the Elatina Fm. diamictites are faceted and striated, with striations paralleling the long axes of clasts, and linear series of chattermarks are seen on some boulders. About 40% of the larger clasts are extrabasinal in origin, comprising granite gneiss, red porphyritic dacite, schist, metaquartzite, vein quartz and iron-formation. Smaller clasts of pebble and cobble size are dominated by dolerite and vesicular basalt, and also include dolostone and heavy-mineral banded sandstone. Faceted and striated basalt clasts that are abundant in the diamictite at the top of the Elatina Fm. in the Central Flinders Zone are ascribed to the glacial erosion of rafts of identical volcanic rocks up to 1 km across in the emergent diapiric islands in the area (Mawson 1949; Coats & Preiss 1987; Lemon & Gostin 1990). Hence grounded glaciers persisted on these islands to the end of Elatina deposition. Some of the extrabasinal clasts may have been derived from the Curnamona Province, and Lemon & Gostin (1990) matched other clasts with basement rocks in the Iron Knob area of the Gawler Craton. Derivation directly from the west of the Central Flinders Zone is unlikely, however, because the presence on the Stuart Shelf of a preglacial palaeosol, a late Cryogenian permafrost regolith and an overlying periglacial –aeolian sand sheet indicates that the craton to the west was free of glaciers throughout the late Cryogenian (Williams et al. 2008).

Hallett Cove area In the Hallett Cove area of the southern Adelaide Geosyncline (Fig. 70.1), sandstone of the preglacial Wilmington Fm. at the top of the Upalinna Subgroup is overlain with an erosional contact by the Reynella Siltstone Member of the Elatina Fm., which is exposed in a 120-m-thick coastal section at Marino Rocks 2.5 km north of Hallett Cove (Coats & Preiss 1987). The section comprises four facies: (1) (2)

(3)

(4)

A lowermost, massive, dark red siltstone that contains rare granules and angular dolostone fragments. Siltstone and fine-grained sandstone, including tidalites displaying herringbone cross-bedding, flaser bedding and tidal rhythmites (Williams 1989, 1991, 2000; Williams et al. 2008). Authigenic carbonates form sub-vertical chimneys 1– 2 m in diameter and branching pipes (Kennedy et al. 2008). Calcareous and dolomitic sandstone and siltstone containing angular, granule- to pebble-sized intraclasts of limestone and dolostone. Tepee-like structures and dolostone interbeds with stromatolitic laminae are also present (Dyson & von der Borch 1986). Acicular crystal pseudomorphs in some beds may record evaporite minerals. Massive, granule-bearing siltstone at the top of the section.

The Reynella Siltstone Member at Marino Rocks displays no conclusive evidence of glaciation, but 7 km to the SSW, the member includes a siltstone with a probable dropstone (Coats & Preiss 1987).

Stuart Shelf Periglacial facies that locally overlie the Cryogenian interglacial Tapley Hill Fm. and are directly followed by the Nuccaleena Fm. occur on the Stuart Shelf west of the Adelaide Geosyncline (Fig. 70.1). These stratigraphic relationships show that the periglacial facies are broadly correlative with the Elatina Fm. and other formations of the Yerelina Subgroup (Coats 1981; Coats & Preiss 1987; Preiss 1993; Preiss et al. 1998). The Nuccaleena Fm. marks marine transgression over most of the Stuart Shelf (Forbes & Preiss 1987), indicating that the late Cryogenian periglacial facies formed near sea level.

THE ELATINA GLACIATION, SOUTH AUSTRALIA

Permafrost regolith. A breccia regolith up to 20 m thick, termed the

Cattle Grid Breccia, that developed on an inlier of flat-lying, silicified sandstone of the Mesoproterozoic Pandurra Formation near Mount Gunson, is interpreted to have formed by in situ frost shattering (Williams & Tonkin 1985; Williams 1986, 1994). The Cattle Grid Breccia is similar to breccias of modern, periglacial block fields (White 1976; Washburn 1980) and in situ brecciated bedrock extending to depths of 11þ m below the ground surface associated with former and present permafrost horizons in SE England, Spitsbergen and the Canadian Arctic (Murton 1996). The Cattle Grid Breccia displays two generations of sand wedges up to 3 m or more deep and 2 –3 m wide that show steeplydipping laminae of coarse-grained sandstone and outline polygons 10– 30 m in diameter. Relict bedding in the Cattle Grid Breccia is upturned adjacent to the wedges. These sand wedges are comparable in dimensions, structure and internal fabric with V-shaped primary sand wedges forming today in rubble produced by frost action on bedrock in the dry valleys of Antarctica (Pe´we´ 1959; Washburn 1980). Additional metre-scale periglacial forms displayed by the Cattle Grid Breccia include anticlines, tepee-like structures, truncated mounds, frost-heaved boulders, diapiric breccia injections, and involutions or sags (Williams & Tonkin 1985; Williams 1986). The in situ Cattle Grid Breccia is capped by a layer of reworked breccia up to 2 m thick, interpreted as an active layer (Washburn 1980). Periglacial aeolianite. The Whyalla Sandstone conformably overlies the Cattle Grid Breccia and covers 25 000 km2 in outcrop and subcrop (Coats & Preiss 1987; Preiss 1993; Williams 1998). The flat-lying formation is up to 165 m thick and comprises mainly medium- to very coarse-grained, well-rounded, commonly bimodal quartzose sandstone that shows regional SSE-ward fining. Low angle strata form the dominant stratification type, with cross-bed sets up to 7 m thick occurring mainly in the central area. The basal few metres of the formation display two generations of primary sand wedges up to 1.5 m deep, periglacial involutions and diapiric injections (Williams & Tonkin 1985; Williams 1998). Red gritty siltstones are intercalated locally. The presence of subcritically climbing translatent strata and grainflow and grainfall deposits in the sandstone facies confirms a predominantly aeolian sand-sheet environment, with the attitude of cross-bedding indicating winds directed towards the present SE (Williams 1998).

Boundary relations with underlying and overlying non-glacial units The Fortress Hill Fm. overlies the Upalinna Subgroup, at least locally with an erosional contact at the base of Marinoan sequence set M2.1 (Fig. 70.2; Preiss et al. 1998). In the Adelaide – Hallett Cove region, the Central Flinders Zone and the western region of the Nackara Arc, the Elatina Fm. unconformably overlies the Upalinna Subgroup. This sequence boundary may equate with that at the base of the Gumbowie Arkose in the eastern part of the Nackara Arc, marking the base of Marinoan sequence set M2.2. The contact between the Yerelina Subgroup and the overlying Wilpena Group is a disconformity to low-angle unconformity marking a sequence boundary (Fig. 70.2; Coats & Blissett 1971; Preiss et al. 1998; Preiss 2000; Knoll et al. 2006). Erosion at this sequence boundary is most pronounced in the North Flinders Zone, where in several areas it cuts down through the Balparana Sandstone, Mount Curtis Tillite and Fortress Hill Fm. (Coats et al. 1969, 1973; Ambrose 1973; Preiss 2000), implying up to 1500 m of erosion prior to deposition of the Wilpena Group. The erosion may indicate post-glacial isostatic rebound of adjoining cratons, particularly the Curnamona Province in the NE, and an

717

appreciable time-gap between the end of the Elatina glaciation and Nuccaleena deposition (Schmidt et al. 2009). In most places in the Adelaide Geosyncline and on the Stuart Shelf the Yerelina Subgroup is overlain by the Nuccaleena Fm., which is a persistent marker in the region (Coats & Preiss 1987; Forbes & Preiss 1987; Preiss 2000). The Nuccaleena Fm. is typified by a laminated, pink, buff and cream micritic dolostone unit, with interbedded dolostone, sandstone and mudstone occurring locally at the base. The dolostone unit is several metres thick near the western margin of the Adelaide Geosyncline and on the Stuart Shelf, and 5 –17 m thick in the Central Flinders Zone. At Hallett Cove in the south, however, the Reynella Siltstone Member of the Elatina Fm. is sharply overlain by the pale red and grey Seacliff Sandstone at the base of the Wilpena Group, and the principal dolostone unit of the Nuccaleena Fm. lies c. 70 m stratigraphically above the Elatina Fm. (Forbes & Preiss 1987). The Seacliff Sandstone attains a thickness of 340 m elsewhere in the southwestern Adelaide Geosyncline and intertongues regionally with the Nuccaleena Fm. and overlying Brachina Fm. (Forbes & Preiss 1987; Preiss 1993). The Ediacaran GSSP in Enorama Creek in the Central Flinders Zone appears to be placed within a conformable succession below prominent dolostone beds of the Nuccaleena Fm. and above a 0.2 –0.3-m-thick bed of pale red sandstone that disconformably overlies the Elatina Fm. This sandstone bed may be a local manifestation of the Seacliff Sandstone (Williams et al. 2008).

Chemostratigraphy The characteristic d13CVPDB profile of the dolostone unit of the Nuccaleena Formation in the Adelaide Geosyncline decreases upwards from –1‰ to –2.5‰ at its base to – 2‰ to –3.5‰ at its top (Calver 2000; McKirdy et al. 2005; Knoll et al. 2006). In its shape and absolute values, the profile is broadly similar to those of basal Ediacaran dolostones elsewhere (Kennedy et al. 1998; Halverson et al. 2004). Authigenic carbonate cements in facies 2 of the Reynella Siltstone Member at Hallett Cove yielded d18OVPDB values from –25‰ to þ12‰ and d13CVPDB values from –10‰ to þ10‰ (Kennedy et al. 2008).

Palaeolatitude and palaeogeography Detailed palaeomagnetic studies have been conducted on red beds from the folded Elatina Fm. in the Central Flinders Zone and unfolded equivalent strata on the Stuart Shelf (Embleton & Williams 1986; Schmidt et al. 1991; Schmidt & Williams 1995; Sohl et al. 1999). Embleton & Williams (1986) determined a stable, high-temperature (c. 680 8C) magnetization carried by hematite (mean declination D ¼ 191.98, mean inclination I ¼ – 9.68, a95 ¼ 3.48; inferred palaeolatitude ¼ c. 58) for six sites in exposed tidal rhythmites in Pichi Richi Pass. Their accompanying study of drill cores of the rhythmites also yielded shallow inclinations with respect to bedding, likewise suggesting a low palaeolatitude, with the spread of declinations resulting from uncertainty in bedding attitudes in the cores. Three positive fold tests were executed on small folds (wavelengths 30– 50 cm) that formed by softsediment gravity sliding of the tidal rhythmites; the truncation and delicate scouring of some fold crests show that the folds formed during deposition (Williams 1996; Williams et al. 2008). Sumner et al. (1987) reported a positive fold test but did not provide a magnetic direction. Two detailed soft-sediment fold tests (Schmidt et al. 1991; Schmidt & Williams 1995) found that directions of remanence for the tightest clustering occurred for 66– 67% unfolding, at 99% level of confidence. The two detailed fold tests showed that the magnetization was acquired essentially coeval with deposition and is mostly a detrital remanent

718

G. E. WILLIAMS ET AL.

magnetization (DRM). The positive fold tests confirmed that the structures are folds and not ripples. These soft-sediment fold tests did not, however, confirm a low palaeolatitude. Because the studies in Pichi Richi Pass sampled the geomagnetic field for ,100 years, there were growing objections that the data provided just a virtual geomagnetic pole, that is, a snapshot of the geomagnetic field, and could indicate a geomagnetic excursion or transition. Confirmation of a low palaeolatitude came only when Schmidt & Williams (1995) studied the full succession of the Elatina Fm. in the Central Flinders Zone and obtained a stable, high-temperature component carried by hematite (dip-corrected D ¼ 197.38, I ¼ –5.38, a95 ¼ 7.48) and interpreted as an early chemical remanent magnetization (CRM). This magnetic direction is concordant with directions obtained previously for the Elatina rhythmites and implied a palaeolatitude of 2.7 + 3.78. The presence of sequential magnetic reversals for the Elatina Fm. in the Central Flinders Zone (Schmidt & Williams 1995; Sohl et al. 1999) and a positive tectonic fold-test (Sohl et al. 1999) are consistent with early magnetization. Combined data for 205 samples from the Elatina Fm. (Schmidt & Williams 1995; Sohl et al. 1999, Geological Society of America Data Repository item) yielded a direction of D ¼ 208.38 and I ¼ 2 12.98 (a95 ¼ 4.28), indicating a palaeopole at 43.78S, 359.38E (dp ¼ 2.18, dm ¼ 4.28) and a palaeolatitude of 6.5 + 2.28. The following observations indicate only minor compaction-related inclination shallowing (Williams 2008; Williams et al. 2008): (i) the typical shallow inclination of the Elatina palaeomagnetic remanence obtained for samples of different lithologies (fine-, medium- and coarse-grained sandstone, with mudstone and muddy diamictite being avoided); (ii) the comparable inclinations determined for samples from flat-lying strata in the condensed succession on the cratonic platform and from folded strata in basinal successions; and (iii) the low anisotropy of magnetic susceptibility (mean ,4% for 65 samples), which indicates only slight magnetic foliation (Enkin et al. 2003). The findings are supported by a palaeolatitude of 8.4 þ 6.2/–5.78 determined for the immediately preglacial Yaltipena Fm. (Sohl et al. 1999) and by the low palaeolatitudes (108) determined for late Cryogenian glaciogenic deposits in the Officer Basin, Western Australia (Pisarevsky et al. 2001, 2007). Schmidt et al. (2009) provide further palaeomagnetic and rock magnetic data and sedimentological evidence for late Neoproterozoic successions in southern Australia that demonstrate that the effects of inclination shallowing for the Elatina Fm. are minor. The Elatina data together satisfy all the palaeomagnetic reliability criteria of Van der Voo (1990), and indicate that the magnetization of the Elatina Fm. was acquired close to the time of deposition and that the Elatina glaciation took place at a palaeolatitude of 108.

Geochronological constraints The Elatina glaciation has not been accurately dated, and only broad age limits can be given. A U – Pb age of 657 + 17 Ma was obtained for a zircon grain of uncertain provenance from the Marino Arkose Member of the underlying Upalinna Subgroup (Preiss 2000). Re –Os dating gave an age of 643.0 + 2.4 Ma for black shale from the Tindelpina Shale Member at the base of the Tapley Hill Fm., which overlies glacial deposits of Sturtian age in the Adelaide Geosyncline (Kendall et al. 2006). Zoned igneous zircon from a tuffaceous layer near the top of the Sturtian-age glaciogenic succession gave a SHRIMP U– Pb age of c. 658 Ma (Fanning & Link 2006). Mahan et al. (2007) reported a Th –U –total Pb age of 680 + 23 Ma for euhedral laths of monazite, interpreted as authigenic, from the Enorama Shale of the Upalinna Subgroup. These data, although in part contradictory, provide maximum age constraints for the Elatina glaciation.

Suggested ages for the Elatina glaciation of 635 + 1.2 Ma (Hoffmann et al. 2004) and near 580 Ma (Calver et al. 2004) are based on U –Pb zircon dating of volcanic rocks associated with the Ghaub Fm. in Namibia and diamictites in Tasmania, respectively, that were thought to be coeval with the Elatina glaciation. Zhou et al. (2004) gave a maximum age of 663 + 4 Ma and Condon et al. (2005) a minimum age of 635.2 + 0.6 Ma for the Nantuo glaciation in China, which they equated with the Elatina glaciation. Zhang et al. (2008) reported a SHRIMP U – Pb zircon age of 636 + 4.9 Ma for a tuff near the base of the Nantuo Fm. Whether Cryogenian glaciations correlate worldwide is unclear, however, and the Tasmanian diamictites may be related to Ediacaran glaciation identified on several continents and dated at c. 580 Ma in Newfoundland (Bowring et al. 2003). The above findings provide only maximum and minimum age limits of c. 640 and 580 Ma, respectively, for the Elatina glaciation. Accepting ages of 643 Ma for the Tindelpina Shale Member and 635 Ma for the Elatina glaciation requires high rates of sedimentation for the .4 km of interglacial strata in the Central Flinders Zone. Alternatively, either the Re –Os shale ages could be too young or the Elatina glaciation could be younger than 635 Ma.

Discussion The late Cryogenian Elatina glaciation in South Australia, which occurred at some undetermined time between c. 640 and 580 Ma, is recorded by a wide range of terrestrial and marine facies: permafrost regolith displaying large-scale cryogenic structures; a periglacial –aeolian sand sheet; littoral and neritic deposits including tidalites and evaporites; continental and inner marine-shelf sandstones; boulder diamictite with associated glaciotectonites; inner marine-shelf diamictite; and outer marine-shelf diamictite and laminated mudstone –siltstone with ice-rafted dropstones. This varied succession throws light on the late Cryogenian depositional environments and climate near sea level in South Australia. Penetrative glaciotectonites affecting the preglacial Yaltipena Fm. beneath the Elatina Fm. in the Central Flinders Zone indicate scouring by ice near sea level, but whether the ice was grounded (Lemon & Gostin 1990) or floating (Eyles & Gammon 2007) is uncertain. The presence of faceted and striated clasts of basalt in diamictite at the top of the Elatina Fm. in the same area indicates that grounded glaciers persisted on diapiric islands to the end of Elatina deposition. The regionally extensive sandstones of the Elatina Fm. (Coats & Preiss 1987; Lemon & Gostin 1990) suggest glaciofluvial and deltaic settings near the margins of the Adelaide Geosyncline passing basinwards to a marine-shelf environment. The widespread occurrence of glaciomarine diamictites and fine-grained laminated facies with dropstones in the Pepuarta Tillite and Mount Curtis Tillite implies the calving of icebergs into open seas and the rainout of ice-rafted debris during the glacial maximum. Extensive and persistent open seas continued during deposition of the Elatina Fm., as indicated by wavegenerated ripple marks and the signature of the annual oscillation of sea level displayed by the Elatina rhythmites (Williams 2008; Williams et al. 2008). These observations are not compatible with a frozen-over ocean like that advocated by Hoffman & Schrag (2002). More than a century of research on periglacial geomorphology and processes indicates that large-scale primary sand wedges like those at Mount Gunson record a strongly seasonal periglacial climate (Pe´we´ 1959; Black 1976; Washburn 1980; Karte 1983). In polar regions, thermal contraction cracks 1–10 mm wide and up to 10 m deep, which outline polygons 10 –30 m across, develop in the upper part of permafrost with rapid drops of temperature during repeated severe winters. In arid periglacial areas the contraction cracks are filled by windblown sand and the resulting

THE ELATINA GLACIATION, SOUTH AUSTRALIA

sand wedges show near-vertical lamination. Lateral pressure during summer expansion causes upturning of adjacent permafrost. The claim of Maloof et al. (2002), based on numerical models, that diurnal fluctuations of temperature produced the 3 þ m deep late Cryogenian sand wedges at Mount Gunson is refuted by (i) the shallow (1 m) influence of diurnal temperature changes on permafrost in modern polar regions (Embleton & King 1975) and (ii) the lack of periglacial wedges at high elevations near the equator, where the mean annual air temperature (MAAT) remained below 0 8C for intervals of several millennia during the Pleistocene and temperature fluctuations are mainly diurnal (Hastenrath 1973, 1981; Williams & Schmidt 2004). Applying an actualistic interpretation of periglacial forms and their climatic significance (Washburn 1980; Karte 1983), the suite of large-scale cryogenic structures in the Cattle Grid Breccia and Whyalla Sandstone implies that the following features characterized the late Cryogenian environment near sea level in South Australia (Karte 1983; Williams & Tonkin 1985; Williams 1986, 1994, 1998; Williams et al. 2008): † frigid climate (MAAT of – 12 to –20 8C or lower), producing a thick (up to 20 m) permafrost regolith of frost-shattered bedrock; † strong seasonality (seasonal temperature range as great as 40 8C, with mean monthly temperatures of –35 8C or lower in midwinter and up to þ4 8C in midsummer), indicated by 3þ m-deep primary sand wedges and adjacent upturned permafrost regolith arranged in polygons 10 –30 m across; † aridity (,100 mm mean annual precipitation), limited snow cover, and windiness causing the infilling of thermal contraction-cracks with windblown sand; † summer temperatures above freezing to produce a 2-m-thick active layer of reworked breccia above the in situ Cattle Grid Breccia; † climate cycles on a ka timescale, producing several generations of sand wedges alternating with episodes of destabilization and erosion of the upper part of the permafrost. The implication that the mean annual air temperature rose above freezing both annually and through long-term temperature changes is consistent with observations indicating long-lived and extensive open seas during the Elatina glaciation. Aridity on the Stuart Shelf accords with the presence of casts and pseudomorphs of evaporites in the Elatina Fm. High quality palaeomagnetic data indicate that this frigid, strongly seasonal, arid climate, under which permafrost, icescouring of preglacial beds, and grounded glaciers on emergent diapiric islands all occurred near sea level, existed within 108 of the palaeoequator. Cross-bedding attitudes for the periglacial– aeolian Whyalla Sandstone record palaeo-westerly to palaeonorthwesterly surface winds near the palaeoequator (Fig. 70.1, inset), employing the geographic polarity indicated for late Neoproterozoic Australia (Li et al. 2008). The lack of evidence for elevated topography and late Cryogenian glaciation on the Gawler Craton in the palaeo-west and palaeo-NW (Coats & Preiss 1987; Preiss 1993; Williams et al. 2008) implies that the palaeowind data record a regional palaeowind direction rather than katabatic winds blowing off mountains or an ice sheet. This palaeowind direction is the reverse of the zonal easterlies in low latitudes today. McKirdy et al. (2005) and Gammon (2006) found that over 90% of the Nuccaleena carbonate comprises dolomicrospar with a radical geochemical zonation which, they argued, formed via early diagenetic organogenic dolomitization mediated by sulphate-reducing bacteria. They concluded that it is unlikely the published d13CVPDB profiles of the Nuccaleena Fm. record the C-isotopic composition of post-glacial seawater. Kennedy et al. (2008) interpreted the d18OVPDB and d13CVPDB values for authigenic carbonate cements in facies 2 of the Reynella Siltstone

719

Member as indicating methane hydrate destabilization that terminated the Elatina glaciation and triggered deposition of the Nuccaleena Formation. However, facies 2 occurs well below the top of the member and has disconformable to unconformable contact with the Seacliff Sandstone. Hence the data of Kennedy et al. (2008) evidently record events during the Elatina glaciation, like the episodic destabilization of permafrost on the Stuart Shelf. Zhang et al. (2008) found an extensive (.500 km) claystone between the late Cryogenian, glaciogenic Nantuo Fm. and its overlying Doushantuo cap carbonate in South China, which they interpreted (p. 293) as indicating ‘a time lag between the end of the Nantuo deglaciation and cap carbonate precipitation’. Their conclusion is consistent with the stratigraphic evidence given here for a time-gap between the end of the Elatina glaciation and deposition of the Nuccaleena Fm. The robustness of the Elatina palaeomagnetic data and the presence of a time gap between Elatina and Nuccaleena deposition indicate that palaeomagnetic data for the Nuccaleena Formation (Schmidt et al. 2009; Evans & Raub 2011) cannot be applied to the Elatina Fm. In conclusion, much geological and palaeomagnetic data indicate an enigmatic late Cryogenian glacial environment in South Australia (Williams et al. 2008). Research in progress aims to provide an accurate age for the Elatina glaciation to test proposed correlations with late Cryogenian glaciogenic successions elsewhere. Photographs of this succession are available in the online companion atlas at http://neoproterozoic-glaciations.weebly.com. We thank N. Lemon for discussions and C. Calver and G. Halverson for reviews. W. V. Preiss publishes with the permission of the Executive Director, Minerals and Energy Resources, Primary Industries and Resources South Australia. This represents a contribution of the IUGS- and UNESCO-funded IGCP (International Geoscience Programme) Project #512.

References Ambrose, G. J. 1973. The geology and geochemistry of Adelaidean sediments, Mount Painter Province, South Australia, with emphasis on the Umberatana Group. BSc Honours thesis, University of Adelaide. Black, R. F. 1976. Periglacial features indicative of permafrost: ice and soil wedges. Quaternary Research, 6, 3– 26. Bowring, S. A., Myrow, P., Landing, E. & Ramezani, J. 2003. Geochronological constraints on terminal Neoproterozoic events and the rise of metazoans. NASA Astrobiology Institute General Meeting, Arizona State University, Tempe, abstract 13045, 113– 114. Calver, C. R. 2000. Isotope stratigraphy of the Ediacarian (Neoproterozoic III) of the Adelaide Rift Complex, Australia, and the overprint of water column stratification. Precambrian Research, 100, 121– 150. Calver, C. R., Black, L. P., Everard, J. L. & Seymour, D. B. 2004. U–Pb zircon age constraints on late Neoproterozoic glaciation in Tasmania. Geology, 32, 893– 896. Coats, R. P. 1973. Copley, South Australia. Explanatory Notes, 1:250,000 Geological Series, Sheet SH/54-9. Geological Survey of South Australia, Adelaide. Coats, R. P. 1981. Late Proterozoic (Adelaidean) tillites of the Adelaide Geosyncline. In: Hambrey, M. J. & Harland, W. B. (eds) Earth’s Pre-Pleistocene Glacial Record. Cambridge University Press, Cambridge, 537– 548. Coats, R. P. & Blissett, A. H. 1971. Regional and Economic Geology of the Mount Painter Province. Geological Survey of South Australia Bulletin 43. Coats, R. P. & Preiss, W. V. 1987. Stratigraphy of the Umberatana Group. In: Preiss, W. V. (compiler) The Adelaide Geosyncline. Late Proterozoic Stratigraphy, Sedimentation, Palaeontology and Tectonics. Geological Survey of South Australia Bulletin, 53, 125–209. Coats, R. P., Horwitz, R. C., Crawford, A. R., Campana, B. & Thatcher, D. 1969. Mount Painter Province 1:125,000 geological sheet. Geological Atlas Special Series, Geological Survey of South Australia, Adelaide. Coats, R. P., Callen, R. A. & Williams, A. F. 1973. Copley 1:250,000 geological sheet, SH 54-9. Geological Survey of South Australia, Adelaide.

720

G. E. WILLIAMS ET AL.

Condon, D., Zhu, M., Bowring, S., Wang, W., Yang, A. & Jin, Y. 2005. U –Pb ages from the Neoproterozoic Doushantuo Formation, China. Science, 308, 95 –98. Cowan, E. A., Seramur, K. C., Cai, J. & Powell, R. D. 1999. Cyclic sedimentation produced by fluctuations in meltwater discharge, tides and marine productivity in an Alaskan fjord. Sedimentology, 46, 1109– 1126. Drexel, J. F. & Preiss, W. V. (eds) 1995. The geology of South Australia, volume 2, The Phanerozoic. Geological Survey of South Australia Bulletin 54. Dyson, I. A. & von der Borch, C. C. 1986. A field guide to the geology of the late Precambrian Wilpena Group, Hallett Cove, South Australia. In: Parker, A. J. (compiler) One Day Geological Excursions of the Adelaide Region. Geological Society of Australia, South Australian Division, Adelaide, 17 –40. Embleton, C. & King, C. A. M. 1975. Periglacial Geomorphology. Edward Arnold, London. Embleton, B. J. J. & Williams, G. E. 1986. Low palaeolatitude of deposition for late Precambrian periglacial varvites in South Australia: implications for palaeoclimatology. Earth and Planetary Science Letters, 79, 419–430. Enkin, R. J., Mahoney, J. B., Baker, J., Riesterer, J. & Haskin, M. L. 2003. Deciphering shallow paleomagnetic inclinations: 2. Implications from Late Cretaceous strata overlapping the Insular/Intermontane Superterrane boundary in the southern Canadian Cordillera. Journal of Geophysical Research, 108, B4, 2186, doi:10.1029/2002JB001983. Evans, D. A. D. & Raub, T. D. 2011. Neoproterozoic glacial palaeolatitudes: a global update. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 93 –112. Evans, D. J. A., Phillips, E. R., Hiemstra, J. F. & Auton, C. A. 2006. Subglacial till: Formation, sedimentary characteristics and classification. Earth-Science Reviews, 78, 115– 176. Eyles, N. & Gammon, P. 2007. Ice scours and storms in the Marinoan (Neoproterozoic) Elatina Formation of South Australia. Geological Society of America Abstracts with Programs, 39(6), 631. Fanning, C. M. & Link, P. 2006. Constraints on the timing of the Sturtian glaciation from southern Australia; ie for the true Sturtian. Geological Society of America Abstracts with Programs, 38(7), 115. Foden, J., Elburg, M. A., Dougherty-Page, J. & Burtt, A. 2006. The timing and duration of the Delamerian orogeny: correlation with the Ross Orogen and implications for Gondwana assembly. Journal of Geology, 114, 189–210. Forbes, B. G. & Preiss, W. V. 1987. Stratigraphy of the Wilpena Group. In: Preiss, W. V. (compiler) The Adelaide Geosyncline. Late Proterozoic Stratigraphy, Sedimentation, Palaeontology and Tectonics. Geological Survey of South Australia Bulletin, 53, 211– 248. Gammon, P. R. 2006. Organogenic dolomitisation of a Marinoan cap carbonate. In: Snowball Earth 2006. Monte Verita`, Ticino, Switzerland, 16 –21 July 2006. Halverson, G. P., Maloof, A. C. & Hoffman, P. F. 2004. The Marinoan glaciation (Neoproterozoic) in northeast Svalbard. Basin Research, 16, 297– 324. Hastenrath, S. 1973. Observations on the periglacial morphology of Mts. Kenya and Kilimanjaro, East Africa. Zeitschrift fu¨r Geomorphologie, Supplementband, 16, 161 –179. Hastenrath, S. 1981. The Glaciation of the Ecuadorian Andes. Balkema, Rotterdam. Hoffman, P. F. & Schrag, D. P. 2002. The snowball Earth hypothesis: testing the limits of global change. Terra Nova, 14, 129–155. Hoffmann, K.-H., Condon, D. J., Bowring, S. A. & Crowley, J. L. 2004. U– Pb zircon date from the Neoproterozoic Ghaub Formation, Namibia: constraints on Marinoan glaciation. Geology, 32, 817– 820. Karte, J. 1983. Periglacial phenomena and their significance as climatic and edaphic indicators. GeoJournal, 7, 329– 340. Kendall, B., Creaser, R. A. & Selby, D. 2006. Re– Os geochronology of postglacial black shales in Australia: Constraints on the timing of ‘Sturtian’ glaciation. Geology, 34, 729–732.

Kennedy, M. J., Runnegar, B., Prave, A. R., Hoffmann, K.-H. & Arthur, M. A. 1998. Two or four Neoproterozoic glaciations? Geology, 26, 1059–1063. Kennedy, M., Mrofka, D. & von der Borch, C. 2008. Snowball Earth termination by destabilization of equatorial permafrost methane clathrate. Nature 453, 642– 645. Knoll, A. H., Walter, M. R., Narbonne, G. M. & Christie-Blick, N. 2004. A new period for the geologic time scale. Science, 305, 621–622. Knoll, A. H., Walter, M. R., Narbonne, G. M. & Christie-Blick, N. 2006. The Ediacaran Period: a new addition to the geologic time scale. Lethaia, 39, 13– 30. Komar, P. D. & Enfield, D. B. 1987. Short-term sea-level changes and coastal erosion. In: Nummedal, D., Pilkey, O. H. & Howard, J. D. (eds) Sea-Level Fluctuation and Coastal Evolution. Society of Economic Paleontologists and Mineralogists Special Publication, 41, 17 – 27. Lemon, N. M. & Gostin, V. A. 1990. Glacigenic sediments of the late Proterozoic Elatina Formation and equivalents, Adelaide Geosyncline, South Australia. In: Jago, J. B. & Moore, P. S. (eds) The Evolution of a Late Precambrian– Early Palaeozoic Rift Complex: The Adelaide Geosyncline. Geological Society of Australia Special Publication, 16, 149–163. Lemon, N. M. & Reid, P. W. 1998. The Yaltipena Formation of the Central Flinders Ranges. MESA (Mines and Energy South Australia) Journal, 8, 37 – 39. Li, Z. X., Bogdanova, S. V. et al. 2008. Assembly, configuration, and break-up history of Rodinia: a synthesis. Precambrian Research, 160, 179– 210. Mahan, K. H., Wernicke, B. P. & Jercinovic, M. J. 2007. Th–U –total Pb geochronology of authigenic monazite near the top of the Sturtian –Marinoan interglacial, Adelaide Rift Complex, South Australia. American Geophysical Union Fall Meeting 2007, abstract V34C-07. Maloof, A. C., Kellogg, J. B. & Anders, A. M. 2002. Neoproterozoic sand wedges: crack formation in frozen soils under diurnal forcing during a snowball Earth. Earth and Planetary Science Letters, 204, 1 –15. Mawson, D. 1949. The Elatina glaciation. A third occurrence of glaciation evidenced in the Adelaide System. Transactions of the Royal Society of South Australia, 73, 117– 121. McKirdy, D. M., Gammon, P. R., Smith, H. D., Hayward, H. R. & Sonter, S. 2005. Biogeochemistry of Neoproterozoic cap carbonates – a speculative hypothesis. In: Organic Geochemistry: Challenges for the 21st Century. 22nd International Meeting on Organic Geochemistry, Seville, Spain, September 2005, Abstracts, 2, 815–816. Murton, J. B. 1996. Near-surface brecciation of Chalk, Isle of Thanet, south-east England: a comparison with the ice-rich brecciated bedrocks in Canada and Spitsbergen. Permafrost and Periglacial Processes, 7, 153–164. Pattullo, J. G. 1966. Mean sea level. In: Fairbridge, R. W. (ed.) The Encyclopedia of Oceanography. Reinhold, New York, 475– 479. Pe´we´, T. L. 1959. Sand-wedge polygons (tesselations) in the McMurdo Sound region, Antarctica – a progress report. American Journal of Science, 257, 545– 552. Pisarevsky, S. A., Li, Z. X., Grey, K. & Stevens, M. K. 2001. A palaeomagnetic study of Empress 1A, a stratigraphic drillhole in the Officer Basin: evidence for a low-latitude position of Australia in the Neoproterozoic. Precambrian Research, 110, 93 –108. Pisarevsky, S. A., Wingate, M. T. D., Stevens, M. K. & Haines, P. W. 2007. Palaeomagnetic results from the Lancer 1 stratigraphic drillhole, Officer Basin, Western Australia, and implications for Rodinia reconstructions. Australian Journal of Earth Sciences, 54, 561– 572. Preiss, W. V. (compiler) 1987. The Adelaide Geosyncline. Late Proterozoic Stratigraphy, Sedimentation, Palaeontology and Tectonics. Geological Survey of South Australia Bulletin 53. Preiss, W. V. 1992. The Ketchowla Siltstone and stratigraphy of the Marinoan glacial Yerelina Subgroup. Geological Survey of South Australia Quarterly Geological Notes, 121, 7– 15. Preiss, W. V. 1993. Neoproterozoic. In: Drexel, J. F., Preiss, W. V. & Parker, A. J. (eds) The Geology of South Australia, volume 1, The

THE ELATINA GLACIATION, SOUTH AUSTRALIA

Precambrian. Geological Survey of South Australia Bulletin, 54, 171– 203. Preiss, W. V. 2000. The Adelaide Geosyncline of South Australia and its significance in Neoproterozoic continental reconstruction. Precambrian Research, 100, 21 –63. Preiss, W. 2005. Global stratotype for the Ediacaran System and Period — the Golden Spike has been placed in South Australia. MESA (Mines and Energy South Australia) Journal, 37, 20 –25. Preiss, W. V., Dyson, I. A., Reid, P. W. & Cowley, W. M. 1998. Revision of lithostratigraphic classification of the Umberatana Group. MESA (Mines and Energy South Australia) Journal, 9, 36– 42. Preiss, W. V., Gostin, V. A., McKirdy, D. M., Ashley, P. M., Williams, G. E. & Schmidt, P. W. 2011. The glacial succession of Sturtian age in South Australia: the Yudnamutana Subgroup. In: Arnaud, E., Halverson, G. P. & Shields-Zhou, G. (eds) The Geological Record of Neoproterozoic Glaciations. Geological Society, London, Memoirs, 36, 701–712. Roden, G. I. 1963. Sea level variations at Panama. Journal of Geophysical Research, 68, 5701– 5710. Schmidt, P. W. & Williams, G. E. 1995. The Neoproterozoic climatic paradox: Equatorial palaeolatitude for Marinoan glaciation near sea level in South Australia. Earth and Planetary Science Letters, 134, 107– 124. Schmidt, P. W., Williams, G. E. & Embleton, B. J. J. 1991. Low palaeolatitude of Late Proterozoic glaciation: early timing of remanence in haematite of the Elatina Formation, South Australia. Earth and Planetary Science Letters, 105, 355– 367. Schmidt, P. W., Williams, G. E. & McWilliams, M. O. 2009. Palaeomagnetism and magnetic anisotropy of late Neoproterozoic strata, South Australia: implications for the palaeolatitude of late Cryogenian glaciation, cap carbonate and the Ediacaran System. Precambrian Research, 174, 35– 52. Smith, N. D., Phillips, A. C. & Powell, R. D. 1990. Tidal drawdown: a mechanism for producing cyclic sediment laminations in glaciomarine deltas. Geology, 18, 10 –13. Sohl, L. E., Christie-Blick, N. & Kent, D. V. 1999. Paleomagnetic polarity reversals in Marinoan (ca. 600 Ma) glacial deposits of Australia: Implications for the duration of low-latitude glaciation in Neoproterozoic time. Geological Society of America Bulletin, 111, 1120–1139. Sumner, D. Y., Kirschvink, J. L. & Runnegar, B. N. 1987. Softsediment paleomagnetic field tests of late Precambrian glaciogenic sediments (abstract). Eos (Transactions of the American Geophysical Union), 68, 1251. Van der Voo, R. 1990. The reliability of paleomagnetic data. Tectonophysics, 184, 1– 9. Washburn, A. L. 1980. Geocryology. A Survey of Periglacial Processes and Environments. Wiley, New York. White, S. E. 1976. Rock glaciers and block fields, review and new data. Quaternary Research, 6, 77 – 97. Williams, G. E. 1986. Precambrian permafrost horizons as indicators of palaeoclimate. Precambrian Research, 32, 233–242. Williams, G. E. 1989. Tidal rhythmites: geochronometers for the ancient Earth –Moon system. Episodes, 12, 162– 171. Williams, G. E. 1991. Upper Proterozoic tidal rhythmites, South Australia: sedimentary features, deposition, and implications for the earth’s paleorotation. In: Smith, D. G., Reinson, G. E., Zaitlin, B. A. & Rahmani, R. A. (eds) Clastic Tidal Sedimentology. Canadian Society of Petroleum Geologists Memoir, 16, 161–177.

721

Williams, G. E. 1994. The enigmatic late Proterozoic glacial climate: an Australian perspective. In: Deynoux, M., Miller, J. M. G., Domack, E. W., Eyles, N., Fairchild, I. J. & Young, G. M. (eds) Earth’s Glacial Record. Cambridge University Press, Cambridge, 146– 164. Williams, G. E. 1996. Soft-sediment deformation structures from the Marinoan glacial succession, Adelaide foldbelt: implications for the palaeolatitude of late Neoproterozoic glaciation. Sedimentary Geology, 106, 165– 175. Williams, G. E. 1998. Late Neoproterozoic periglacial aeolian sand sheet, Stuart Shelf, South Australia. Australian Journal of Earth Sciences, 45, 733– 741. Williams, G. E. 2000. Geological constraints on the Precambrian history of Earth’s rotation and the Moon’s orbit. Reviews of Geophysics, 38, 37– 59. Williams, G. E. 2004. The paradox of Proterozoic glaciomarine deposition, open seas and strong seasonality near the palaeoequator: global implications. In: Eriksson, P. G., Altermann, W., Nelson, D. R., Mueller, W. U. & Catuneanu, O. (eds) The Precambrian Earth: Tempos and Events. Developments in Precambrian Geology, vol. 12. Elsevier, Amsterdam, 448– 459. Williams, G. E. 2008. Proterozoic (pre-Ediacaran) glaciation and the high obliquity, low-latitude ice, strong seasonality (HOLIST) hypothesis: Principles and tests. Earth-Science Reviews, 87, 61– 93. Williams, G. E. & Schmidt, P. W. 2004. Neoproterozoic glaciation: reconciling low paleolatitudes and the geologic record. In: Jenkins, G. S., McMenamin, M., Sohl, L. E. & McKay, C. P. (eds) The Extreme Proterozoic: Geology, Geochemistry and Climate. American Geophysical Union Geophysical Monograph, 146, 145– 159. Williams, G. E. & Tonkin, D. G. 1985. Periglacial structures and palaeoclimatic significance of a late Precambrian block field in the Cattle Grid copper mine, Mount Gunson, South Australia. Australian Journal of Earth Sciences, 32, 297– 300. Williams, G. E., Gostin, V. A., McKirdy, D. M. & Preiss, W. V. 2008. The Elatina glaciation, late Cryogenian (Marinoan Epoch), South Australia: Sedimentary facies and palaeoenvironments. Precambrian Research, 163, 307–331. Wingate, M. T. D., Campbell, I. H., Compston, W. & Gibson, G. M. 1998. Ion-probe U –Pb ages for Neoproterozoic basaltic magmatism in south-central Australia and implications for the breakup of Rodinia. Precambrian Research, 87, 135– 159. Wunsch, C. 1972. Bermuda sea level in relation to tides, weather, and baroclinic fluctuations. Reviews of Geophysics and Space Physics, 10, 1– 49. Zhang, S., Jiang, G. & Han, Y. 2008. The age of the Nantuo Formation and Nantuo glaciation in South China. Terra Nova, 20, 289– 294. Zhao, J.-X. & McCulloch, M. T. 1993. Sm– Nd mineral isochron ages of Late Proterozoic dyke swarms in Australia: evidence for two distinct events of mafic magmatism and crustal extension. Chemical Geology, 109, 341–354. Zhao, J.-X., McCulloch, M. T. & Korsch, R. J. 1994. Characterization of a plume-related 800 Ma magmatic event and its implications for basin formation in central – southern Australia. Earth and Planetary Science Letters, 121, 349–367. Zhou, C., Tucker, R., Xiao, S., Peng, Z., Yuan, X. & Chen, Z. 2004. New constraints on the ages of Neoproterozoic glaciations in South China. Geology, 32, 437– 440.

Index Page numbers in italics refer to Figures; page numbers in bold refer to Tables. Abenab Subgroup, 201 stratigraphy, 201 abiotic dolomite, 76 Abu Mahara Group, Jabal Akhdar, 259, 251 –260 boundary relations, 258 carbon-isotopic data, 256 chemical and mineralogical indices of alteration, 258 chemostratigraphy, 256 CIA, 256 geochronological constraints, 258– 260 geological map, 253 glaciogenic deposits and associated strata, 253–256 Hadash Formation summary log, 257 histograms of zircon ages derived by LA-ICPMS, 259 Neoproterozoic outcrops and salt basins of Oman, 252 palaeolatitude and palaeogeography, 258 sedimentological logs through Fiq Formation, 254 stratigraphy, 253 structural framework, 252 –253 zircon geochronology, 259 Acaciella australica Stromatolite Assemblage, 122, 125 Acraman impact structure, 674 Adelaide Geosyncline map, 714 relationship of major lithostratigraphic units, 703 Adelaide Geosyncline, South Australia, 673–675 Acraman impact structure, 674 basement geology, 703 biostratigraphy, 674 chemostratigraphy, 673– 674 early rifting, 703 geochronological constraints, 674 glaciogenic deposits and associated strata, 673 Mid-Ediacaran ice-rafting, 673–675 palaeolatitude and palaeogeography, 674 stratigraphy, 673 structural framework, 702 –704 Sturtian rifting, 703–704 tectonic subdivision, 702– 703 Adelaide Geosyncline Cryogenian glaciogenic formations, 715 Adelaide Rift Complex locality maps, 694 Adrar area stratigraphy, 165 Africa geological sketch map, 186 Kaigas and Numees formations, Port Nolloth Group, 223– 230 Karoetjes Kop and Bloupoort formations, Gifberg Group, 233–237 Katanga Supergroup, 173–182 knowledge base, 12 Neoproterozoic glacial palaeolatitudes, 104– 105 Neoproterozoic ice age data set, 3 Port Nolloth Group, 223– 230 Taoudeni Basin, 163– 169, 167 West Congo Supergroup, 185–192 Agassiz, L., 20, 29 Aisa Formation stratigraphy, 321 Aitken, J. D., 25 Akademikerbreen Group, 577 Akkajaure Nappe Complex, 609 Aksu–Wusi area composite stratigraphic log, 371 geological map, 370 glaciogenic diamictites, 369– 370 Akwokwo Tillite, 191

Alaska, 389–395 Hula Hula diamictite, 379– 385 Aldan Shield, 302 Altaids, 100–101 Altungol diamictites, 372 Altungol Formation stratigraphy, 370 Amadeus Basin, 688 glaciogenic deposits and associated strata, 684–685 Inindia beds, 685 Pioneer Sandstone, 685 Amazonia and environs, 105 Andre´e Land Group glaciogenic deposits and associated strata, 586–587 stratigraphy, 583–584 Antelope Island, 430 Anti-Atlas Mountains of Morocco, 105 Arac¸uaı´ orogen, SE Brazil, 523–531 aragonite cements, 76 Araras group, 489, 495 glacially influenced sedimentation and carbonates, 487–496 northern Paraguay Belt stratigraphy, 491 Arctic Alaska–Chukotka Plate (AACP), 390 Arena Formation, 587 Areyonga Formation, 684–685 Argentina Tandilia System, 565–568 Arroyo del Soldado Group stratigraphic chart, 550 Aruwimi Group, 189 Assem Limestone, 273 astronomical theories on Neoproterozoic glaciation, 26 Atar Group, 105 Aties Formation, 234 Atud diamictite boundary relations, 281 geochronological constraints, 282 glaciogenic deposits and associated strata, 281 location, 280 Australia, 673–675, 693–698. See also biostratigraphy of Australia biostratigraphy, 113–130 cap-carbonate sequences, 73–74 glacial succession of Sturtian age, 701–710 glaciogenic succession map, 12 global correlation, 126 –129 Kimberley Region, 659–671 knowledge base, 14–15 Nackara Arc, 69, 706, 715–716 Neoproterozoic ice age data set, 13 palaeoequators, 103 palaeomagnetic constraints on Neoproterozoic glacial palaeolatitudes, 102–103 stromatolites, 121– 125 Sturt glaciation stratigraphy, 124 –125 Australia, central, 677–688 boundary relationships, 686 cap dolomites, 678– 681 characteristics, 687 chemostratigraphy, 686– 687 geochronological constraints, 687 glaciogenic deposits and associated strata, 682–686 Neoproterozoic basins, 682 Neoproterozoic glacial deposits, 677 –688, 678–681 Neoproterozoic stratigraphy, 682, 683 palaeolatitude and palaeogeography, 687 structural framework, 677–682 autochthon, 606–607 Avalon assemblage, 146 Avalonia palaeomagnetic constraints, 105

Avalon Peninsula map, 468 rocks, 468 stratigraphy, 469 Ayn Formation, Dhofar, Oman, 239– 247 boundary relations, 243 carbon-isotope composition of post-glacial carbonate, 243 chemostratigraphy, 243– 245 frequency distribution of detrital zircons from Mirbat Group, 246 geochronological constraints, 246– 247 glaciogenic deposits and associated strata, 241–243 carbonate facies, 243 carbonate-filled fissures, 243 carbonate mass-flow facies association, 243 distal glaciomarine facies association, 241–243 fluviodeltaic facies association, 241 post-glacial carbonate, 243 proximal glaciomarine facies association, 241 shallow-water facies association, 243 subaerial facies association, 241 MIA in Mirbat Group, 245 Neoproterozoic glaciation, nature, 247 Neoproterozoic weathering, chemical index of alteration (CIA), 243– 245 outcrop areas of Neoproterozoic basement, 240 palaeolatitude and palaeogeography, 245– 246 sedimentological logs of Ayn Formation, 242 stratigraphy, 241 structural framework and basement geology, 239–241 summary log data, 244 Backbone Ranges Formation, 399 Baicalia burra Stromatolite Assemblage, 123 Baikal Group, 325 Bakouma Basin, 190 Bakoye Group, 165 Baltica, 100 Baltoscandian craton, 620 Baltoscandian ice sheet, 619 Bambuı´ formation correlations, 511 Bambuı´ Group, Southern Sa˜o Francisco Basin glaciogenic deposits and associated strata, 512 lithostratigraphic and chemostratigraphic correlations, 516 Neoproterozoic successions, 509– 519 stratigraphy, 537 banded iron formation (BIF), Egypt boundary relations, 281 geochronological constraints, 282 glaciogenic deposits and associated strata, 281 Ba˚ngona˚ive window, 626 Bangui Basin, 190 Barakun Formation, 298 Baratta Trough, 706– 707 Barents Shelf, 571 Barite mineralization, 67– 77, 169 Basal Mahd Group diamictite, central Arabian Shield, 277, 280 boundary relations, 281 geochronological constraints, 282 Bas-Congo Basin, 187–188 Baydaric microcontinent, 331 Bayixi Formation glaciogenic diamictites, 372 stratigraphy, 370 Baykonur Formation, 303–306 age, 306 boundary relations, 305

724

Baykonur Formation (Continued) characteristics, 305 chemostratigraphy, 305 conglomerates, 305 dolomite, 305 geochronological constraints, 305 glaciogenic deposits and associated strata, 303– 306 outcrops, 304 palaeolatitude and palaeogeography, 305 sandstones, 305 sedimentary environments, 305 –306 shales, 305 stratigraphy, 303 structural framework, 303 type sections, 304 Bebedouro Formation geological map, 504 ice-contact glaciomarine system, 505 schematic representation of glacial lithofacies, 506 Una Group, Bahia (Brazil), 503– 507 Beck Spring Dolomite, 464 Beck Spring Formation, 464 Bedgroup 19, 586 Bedgroup 20, 586 –587 Bethanis Member, 199 bidirectional climate change, 20 Biexibastao Formation glaciogenic diamictites, 374 stratigraphy, 373 Big Cottonwood Formation, 428 Big Creek roof pendant map, 440 Billy Springs Formation, 697 Billy Springs glaciation, South Australia, 693–698 Adelaide Rift Complex locality maps, 694 boundary relations, 696 chemostratigraphy, 696 geochronological constraints and biostratigraphy, 696– 697 glaciogenic deposits and associated strata, 694– 696 Mid Ediacaran palaeogeography, 697 palaeolatitude and palaeogeography, 696 stratigraphy, 693– 694 structural framework, 693 Bimbo sandstones, 190 bimodal volcanic rock, 439 biogenic dolomite, 76 biological theories, 26 biostratigraphy of Australia, 113 –130 Bitter Springs anomaly, 55, 57 Blackrock Canyon Limestone stratigraphy, 429 Black Stump Arkose, 686 Blackwelder, Eliot, 21 Blaini Formation of Lesser Himalaya, India, 347– 353 age plot for detrital zircons, 353 boundary relations, 352 chemostratigraphy, 352–353 diamictites, 350 dolomite, 350–352 geochronological constraints, 353 geological map of fold-and-thrust belt, 348 glaciogenic deposits and associated strata, 348– 352 interpretation, 349– 350, 352 lithostratigraphic and biostratigraphic subdivision, 349 lithostratigraphic sections and chemostratigraphy, 351 locations of measured sections, 350 massive and laminated diamictite lithofacies, 348– 349 Neoproterozoic–Cambrian succession, 349 palaeolatitude and palaeogeography, 353 sandstones, siltstones and shales, 350 significance of isotopic analyses, 352– 353 stratigraphy, 348 structural framework, 347– 348

INDEX

Blaubeker Formation, 213, 215 glaciogenic deposits and associated strata, 213 Bloeddrif Member, 227–228 Bloupoort Formation, 233–237 boundary relations, 235 chemostratigraphy, 235–236 correlation of diamictite, 237 diamictite-hosting, 235 glaciogenic deposits and associated strata, 234 –235, 235 palaeolatitude and palaeogeography, 236 structural framework and tectonic evolution, 233 –234 Bocaina Formation, 493 Bogenfels Formation, 220 Bokson Group age, 285–288 boundary relations, 286 characteristics, 286 chemostratigraphy, 286 diamictites, 287 geochronological constraints, 287 geographic position, 286 glaciogenic deposits and associated strata, 285 –288 lithological and chemostratigraphic succession, 287 palaeolatitude and palaeogeography, 286– 287 sedimentary environments, 287–288 stratigraphy, 285– 286 structural framework, 285 Bolla Bollana Tillite, 707 Bol’shoy Patom Formation, Lena River, central Siberia, 309–315 associated strata, 310– 312 boundary relations, 313 carbonate conglo-breccias, 312 chemostratigraphy, 313 composite sections, 311 conglomerates, 312 correlation between Vendian and Ediacaran Systems, 313 diamictites, 311–312 massive, 311–312 stratified, 312 geochronological constraints, 313– 315 geographic distribution, 310 glaciogenic deposits, 309– 315 lithostratigraphic section, 310 mudstones and siltstones, 312 palaeolatitudes and palaeogeography, 313 sandstones and grits, 312 stratigraphy, 309 structural framework, 309, 310 subordinate rocks, 312 succession of Patom Supergroup in Ura Uplift, 311 typical structures in diamictites, 312 Upper part of Mariinskiy Formation, 310 Bonahaven Formation, 637–638 Boonall Dolomite, east Kimberley, 666– 671 glaciogenic deposits, 668 Boondawari Formation, 684 Boord Formation, 685 Børgefjellet window, 626 boron isotopes, 203 Boston Basin, Massachusetts, USA, 475– 479 boundary relations, 477 characteristics, 478 chemostratigraphy, 477 geochronological constraints, 478 glaciogenic deposits and associated strata, 477 map of outcrops, 476 palaeolatitude and palaeogeography, 478 Squantum Member, 475–479 stratigraphic framework of lithostratigraphic units, 476 stratigraphy, 476 structural framework, 475 Boston Bay Group, 478

Brasilia Fold Belt chemostratigraphy, 514– 515 geological map, 513 Bra˚vika member, 576 Brazil Jequitaı´ Formation, 541–545 Neoproterozoic Macau´bas Group, 523– 531 Una Group, 503– 507 Brigham Group stratigraphy, 429 stratigraphy showing available carbon isotope data, 429 Brookline Member, 477 Buckland, William, 20 Buenos Aires Complex, 565 Buldya Group palynomorphs, 118 Buryatian Republic, 285– 288 Buschmannsklippe Formation, 213 glaciogenic deposits and associated strata, 214 Cadomia, 105 calcium isotopes Canadian Cordillera Mackenzie Mountains, 404–405 Otavi carbonate platform and foreslope, northern Namibia, 203 Caledonian age position of Hedmark, Valdres and Engerdalen basins, 618–619 Caledonian mountain belts, 583 Caledonides of Scandinavia, 603–609 boundary relations, 608 chemostratigraphy, 608 evidence of late Neoproterozoic glaciation, 603–609 geochronological constraints, 608 glaciogenic and associated deposits, 604, 606–608 palaeolatitude and palaeogeography, 608 palinspastic cross-section showing development of inboard and outboard basins, 604 regional stratigraphic profiles, 605 stratigraphy, 605 –606 structural framework, 603 –605 Cambridge Formation, 477 Canada cap-carbonate sequences, 74 Cordillera Mackenzie Mountains, 397– 409 global correlation, 126– 129 Windermere Supergroup, 413 –421 Canadian Cordillera, 413–421 Canyon Formation, 587 carbon isotopes, 51–62 current knowledge base, 6– 14 Cariboo– Purcell–Rocky Mountains stratigraphic columns, 415 Carrancas diamictites, 531 Central Australia, 677– 688 Central Flinders Ranges, Holowilena, South Australia glaciogenic deposits and associated strata, 695–696 iron and manganese deposits, 69 Central Flinders Zone, 716 glaciogenic deposits and associated strata, 715–716 Central Scandinavian Caledonides, 623– 627 Cerebrosphaera buickii from Spitsbergen, 119 Cerro Largo Formation, 566 Cerro Negro Formation, 566 Chambers Bluff Tillite, 684 Chambishi Basin, 178 Chameis Gate Member, 217– 221, 218 boundary relations, 220 characteristics, 220 chemostratigraphy, 220 generalized lithostratigraphy of Marmora Terrane, 219 geochronological constraints, 220

INDEX

glaciogenic deposits and associated strata, 219–220 Bogenfels Formation, 220 Chameis Gate Member, Dernburg Formation, 219 palaeolatitude and palaeogeography, 220 stratigraphy, 219 structural framework and tectonic evolution, 217–219 tectonic subdivision of Marmora Terrane in western Gariep Belt, 218 Chameis Group, 217 –221, 218 Chameis Subterrane, 218 Chang’an Formation, 359 Chapada Acaua˜ Formation, 528 Chara River basin, 297– 302 boundary relations, 299– 300 chemostratigraphy, 300– 301 glaciogenic deposits and associated strata, 298–299 palaeolatitude and palaeogeography, 301 stratigraphy, 297 –298 Upper Precambrian diamictites of Central Siberia, 297– 302 chemical index of alteration (CIA), 81–90 A– CN–K diagram, 85 evidence of glaciomarine deposits, 86– 89 implications for study of Neoproterozoic climate change, 89 Neoproterozoic glacial deposits and climate transitions, 81– 90 Oman Masirah Bay Formation, 89 oxides in molecular ratios, 85 provenance and reconstruction of original compositions, 84–86 relationship between two weathering proxies, 82 sampled formations and compositional groupings, 83 values of Neoproterozoic Port Askaig Formation, 88 weathering and weathering indices, 81– 83 weathering trends, 86 whole-rock compositions and weathering indices, 87, 88 Weathering Index of Parker (WIP), 83– 84 Yangjiaping section in South China, 84 zirconium v. thorium diagram, 86 chemical sediments, 67–74 average thicknesses and idiosyncratic features, 70 cap-carbonate sequences, 70–74 barite in Marinoan-type (based Ediacaran) cap dolostones, 73– 74 distribution and thickness, 71 early diagenetic barite in Taoudeni Basin, 74 early diagenetic barite on Dzabkhan Platform, western Mongolia, 74 early diagenetic barite on Yangtze Platform, South China, 74 expanded and condensed sequences, 71 giant wave ripples, 72 highstand deposits of cap-carbonate sequences, 73 Marinoan-type (basal Ediacaran) cap-carbonate sequences, 70– 73 Mid-Ediacaran cap carbonate sequences, 73 phosphorites, 74 primary and early diagenetic types of barite in cap dolostones, 73 seafloor barite in central Australia, 73–74 seafloor barite in northwestern Canada, 74 seafloor carbonate cement, 72–73 sheet-crack cements, 72 Sturtian-type (Cryogenian) cap-carbonate sequences, 73 transgressive cap dolostones in condensed sequences, 73

transgressive cap dolostones in expanded sequences, 71–72 tubestone (geoplumb) stromatolites, 72 iron and manganese deposits, 67–70 Central Flinders Ranges (Holowilena) and Nackara arc (Braemar), South Australia, 69 distribution in time and space, 67– 68 geochemical characteristics, 69 Jakkalsberg Member (Numees Formation, Port Nolloth Group), Gariep Belt, Namibia and South Africa, 69 lithological associations, 68–69 notable examples, 69 Rapitan Group, northern Canadian Cordillera, 69 genesis and significance of glacial-associated chemical sediments, 75–77 abiotic or biogenic dolomite, 76 accumulation rates for syndeglacial cap dolostones, 75 barite in cap dolostones, 76– 77 cap-carbonate sequences, 75 early diagenetic (void-filling) barite, 76–77 ferrous v. euxinic anoxia, 75 iron and manganese oxide deposits, 75 localization of oxidative titration, 75 non-marine syndeglacial cap dolostones, 75–76 significance of seafloor aragonite cements, 76 sources alkalinity for cap dolostones, 76 subglacial sulphate-rich ferrous waters, 75 chemostratigraphy, 51–62 carbon isotopes, 52 Ediacaran Period, 59–61 iron speciation, 53 pre-glacial Neoproterozoic, 53–57 spanning Cryogenian glaciations, 57–59 stratigraphic plots of carbon and sulphur trends, 59 strontium isotopes, 53 sulphur isotopes, 52– 53 summary of carbon and sulphur signatures, 56– 57 working chemostratigraphic compilations, 54 China cratons, 101 Chiquerı´o Formation, Peru, 481–486 boundary relations, 484 chemostratigraphy, 484 economic deposits and biomarkers, 484 geochronological constraints, 484– 485 glaciogenic deposits and associated strata, 482–484 global correlation, 485 –486 laminated dolostone facies, 485 location map and geological map of western coast, 482 palaeolatitude and palaeogeography, 484 stratigraphic section and carbon isotope trends, 483 stratigraphy, 482 structural framework, 482 zircon probability density distribution diagrams, 485 Chuos Formation boundary relations, 201 glaciogenic and associated strata, 198 Churochnaya Formation, 294 boundary relations, 290 chemostratigraphy, 290 glaciogenic deposits and associated strata, 289–290 climate physicists, 23 Coal Creek inliers, 390 Coates Lake group chemostratigraphy, 403– 404 stratigraphy, 400 Congo. See also West Congo Supergroup palaeomagnetic constraints, 104

725

Congo Craton, Central Africa, 185– 192 Cordillera Mackenzie Mountains, 397– 409 boundary relations, 403 boundary relations Rapitan Group, 403 boundary relations Stelfox Member, 403 calcium isotope record of Stelfox glaciation, 408 calcium isotopes, 404–405 carbon isotope data from dolomite of Ravensthroat Formation, 404 carbon isotopes, 403 characteristics, 406 chemostratigraphy, 403– 405 composite columnar section of Neoproterozoic strata, 399 correlation and palaeogeography, 408–409 depositional strike of Neoproterozoic passive margin, 401 geochronological constraints, 406 giant wave ripples in Ravensthroat Formation, 402 glaciogenic and associated strata, 400–403 iron isotopes cerium anomaly data, 405 location map, 390 lower Rapitan Group near Iron Creek, 399 Neoproterozoic glacial record, 397–409 origin and stratigraphic localization of basin-facies, 407 origin of transitional-facies iron-formation, 407 outcrop distribution, 398 oxygen isotopes, 404, 404 palaeoenvironmental interpretation, 407– 408 palaeoenvironmental interpretation of Rapitan Group, 406– 407 palaeolatitude and palaeogeography, 406 palaeontology, 405–406 Rapitan Group type section in Hayhook Lake area, 399 stratigraphic nomenclature, 398 stratigraphic relations, 401 stratigraphy, 400 structural framework, 398–400 synglacial hematite jaspilite Rapitan Group, 405 upper Sayunei and Shezal formations, 399 Windermere Supergroup in Mackenzie Mountains, 398 Corumba´ Group, 495 composite record of carbon and strontium isotopes, 494 formations, 489 glacially influenced sedimentation and carbonates, 487– 496 stratigraphy, 492 Cotcase Creek Formation, 654–656, 656 Cottons Breccia, 649– 651 Court Formation, 213– 214 Croles Hill Diamictite, 651–654 boundary relations, 653 glaciogenic deposits and associated strata, 653 stratigraphy, 651–652 structural framework, 651 Cryogenian, 1 –15, 145, 146 composite stratigraphic log for Tarim Block, 375, 376 Elatina glaciation, 718 glacial erosional forms, 323 glacial event, 489 glacial ocean, 29 glaciation and Northern Arabian–Nubian Shield, 277–283 glaciogenic deposits of Marnya Formation, 317–327 glaciogenic strata, 398 Idaho and Utah, 425–433 Brigham Group, 429 characteristics, 430–431 chemostratigraphy, 430 depositional settings and climatic controls, 431–432

726

Cryogenian (Continued) diamictite and volcanic succession, 428 geochronological constraints, 431 glaciogenic deposits and associated strata, 429 map showing areas of outcrop, 426 Neoproterozoic strata, 425–433 palaeolatitude and palaeogeography, 431 Pocatello area, 429 Pocatello Formation and correlative units, 427 –428 regional correlations, 432– 433 rifting and glaciation, 425– 433 structural framework, 425– 426 tectonic setting, 432 Uinta Mountain Group, 427 Uinta Mountain Group and Big Cottonwood Formation, 427 Utah and Idaho Neoproterozoic correlation chart, 426 stratigraphy, 373, 426–429 Yerelina Subgroup map, 714 Cryogenian biostratigraphy of Australia, 113–130 Acaciella australica Stromatolite Assemblage, 122 Australian Neoproterozoic stratigraphy and correlations, 116 Baicalia burra Stromatolite Assemblage, 123 biostratigraphy, 118 –125 Cerebrosphaera buickii from Spitsbergen, 119 chemostratigraphy, 125–126 distribution of Acaciella australica stromatolite, 114 distribution of Baicalia burra stromatolite, 115 distribution of Neoproterozoic basins, 114, 115 geochronological constraints, 117–118 global biostratigraphical correlation, 128 global correlation, 126– 129 palynology, 118–121 pre-Sturt glaciation stratigraphy of Australia, 124– 125 range chart of Neoproterozoic stromatolite distributions, 120 stromatolite range chart, 121 stromatolites, 121–125 Sturt glaciation, 127 taxonomic citations, 130 filamentous microfossils, 130 spheroidal microfossils and acritarchs, 130 stromatolites, 130 Cuiaba´ Group, 487 –496 boundary relations, 493 characteristics, 494 chemostratigraphy, 493–494 composite record of carbon and strontium isotopes, 494 glaciogenic deposits and associated strata, 492– 493 palaeolatitude and palaeogeography, 494 schematic cross-section, 489 structural framework, 489– 490 Dabbsjo¨n–La˚ngvattnet areas, 625 Dahomeyide belt synthetic cross-section, 167 Dalradian Supergroup, 18 stratigraphic framework, 644 Dalradian Supergroup, Scotland, 19, 643– 644, 644, 646 Damara foreland geological map, 212 Datangpo Formation glaciogenic deposits and associated strata, 360 interglacial depositional environment, 364 isotope stratigraphy, 362 Daugherty Gulch diamictite, 442 volcaniclastic diamictite, 441

INDEX

David, Edgeworth, 21 Death Valley region, 449–455, 459– 464 Kingston Peak Formation, 449– 455, 459 –464 Delamerian Orogen subdivision, 704 Denison Ranges lithofacies description, 708 Dernburg Formation, 219 detrital zircon, 246 Dhofar, Oman, 239–247 diamictite characteristics and origin, 42– 43 palaeoclimatic significance, 45–46 Didikama Formation, 273 discoidal impression, 168 disrupted beds, 639 Dom Feliciano Belt, 106 geology, 549 Dongqiaoenblaq Formation, 369 Doornpoort, 213 Dorchester Member, 477 Doushantuo fauna, 146 Doushantuo Formation, 101, 143 isotope stratigraphy, 362 Drook Formation Ediacaran biota, 472 glaciogenic deposits and associated strata, 470 stratigraphy, 469 Duas Barras Formation, 527 Duerdin Group, 670 Dugub Formation, 267, 268 Dutch Creek Formation, 416 Duurwater trough and moraine, 199, 200 –201 Dzabkhan Platform, western Mongolia, 74, 343 Dzabkhan terrane, 331 Earth’s time-averaged magnetic field, 98 East-Central Namibia, 211– 215 Eastern Death Valley region, 449–455 Eastern Officer Basin, 684 Eastern Sadlerochit Mountains geological map, 382 East Sayan Mountains, 285–288 East Sayan Range, 101, 317 –327 Ediacaran ice-rafting in South Australia, 673–675 palaeogeography in South Australia, 697 Ediacaran Period acritarch successions, 121 biota and Drook and Mistaken Point formations, 472 cap carbonate sequences, 73 composite stratigraphic log for Tarim Block, 375, 376 geochronology dating Neoproterozoic, 144– 145 Kailaketik Group stratigraphy, 373 Neoproterozoic chemostratigraphy, 59–61 successions and chronometric constraints, 118 sulphur-isotope record, 61 Edwardsburg Formation boundary relations, 442 glaciogenic deposits and associated strata, 441 volcanic and diamictite rocks, 439 Windermere Supergroup, Idaho, USA, 437 –446 Egan Formation, 662– 666, 664 boundary relations, 666 depositional interpretation, 665–666 palaeolatitude and palaeogeography, 666 stratigraphy, 664– 665 structural framework, 663 Egypt BIF, 281, 282 Elatina Formation, 26, 113, 117, 684 Elatina glaciation, South Australia, 713–719 boundary relations, 717 chemostratigraphy, 717 deposits, 102 geochronological constraints, 718 glaciogenic deposits and associated strata, 714 –717 map, 714 palaeolatitude and palaeogeography, 717– 718

stratigraphy, 713 –714, 715 structural framework, 713 Eleonore Bay Supergroup, 586 Engerdalen Basin composite section, 620 Eritrea glaciogenic deposits, 267– 268 Eskadron glaciogenic deposits, 213 Ethiopia Tambien Group, 102, 263–275 Eurasia glaciogenic succession map, 4 knowledge base, 12– 14 Neoproterozoic ice age data set, 5 Europe glaciogenic succession map, 10 knowledge base, 14 Neoproterozoic ice age data set, 11 Fadnuvag’gi Formation, Tanafjord, 595 Fargoo Formation, 666–671 boundary relations, 669 chemostratigraphy, 669 depositional interpretation, 668–669 geochronological constraints, 669 glaciogenic deposits, 668– 669 Kimberley Region, 666–671 palaeolatitude and palaeogeography, 669 stratigraphy, 667 –668 structural framework, 666 –667 Fedo potassium-metasomatic effect, 85 Feforkampen outlier, 616 Finnmark Mortensnes Formation, 593–600 Nyborg Formation, 597, 599, 600 Smalfjord Formation, 20, 593–600 Fiq diamictite, 102 Fiq Formation, 251–260. See also Abu Mahara Group, Jabal Akhdar, Oman depositional history, 255– 256 geochronological constraints, 258– 260 glaciogenic deposits and associated strata, 253–256 potassium-metasomatism, 88 repeated glacial advance and retreat, 260 sedimentology, 254– 256 distal glaciomarine facies association, 255 facies associations, 254–255 glaciogenic deposits and associated strata, 254–255 non-glacial sediment gravity flow facies, 255 non-glacial shallow marine facies association, 255 proximal glaciomarine facies association, 255 stratigraphic height, 89 structural framework, 252 –253 turbiditic sandstone, 259 volcaniclastic Saqlah unit, 259 Formiga area, 536– 537 Formiga outcrop, 537 Fouroumbala Basin, 190 Fourth Range geological map, 380 Frank River Sandstone glaciogenic deposits, 668 Franni-aus Member boundary relations, 201 –202 Fransfontein foreslope differentiation, 198 Fransfontein Ridge, 197 foreslope, 200 Fulu Formation glaciogenic deposits and associated strata, 359–360 isotope stratigraphy, 362 Yangtze Region, China, 357– 358 Fungurume Group, 176 Gabon Basin stratigraphy, 188 Gairdner Dyke Swarm, 117 Gaissa Thrust Belt, 604 Ga¨rdsjo¨n Formation, 626 Gariep belt, 104, 224 Gariepian stratigraphy, 225

INDEX

Garvellach Islands, 640 Gaskiers glacial event, 620 glaciation, 105 Gaskiers Formation, Newfoundland, Canada, 467 –473 boundary relations, 470 characteristics, 471 chemostratigraphy, 471 deep-marine glaciogenic, 467– 472 general map of Avalon Peninsula, 468 geochronological constraints, 471 glaciogenic deposits and associated strata, 470 associated volcanic rocks, 470 Drook Formation, 470 Gaskiers Formation, 470 Mall Bay Formation, 470 outcrop map, 468 palaeolatitude and palaeogeography, 471 stratigraphy, 469 –470, 469 structural framework, 467 –469 Gearbelja´vri Formation, 607 geocentric-axial dipole (GAD), 98 geochronology. See Neoproterozoic geochronology geodynamic theories for Neoproterozoic glaciation, 26 geomagnetic field, 106 Ghaub Formation, 117 boundary relations, 201 Duurwater trough and moraine, 200– 201 facies association, 200 facies associations and stratal architecture, 199–200 glaciation, 197 glaciogenic and associated strata, 199–201 palaeoenvironmental setting, 205 stratigraphy, 201 Ghubrah Formation, 251–260. See also Abu Mahara Group, Jabal Akhdar, Oman geochronological constraints, 258– 260 glaciogenic deposits and associated strata, 253–256 palaeolatitude and palaeogeography, 258 structural framework, 252 –253 Gifberg Group, 233–237 areal distribution of individual formations, 235 boundary relations, 235 chemostratigraphy, 235– 236 glaciogenic deposits and associated strata, 234–235 lithostratigraphy, 236 map showing location of Vredendal Outlier, 234 structural framework and tectonic evolution, 233–234 glaciation at sea level, 24 glaciogenic sedimentary successions, 45 glaciomarine deposits, 24 chemical index of alteration, 86– 89 glaciomarine successions, 40 Gondwana distribution of cratonic blocks, 556 Gospel Peaks, 437 Grand Conglomerat Formation, Nguba Group, 176 –177 facies, 177– 179 Grandfather Mountain Formation, 99 Grasdalen Formation, Tanafjord, 594 Great Breccia, 639 greenhouse effect, 20 Greenland Neoproterozoic deposits, 581– 590 Gropbreen Member, 575 Guozigou– Keguqingshan area composite stratigraphic log, 375 geological map, 374 glaciogenic diamictites, 373– 374 Gwna Group, 105

Hadash carbonate, 102, 256 Hadash Formation summary log, 257 Hallett Cove area, 716 Hankalchough diamictite, 372–373 Hankalchough Formation, 306 stratigraphy, 371 Hank-Fersiga area stratigraphy, 165 Harbour Main volcanic rocks, 468 Hard Luck Creek, 391 Harland, Brian, v, 1, 6, 14, 15, 17, 22– 24, 93 Hay Creek Group diamictite, 395 glaciogenic deposits and associated strata, 391–392 Hayhook Formation post-glacial carbonate couplet, 402– 403 stratigraphy, 400 strontium-rich aragonite and barite cements, 58 Hedmark Basin, 617 Hedmark Group formations lithology, 617 rift basins, 620 Hilda Subgroup, 228 Hoggar, 104 –105 Holgat Formations, 227 –228 Hospers, Jan, 22 Howchin, Walter, 21 Huangyanggou Formation stratigraphy, 371 Hula Hula diamictite, Arctic Alaska, 379– 385 boundary relations, 383 chemo- and lithostratigraphy Katakturuk Dolomite, 381 Mt. Copleston, 383 Nularvik dolomite, 384 chemostratigraphy, 383– 384 geochronological constraints, 384 geological map Fourth Range, 380 Kikitak Mountains, 380, 382 Sadlerochit Mountains, 380, 382 Shublik Mountains, 380 glacial origin, 384 glaciogenic deposits and associated strata, 382–383 palaeolatitude and palaeogeography, 384 stratigraphy, 381–382 structural framework, 379–380 volcanic rocks, 383 Huqf Supergroup, 251, 252 Ice Brook Formation glaciogenic and associated strata, 402 stratigraphy, 400 Idaho Cryogenian rifting and glaciation, 425–433 Edwardsburg Fm., central Idaho, 437–446 geological record, 444– 446 Neoproterozoic correlation chart, 426 Neoproterozoic rocks in roof pendants, 439 Pocatello Fm., southeastern Idaho, 425– 436 Windermere Supergroup, 437–446 India, 347–353 Infracambrian stratigraphic succession, 101 Inindia beds, 685 Inishowen Bed, 645– 646 interbasinal correlations, 61 Ireceˆ Basin chemostratigraphy, 514 Ireland, 643 –647 boundary relations, 645– 646 carbon isotope trends for Dalradian Supergroup, 646 chemostratigraphy, 646 economic deposits and biomarkers, 646 geochronological constraints, 646– 647 Neoproterozoic glaciogenic deposits, 643–647 palaeolatitude and palaeogeography, 646 iron and manganese deposits Central Flinders Ranges and Nackara arc, 69

727

distribution in time and space, 67– 68 glacial-associated chemical sediments, 75 Jakkalsberg Member and Gariep Belt, 69 lithological associations, 68–69 Neoproterozoic glaciation iron formation, 67–70, 68 Otavi carbonate platform and foreslope, 203 Rapitan Group, northern Canadian Cordillera, 69 Iron Creek area, 401 iron isotopes, 405 Hayhook Lake, 405 iron speciation, 53 water column redox conditions, 62 working chemostratigraphic compilations, 54 Islay anomaly, 56, 58 Isochron techniques analytical methodologies, 139–140 Itabaiana dome area stratigraphy, 515 Itawa area, 178 Ituri Group, 191 Jabal Akhdar, Oman, 251–260 geological map, 253 Jacadigo Group, 487–496 boundary relations, 493 chemostratigraphy, 493– 494 composite record of carbon and strontium isotopes, 494 depositional settings, 494–495 geochronological constraints, 494 glaciogenic deposits and associated strata, 492–493 lithostratigraphy, 491 palaeolatitude and palaeogeography, 494 schematic cross-section, 489 stratigraphy, 490–493, 491 structural framework, 489–490 Jacoca Formation, 513 Jakkalsberg Member, 69 Ja¨mtland Supergroup stratigraphy, 625 Jbe´liat area, 166 Jbe´liat Group, 72 Jequitaı´ Formation, southeastern Brazil, 541– 545 boundary relations, 544 chemostratigraphy, 544 geochronological constraints, 544 geological map, 543 glaciogenic deposits and associated strata, 542–544 measured vertical sections, 543 mineralization and characteristics, 544 palaeolatitude and palaeogeography, 544 simplified geological map, 542 stratigraphy, 541–545 structural framework, 541–545 Jerta Formation, 606 Jiangkou glaciations, 364 depositional environment, 364 geochronological constraints, 363 Yangtze Region, China, 357 Jiangkou Group, 357 boundary relations, 361 distribution, 360 Julius River Member, 651–654, 653 boundary relations, 653 glaciogenic deposits and associated strata, 652–653 stratigraphy, 651–652 structural framework, 651 Kaigas Formation, 223–230, 224 boundary relations, 228 characteristics, 229 chemostratigraphy, 228– 229 diamictite sedimentary features, 229 dolostone, 226 geochronological constraints, 229

728

Kaigas Formation (Continued) glaciogenic deposits and associates strata, 226– 228 holostratotype, 225 lithostratigraphy, 225 palaeolatitude and palaeogeography, 229 Port Nolloth Zone within Gariep Belt, 224 stratigraphy, 225– 226 structural framework and tectonic evolution, 223– 226 Kailaketik Group stratigraphy, 373 Kalahari and environs, 103–104 Kamtsas Formation, 213 Karagassy Group, 319 Karoetjes Kop Formation, 233–237 boundary relations, 235 characteristics, 236 chemostratigraphy, 235–236 depositional environment, 236 generalized lithostratigraphy of Gifberg Group, 236 geochronological constraints, 236 glaciogenic deposits and associated strata, 234– 235 map showing location of Vredendal Outlier, 234 palaeolatitude and palaeogeography, 236 stratigraphy, 234 structural framework and tectonic evolution, 233– 234 Katakturuk Dolomite, Arctic Alaska, 379 –385 composite carbon chemo- and lithostratigraphy, 381 glaciogenic deposits and associated strata, 382 metamorphic grade, 380 stratigraphy, 381 Katangan Series, 21 Katanga Supergroup, Central Africa, 173–182 boundary relations, 180 carbonate rocks, 180–181 characteristics, 180– 181 chemostratigraphy, 180 cross-section, 178 distribution, 175 economic deposits, 180–181 geochronological constraints, 181 glaciogenic deposits and associated strata, 176– 180 lithostratigraphic section, 177 Lufilian arc in Pan-African orogenic belts system, 174 Neoproterozoic glaciogenic diamictites, 173– 182 palaeolatitude and palaeogeography, 181 regional geology of Lufilian belt, 174 stratigraphy based on syntectonic conglomerate complexes, 175 structural and stratigraphic framework, 173– 176 Kazakhstan, 303– 306 Keele formation stratigraphy, 400 Keilberg Member boundary relations, 202 cap dolostone, 71 Keyindi Formation stratigraphy, 374 Khesen Formation, 339, 344 glaciogenic deposits and associates strata, 340 Khubsugul Group, Northern Mongolia, 343 stratigraphy, 343 Khesen Gol, 344 Khongoryn Member, 100 diamictite, 335 glaciogenic deposits and associated strata, 333 stratigraphy, 335 Khubsugul basin, 331 Khubsugul Group, Northern Mongolia, 339–344 boundary relations, 340–342 chemostratigraphy, 342 detailed stratigraphy of Khesen diamictite, 343 detailed stratigraphy of Ongoluk diamictite, 343

INDEX

geochronological constraints, 342– 343 geology of western shores of Lake Khubsugul, 341 glaciogenic deposits and associates strata, 340 –342 Khesen and Ongoluk Gols, 343 palaeolatitude and palaeogeography, 342 stratigraphy, 340, 342 structural framework, 339– 340 tectonic map, 340 Kikitak Mountains geological map, 380, 382 Kimberley Region, Australia, 103, 659–671 chemostratigraphy, 671 Duerdin and Albert Edward Groups, 667 Fargoo Formation and Moonlight Valley Formation, 666–669 glaciogenic correlation, 671 glaciogenic stratigraphy of lower Duerdin Group, 670 Kuniandi Group and Landrigan Fm., 663 Landrigan and Egan formations, Mount Ramsay area, 662–666 Louisa Downs Group and Egan Fm., 664 Mount House Group and Walsh Formation, 661 Neoproterozoic glacial deposits, 659– 671 tectonic units, 660 Walsh Formation, 659–662 boundary relations, 662 depositional interpretation, 662 geochronological constraints, 662 glaciogenic deposits and associated strata, 661–662 palaeolatitude and palaeogeography, 662 stratigraphy, 661 structural framework, 661 King Island, 656 geochronological constraints, 651 glaciogenic and associated strata, 650 structural framework, 649– 651 Kingston Peak Formation, 449–455, 459–464. See also Panamint Range boundary relation, 453, 463 carbon and oxygen isotopic values for carbonate, 453 characteristics, 454, 463 chemostratigraphy, 453–454, 463 contact with overlying Noonday Dolomite, 453 contact with underlying units, 453 cross-section of four measured sections, 452 eastern Death Valley region, 449– 455 Eastern Facies Assemblage stratigraphy, 451 evidence for glaciation, 455 facies, 459 geochronological constraints, 454, 464 glaciogenic deposits and associated strata, 451 –453, 462–463 glaciogenic deposits of Southern facies, 451 –452 map, 461 map of facies assemblages, 450 Noonday Dolomite, 459, 462– 464 northern facies, 451, 452 palaeolatitude and palaeogeography, 454, 464 Panamint Range, 459 –464 regional unconformity and diamictite deposition, 451 stratigraphy, 450– 451, 460 –462 structural framework, 449– 450, 459 –460 Virgin Spring limestone, 453 Komagfjord Antiformal Stack, 606 Komagfjord Window, 609 Koppang Formation, 613 facies and stratigraphic context, 615 outcrop area location for Neoproterozoic glacial, 614 stratigraphy and sedimentology, 618 structural setting, 615

Koyva Formation, 294 boundary relations, 292 characteristics, 292 chemostratigraphy, 292 glaciogenic deposits and associated strata, 291–292 stratigraphy, 294 Koyva River stratigraphy, 294 Krokvatn Diamictite, 599 Krokvatn Palaeovalley glaciogenic deposits and associated strata, 595–596 schematic profile, 596 Kulutieliekti Formation glaciogenic diamictites, 374 stratigraphy, 373 Kundelungu Group, 179–181 Kuokkel Windows, 604 Kurtun Formation, 325 Kuruktag Range, 306 Kyrgyzstan, 303 –306 Kzisuhum Formation, 367 Laksefjordvidda, 595–596 Landrigan Formation, 662 –666, 669 biostratigraphy, 666 boundary relations, 666 chemostratigraphy, 666 depositional interpretation, 665–666 geochronological constraints, 666 glaciogenic deposits, 665 palaeolatitude and palaeogeography, 666 stratigraphy, 664 –665 structural framework, 663 La˚ngmarkberg Formation, 625, 626–627, 627 glaciogenic deposits and associated strata, 625–626 Lapa Formation, 518 Las Ventanas Formation, 555–563 boundary relations, 561 characteristics, 561 chemostratigraphy, 561 distribution of cratonic blocks of west Gondwana, 556 geochronological constraints, 562 geochronological data, 561 geological map, 558 glaciogenic deposits and associated strata, 560–561 map of formations, 557 palaeolatitude and palaeogeography, 561–562 simplified stratigraphic column, 560 stratigraphic column, 559 stratigraphy, 556 –560 structural framework, 555 –556 Laurentia and environs, 98–100 Leger Granite, 246 Lena River, central Siberia, 309– 315 Lenda Formation, 188 Lesser Himalaya, India, 347–353 Lillevannet Member, 598 Lillfja¨llet Formation, southern Swedish Caledonides, 629 –633 boundary relationships, 631–632 clastic wedges, 632 geochronology, 632 geological map of Stor-Lo¨vsjo¨n area, 631 glaciogenic and associated deposits, 630– 632 glaciogenic origin, 633 map of main outcrop area, 630 palaeolatitudes, 632 profile along northern hill side, 632 stratigraphy, 629 –630, 630, 631 structural relationships, 629 Limekiln Spring Member boundary relations, 463 stratigraphy, 460 –461 Lindian Basin stratigraphy, 188 –190

INDEX

Lindi Supergroup, 185– 192 Congo Craton, Central Africa, 185– 192 geological sketch map, 186 palaeolatitude and palaeogeography, 191 stratigraphy, 186 –188, 188–190 structural framework and basin setting, 185– 186 Little Burke Tillite, 686 Little Dal Group chemostratigraphy, 403– 404 stratigraphy, 400 Loch na Cille Boulder Bed, 100, 645 –646 Lokoma Group, 189 Lossit Limestone glaciogenic deposits and associated strata, 637 Louisa Downs Group, 664 Lower Doushantuo Formation samples, 143 Lower Krokvatn Diamictite, 599 Lower Starye Pechi Subformation, 292 Lower Vendian Bol’shoy Patom Formation, 314 Lufilian arc, 174 Lufilian belt, 176 regional geology, 174 Lupton Formation, 684 Lyell, C., 20 Lyell Land, 586, 587 Lyndhurst Formation, 707 Macau´bas Group, 523– 531 boundary relations, 529– 530 characteristics, 530 chemostratigraphy, 530 composite stratigraphic columns, 528 diamictite formations, 523–531 general stratigraphic scheme, 527 geochronological constraints, 530 glaciogenic deposits and associated strata, 528–529 location in relation to Sa˜o Francisco craton, 524 map showing distribution of different formations, 526 palaeolatitude and palaeogeography, 530 simplified geological map, 525 stratigraphy, 527 structural framework, 524 –527 MacDonaldryggen Member, 574 MacDuff Bed, 645–646 Mackenzie Mountains Supergroup stratigraphy, 400 strontium-isotope ratios, 60 Maieberg anomaly, 56 Maieberg Formation, 198, 202 –205 Mai Kenetal Synclinorium, 268 composite chemostratigraphic reference section, 273 Tambien Group, Northern Ethiopia (Tigre), 264 Maikhan Ul Member, 100 diamictite units, 335 glaciogenic deposits and associated strata, 332–333 stratigraphy, 334 Makonga–Kibambale area, 179 Maldonado Group, 555 –563 boundary relations, 561 palaeolatitude and palaeogeography, 561– 562 radiometric ages, 563 stratigraphy, 556 –560 structural framework, 555 –556 Mali Group, 165, 169 depositional environments, 166 Mall Bay Formation glaciogenic deposits and associated strata, 470 stratigraphy, 469 Marinoan age, 100 cap-carbonate sequences, 70–73 glaciations dating Neoproterozoic, 145 Marmora Terrane, 217–221, 218, 218 generalized lithostratigraphy, 219

Marnya Formation, 101, 317– 327 boundary relations, 323– 324 chemostratigraphy, 324, 325 Cryogenian glacial deposits, 317–327 Cryogenian glacial erosional forms, 323 depositional systems, 320 geochronological constraints, 324 glaciation, 327 glaciogenic deposits, 321 –323 Late Neoproterozoic deposits, 326 outcrop logs and field sketches, 322 palaeolatitude and palaeogeography, 324 Sayan region geological map, 318 stratigraphy, 319–321, 320 Uda Formation, 321 structural elements and lithostratigraphy, 318 structural framework, 317–319 Masirah Bay Formation, 89 Mata˜o Formation, 527 Matheos Formation, 267, 268 Mawson, Douglas, 22 Mawsonland, 102– 103 palaeoequators, 103 mean annual air temperature (MAAT), 719 Mechum River succession, 99 Meltout tillite, 42 Meritri Group metaconglomerate, E. Sudan boundary relations, 281 geochronological constraints, 282 glaciogenic deposits and associated strata, 280–281 Miaba Group chemostratigraphy, 518 glaciogenic deposits and associated strata, 513–514 and Neoproterozoic successions, 509– 519 Middle Urals, 289–295 Mid Ediacaran ice-rafting in South Australia, 673–675 palaeogeography in South Australia, 697 Mineral Fork Formation, 425–433 Mirassol d’Oeste Formation, 492– 493 Mirbat Group, Dhofar, Oman, 239–247 CIA and MIA, 245 Misinchinka Group, 416 Mistaken Point formation, 472 Moelv Formation, 613, 615–619 facies and stratigraphic context, 615 Hedmark Basin, 617 location of outcrop area, 614 stratigraphy and sedimentology, 615–619 structural setting, 613–615 Valdres Basin, 617–618 Moema laminites, 535– 540 boundary relations, 538 characteristics, 538 chemostratigraphy, 538 geochronological constraints, 539 geological map of Sa˜o Francisco Basin, 536 glaciogenic deposits and associated strata, 536–537 Neoproterozoic glaciogenic unit, 535–540 palaeolatitude and palaeogeography, 539 partial diagram of Formiga outcrop, 537 Sa˜o Francisco Basin, Brazil, 535 –540 stratigraphy, 535–536, 537, 538 structural framework, 535 Mongolia global correlation, 129 Khubsugul Group, 339–344 Tsagaan Oloom Formation, 331– 336, 343 Monkman Pass area stratigraphic columns, 415 Moonlight Valley Formation, 666– 671 boundary relations, 669 depositional interpretation, 668– 669 glaciogenic deposits, 668 –669 palaeolatitude and palaeogeography, 669 stratigraphy, 667–668 structural framework, 666–667 Moores Lake Formation, 442

729

Moores Station Formation, 439 glaciogenic deposits and associated strata, 441–442 Mortensnes Formation, 593–600, 599 boundary relations, 598 chemostratigraphy, 598 glaciogenic deposits and associated strata, 594–598, 597–598 palaeolatitude and palaeogeography, 598 stratigraphy, 594 structural framework, 593–594 Mount Copleston volcanic rocks stratigraphy, 381 Mount Cornish Formation, 686 Mount Davenport Diamictite Member, 268, 686 Mount Doreen Formation, 686 Mount House area, 661 Mount House Group, 661 Mount Lofty Ranges lithofacies description, 706 Mount Nelson, 416 Mount Ramsay area, 662 –666, 664 boundary relations, 666 depositional interpretation, 665– 666 palaeolatitude and palaeogeography, 666 stratigraphy, 664–665 structural framework, 663 Naburula Formation, 685 Nackara Arc, South Australia glaciogenic deposits and associated strata, 715–716 iron and manganese deposits, 69 lithofacies description, 706 Nafun Group, 258 Nakfa terrane, 265 Nama assemblage dating Neoproterozoic, 146 Namibia Chameis Group, 217–221, 218 global correlation, 129 Kaigas and Numees Formations, 223–230 Otavi Group, 195–206 Port Nolloth Group, 223–230 Witlev Group, 211– 215 Nanhuan system and period, 357 Nanook Limestone composite carbon chemo- and lithostratigraphy, 381 stratigraphy, 381–382 Nantuo Formation, 314 boundary relations, 361– 362 claystone, 719 glaciations, 364 glaciogenic deposits and associated strata, 360 depositional environment, 364 geochronological constraints, 363 Negash Synclinorium composite chemostratigraphic reference section, 273 Tambien Group, Northern Ethiopia (Tigre), 264, 268 Neoproterozoic chemostratigraphy, 51–62 Neoproterozoic geochronology, 138– 140 age of primary standards, 141 analytical methodologies, 138–140 isochron techniques, 139–140 uranium lead methodologies, 138 –139 uranium lead microbeam techniques, 139 calculating age from multiple dates, 142 calibrating tracers for isotope-dilution, 141–142 complex uranium lead zircon systematic, 142–144 dating, 135–137, 144–147 accessory minerals from volcanic rocks, 135–136 Avalon assemblage, 146 chemical precipitates and organic residues, 136 Doushantuo fauna, 146 earliest fossils, 146 Ediacaran glaciations, 144– 145

730

Neoproterozoic geochronology (Continued) future directions, 146 –147 glacial intervals, 144– 145 Marinoan glaciations, 145 maximum and minimum age constraints, 137 Nama assemblage, 146 Neoproterozoic events, 146 Pre-Sturtian glaciations, 145 Re–Os dating of organic-rich sediments, 136 Shuram– Wonoka carbon isotope excursion, 146 Sturtian glaciations, 145 uranium lead dating of carbonate, 136 uranium lead dating of phosphates, 136–137 White Sea assemblage, 146 decay constants, 141 isochron diagram, 138 linear arrays v. isochrons, 144 microbeam uranium lead standardization, 140 radio-isotopic dating techniques, 147 radio-isotopic geochronometers, 137– 138 Uranium-lead, 137– 138 whole-rock geochronometers, 138 radiometric decay systems used in geochronology, 136 random/internal uncertainties, 140 sources and types of uncertainty, 140–147 subjective interpretation of dates, 147 uncertainties from geologic complexity, 142– 144 uranium lead concordia diagram, 137, 138 uranium lead zircon data, 143, 144 lower Doushantuo Formation samples, 143 Pocatello Formation samples, 144 Scout Mountain Member samples, 144 users guide, 135–147 Neoproterozoic glacial palaeolatitudes, 93–107 depositional palaeolatitudes for Neoproterozoic strata, 94–97 global update, 93–107 methods, 93–98 palaeoequators plotted across Australia and Mawsonland, 103 palaeoequators plotted across Laurentia, 99 palaeomagnetic constraints, 98– 106 published assessment of glacial influences, 94– 97 reliability of palaeomagnetic depositional latitudes, 98 Neoproterozoic glaciations chemical index of alteration, 81–90 chemical sediments, 67– 74 chemostratigraphy, 51–57 history of research, 17– 29 geochronology, 135–147 Ediacaran glaciations, 144 Marionoan glaciations, 145 Pre-Sturtian glaciation, 145 Sturtian glaciation, 145 iron formation, cap carbonate, barite and phosphorite, 67–77 palaeolatitudes, 93– 107 Neoproterozoic glaciogenic successions, 39–47 classification, 40 diamictite units, characteristics, 42–43 glacial environments characteristics, 39– 45 glacial influence recognition, 39– 47 glaciolacustrine and glaciomarine settings, 44– 45 glacial sedimentary indicators, 41 historical development of terminology, 39 palaeoclimatic significance clast characteristics, 46 diamictite, 45– 46 outsized clasts in bedded sediments, 46 stratigraphic trends and sequence boundaries, 46– 47 reconstructing palaeoenvironmental conditions, 45– 47

INDEX

subglacial settings, 40– 44 terrestrial proglacial settings, 44 Neoproterozoic ice ages, 1– 16 available datasets, Africa, 3 available datasets, Australia, 13 available datasets, Eurasia and Nubian Shield, 5 available datasets, Europe, 11 available datasets, North America, 7 available datasets, South America, 9 calibrating change, 15 current knowledge base, 6– 15 Africa, 12 Australia, 14 Eurasia-Nubian Shield, 12 Europe, 14 North America, 14 South America, 14 glaciogenic succession, Africa, 2 glaciogenic successions, Australia, 12 glaciogenic successions, Eurasia and Nubian Shield, 4 glaciogenic successions, Europe, 10 glaciogenic successions, North America, 6 glaciogenic successions, South America, 8 Neoproterozoic research, glacial geology, 17– 30 1871– 1908, pioneering discoveries, 17– 21 1909– 1941, period of globalization, 21–22 1942– 1964, rebutting challenges, 22– 23 Newcastle and Ko¨ln palaeoclimate conferences, 22–23 palaeomagnetism and meridional extent of ice sheets, 22 1965– 1981 in wake of plate tectonic revolution, 23– 24 associated chemical sediments banded iron- and manganese-formation and cap carbonates, 23 climate models and white Earth instability, 23– 24 earth’s Pre-Pleistocene Glacial Record volume, 24 1982– 1997, gathering storm, 24– 28 biocatalysed weathering, 28 cap carbonates, 25–26 carbonate burial, 26 causative theories for glaciation, 26–28 continental break-up, 28 continental distribution, 27 ice-ring collapses, 27 impact ejecta, 27 large orbital obliquity, 26–27 ocean stagnation, 27– 28 bibliographic history of Pleistocene glacial controversy, 29 bidirectional climate change, 20 causative theories for glaciation, 26 astronomical theories, 26 first reported occurrences of glaciogenic deposits by palaeocontinent, 19 glaciation papers growth in annual number, 18 glaciogenic deposits, cumulative discovery by palaeocontinent, 18 greenhouse effect, 20 iconic sketch of glacially striated pavement beneath end-Cryogenian Smalfjord diamictite, 19 idiosyncratic sedimentary and early diagenetic features in cap dolostones, 25 long road to consensus, 28– 30 Pleistocene glacial controversy, 20 Port Askaig Tillite discovery and historical scientific context, 18– 21 boulder beds, 18– 19 Ngalia Basin, 687 glaciogenic deposits and associated strata, 685 –686 Nguba groups, 173 Niari diamictites, 188 Nichatka Formation, 297– 302 age, 301

boundary relations, 299– 300 chemostratigraphy, 300– 301 geochronological constraints, 301 geographical position, 298 glaciogenic deposits and associated strata, 298–299 palaeolatitude and palaeogeography, 301 regional palaeogeography, 301– 302 sedimentary environments, 301 stratigraphic position, 299 stratigraphy, 297 –298 structural framework, 297 type sections, 300 Upper Precambrian diamictites of Central Siberia, 297– 302 Noonday Dolomite, 453 carbonate platforms, 454 map, 461 stratigraphy, 451, 462 Nordaustlandet, 576 stratigraphic nomenclature and correlations, 573 North America glaciogenic succession map, 6 knowledge base, 14 Neoproterozoic ice age data set, 7 North American Cordillera correlations, 444 Northern Arabian–Nubian Shield, 277– 283 Atud and Nuwaybah diamictite location, 280 boundary relations, 281 characteristics, 281– 282 chemostratigraphy, 281 evidence for Early and Mid-Cryogenian glaciation, 277– 283 geochronological constraints, 282 locality map showing geologic units, 278 location of Neoproterozoic deposits of possible glacial origin, 278 palaeolatitude and palaeogeography, 282 possible glaciogenic deposits and associated strata, 280– 282 stratigraphic summaries of possibly glaciogenic Cryogenian units, 279 stratigraphy, 278 –280 structural framework, 278 Northern Paraguay Belt Puga Formation, 487–496 Serra Azul Formation, 499– 501 North Flinders Zone glaciogenic deposits and associated strata, 714–715 lithofacies description, 707–708 North Urals, 289–295 Norway. See also South Norway Baltoscandian craton and Hedmark and Valdres rift basins and Engerdalen Basin, 620 Smalfjord and Mortensnes formations, Vestertana Group, E. Finnmark, 593–600 Norwegian sparagmite basins, 100 Nova Aurora Formation, 529 Nubian Shield, 239, 251, 263–275, 277–283 glaciogenic succession map, 4 knowledge base, 12– 14 Neoproterozoic ice age data set, 5 Nularvik dolomite chemo- and lithostratigraphy, 384 crystal fans, 385 glaciogenic deposits and associated strata, 382–383 Numees Formations, 223– 230 boundary relations, 228 chemostratigraphy, 228– 229 distinct carbonate unit, 227 glaciogenic deposits and associated strata, 226–228, 227 palaeolatitude and palaeogeography, 229 radiometric age control, 230 stratigraphy, 225 –226 structural framework and tectonic evolution, 223–226

INDEX

Nuwaybah Formation diamictite, NW Saudi Arabia geochronological constraints, 282 Northern Arabian– Nubian Shield, 280 possible glaciogenic deposits and associated strata, 281 NW Tasmania, 651–654 geochronological constraints, 654 glaciogenic and associated strata, 652–653 Nyanga-Niari Basin, 188 Nyborg Formation, Finnmark, 599 dolostones, 600 glaciogenic deposits and associated strata, 597 oceanographic theories for Neoproterozoic glaciation, 26 ocean stagnation Neoproterozoic glacial geology, 27– 28 Officer Basin, South Australia, 673 –675, 688 chemostratigraphy, 673– 674 geochronological constraints, 674 glaciogenic deposits and associated strata, 673 Mid-Ediacaran ice-rafting, 673–675 palaeolatitude and palaeogeography, 674 stratigraphy, 673 structural framework, 704 Old Fort Point (OFP) Formation boundary relations, 419 glaciogenic deposits and associated strata, 418 stratigraphic columns, 418 Windermere Supergroup, southern Canadian Cordillera, 414– 415, 420 ´ gua Formation, 514 Olhos d’A Olympic Formation, 685 Oman Member Formation, 239–247, 251–260 CIA, 89 Neoproterozoic outcrops and salt basins, 252 Ombaatjie Formation boundary relations, 202 Ongoluk Diamictite glaciogenic deposits and associated strata, 340–342 stratigraphy, 343 Ongoluk Formation, 343 Ongoluk Gol stratigraphy, 343 Oorabra, 686 Oselok Group, 101, 317–327 boundary relations, 323– 324 correlation of Late Neoproterozoic deposits, 326 palaeolatitude and palaeogeography, 324 stratigraphy, 319 –321 structural framework, 317 –319 Otavi carbonate platform and foreslope, northern Namibia, 195 –206 associated carbonates, 206 boundary relations, 201– 202 chemostratigraphy, 202– 204 boron and calcium isotopes, 203 carbon isotopes, 202– 203 oxygen isotopes, 203 post-glacial negative anomalies, 202 pre-glacial negative anomalies, 202 reactive iron and manganese concentrations, 203 strontium isotopes, 203 sulphur isotopes, 203 diagnostic glacial indicators, 204 Duurwater trough and moraine, 200 foreslope of Fransfontein Ridge, 200 Fransfontein foreslope differentiation, 198 geochronological constraints, 204 geological map of Otavi Group fold belt, 196 Ghaub Formation facies association, 200 glaciogenic and associated strata, 195–206 grounding-line oscillations, 206 ice-rafted dropstones for maximum ice-shelf extent, 205–206 magnitude of base-level and glacioeustatic changes, 206

palaeoenvironmental setting of Ghaub Formation, 205 palaeolatitudes and palaeogeography, 204 role of rift faulting, 205 stratigraphy, 196–197 cross-section, 198 crustal stretching and thermal subsidence, 196–197 Otavi Group subgroups, 196 palaeogeography of platform and southern foreslope, 197 relations in Abenab Subgroup and Ghaub Formation, 201 restoration in north-south cross-section, 197 structural framework, 195–196 Otavi Mountains, 195 oxygen isotopes current knowledge base, 6 –14 Pahrump Group, 432, 450, 459 Palaeozoic succession in Greenland, 588 Panamint Range, Death Valley, California, 459–464. See also Kingston Peak Formation boundary relations Limekiln Spring/Surprise Members, 463 overlying and underlying non-glacial units, 463 Wildrose sub-member, 463 characteristics, 463 chemostratigraphy, 463 composite stratigraphic column of Neoproterozoic units in Death Valley region, 460 geochronological constraints, 464 glaciogenic and related strata of Neoproterozoic, 459–464 glaciogenic deposits and associated strata, 462–463 Kingston Peak Formation, 459–464 map showing distribution, 461 palaeolatitude and palaeogeography, 464 stratigraphy, 460–462 Limekiln Spring Member, 460–461 Noonday Dolomite, 462 Sourdough Member, 462 South Park Member (exclusive of Wildrose Sub-member), 462 Surprise Member, 461–462 Wildrose sub-member of South Park Member, 462 structural framework, 459–460 pan-global infra-Cambrian glaciation, 577 Paraguay Belt, Brazil, 487–496, 499– 501. See also Puga Formation, Serra Azul Formation Patom region, 101 Patom SGr in Ura Uplift succession, 311 Peake Ranges lithofacies description, 708 Penge Formation, 188 Pepuarta Tillite dropstones, 718 periglacial aeolianite, 717 permafrost regolith, 717 Perry Canyon, 428 Pertatataka Formation age, 314 Petit Conglomerat Formation clasts, 181 glaciogenic deposits and associated strata, 179–180 petrographic studies, 180 radiometric data, 181 Petrovbreen Member characteristic stratigraphic sections, 575 diamictite, 577 glaciogenic deposits and associated strata, 573–574 phosphorite, 67– 77 Pickelhaube Formation, 227 Pioneer Sandstone, 685 Pirrilyungka Formation, 684 Playa Hermosa Formation, Uruguay, 547–552 boundary relations, 550

731

chemostratigraphy, 550 comparative stratigraphic chart, 550 Dom Feliciano Belt geology, 549 geochronological constraints, 551 glacially influenced environment, 551 glaciogenic deposits and associated strata, 548–550 gravity-driven deposits, 552 logged section, 550 morphotectonic divisions, 548 palaeogeographic reconstruction, 551 palaeolatitude and palaeogeography, 550– 551 stratigraphy, 548 structural framework, 547–548 Playa Verde Basin, 547– 552 stratigraphic chart, 550 Pleistocene glacial controversy bibliographic history, 29 Neoproterozoic glacial geology, 20 Pocatello area stratigraphy, 429 Pocatello Formation, 425 –433 stratigraphy, 427–428, 429 uranium lead zircon data, 144 Pod’em Formation, 101 Polarisbreen Group, Svalbard, 571– 578 biostratigraphy, 577 chemostratigraphy, 576 composite stratigraphic column, 574 geochronological constraints, 577 geological sketch map, 572 glaciogenic deposits and associated strata, 573–576 palaeolatitude and palaeogeography, 576– 577 stratigraphic nomenclature and correlation, 573 stratigraphic sections, 575 stratigraphy, 573 structural framework, 571–578 Port Askaig Formation, 19, 635– 640 boundary relations, 638 chemostratigraphy, 638 CIA and WIP values, 88 economic deposits and biomarkers, 638 environmental conditions, 639 geochronological constraints, 639 glaciogenic deposits and associated strata, 636–638 outcrops map, 636 palaeolatitude and palaeogeography, 638–639 potassium-metasomatism, 88 stratigraphy, 636, 636, 637 structural framework, 635–636 Port Askaig Tillite centenary of Thomson’s (1891) study of, 23 discovery and historical scientific context, 18–21 Port Nolloth Group, 223 –230 lithostratigraphy, 225 Port Nolloth Zone within Gariep Belt, 224 potassium feldspar Chinese glacial and non-glacial deposits, 84 sedimentary enrichment, 87 potassium metasomatism, 85 Port Askaig and Fiq formation, 88 Precambrian diamictites of Central Siberia, 297–302 Proterozoic evaporites, 106 Puga Formation, 487–496 boundary relations, 493 characteristics, 494 chemostratigraphy, 493– 494 composite record of carbon and strontium, 494 correlations and geotectonic evolution, 495–496 geochronological constraints, 494 geological map of Paraguay Belt, 488

732

Puga Formation (Continued) glacially influenced sedimentation and carbonates, 487–496 glaciogenic deposits and associated strata, 492– 493 lithostratigraphy of Jacadigo Group, 491 Neoproterozoic lithostratigraphy nomenclature, 489, 490 Northern Paraguay Belt depositional settings, 494– 495 palaeolatitude and palaeogeography, 494 radiometric constraints, 496 schematic cross-section, 489 Southern Paraguay Belt depositional settings, 495 stratigraphic section and variations, 493 stratigraphy, 490– 492 structural framework, 489– 490 Qakmaklik Group, 367 Quruqtagh area geological map, 372 glaciogenic diamictites, 370–373 stratigraphic log, 373 stratigraphy, 370– 372 Quruqtagh region, 101 radiative energy-balance equations, 23 Rainout deposits, 42 Rapitan Group boundary relations, 403 Canadian Cordillera Mackenzie Mountains, 398, 399 geochronological constraints, 406 glaciogenic and associated strata, 400 –402 iron and manganese deposits, 69 palaeolatitude and palaeogeography, 406 palaeontology, 405– 406 section in Hayhook Lake area, 399 stratigraphy, 393, 400 synglacial hematite jaspilite, 405 Tatonduk Inlier, Alaska-Yukon, 391 rare-earth elements, 405 Rasthof anomaly, 56 Ravensthroat Formation carbon and oxygen isotope data from dolomite, 404 giant wave ripples, 402 glacial carbonate, 402– 403 stratigraphy, 400 Red Pine Shale, 430 Red Sea Hills, 274 Re–Os dating organic residues, 136 whole rock geochronometers, 138 Re–Os ages, 136, 420, 517, 519, 654, 687, 710, 718 Rhynie Sandstone, 119 Rio de la Plata Craton morphotectonic divisions, 548 Neoproterozoic palaeomagnetic poles, 567 palaeomagnetic constraints on Neoproterozoic glacial palaeolatitudes, 105–106, 561 palaeomagnetic poles, 568 Rio Peixe Bravo Formation, 527 Risba¨ck Group glaciogenic deposits and associated strata, 626 glaciogenic diamictite, 627 Ritsemjaure Window, 607 Roan Group, 173, 174 Rosh Pinah Formation glaciogenic deposits and associated strata, 227 volcanic lithotypes, 227 Roxbury Conglomerate, 105 Roxbury gravels, 478 Russian Federation Bokson Group, 285–288 Bol’shoy Patom Formation, 309–315 Marnya Formation, 317– 327

INDEX

Nichatka Formation, 297– 302 North and Middle Urals, 289– 295 Russøya Member, 573 Sadlerochit Mountains geological map, 380, 382 SAFFRAN quarry glaciogenic deposits and associated strata, 537 stratigraphy, 538 Salitre Formation, 506 San Carlos Formation, 555– 563 boundary relations, 561 chemostratigraphy, 561 fossil content support, 562 glaciogenic deposits and associated strata, 560 –561 palaeolatitude and palaeogeography, 561– 562 stratigraphy, 556– 560 structural framework, 555– 556 whole succession, 557 Sansikwa Subgroup, 188 Sa˜o Francisco Basin, Brazil, 509– 519, 524, 535– 539, 541 –545 associated strata, 512– 513 basal glacial deposit, 518 Brasilia Belt geological map, 513 chemostratigraphy, 514–517 correlations, 511, 516 geochronological constraints, 516– 517 geological map, 536 glaciogenic deposits, 509– 519 lithostratigraphic and chemostratigraphic correlations, 516 megasequences of Neoproterozoic successions, 511 mineralization, 517 Moema laminites, 535– 540 Neoproterozoic cover, 510 Neoproterozoic successions, 509–519 ´ gua Formation, 514 Olhos d’A palaeomagnetic constraints, 104 Sergipano Belt geological map, 515 stratigraphy, 510– 512 carbonate platform megasequence (marine), 511 dominant continental siliciclastic megasequence, 512 glaciogenic megasequence, 510– 511 Itabaiana dome area, 515 structural and geotectonic framework, 509– 510 Sa¨rv nappe, 630 Saudi Arabia, 282 Sayan Mountains, 285– 288, 317– 327 Sayan region, 318 Sayunei Formation, 399 Scandinavian Caledonides, 623 –627 boundary relations, 626– 627 chemostratigraphy, 627 diamictite units older than Risba¨ck Group, 627 geological map, 624 glaciogenic deposits and associated strata, 625 –626 glaciogenic diamictite, 627 glaciogenic lithologies map, 594 Ja¨mtland Supergroup stratigraphy, 625 La˚ngmarkberg Formation, 626– 627, 627 sections through La˚ngmarkberg Fm., 625 stratigraphy, 624– 625 structural framework, 623– 624 Schisto-Calcaire Group, 188 Schwarzbach, Manfred, 22 Scotland, 635– 640, 643– 647 boundary relations, 645– 646 carbon isotope trends for Dalradian Supergroup, 646 chemostratigraphy, 646 CIA and WIP values, 88 economic deposits and biomarkers, 646

geochronological constraints, 646– 647 geological map, 644 geological map of Scottish–Irish Dalradian Highlands, 644 glaciogenic deposits and associated strata, 644–645 MacDuff, Inishowen and Loch na Cille Boulder Beds, 645 –646 palaeolatitude and palaeogeography, 646 Port Askaig Formation, Dalradian Supergroup, 19, 635–640 Stralinchy– Reelan formations and Cranford and Whiteness cap carbonates, 645 stratigraphic framework of Dalradian Supergroup, 644 stratigraphy, 644 structural framework, 643 –644 Scottish– Irish Dalradian Highlands, 647 geological map, 644 Scout Mountain Member, 144, 428– 429 Sen Formation, 298 Sergipano Fold Belt chemostratigraphy, 515– 516 geological map, 515 Serra Azul Formation, Brazil, 499– 501 boundary relations, 500– 501 chemostratigraphy, 501 geochronological constraints, 501 geological maps, 500 glaciogenic deposits and associated strata, 499–500 stratigraphy, 499, 500 structural framework, 499 Serra da Bodoquena, 494 Serra do Catuni Formation, 528 Sheepbed formation stratigraphy, 400 Sheldon, Richard P., 27 Shezal Formation, 399 diamictite, 402 Shiquan Formation stratigraphy, 371 Shiraro area, 264 Shublik Mountains geological map, 380 Shuram anomaly, 60, 258 Shuram– Wonoka anomaly, 56 carbon isotope excursion dating Neoproterozoic, 146 Siberia Bokson Group, 285– 288 Bol’shoy Patom Formation, 309–315 Marnya Formation, 317 –327 Nichatka Formation, 297– 302 palaeomagnetic constraints, 100–101 Siberian Craton, 101, 317 –327 Siberian Platform, 327 Sierra del Volca´n Formation logged section, 567 outcrops, 566 Silasia Formation BIF, NW Saudi Arabia boundary relations, 281 geochronological constraints, 282 glaciogenic deposits and associated strata, 281 Sinian system and period, 8, 357 sedimentary basins, 101 Yangtze Region, China, 357 Slangen Member, 574 Smalfjord Formation, 20, 593–600 boundary relations, 598 characteristics, 598– 599 chemical index of alteration, 81– 90 chemostratigraphy, 598 diamictite, 19 geochronological constraints, 599 glaciogenic deposits and associated strata, 594–598 map showing glaciogenic lithologies, 594 Marinoan-type cap dolostone, 608 palaeolatitude and palaeogeography, 598 schematic profile, 596 stratigraphic profiles, 595

INDEX

stratigraphy, 594 structural framework, 593 –594 Snowball Earth, 151–159 aftermath, 157–158 carbon dioxide consumption, 153 climatic modelling argument, 156–157 duration as function of carbon dioxide, 157 geochemical modelling argument, 157 hypothesis, 28, 151, 247 ice and climate, 154–156 clear equatorial thin-ice as solution for life, 154–155 glacial deposits and continental ice behaviour, 155–156 onset of snowball, ice-albedo instability, 154 slushball theory, 154 initiation, 151–154 melting, 156, 158 modelling, 151– 159 modelling studies of Neoproterozoic ice ages, 152 onset, 158 sea-ice thickness, 155 Sourdough Member stratigraphy, 449, 455, 462 chemostratigraphy, 453, 455 South America glaciogenic succession map, 8 knowledge base, 14 Neoproterozoic ice age data set, 9 Southern Georgina Basin glaciogenic deposits and associated strata, 686 Oorabra, Sun Hill and Black Stump Arkose, 686 Southern Mount Freeling syncline, 695 Southern Paraguay Belt, 487–496 chemostratigraphy, 494 depositional settings, 495 glaciogenic deposits and associated strata, 492– 493 Southern Tasmania, 654–656 geochronological constraints, 656 glaciogenic deposits and associated strata, 655 structural framework, 654 –655 South Norway, 613–621, 615–619 Baltoscandian ice sheet, 619 chemostratigraphy, 618 composite section across west, 620 correlation and age, 619–620 examples of facies and stratigraphic context, 615 geochronological constraints, 619 geological map of Sparagmite region, 614 Hedmark, Valdres and Engerdalen basins, 618–619 Koppang Formation, 618 Moelv Formation, 615– 619 Neoproterozoic glaciation, 613–621 outcrop area location, 614 palaeogeographical reconstruction, 616 palaeogeography, 618–619 Pre-Caledonian position, 618–619 structural framework, 613 –615 South Park Member, 450, 464 stratigraphy, 462 Southwestern Umberatana syncline, 694–695 Sparagmite region geological map, 614 Spiral Creek Formation, Tillite Group, 587 Spitsbergen, 571–578, 573 Squantum Member, 475–479 glaciogenic deposits and associated strata, 477 Stappogiedde Formation, 598 Starye Pechi Formation stratigraphy, 295 Starye Pechi Subformation, 292 Stelfox Member, 403 glaciogenic and associated strata, 402 Storeelv Formation, 581 glaciogenic deposits and associated strata, 587 Stor-Lo¨vsjo¨n diamictite, 633

geological map, 631 UNU, 633 Støvfanget Formation para-autochthonous deposits, 587 Stralinchy– Reelan formation, 645 stromatolites, 25 range chart, 121 successions of Australia, 121– 125 taxonomic citations, 130 strontium isotopes, 53 current knowledge base, 6 –14 ratios, 55, 60 Stuart Shelf glaciogenic deposits and associated strata, 716 map, 714 Stuart Shelf glaciogenic formations, 715 Sturtian age cap-carbonate sequences, 73 glacial and post-glacial assemblages, 121 glacial succession in Australia, 701–710 glaciations dating Neoproterozoic, 145 glaciation stratigraphy of Australia, 124– 125, 126 lithostratigraphic units, 708 rifting, 703–704 Sturtian glaciogenic succession, 710 successions of Australia, 127 Sturt Tillite, 113 subaqueous debris flows, 43 Sub-Saharan Africa. See also Africa geological sketch map, 186 Sudan Meritri Group, 280–282 sulphur isotopes, 52– 53 current knowledge base, 6 –14 Ediacaran, 61 Sun Hill Arkose, 686 Supergroup in Mackenzie Mountains, 398 Surprise Member, 449, 461–462 boundary relations, 463 glaciogenic deposits and associated strata, 462 –463 stratigraphy, 461–462 Svalbard, 571–578 archipelago, 571 global correlation, 126 –129 Swartleikrans Bed of Buehrmann, 235 Table Mountain Group, 88 Talisayi Formation stratigraphy, 374 Tambien Group, Northern Ethiopia, 102, 263– 275 boundary relations, 268 chemostratigraphy, 268– 270 chemostratigraphy data sets, 269 composite chemostratigraphic reference section, 272, 273 diamictite member, 267 Dugub Formation, 267, 268 geochronological constraints, 271– 272 glacial influence on sedimentation, 274– 275 glaciogenic deposits and associated strata, 266–268 glaciogenic deposits in Eritrea, 267 –268 global palaeogeographic reconstruction, 270 key exposures, 264 lithostratigraphic subdivision, 265 magmatism, 272 Mai Kenetal Synclinorium, 264, 268 Matheos Formation Diamictite Member, 268 Negash Synclinorium, 264, 268 palaeoenvironmental changes, 272–274 palaeolatitude and palaeogeography, 270– 271 Post-Tambien Group magmatism, 272 Pre-Tambien Group magmatism, 271–272 radiometric age constraints, 271 Shiraro area, 264 stratigraphy, 266 structural framework, 264–266 tectonic and palaeogeographic setting, 275 timing of prospective glacial intervals, 274

733

Upper Werri Slate, 266 –267 Tanafjord area, 597 Tandilia System, Argentina, 565– 569 boundary relations, 567 chemostratigraphy, 567 geochronology, 568 glaciogenic deposits and associated strata, 566 integrated stratigraphy, 566 Neoproterozoic palaeomagnetic poles for Rio de la Plata, 567 palaeolatitude and palaeogeography, 567– 568 palaeomagnetic poles for Rio de la Plata, 568 Sierra del Volca´n Formation logged section, 567 Sierra del Volca´n Formation outcrops, 566 stratigraphic chart, 550 stratigraphy, 565–566 structural framework, 565 Tany Formation stratigraphy, 293 Taoudeni Basin, Africa, 163– 169 boundary relations, 167– 168 characteristics, 168 chemostratigraphy, 168 early diagenetic barite, 74 geochronological constraints, 168 geological map, 164, 167 glaciogenic deposits and associated strata, 166–167 palaeolatitude and palaeogeography, 168 phosphorite in cap-carbonate sequences, 74–75 stratigraphy, 163–166 structural framework and basin setting, 163 synthetic cross-section, 167 Tarim area composite stratigraphic log, 376 Tarim Block, NW China, 367– 377 composite stratigraphic log Aksu-Wusi area, 371 Guozigou-Keguqingshan area, 375 Quruqtagh area, 373 Tarim area, 376 Tielikeli area, 369 geochronological constraints on diamictites, 375–376 geological map, 368 Aksu-Wusi area, NW Tarim Basin, 370 Guozigou-Keguqingshan area, 374 Quruqtagh area, 372 Tielikeli area, 368 glaciogenic diamictites, 367– 374 description, 368–369, 369– 370, 371– 373, 374 palaeolatitude data, 377 palaeolatitudes and palaeogeography, 376–377 stratigraphy, 373–374 Tarqat Formation stratigraphy, 374 Tasmania, 649– 656 bedrock geological map of Smithton Synclinorium, 652 boundary relations, 653 chemostratigraphy, 653 Cotcase Creek Formation, 654–656 Cotton Breccia, 649– 651 distribution of Proterozoic rocks and major tectonic elements, 650 geochronological constraints, 654 glaciogenic deposits and associated strata, 652–653 Julius River Member and Croles Hill Diamictite, 651–654 King Island, 649–651 boundary relations, 651 characteristics, 651 chemostratigraphy, 651 geochronological constraints, 651 glaciogenic deposits and associated strata, 650 palaeolatitude and palaeogeography, 651 stratigraphy, 650 structural framework, 649 palaeolatitude and palaeogeography, 653– 654 Southern Tasmania, 654–656 boundary relation, 655

734

Tasmania (Continued) chemostratigraphy, 655 geochronological constraints, 656 glaciogenic deposits and associated strata, 655 stratigraphy, 655 structural framework, 654– 655 stratigraphy, 651– 652 structural framework, 651 Wedge River beds, 654– 656 Tatonduk inlier, Alaska– Yukon border, 389– 395 boundary relations, 392–393 chemo- and lithostratigraphy, 392 chemostratigraphy, 393 depositional setting, 394– 395 geochronological constraints, 393–394 geological map of exposures, 391 glaciogenic deposits and associated strata, 391– 392 location map, 390 Neoproterozoic stratigraphy, 390, 394 nomenclature chart, 390 palaeolatitude and palaeogeography, 393 regional correlations, 395 stratigraphy, 390, 393 structural framework, 389– 390 Tatonduk River geological map, 391 Tayga Group, 309 Tayshir Member, 335 Te´niagouri Group, 169 Tereeken Formation glaciogenic diamictites, 372 stratigraphy, 371 Terra-Wasserburg plots of SHRIMP data, 443 terrestrial debris flows, 43 thermal-ionization mass spectrometry (TIMS), 117, 138 Thomson, James, 18 Tielikeli area composite stratigraphic log, 369 geological map, 368 glaciogenic diamictites, 367–369 stratigraphy, 367– 368 Tillite Group, Greenland, 581–590 boundaries relations, 588 chemostratigraphy, 588 depositional setting, 589 detailed stratigraphic logs, 585 geochronological constraints, 589 glaciogenic deposits and associated strata, 586– 587 location map, 582 Lower Palaeozoic succession, 588 other characteristics, 588 palaeogeographical setting, 590 palaeolatitude and palaeogeography, 588– 589 sequence stratigraphy, 589 stratigraphical nomenclature, 583 stratigraphy, 583– 586 Andre´e Land Group, 583–584 associated para-autochthonous strata, 586 Tillite Group, 584–586 stratigraphy of uppermost Andre´e Land Group, 584 structural framework, 582– 583 timing of glaciation, 589–590 Timan Basin, 593 Toby Formation, 413, 417, 420 boundary relations, 419 glaciogenic deposits and associated strata, 416, 417 stratigraphic columns, 417 Tonian successions, 117 Tossa˚sfja¨llet Group stratigraphy, 630 Trezona anomaly, 56, 58 Tsagaan Oloom Formation, southwestern Mongolia, 331–336, 343 associated carbonate rocks, 333– 334

INDEX

chemo- and lithostratigraphy of Dzabkhan basin, 333 chemostratigraphy, 334 geochronological constraints, 335 glaciogenic deposits and associated strata, 332 –334 palaeolatitude and palaeogeography, 334– 335 stratigraphy, 332, 334, 335 structural framework, 331– 332 tectonic map, 332 Tulasu Formation stratigraphy, 373 Turkey Hill Formation, 684 Twitya Formation, 400 Tyndall, John, 20 Ubangi Supergroup, 185–192 geological sketch map, 186 palaeolatitude and palaeogeography, 191 stratigraphy, 186– 188, 188 –190 structural framework and basin setting, 185– 186 Uda Formation depositional system, 320 stratigraphy, 320, 321 Uinta Mountain Group stratigraphy, 427, 427 Ulvesø Formation, 581 glaciogenic deposits and associated strata, 587 Ulyakha Member, 323 tillite, 325 Una Group, Bahia, Brazil, 503 –507, 509–519 boundary relations, 506– 507 chemostratigraphic correlations, 516 chemostratigraphy, 507 correlations, 511 geochronological constraints, 507 geological map showing Bebedouro Formation, 504 glacial lithofacies associations, 506 glaciogenic deposits and associated strata, 504 –506, 512 ice-contact glaciomarine system, 505 lithostratigraphic correlations, 505, 516 lithostratigraphic successions, 505 mineralization, 507 stratigraphic variation in carbonates, 506 stratigraphy, 503– 504 structural framework, 503 United States of America global correlation, 129 Windermere Supergroup, 437– 446 Upper Chapada Acaua˜ Formation, 528 Upper group Tatonduk inlier in east-central Alaska, 392 Urals, 289– 295 glaciogenic deposits and associated strata, 289 –293 late Mesoproterozoic to Neoproterozoic, 290 North and Middle, 289– 295 palaeolatitudes and palaeogeography, 293 possible correlations, 294 regional palaeogeography, 294–295 sedimentary environments, 293 stratigraphy, 289, 291, 293, 294, 295 structural framework, 289 uranium lead dating of chemical precipitates, 136– 137 microbeam techniques, 139 sources and types of uncertainties, 140– 144 U–Pb methodologies, 138 U–Pb radio-isotopic geochronometers, 137 Uruguay Las Ventanas and San Carlos formations, 555 –563 Playa Hermosa Formation, Playa Verde Basin, 547–552 Utah Cryogenian glacial deposits, 425 –433 Neoproterozoic correlation chart, 426 rifting and glaciation, 425– 433

Vakkejokk Breccia, 606 tillite, 609 Valdres Basin, 617–618 composite section, 620 Varangerfjord, 596 Varangeristiden, 609 Vaza Barris Group chemostratigraphy, 515– 516, 518 glaciogenic deposits and associated strata, 513–514 Neoproterozoic successions, 509–519 Vazante Group, 518 chemostratigraphy, 514– 515, 514, 517 glaciogenic deposits and associated strata, 512–513 Vendian Bol’shoy Patom Formation, 314 Vendian glacials stratigraphy, 291 Vestertana Group, 593– 600 boundary relations, 598 glaciogenic deposits and associated strata, 594–598 palaeolatitude and palaeogeography, 598 stratigraphy, 594 structural framework, 593 –594 Villa Monica formation diamictite, 566 Vindela¨lven River, 626 Virgin Spring limestone, northern facies, 453 Volta Basin succession, 166 synthetic cross-section, 167 Vredefontein Formation, 226 Vredendal Outlier map, 234 Vreeland Formation, 413 –414, 415, 417, 420 boundary relations, 419 glaciogenic deposits and associated strata, 417–418, 417 stratigraphic columns, 417 Wahlgu Formation, 684 Walsh Formation, 659–662, 661, 669 Walsh Tillite, 103 Wanapi Dolomite Member, 686 Warren Gorge, 716 Weathering Index of Parker (WIP), 82 CIA, 83–84 Wedge River Beds, 654–656 Wegener’s theory of continental displacement, 19 West Congo Belt (WCB), 185– 192 diamictites, 190–191 West Congo Supergroup, 185–192 chemostratigraphy, 191 Congo Craton, Central Africa, 185– 192 correlations, 191– 192 geochronological constraints, 191 glaciogenic character of diamictites, 190 map of Neoproterozoic sedimentary basins, 186 palaeolatitude and palaeogeography, 191 stratigraphy, 186 –190 structural framework and basin setting, 185–186 Western Officer Basin Boondawari Formation, 684 glaciogenic deposits and associated strata, 684 White Earth problem, 23 White Sea assemblage, 146 Widouw Formation glaciogenic deposits, 234 Wildrose sub-member boundary relations, 463 stratigraphy, 462 Williams, George, 26 Willis, Bailey, 21 Wilsonbreen Formation characteristic stratigraphic sections, 575 glaciogenic deposits and associated strata, 575–576 Windermere Supergroup, Idaho, USA, 437– 446 boundary relations, 442

INDEX

broader implications of Idaho geological record, 444–446 Canadian Cordillera Mackenzie Mountains, 398 characteristics, 442 chemostratigraphy, 442 correlation chart, 445 correlations along North American Cordillera, 444 depositional setting, 443 –444 Edwardsburg Formation, 437–446 geochronological constraints, 443, 444 geological map, 439, 440 glaciogenic deposits and associated strata, 441–442 palaeolatitude and palaeogeography, 443 related rocks, 437–446 stratigraphy, 439 –441 structural framework, 438 –439 Terra-Wasserburg plots of SHRIMP data, 443 Windermere Supergroup, southern Canadian Cordillera, 413–421 boundary relations, 419 characteristics, 419 chemostratigraphy, 419 comparative stratigraphic columns, 415 correlatives, 397 geochronological constraints, 420 geological map showing outcrop distribution, 414 glaciogenic deposits and associated strata, 416–419 Old Fort Point (OFP) Formation, 414 –415, 420 palaeolatitude and palaeogeography, 419– 429 regional correlations, 421 stratigraphy, 416 structural framework and basin setting, 415– 416 Toby Formation, 413, 417, 420 Vreeland Formation, 413–414, 417, 420 Window Allochthon, 607 Witvlei Group, 211–215

boundary relations, 214 chemostratigraphy, 214 economic deposits, biomarkers, 214 geochronological constraints, 214– 215 geological map of Damara foreland, 212 glaciogenic deposits and associated strata, 213–214 palaeolatitude and palaeogeography, 214 stratigraphy, 212–213, 212 structural framework, 211–212 Wonnadinna Dolostone, 686 Wushinanshan Group, 369 Yalaguz Formation, 367 Yangjiaping section in South China, 84 Yangtze Platform, South China, 74 Yangtze Region, China, 357– 364 BIF, 362 biomarkers, 363 boundary relations, 361– 362 cap carbonate, 362 characteristics, 362– 363 comparison of glaciations, 364 depleted mantle model age, 362 depositional environment, 363–364 distributions of Jiangkou Group, 360 Fulu Formation, 357– 358 geochronological constraints, 363 glaciogenic deposits and associated strata, 359–360 isotope stratigraphy, 362 Jiangkou Group and Jiangkou glaciation, 357 manganese deposits, 363 Nanhuan system and period, 357 Neoproterozoic lithostratigraphy, 359 outcrops of Neoproterozoic glacial deposits, 358 palaeolatitude and palaeogeography, 363 Sinian system and period, 357 stratigraphic columns, 361

735

stratigraphy, 358 structural framework, 358 Yardida Tillite, 686 Yerelina Subgroup map, 714 Yudnamutana Subgroup, South Australia, 701– 710 Delamerian Orogen subdivision, 704 facies architecture and depositional environments, 709 geochronology, 709– 710 glaciogenic deposits and associated strata, 704–708 ironstone, 705 isotope chemostratigraphy, 709 lithofacies, 710 lithofacies description, 704– 705 lithostratigraphic units, 703 lithostratigraphy of Yudnamutana Subgroup, 706 outcropping deposits, 702 palaeolatitude and palaeogeography, 709 stratigraphic and tectonic relationships, 708 stratigraphy, 701–702 stratigraphy and suggested correlations, 715 structural framework, 702–704 Yudnamutana Trough lithofacies description, 707–708 Yukkengol Formation stratigraphy, 371 Yukon Tatonduk inlier, 389– 395 Za’am Group diamictite, 281 –282 Zabit Formation, 288 diamictites origin, 344 Zambian Copperbelt, 178 Zhamoketi Formation stratigraphy, 371 Zhaobishan Formation stratigraphy, 370 Zhuya Group, 309 age, 314 zircon, 135 dating rocks, 135–147 detrital zircons, 137

In recent years, interest in Neoproterozoic glaciations has grown as their pivotal role in Earth system evolution has become increasingly clear. One of the main goals of the IGCP Project No. 512 was to produce a synthesis of newly available information on Neoproterozoic successions worldwide similar in format to Hambrey & Harland’s (1981) Earth’s pre-Pleistocene Glacial Record. This Memoir therefore consists of a series of overview chapters followed by site-specific chapters. The overview chapters cover key topics including the history of research on Neoproterozoic glaciations, identification of glacial deposits, chemostratigraphic techniques and datasets, palaeomagnetism, biostratigraphy, geochronology and climate modelling. The site specific chapters for 60 successions worldwide include reviews of the history of research on these rocks and up-to-date syntheses of the structural framework, tectonic setting, palaeomagnetic and geochronological constraints, physical, biological, and chemical stratigraphy, and descriptions of the glaciogenic and associated strata, including economic deposits.

This document is available for free on the internet. You should not pay for it or to gain access to it.