"The contributors to this timely volume are top-notch. The book marks the formal opening step in recognizing
Contribut...
58 downloads
1248 Views
57MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
"The contributors to this timely volume are top-notch. The book marks the formal opening step in recognizing
Contributors C. Geoffrey Adams,
OBE
Salim Al-Busaidi Zaher Al-Sulaimani
a new and rapidly developing area
Peter Andrews
of research."
Mustafa Latif As-Saruri
-John A. Van Couvering, editor in chief, Micropaleontology Press
John C. Barry Deryck D. Bayliss Laura Bishop Charlie S. Bristow France de Lapparent de Broin Hans de Bruijn Diana Clements Margaret E. Collinson Peter Paul van Dijk Peter W. Ditchfield William R. Downs Vera Eisenmann Hamed A. El-Nakhal La�Tence J. Flynn Peter L. Forey Eberhard "Dino" Frey Peter
Friend
Alan W. Gentry Emmanuel Gheerbrant Ken W. Glennie Tom Gundling Ernie A. Hailwood Andrew Hill Louis L. Jacobs Paul A. Jeffery John D. Kingston "Norman MacLeod Sally McBrearty Peter B. Mordan Phillip A. Murry His Excellency Sheikh Nahayan bin Mubarak AI Nahayan Ross G. Peebles
Yale University Press New Haven and London http:/ jwww.yale.edu/yup/
Daniel S. Pemberton Martin Pickford Michael Rauhe Jack Roger Fred Rogl Torsten Rossman Sevket Sen Pascal Tassy Herbert Thomas John E. Whittaker Peter J. vVhybrow Walid Yasin Sally V. T. Young
Fossil Vertebrates of Arabia
Continued from front flap
\Vith Emphasis on the Late Miocene Faunas, Geology, and invertebrates, fish, reptiles, and mam
Palaeoenvironments of the
mals from the Emirate of Abu Dhabi,
Emirate of Abu Dhabi, United
including several new species. Part IV
Arab Emirates
reviews taphonomy, carbon isotopes, ancient Arabian environments, and the
Edited by Peter J. Whybrow and
earliest evidence of the genus Homo
Andrew Hill
in the region. The fifth section links research findings in Arabia to others
In collaboration with The Abu Dhabi
in Asia and Africa, and the final sec
Company for Onshore Oil Operations
tion looks at Arabia in the larger con
and The Ministry for Higher Educa
text of Old World Tertiary faunas and
tion and Scientific Research, United
the world's Tertiary oceans.
Arab Emirates
This extensively
illustrated
volume
Peter J. Whybrow
is a seruor re
brings together for the first time the
searcher and leader of AJ:abian iVlio
results of research on Arabian conti
cene Biotic Research for the Depart
nental vertebrates discovered in the
ment of Palaeontology, The Natural
United Arab Emirates, the Sultanate
History Museum, London.
of Oman, and the Republic of Yemen. Eminent scientists from Arabia, Eur
Andrew
ope, and the United States provide
pology at Yale UniYersity and curaror
up-to-date information on Arabian
of
paleontology as well as on Arabian
Museum of :\'atural
stratigraphical,
Uni,·ersin·.
geological,
isotopic,
Hill is professor of anthro
anthropologY
at
the
Peabod,
Riston·. Y2.!e
and paleomagnetic topics. The book presents
new
fossil
records· from
Arabia and Pakistan and discusses the closing of the ancient Middle East Tethy s seaway.
Printed in the U.S.A.
The first section of the book provides a
history of the Abu Dhabi Miocene
proj ect, and the second describes the
jacket illustration: View
of Az Zabbu
from Shuwaihat, Abu Dhabi, Cn.ited
local geology and stratigraphy. Part
Arab Emirates (photograph b�- Peter J.
III details studies on Late Miocene
Whybrow)
Continued on back flap
His Highness Sheikh Zayed bin Sultan Al Nahyan, President of the United Arab Emirates, Ruler o- C:_ of Abu Dhabi.
OSSIL VERTEBRATES OF With Emphasis on the Late Miocene Faunas) Geology) and Palaeoenvironments of the Emirate ofAbu Dhabi) United Arab Emirates
Edited by Peter]. 1/Vhybrow and Andrew Hill In collaboration with The Abu Dhabi Company for Onshore Oil Operations The Ministry for H¥fher Edztcation and Scientific Research, United Arab Erttirates
YALE 1JNIVERSI1Y PRESS
NEW HAVE� AND LoNDON
-=--�
-
-• -
�.
--
�
.. �..
-·� ._ __ a .,.
--
-
� -
Pul:>lished with generous assistance from The Abu
•
'-•
•
-
•
Library of Cong1ress Cataloging-in-Publication Data
Dhabi Company for Onshore Oil Operations Fossil vertebrates of Arabia : late Miocene faunas, geology, and palaeoenvironments of the Emirate of Copyright© 1999 by Yale University.
All rights reserved. This book may not be reproduced, in whole or in par t, including illusu·ations, in any form (beyond that copying permitted by Sections 107 and 108 of
Abu Dhabi, United Arab Emirates/edited by Peter
J. Whybrow and Andrew Hill, in collaboration with the Abu Dhabi Company for Onshore Oil Opera tions the Ministry for Higher Education and Scien ,
tific Research, United Arab Emirates. p.
the U.S. Copyright Law and except by reviewers for
em.
Includes bibliographical references and index.
the public press), without written permission from
ISBN 0-300-07183-3 (cl. : alk. paper)
the publishers.
l. Vertebrates, Fossil--Abu Z,aby (United Arab Emirates: Emirate)
Designed by In House Production Company, Min neapolis, Minnesota, and set in Galliard type by The Clarinda Company, Clarinda Iowa. Printed in the ,
United States of America by Edwards Brothers, Ann A.rbor, Michigan.
2. Animals, Fossil--Abu Z,aby
(United Arab Emirates : Emirate) I. Whybrow, Peter J.
II. Hill, Andrew P.
III. Sharikat Abu Z,aby lil-' An1allyat al-Batrillfyah al-Barrfyah.
IV. United Arab Emirates. Ministry
for Higher Education and Scientific Research. QE84l.F653
1999
566'.095357--dc21
97-41809 CIP
The paper in this book meets the guidelines for per
A catalogue record for this book is available from
manence and durability of the Committee on Pro duction Guidelines for Book Longevity of the Council on Library Resources.
the British Library. 10
9
8
7
6
5
4
3
2
1
To Jonathan) Alex) and Valerie Whybrow)· to May Hill) and to the memory of Rowland Hill)· and also to the memory of Roger Hamilton and Colin Patterson) FRS
CONTENTS
Foreword
6 Aeolian and sabkha sediments in the Miocene Shuwaihat Formation, Emirate of Abu Dhabi, United Arab Emirates 50 Charlie S. Bristow
by His Excellency Sheikh Nahayan bin Mubarak Al Nahayan Minister for Higher Education and Scientific Research, United Arab Emirates
xi
Preface
by Ken W. Glennie
Xlll
Acknowledgements
:xvii
Contributors Abbreviations
XIX
XXlll
Map of the Western Region, Emirate of Abu Dhabi, United Arab Emirates
x:xv
Part I Introduction, summary, overview, and history of palaeontological research in the Emirate of Abu Dhabi, United Arab Emirates
Peter]. VVhybrow and Andrew Hill
3
2 Summary and overview of the Baynunah fauna, Emirate of Abu Dhabi, and its context
Andrew Hill and Peter]. Whybrmv
7
3 Histor y of palaeontological research in the Western Region of the Emirate of Abu Dhabi, United Arab Emirates 15
Part II Miocene geology of the Western Region, Emirate of Abu Dhabi, United Arab Emirates Introduction
26
4 Local stratigraphy of the Neogene outcrops of the coastal area: vVestern Region, Emirate of Abu Dhabi, United Arab Emirates
Peter]. lVhybrow, Peter F. F1'iend) Peter W. Ditchfield, and Charlie S. Bristow
Peter W Ditchfield
61
8 Palaeomagnetic correlation and dating of the Baynunah and Shuwaihat Formations, Emirate of Abu Dhabi, United Arab Emirates Ernie A. Hailwood and Peter]. vilhybrow 75 9 St�ble isotope analyses and dating of the M1ocene of the Emirate of Abu Dhabi, United Arab Emirates Ross G. Peebles 88
Part III Miocene fossil fauna from the Baynunah Formation, Emirate of Abu Dhabi, United Arab Emirates
l Introduction to fossil vertebrates of Arabia
Andrew Hill, Peter]. Whybrow, and Walid Yasin
7 Diagenesis of the Baynunah, Shuwaihat, and Upper Dam Formation sediments exposed in the Western Region, Emirate of Abu Dhabi, United Arab Emirates
28
5 Rivers of the Lower Baynunah Formation, Emirate of Abu Dhabi, United Arab Emirates Peter F. Friend 38
108
Introduction
l 0 Late Miocene S\Van mussels from the Baynunah Formation, Emirate of Abu Dhabi, United Arab Emirates Paul A. Jeffery Ill l l A terrestrial pulmonate gastropod from the late Miocene Baynunah Formation, Emirate of Abu Dhabi, United Arab Emirates Peter B. Mordan 116 12 Late Miocene fishes of the Emirate of Abu Dhabi, United Arab Emirates
Peter L. Forey and Sally
V. T
Young
120
13 Chelonia from the late Miocene Bavnunah ' Formation, Emirate of Abu Dhabi, United Arab Emirates: palaeogeographic implications
France de Lapparent de Broin and Peter Paul van Dijk
136
14 Fossil crocodilians from the late Miocene Baynunah Formation of the Emirate of Abu Dhabi, United Arab Emirates: osteology " and palaeoecology
lvfichael Rauhe, Eberhard ((Dino" Frey, Daniel S. Pemberton, and Torsten Rossmann
163
miD
CONTENTS
15 A late Miocene insectivore and rodent fauna from the .Bavnunah Formation, Emirate of Abu Dhabi, United Arab Emirates 186 Hans de Bruijn
25 Isotopes and environments of the Baynunah Formation, Emirate of Abu Dhabi, United Arab Emirates 354 John D. Kingston
16 A monkey (Primates; Cercopithecidae) from the late Miocene of Abu Dhabi, United Arab Emirates 198 Andrew Hill and Tom Gundling
26 Earliest stone tools from the Emirate of Abu Dhabi, United Arab Emirates Sally McBrearty 373
17 Late Miocene Carnivora from the Emirate of Abu Dhabi, 1Jnited i\rab Emirates ]oh1� C. Barry 203 18 Miocene elephantids (Mammalia) from the Emirate of Abu Dhabi, 1Jnited Arab Emirates: palaeobiogeographic implications Pascal Tassy 209 19 Hipparions from the late Miocene Baynunah Formation, Emirate of Abu Dhabi, United Arab Emirates Vera Eisenmann and Pete1·]. Whybrow 234 20 Fossil Suidae from the Baynunah Formation, Emirate of Abu Dhabi, United Arab Emirates 254 Laura Bishop andAndrew Hill 21 A fossil hippopotamus from the Emirate of Abu Dhabi, United Lt\rab Emirates 271 Alan W Gentry 22 Fossil pecorans from the Baynunah Formation, Emirate of Abu Dhabi, United Arab Emirates Ala1� W Gentry 290 23 Late Miocene Baynunah Formation, Emirate of Abu Dhabi, United Arab Emirates: fauna, flora, and localities Peter]. Whybrow and Diana Clements 317
Part IV Proboscidean taphonomy, isotopes, and environments of the Baynunah Formation; artifacts from the Western Region, Emirate of Abu Dhabi; and Arabian palaeoenvironments Introduction
336
24 Taphonomy of the Shuwaihat probosddean, late Miocene, Emirate of Abu Dhabi, United Arab Emirates 338 PeterAndrews
27 Late Miocene palaeoenvironments in Arabia: a synthesis 389 john D. Kingston andAndrew Hill
PartV Regional faunas and floras from the Sultanate of Oman, the Republic of Yemen, Mrica, and Asia Introduction
410
28 Late Miocene small-mammal faunal dynamics: the crossroads of the Arabian Peninsula 412 Lawrence]. Flynn and Louis L. Jacobs 29 Late Miocene sub-Saharan African vertebrates, and their relation to the Baynunah fauna, Emi rate of Abu Dhabi, United Arab Emirates 420 Andrew Hill 30 Oligocene and Miocene terrestrial vertebrates in the southern Arabian Peninsula (Sultanate of Oman) and their geodynamic and palaeogeo graphic settings
Herbert Thomas, Jack Roger, Sevket Sen, 1i1artin Pickford, Emmanuel Gheerbrant, ZaherAl-Sulaimani, a1�d SalimAl-B�Haidi 430
31 Geology, fruits, seeds, and vertebrates ( ?Sirenia '> from the Kaninah Formation (middle Eocene), Republic of Yemen
Mustafa LatifAs-Saruri, Peter]. VVhybrow, and 443 Margaret E. Collinson
32 A dinosaur from the Republic of Yemen
Louis L. jacobs, Phillip A. Murry, William R. Downs, and HamedA. El-Nakhal
45-1
33 Arabian Tertiary fauna, flora, and localities
Peter]. Whybrow and Diana Clements
460
36 Oligocene and Miocene palaeoceanography-a
. Y1 - :. --:-t:thyan
Arabian Gulf, the Mediterranean, and -)rid's Tertiary oceans
-=-oduction
476
--= --:-:;,e terminal Tethyan event: a critical review of :::e conflicting age determinations for the connection of the Mediterranean from the �dian O cean :_ Geoffrey Adams, Deryck D. Bayliss, and 477 ·m E. Whittaker _
-
·
-=
..J:.igocene and Miocene palaeogeography and :ratigraphy of the circum-Mediterranean =-�!!ion ��-d Rigl j 48 5
review
Norman MacLeod
SOl
Index
509
Arabic contents
566
FOREWORD HIS EXCELLENCY SHEIKH NAHAYAN BlN MUBARAK AL NAHAYAN , MINISTER FOR HIGHER EDUCATION AND SciENTIFIC RESEARCH, UNITED ARAB EMIRATES
For centmies, we who have lived in the Arabian Gulf have drawn strength and inspiration from our natural environment. Inhospitable in some ways, very supportive of human existence in other ways, our natural environment has always been our main natural resource. And, indeed, d1e distinctive envi ronmental and natural conditions of the region have shaped our history and will continue to shape our future. The President of the United Arab Emirates, His Highness Sheikh Zayed bin Sultan Al Nahyan, continually expresses our strong national commit ment to d1e study of our past and the importance of studying our human and natural history. It was
with his support and encouragement that the first ever conference on d1e "Fossil Vertebrates of Ara bia" was held at Jebel Al-Dhanna, United Arab Emirates, in March 1995. The conference provided an opportunity for prominent scientists from around the world to present and discuss the results of their work on the Miocene fossils from Abu Dhabi's Baynunah Formation. The findings of sev eral of the participating distinguished scholars are quite significant. The fossils themselves, we are pleased to note, have proved to be the most impor tant source of terrestrial vertebrates to be found anywhere in Arabia and they are of international significance.
Inauguration by His Excellency Sheikh Nahayan bin Mubaral( Al Nahayan, Minister for Higher Education and Scientific Research, United Arab Emirates (fifth from left) of the First International Conference on the Fossil Vertebrates of Arabia held at the Dhafra Beach Hotel, Western Region, Emirate of Abu Dhabi, United Arab Emirates in March 1995. Back row, from the left, are Hans de Bruijn, Jes de Bruijn, guest, Herbert Thomas, Peter Friend, Vera Eisenmann, Peter Forey, Norman MacLeod, Pascal Tassy, Andrew Hill, John Kingston, Peter Ditchfield, Walid Yasin, Fred Rogl, and Peter Andrews. Front row, from the left, guest, Saif Rashed al Swedi, Finance and Administration Manager, Ministry of Higher Education and Scientific Research, United Arab Emi rates, Peter Whybrow, His Excellency Yousef Omair Bin Yousef, Chairman Abu Dhabi National Oil Company, His Excellency Sheikh Nahayan bin Mubarak al Nahayan, Minister for Higher Education and Scientific Research, United Arab Emirates, Kevin Dunne, General Manager, Abu Dhabi Company for Onshore Oil Oper ations, Sally McBrearty, and other guests. Photograph courtesy of Sah el Baz, Emirates News .
mil
FOREWORD
The fossil record of the Arabian Gulf region provides a very long history of important environ mental changes. Concerted and collaborative efforts of local and international institutions and scholars, as embodied in this volume, are bringing about a proper understanding of Arabia's past natural his tory. In this fossil-rich area, we have alway s known of plants and animals from eras long before humans walked the earth. We are also learning from the diverse fossil record that the area has been a palaeontological bridge between East and West, and we are beginning to understand the processes by which life in Arabia, and in particular human life,
evolved in its changing environment. From this understanding, an appreciation emerges of the unique contribution of our region to the natural history of the world. I am pleased to introduce Fossil Vertebrates of Arabia. The chapters contained herein, written by many of the world's top scholars in the field, con stitute an important contribution to palaeontology and geology, as well as to the scientific studies of the natural history of the United Arab Emirates. I am confident that these proceedings will stand as a significant work of reference and as a stimulus to further research for many years to come.
PREFACE KEN W. GLENNIE
One does not normally associate a land covered with sand dunes and salt-covered sabkhas with a wide vari ety of fossil vertebrates, especially when many of those described between the covers of this book crocodiles and hippopotamuses, for example-obvi ously needed water on a scale that is not found in Arabia today. Fossil Vertebrates ofArabia represents an important compilation of palaeontological data covering an association of vertebrates, both large and small, aquatic and terrestrial. Most of these fossils were found in one rock unit of Miocene age, the Bay nunah Formation of the Western Region, Emirate of Abu Dhabi, United Arab Emirates. Although multiauthored, the volume is not the usual compendium of isolated articles, but repre sents the result of carefully planned co-operation between scientists from Arabia, Europe (mostly the United Kingdom), and the United States of Amer ica. The driving force in both authorship and in ensuring that the work was undertaken in an effi cient manner was the combination of Peter J. Why brow of The Natural History Museum, London, and Andrew Hill of Yale Un.iversity, USA; between them, they also authored or coauthored almost a third of the 36 chapters . Both Whybrow and Hill had visited the origi nal discovery site at Jebel Barakah separately with out knowing of the other's work in tl1e area. Once they realised their common interest, co-operation between them was automatic, and further expedi tions were undertaken to find more fossils and to study them adequately. They also ensured that the sedimentology of the exposed Miocene host rocks was properly described and evaluated to provide a sound palaeogeographic fi:amework for their fossils. Despite the emphasis on vertebrate fossils, they also had the collaboration of palaeontologists working on associated nonvertebrate fossils of the Baynunah Formation, and of palaeobotanists working in other parts of Arabia where tl1is contributed to the over all palaeoecology. The importance of the Abu Dhabi vertebrates in further unravelling the migrational pattern of differ ent animal types between Mrica, Europe, and Asia during the Tertiary cannot be overemphasised. In
this process, Arabia played a pivotal part during the Miocene. First, a partial marine barrier to migration between Mrica and Arabia was created by the open ing of the Red Sea early in the Miocene; and, sec ond, also during the early Miocene, a land barrier to the migration ofTethyan marine faunas between what is now the Indian Ocean and the Mediter ranean Sea was created by the collision of Arabia and Asia, thereby permitting the interchange of terres trial vertebrate faunas bet\veen those two areas for the :first time. The newly formed migrational route to Asia was no doubt broadened by a major fall in global sea level during the later Miocene, probably because of the rapidly increasing ice cover of Antarc tica. The key role taken by the Miocene vertebrate faunas from the Western Region of the Emirate of Abu Dhabi in understanding the migrations between especially Mrica and Asia, will probably not be fully realised until tl1e contents of this volume have been thoroughly digested and compared with studies elsewhere in the Middle East. It is perhaps unusual for a nonpalaeontologist to be invited to write the preface to a major book on 1 fossil vertebrates. It seems, however, that by men tioning in a 1968 publication the discovery of a pro boscidean tooth in gravels at Jebel Barakah, my col league Brian Evamy and I led Peter Whybrow to visit the site in 1979 and to find evidence of other fossil vertebrates; and as Hill and Whybrow record in Chapter 3, in 1982, Whybrow and I were junior authors in a reappraisal of that tooth by Madden et al. ( 1982); the rest is history (see Chapter 3). When Evamy and I wrote our short 1968 paper entitled "Dikaka", we had no idea that it included the first published identification of a Miocene verte brate fossil in Arabia, and would e\'entually be fol lowed by the present treatise. And earlier, in 1965, during my first field trip to southeastern Arabia, I certainly had no idea that for the next 30 years or more I would be involved intermittently in trying to unravel some of its geological secrets. As a Shell research geologist in 1965, my task was to study modern deserts to better understand the Permian (Rotliegend) gas-bearing reservoirs of the Dutch Groningen gasfield, whose great size had
ml!J
PREFACE
only recently been realised; this knowledge was also
not only indicate a much wetter climate in the later
applied to exploration in the southern North Sea,
Miocene than the area experiences today but,
which then was beginning. That 1965 field trip had
a probable time gap of some 9 million years, is
\\ith
already tal<en Evamy and me through much of inte
underlain by dune sands and sabkha sediments of
rior Oman and the Trucial States (now the United
the early Miocene Shuwaihat Formation, which
Arab Emirates).
more akin to the product of today's climate. It is
are
Our direct objective in the western part of the
perhaps pertinent that the late Quaternary climate
Emirate of Abu Dhabi was to gain a better under
in the Emirates fluctuated between hyperariditv at
standing of Sabkh:t Matti, an area of widespread
the peaks of high-latitude glaciations and one that
saltflats, for comparison with the coastal sabkhas
is more humid than today's during interglacials
that were being studied by other geologists of
example, the so-called Climatic Optimum of about
(tor
Imperial College London, and Shell (see, for exam
l 0 000 to 5000 years ago) In this respect, the
ple, Purser, 1973). Jebel Barakal1 acted like a bea
Shuwaihat dune sands apparently migrated south
con, drawing us to make a brief geological diver
wards under the influence of a Miocene northern
sion after obtaining a much-needed shower and
(Shamal) wind, much like the prevailing sand-trans
fresh supplies of water and fuel from the Iraq
porting v;rinds of today.
Petroleum Company (predecessor to the modern Abu Dhabi Company for Onshore Oil Opera
Apart from their importance in terms of verte brate evolution and migration in the area, the con
tions-ADCO) base at Jebel Dhanna. What immedi
tributions to this volume provide information of
ately caught our eye was the lightly cemented red
immense value to geologists like me, with an inter
dened dune sand riddled with rhizoconcretions
est in the Neogene history of Arabia in general,
And b,
(dikaka, tl1e main topic of our 1968 paper) in the
and the Emirate of Abu Dhabi in particular.
coastal cliff, evidence of the close proximity of the
including a study of artifacts from the area associ
water table in an otherwise arid environment.
ated with the most destructive of all vertebrates,
Climbing to the gravels at the top of the cliff
human beings, the book directly impinges on mY
resulted in the discovery (by Evamy, if I recall cor
own special interest in tl1e late· Quaternary historY
rectly) of the proboscidean tooth referred to above.
of the area.
The single event that really brought me back
It is good to see that the United Arab Emirates
into active Middle East geology, however, was an
is in the forefront of several aspects of geological
invitation in 1990 from Dr Terry Adams (well
research in Arabia. This is in no small measure
known to Whybrow and Hill), then General Man ager of
ADCO,
to give a talk on the geology of the
because of support by the Government of tl1e United Arab Emirates (see the Foreword, by His
Al
Oman Mountains to the Society of Explorationists
Excellency Sheikh Nahayan bin Mubarak
in tl1e Emirates. This led to field studies in both the
Nahayan) and by local industry. My ovvn recent
United Arab Emiratq and in the Sultanate of
work and that of Whybrow and Hill would have
Oman (although I could not return to Jebel
been impossible without support from the manage
Barakah until it was vacated by an artillery battery
ment of
after tl1e Gulf War), to the supervision of Ph.D.
ADCO.
The contributions to this book are evidence of
froiT'
students studying desert sediments in the Emirates
today's strong collaboration between scientists
and Permian glaciogenic rocks in Oman, and to the
many scientific disciplines. Such interdisciplinan·
co-convenorship (and leader of two field trips) of
research is now a prerequisite for unravelling the
an international conference on "Quaternary Deserts and Climatic Change" in
Al Ain, Emirate of Abu
Dhabi, in December 1995. Here, I seem to have come full circle, for the rocks of the Baynunah Formation at Jebel Barakah
history of tl1e evolving biosphere and lithosphere i.:-. many parts of the world-especially Arabia, \Yhi.:: i:; is now becoming an important region for studies o:'" past and present climate change.
PREFACE
NOTE
REFERENCES
l. Ken Glennie was educated at the University of
Glennie, K. W., and Evamy, B. D. 1968 . Dikalca:
�
Edinburgh (D.Sc. , 1984) and spent over 32 years
plants and plant-root structures associated with aeo
'''orking as an exploration geologist for Shell in
lian sand.
Palaeogeography, Palaeoclimatology, Palaeo
New Zealand, Canada, Nepal, India, the Middle
ecology 4:
77-87.
East, London, and The Hague. His main research interests comprise desert geology (present and
Madden, C. T. , Glennie, K. W., Dehm, R., Whit
past), geology of the Oman Mountains, and geol
more, F. C., Schmidt, R. J., Ferfoglia, R. J. , and
ogy of the North Sea. Since his "retirement" i n
Whybrow, P. J . 1982.
1987, he has continued t o b e active in these areas.
Stegotetrabelodon (Proboscidea, Gomphotheriidae) from the Miocene ofAbu Dhabi.
He is an Honorary Lecturer at the University of
United States Geological Survey, Jiddah.
Aberdeen, Department of Geology and Petroleum Geology.
The Persian Gulf Holocene Car bonate Sedimentation and Diagenesis in a Shallow Epi continental Sea. Springer-Verlag, Berlin.
Purser, B. H. 1973.
ACKNOWLEDGEMENTS
The editors and other specialists who have carried out fieldwork in the Emirate of Abu Dhabi have
The early work for this project received great help from the Department of Antiquities and
received immense and welcome support from
Tourism, AlAin, and we thank the Secretary, His
numerous organisations in the Emirate. Rarely dur
Excellency SaifAli Dhab'a al Darmaki, for the hos
ing the history of discovery of Miocene terrestrial
pitality and kindness shown by his department at
faunas and floras from the Old World has such sup
that time, especially from Dr Walid Yasin. We are
port and interest been forthcoming. We offer our most sincere thanks to The Pres ident of the United Arab Emirates, His Highness
also grateful to the staff of the Dhafra Beach Hotel, Jebel Dharma, and its General Manager, Mr Sashi Panild<ar, for their logistic help over many years and
Sheikh Zayed bin Sultan Al Nahyan, for his
for their efforts in making the First International
enlightened support and continued interest in our
Conference on Arabian Fossil Vertebrates, March,
work that can now be added to local information
1995, such a great success.
concerning the ancient river systems of eastern Arabia. We are also most grateful to the UnitedArab Emirates Minister for Higher Education and Scien
A book such as this could not have been pro duced without the assistance of many people. We thank Valerie Whybrow, formerly of The Natural History Museum, London, for her initial work on
tific Research, His Excellency Sheikh Nahayan bin
the electronic formatting of manuscripts. To Diana
Mubarak Al Nahayan, for agreeing to be Patron of
Clements (NHM) we are especially indebted for
the First International Conference on the Fossil
her diligent and sustained work on texts, figures,
Vertebrates of Arabia held in the Emirate of Abu
and, especially, references. Other colleagues from
Dhabi during March 1995, and to his ministry,
The Natural History Museum who have assisted are
especially Saif Rashed al Swedi, for the organisation
Norman MacLeod, Jeremy Young, Alan Gentry,
of the conference in collaboration with the Abu
Mike Howarth, Peter Forey, Phil Crabb, Harry
Dhabi Company for Onshore Oil Operations
Taylor, and Paul Lund, the last three of the NHM
(ADCO).
Photographic Unit.
From ADCO itself, our work could not have
Lastly, we thank the Yale University Press team
been carried out without the encouragement of
of Jean Thomson Black, Science Editor, Mary
successive General Managers, Terry Adams, David
Pasti, Senior Manuscript Editor, and Joyce Ippolito,
Woodward, and Kevin Dunne, and their approval of
Production Editor, for their advice concerning the
ADCO's grant to The Natural History Museum to
timely production of this book. To our copyeditor,
support the project. In addition we have received
Sarah Bunney, we are especially grateful. Her long
enormous help from ADCO's Public Mfairs Depart
standing experience of books about palaeontologi
ment, Nabil Zakhour; General Relations, Hassan
cal research and her first-hand knowledge ofArabia
M. Al Saigal; Geodectics, El Badri Khalafalla; and
has greatly improved the content of the volume.
Government Relations, Nasser M. Al Shamsi. Nasr
We also thank Jean Macqueen for her hard work
M. Salameen,ADCO's Senior Translator, kindly
preparing the index.
prepared theArabic section of the volume. We also thank ATA Translations, London, for formatting the Arabic text.
� -------
CONTRIBUTORS
t C. Geoffrey Adams, OBE
France de Lapparent de Broin
The N atural History Museum, Depar tment of Palaeontology, Cr omwell Road, London SW7 5BD, U.K.
Museum National d'Histoire Naturelle, Laboratoire de Paleontologie , 8 rue Buffon, 75005 Paris Cedex 05, France
Salim Al-Busaidi
Hans de Bruijn
:Vlinistry of Petroleum and Minerals, Directorate General of Minerals, P.O. Box 551, Muscat, Sultanate of Oman
Institute of Earth Sciences, Utrecht L'ni\-ersitv, P.O . Box 80021 , 3508 TA U trecht, The Netherlands
Zaher Al-Sulaimani
Diana Clements
Ministry of Petroleum and Minerals, Directorate General of Minerals, P.O. Box 551, Muscat, Sultanate of Oman
The Natural History Museum, Department of Palaeontology, Cromwell Road, London SW7 SBD, C.K.
Peter Andrews
Margaret E. Collinson
T he Katural History Museum, Department of Palaeontology, Cromwell Road , London SW7 5BD, U .K.
Depar tment of Geology, Royal Holloway College , University of London, Egham, Surrey TW20 OEX , U . K.
Mustafa Latif As-Saruri
Ministry of Petroleum and Mineral Resources, Mineral Exploration Board, Aden Branch, P.O. B ox 5252, Ma'alla, Aden, Rep ublic of Yemen
Department of Zoology, University College Galway, Galway, Ireland
John C. Barry
Peter W. Ditchfield
Department of Anthropology, Harvard University, Peabody Museum, Cambridge , Massachusetts 02138 , USA
Department of Geology, University of Bristol, Wills Memorial B uilding, Queens Road, Bristol BS8 1 RJ
Peter Paul van Dijk
Deryck D. Bayliss
10 , The Fair way , Northwood, Middlesex , HA6 3DY, U . K. Laura Bishop
Department of Human Anatomy and Cell Biology, University of Liverpool, P.O . Box 147, Liverpool L69 3BX, U.K. Charlie S. Bristow
Research School of Geological and Geophysical Sciences, Birkbeck College London, Gower Street, London vVCl 6BT, U . K. tneceased
William R. Downs
Bilby Research Center, Nor thern Arizona University , Flagstaff, Arizona 86011 , USA
Vera Eisenmann
Museum National d'Histoire Naturelle, Laboratoire de Paleontologie, 8 rue Buffon, 75005 Paris Cedex 05, France Hamed A. El-Nakhal
Depar tment of Environmental Earth Science , The Islamic University of Gaza, Gaza
m
CONTRIBUTORS
Lawrence J. Flynn
Ernie A. Hailwood
Department of Anthropology, Harnrd UniYersity, Peabody Museum, Cambridge, Massachussets 02138, USA
Department of Oceanography, Palaeomagnetism Laboratory, University of Southampton, Southampton 509 SNH, U.K., and Core Magnetics, The Green, Sedbergh, Cumbria LAlO 5JS, U.K.
Peter L. Forey The Natural History Museum, Department of Palaeontology, Cromwell Road, London SW7 SBD, U.K.
Eberhard "Dino" Frey Staatliches Museum fur Naturkunde Karlsruhe, Geowissenschaftliche Abteiling, Erbprinzenstrasse 13, D-76133 Karlsruhe, Germany
Peter F. Friend Department of Earth Sciences, University of Cambridge, Downing Street, Cambridge CB2 3EQ, U.K.
Andrew Hill Department of Anthropology, Yale University, P.O. Box 208277, New Haven, Connecticut 06520, USA
Louis L. Jacobs Department of Geological Sciences and Shuler Museum of Paleontology, Southern Methodist University, Dallas, Texas 75275, USA
Paul A. Jeffery Alan W. Gentry The Natural History Museum, Department of Palaeontology, Cromwell Road, London SW7 SBD, U.K.
The Natural History Museum, Department of Palaeontology, Cromwell Road, London SW7 5BD, U.K.
John D. Kingston Emmanuel Gheerbrant Laboratoire de Paleontologie des Vertebres (URA CNRS 1433), Universite Pierre et Marie Curie, 4 place Jussieu, 75252 Paris Cedex 05, France
Departments of Anthropology and Geology and Geophysics, Yale University, P.O. Box 208277, New Haven, Connecticut 06520, USA
Norman MacLeod Ken W. Glennie 4, Morven Way, Battater, Aberdeenshire AB35 SSF, Scotland, and the Department of Geology and Petroleum Geology, King's College, University of Aberdeen, Aberdeen AB9 2UE, Scotland Tom Gundling Department of Anthropology, Yale University, P.O. Box 208277, �e"· Haven, Connecticut 06520, USA
The Natural History Museum, Department of Palaeontology, Cromwell Road, London SW7 SBD, U.K.
Sally McBrearty Department of Anthropology, University of Connecticut, U-176, Storrs, Connecticut 06269, USA
Peter B. Mordan The Natural History Museum, Department of Zoology, Cromwell Road, London SW7 SBD, U.K.
CoNTRIBUTORS
m
Phillip A. Murry
Torsten Rossman
Department of Physical Sciences, Tarleton State University, Stephenville, Texas 76402, USA
Sandstrasse 98, D-64319 Pfungstadt, Germany
Ross G. Peebles
Department of Geological Sciences, University of Durham, Durham DH1 3LE, U.K., and Halliburton Energy Services, 800 Halliburton Center, 5151 San Felipe Boulevard, Houston, Texas 77056, USA Daniel S. Pemberton
Staatliches Museum ftir Naturkunde Karlsruhe, Geowissenschaftliche Abteiling, Erbprinzenstrasse D-76133 Karlsruhe, Germany Martin Pickford
Laboratoire de Paleoanthropologie et Prehistoire (URA CNRS 49), College de France, ll place Marcelin-Berthelot, 75231 Paris Cedex 05, France Michael Rauhe
Staatliches Museum ftir Naturkunde Karlsruhe, Geowissenschaftliche Abteiling, Erbprinzenstrasse 13, D-76133 Karlsruhe, Germany
Sevket Sen
Laboratoire de Paleontologic des Vertebres (URA CNRS 1433), Universite Pierre et Marie Curie, 4 place Jussieu, 75252 Paris Cedex 05, France Pascal Tassy
Laboratoire de Paleontologic des Vertebres (UR.c'\ CNRS 12), Universite Pierre et Marie Curie, 4 place Jussieu, 75252 Paris Cedex 05, France Herbert Thomas
Laboratoire de Paleoantl1ropologie et Prehistoire (URA CNRS 49), College de France, 11 place Marcelin-Berthelot, 75231 Paris Cedex 05, France, John E. Whittaker
The Natural History Museum, Department of Palaeontology, Cromwell Road, London SW7 5BD, U.K. Peter J. Whybrow
The Natural History Museum, Department of Palaeontology, Cromwell Road, London SW7 5BD, U.K.
Jack Roger
Bureau de Recherches Geologiques et Minieres, Orleans-la-Source, BP 6009, 45060 Orleans Cedex 2, France
Walid Yasin
Departu1ent of Antiquities and Tourism, P.O. Box 15715,AlAin Museum, Al Ain, Emirate of Abu Dhabi, United Arab Emirates
Fred Rogl
Naturhistorisches Museum Wien, Geologisch-PaHiontologische Abteilung, Burgring 7, A-1014 Wien, Postfach 417, Austria
Sally V. T. Young
The Katural History Museum, Department of Palaeontology, Cromwell Road, London SW7 5BD, U.K.
ABBREVIATIONS
AABW
Antarctic Bottom Water
CAM
crassulacean acid metabolism
.wco
Abu Dhabi Company for Onshore Oil Operations
CENOP
Cenozoic Palaeoceanography Project
ADNOC
Abu Dhabi National Oil Company
cf.
referrable to (a species)
AF
alternating field
ChRM
characteristic remanent magnetism
aff.
affinities with (a species)
chron
unit of time; magnetic time scale
A-horizon
a mineral horizon with an accumulation of organic matter formed or forming at or near the surface of a soil profile; typically has lost iron, aluminium, or clays and is enriched in resistant sandor silt-sized minerals
CL
cathodoluminescence
dM
deciduous molar tooth
dP
deciduous premolar tooth
DSDP
Deep Sea Drilling Project
AMNH
American Museum of Natural History, New York
sp. indet.
gen. et
genus and species indeterminate
apw
apparent polar-wander curve
GPS
Global Positioning System
asl
above sea level
HMC
Holocene Marine Carbonate
AUH
Emirate of Abu Dhabi (in fossil catalogue)
I
incisor tooth
B/L
breadth/length
lAS
Isotopic Analytical Services Ltd
The Natural History Museum, London; formerly the British Museum (Natural History) (in fossil catalogue)
IGCP
BMNH
International Geological Correlation Project
KNM
National Museums of Kenya
ka
thousand years
It
left
M
permanent molar tooth
b.p.
before present
BRGM
Bureau de Recherche Geologiques et Minieres
c.
orca
Ma
millions of years
c
canine tooth
mAjm
milliamperes per metre
c3
generally, a photosynthetic pathway that reflects wooded/forested habitats
MCZ
Museum of ComparatiYe Zoology, Harvard Universitv
generally, a photosynthetic pathway that reflects tropical grassland habitats
MHNT
c4
Museum d'Histoire Naturelle de Toulouse
-
-
---
�_
-_
__ ---
--::::
-
�.
--:.::___:.._.=--��
-
-
-
-
-ABBREVIATIONS MN
Mammals Neogene zones: a series of fossil mammal assemblages from single localities placed in a chronological sequence on the basis of evolutionary stage, entries by migration and exits by extinction of specific ta.''I FOSSIL
VERTEBRATES
��
Figure 1.3. Miocene siliceous cap-rock, Shuwaihat, Emirate of Abu Dhabi.
history of the Abu Dhabi Project, and Part II pre
tors necessary for an understanding of climatic
sents the geological context from which the verte
change during the Miocene.
brate fossils were collected and the local stratigra
Arabia does not readily yield evidence of its
phy. The chapters in Part III describe systematic
past terrestrial animal and plant life to the palaeon
studies on late Miocene invertebrates, reptiles, and
tologist. Apart from the rigours of exploration,
mammals (several of them new species) from the
where serendipity rules, almost all its rocks are
Emirate of Abu Dhabi and discusses the palaeogeo
marine. 2 Where rare continental sediments are nmv
graphical relationships of the fossils. The four chap
exposed, there are no rivers and little rain to erode
ters in Part IV, deal with topics associated with the
fossils from the rocks. When erosion exposes a fossil
fossil biota-taphonomy, carbon isotopes, and Ara
vertebrate, its bones can be fractured and frag
bian palaeoenvironments-and we include a study
mented by extremes of temperature and subse
on lithic artifacts and the earliest evidence of the
quently scattered and abraded by sand-laden winds.
genus Homo in the region. Part V then links the
The rarity of Arabian fossil vertebrates is shown by
Abu Dhabi Project to other research in Asia and
the fact that despite the size of the Arabian geolog
Mrica and includes studies of older faunas from the
ical plate-which is slightly larger than the Indian
Sultanate of Oman and the Republic of Yemen.
subcontinent-there are only 11 locations from
Finally, in Part VI, a broader picture of Arabia in an
which significant vertebrate fatmas of any geological
Old World context is presented. The timing of the
age have so far been collected.
disconnection of Tethys is examined as are events in the Mediterranean and Paratethys that relate to the
As we previously mentioned, studies of these faunas and floras originate from the discovery in
dispersal of mammals. The last chapter provides an
1974 by palaeontologists from The Natural History
up-to-date review of the palaeo-oceanographic fac-
Museum, London-then called the British Museum
m
P. J. WHYllROW
AND
A. HILL
(Natural History)-of the first Miocene terrestrial vertebrate fauna from the peninsula, in the eastern part of the Kingdom of Saudi Arabia. Since the 1970s work has been carried out principally by two European teams collaborating with Arab organisa tions-the oil companies of the countries where the research has been undertalcen. The first team, led by Peter Whybrow of The Natural History Museum, 3 London, and Andrew Hill of Yale University (in collaboration with the Abu Dhabi Company for Onshore Oil Operations), has focused on the late Miocene of the Emirate of Abu Dhabi; the second team, led by Herbert Thomas of the Laboratoire de Paleoanthropologie et Prehistoire, College de France, Paris (in collaboration with the Ministry of Petroleum and Minerals, Directorate General of Minerals, Sultanate of Oman), has concentrated on Oligocene and Miocene rocks in the Sultanate of Oman. In addition, work by various research groups in the Republic of Yemen (collaborating with the Ministry of Oil and Mineral Resources and the Uni versity of Sana'a) has produced the first hints of a Tertiary terrestrial fauna and flora, and the late Jurassic sequence has produced what is believed to be the second record of a dinosaur from Arabia. Palaeontology has changed dramatically since the first vertebrates were discovered in Arabia. As readers of this book will recognise, vertebrate palaeontology is no longer an independent science. Today, to understand-as far as we ever can-the habitats of extinct biotas, palaeontologists collabo rate with other geological specialists; a project such as our research in the Emirate of Abu Dhabi becomes the work of a multidisciplinary team. While the systematic identification of fossils will always be the backbone of vertebrate palaeontology, other specialist studies now provide a multitude of scientific themes that range from evolutionary biol ogy, through palaeobiogeography and the move ment of continental plates, to the diagenesis and isotopes within the rocks and fossils themselves. Explorations for Arabia's palaeontological her itage continue, thanks to the enlightened support of government organisations within the countries of the peninsula. This book is not only the first step in publicising the results of Arabian palaeontological
and geological researches to the benefit of the Ara bian peoples, but it also provides an in-depth testi monial for the emerging academic role that Arabia now provides for linking studies of Old World Ter tiary faunas and environments.
NOTES 1. The impetus for this book came from the First International Conference on the Fossil Verte brates of Arabia held in the Emirate of Abu Dhabi, United Arab Emirates, March 1995. 2. "Serendipity" was coined by Horace Walpole, fourth Earl of Oxford (1717-97), from the Per sian fairy tale The Three Princes of Serendip, in which the heroes possess this gift. 3. The Abu Dhabi Project, continuing until the year 2000, forms part of The Natural History Museum's Global Change and the Biosphere research programme.
REFERENCES Anon. 1975. Mammalian remains from Saudi Arabia. In Report on the British Museum (Natural History)) 1972-1974. Trustees of the British Museum (Natural History), London. Thomas, 0. 1894. On some specimens of mammals from Oman. Proceedings of the Zoological Society of London 1894: 448-55. . 1900. On the mammals obtained in South western Arabia by Messrs Percival and Dodson. Pro ceedings of the Zoological Society of London 1900: 95-104.
---
Wallace, A. R. 1876. The Geographical Distribution of
Animals with a Study of the Relations of Living and Extinct Faunas as Elucidating the Past Changes of the Earth)s Surface. Macmillan, London. Yerbury, J. W., and Thomas, 0. 1895. On the mam mals of Aden. Proceedings of the Zoological Society of London 1895: 542-55.
Summary and Overview of the Baynunah Fauna, Emirate of Abu Dhabi, and Its Co11text ANDREW
H1LL
AND
PETER J . WHYBROW
How impossible must it be for us to guess, in most cases, at the exact nature of the forces that limit the range of some species and cause others to be rare or to become extinct! All that we can in general hope to do is, to trace out, more or less hypothetically, some of the larger changes in physical geography that have occurred during the ages immediately preceding our own , and to estimate the effect they will probably have produced on animal distribution. We may then, by the aid of such knowledge as to past organic mutations as the geological record supplies us with, be able to determine the probable birthplace and subsequent migrations of the more important genera and families
-Wallace (1876)
One of the most interesting biological features of the Arabian Peninsula is the part it has played in Old World biogeography. Arabia lies at the junction of the classic Old World biogeographic divisions the Ethiopian, Palaearctic, and Oriental regions (Wallace, 1 876) . The Arabian Peninsula is also a 1 large global area. Situated between 12° and 30°N, and between 35° and 60°E, it occupies an area of just over 3 million square kilometres-almost all of the Arabian continental plate. Arabia is therefore about as large as the Indian subcontinent. The peninsula today provides a diverse set of habitats that range from the mountainous regions of the southwest with their high plateaus and where some peaks reach nearly 3800 metres, to the low-lying sand deserts that occupy most of the eastern area. In the north, bordering the Arabian Gulf, are salt flats, some of which are below sea level . The region also has a variable climate, within the arid to hyperarid range (Takahashi and Arakawa, 198 1 ) . For example, in the high south west there are 5 0 0 mm of rain a year, with a low temperature range and snow on some high moun tains in the cold season. In the central and eastern areas, which are occupied by some of the hottest and most arid deserts on earth, daily mean temper-
Copyright © 1999 by Yak University.
atures reach close to 40 oc and rainfall is slight. Descriptions of the arduous journeys under talf.-\rabia, held in the Emirate of Abu Dhabi during
Palaeontology Department at The Natural His
:\larch 1 995, and to his ministry and ADCO tor their
tory Museum, and Dr Neil Chalmers, Director of
;.p onsorship of the meeting and for their continuing
The Natural History Museum, as well as by rep
:-esearch support. We are also grateful to the excel
resentatives of British business in Abu Dhabi.
'ent support and sponsorship received from the Abu Dhabi National Hotels Company (especially the Dhafra Beach Hotel, Jebel Dhanna, and its General .\ lanager, Mr Sashi Panild pp. 209-33
18
in
(ed. P. J.
V\'hybrow and A. Hill) . Yale University Press, New Hailwood, E . A. , and vVhybrow, P. J .
1 999.
Palaeo
Haven.
magnetic correlation and dating of the Baynunah and Shuwailiat Formations, Emirate of Abu Dhabi, United Arab Emirates. Chap .
Arabia,
pp.
75-87
8
in
Fossil Vertebrates of
(ed. P. J . \Vhybrow and A.
Yale University Press, New Haven. Hill , A., and Gundling, T.
1 999. A monkey
mates; Cercopithecidae) from the late .Miocene of
Vogt, B . , Gockel, \V., Hofbauer, H . , and Al- Haj, A. A. 1 989. The coastal survey in the Western Province of Abu Dhabi. Archaeology itt the United Arab Emirates V: 49-60. Whybrow, P. J.
1 989.
New stratotype; the Baynunah
Formation (Late Miocene), United Arab Emirates:
PALAEONTOLOGICAL RESEARCH IN THE WESTF.!U'-: REGION
Litl1ology and palaeontology. Newsletters on Stratigra phy 2 1 : 1-9.
Whybrow, P. J., Hill, A., and Yasin al-Tikriti , W. 199 1 . Miocene fossils from Abu Dhabi. Tribulus: Bulletin of the Emirates Natural History Group 1 :
Whybrow, P. J., and Bassiouni, M. A. 1986. The Ara bian Miocene; rocks, fossils, primates and problems. In Primate Evolution, pp. 85-9 1 (ed. J. G. Else and P. C. Lee ). Cambridge University Press, Cambridge.
4-9 .
PI]
Whybrow, P. ]., Hill, A., Yasin al-Tilcriti, W. , and Hailwood, E. A. 1990. Late Miocene primate fauna, flora and initial palaeomagnetic data from the Emirate of Abu Dhabi, United Arab Emirates. Journal of
Whybrow, P. J., and McClure, H. A. 198 1 . Fossil mangrove roots and palaeoenvironments of me Miocene of me eastern Arabian peninsula. Palaeogeog raphy, Palaeoclimatology, Palaeoecology 32: 2 1 3-25 .
Human Evolution 19: 583-88.
>J,: .
·
.....
''· -
>·•
.
.: . .;,
>�-�
'j . . •
..
' � :
Plate 3 . 1 . Delegates to the First International Conference on the Fossil Vertebrates of Arabia enjoying a field excursion to Miocene localities in the Western Region, Emirate of Abu Dhabi .
Peter Whybrow gazing at a sea-cliff exposure of the B aynunah Formation, Jebel B arakah , Western Region, Emirate of Abu Dhabi.
Miocene Geology of the Western Region, Emirate of Abu Dhabi, United Arab Emirates
PART
-
When Whybrow started his first geological studies of the Miocene exposures in the Western Region i n
ment. These appear to have been variable in flow, and probably some 3-10 metres deep at time of
1 9 79-84 the only publications available were large
flood. They were braided in the sense that they con
scale geological maps of the area and a brief report
tained many sediment bars, and channels of varied
by petroleum geologists. Much of his work was con
size and form. Flow directions appear to have been
fined to the coastal exposures where the complex
generally towards the ESE. The river system may
lithologies of the rocks were, in general, more easily seen. Miocene exposures farther inland from the coast, both then and now, are mostly covered with
have been ancestral to the presen t-day Tigris Euphrates system, but this is a speculative sugges
wind-blown sand and eroded Miocene sediments. Since the 1980 s, more detailed studies by The Nat
tion. Sediments of the Shuwaihat Formation that lie
ural History Museum, London/Yale University
beneath the river sediments of the Baynunah For mation have been studied by Charlie S . Bristow
team have obtained the basic data for any future Miocene researches in the area. Mention must be
nate the Shuwaihat Formation, formerly part of the
( Chapter
6).
Aeolian and sabkha sediments domi
made of the fact that, although surface exposures
Baynunah Formation. Aeolian dune bedforms are
reveal only about 60 metres of Miocene sediments, the succession in the Western Region falls into a period of Miocene time from about 19 to 6 million
well exposed and indicate that dune morphology was transverse and barchanoid. The dune sands are
years (Ma) ago . The age of the vertebrate-bearing
often interbedded with, and truncated by, sabkha sediments that can be correlated for several kilome
sediments of the lower p art of the Baynunah Forma
tres along strik e . Two types of sabkha are recog
tion is believed to be between 8 and 6 M a . T h e first step in a n y geological study of a previ ously unknown area is to define the succession and provide names for stratigraphical units if they can not be easily linked with other regional units. Peter J. Whybrow, Peter F . Friend, Peter W. Ditchfield, and Charlie S. Bristow ( Chapter 4) present geologi cal information supporting a new stratigraphic
rused: small interdune sabkhas ( 1 0-100 metres wide ) , and more extensive playa lake or coastal sabkhas ( more than 1 0 00 metres wi de ) . These sedi ments are occasionally interbedded with and over lain erosionally by fluvial sediments. Palaeocurrent measurements from the aeolian sandstones indicate that the dominant palaeowind was from nortl1 to south. Palaeocurrent directions from fluvial sedi
name, Shuwaihat Formation, for sediments pre
ments indicate a much wider current dispersion
dominantly of aeolian origin that overlie the marine
vvi th rivers flowing from west to east and south tO
Dam Formation, previously described from the
north, across or opposed to the aeolian transport direction. Palaeogeographic reconstructions indicate
Kingdom of Saudi Arabia. The previously defined name, Baynunah Formation, is now restricted to the overlying, predominantly fluvia l , unit that con tains the important nonmarine fossil fauna
that the Sbuwaihat Formation was deposited in an intracratonic basin with a semi-enclosed drainage system where rivers debouched into an inland
described in Part III. Sedimentary logs measured at
sabkha o r p laya lake. Changes in lake level that
1 4 localities are presented.
strongly affected the dune systems are attributed to
Peter F. Friend ( Chapter
5 ) describes the sedi
ments deposited by the river system that once flowed through the area known today as the Baynunah. The lower Baynunal1 Formation consists mainly of fine grained sands, but also contains numerous gravel beds and some distinct mud units. Its important fos sil vertebrate fauna, along with some molluscs, abun dant root marks and soils, con firms that the unit was formed predominantly by river deposition . There is no positive evidence of marine influence . Detailed examination of unusually clean expo sures at Jebel Dhanna Dalma Ferry Terminal and a t Jebel Mimiyah (AI Mirfa ) provides information about the rivers that deposited some of the sedi-
climatic fluctuations during the late Miocene. The presence of reworked foraminifera in some of the aeolian sands suggests that marine conditions were present within the basin, so the overall setting may have been a low-relief coastal plain although no evi dence of marine sediments in the Shuwaihat For mation has been fo und at outcrop. To understand the palaeoenvironment in which sediments were deposited, it is necessary to under take detailed analyses of the rocks formed from such sediments. Peter W. Ditchfield's collection of samples ( Chapter 7 ) from the exposed Miocene for mations from the \tVestern Region have been analysed petrographically and geochemically to
INTRODUCTION
determine the degree of diagenetic alteration and
rl.4
a long-reverse polarity magnetozone in the Jebel B aral_:1 � � � l- w a: w �5 tii t;
INTERPRETATION
::;; :0 o a: W f(f)CJJ
LOG 2
LOG 1
Fluvial channel sandstones with reworked calcrete clasts
.,
Poorly sorted granule conglomerate of calcrete clasts
Fl uvial sandstones
-
lill
grey
Fluvial sandstones
'Wasp-nest' concretions become dominant
Palaeo-watertable ?
Aeolian
Aeolian sandstone
Aeolian
Sabkha
1- -
-
-
- -
Sabkha
- 225 metres approx.
- - - --/
Figure 6 . 3 . Sedimentary logs of the coastal cliff at Jebel Barakah showing basal sabkha facies overlain by aeolian �andstones, which are truncated by fluvial sediments. The fluvial incision has cut out 6 metres of aeolian sands L"1 log 2 . For key to lithologies and sedimentary structures, see figure 6.2.
SHUWAIHAT (JAZIRAT SHUWAIHAT) Shuwaihat is about 10 km east of Jebel Barakah and provides the best exposures of aeolian and sabkha sediments in the Shuwaihat Formation. This rrea is proposed as the stratotype for the Shuwaihat Formation. The exposures are described from two
locations, on the west and east sides of the island, that illustrate the typical facies and lateral variations seen in the Shuwaihat Formation (figs 6 . 5 and 6. 6 ) . On the west coast there i s continuous exposure for more than l km in the low 5 metre-high cliff, which provides an excellent exposure of aeolian cross-stratification ( see fig. 6.7). On the east coast
�
C. S. BRISTOW
NE
-�� Fl uvial
Grey d•y
C ong l omerate
J EBEL BARAKAH
�
,=�'- ,:;- white limestone
�-� ----
Rhi
sw
l ilh'
E� io" '""'"
z;·· }�=;����;�� � ��;����;� Beach
�
LOG 2
Sabkha sandstones
_
_ __ _ __
� _-
Sand bar
0
1 00
(
(':- - c:,- ' -
-'
200
_- ' /,_-
(
:
Ce 1 est1te . cemented
Aeolian
-
san dstone
-
/"... 1r LOG 1
sand
300
400
metres
Figure 6.4. Outcrop sketch of Jebel Barakah showing the location of logs 1 and 2 (fig. 6 . 3 ) and the fluvial erosion surface that cuts into aeolian sands. the equivalent section shows more variation with less aeolian cross-stratification, more sabld1a sedi ments, and more fluvial deposits. Four sedimentary logs of the eastern outcrop are shown in figure 6.6; the sections are correlated at the base of a grey claystone with gypsum veins that can be traced along the outcrop.
Western Outcrops of Shuwaihat On the western side of Shuwaihat dune sandstones are exposed at very low tide; these are overlain by sabkha deposits exposed at beach level (fig. 6.5 ) . The overlying aeolian dune sandstones are excep tionally well exposed in low cliffs and platforms along the coast ( Bristow and Hill, 1 99 8 ) . The sand stones are quartz dominated ( 60% ) , with fine, sub rounded grains of moderate sphericity. They also contain 5% feldspar and 3 0% dolomite clasts, some of which appear to contain a relict fabric, indicating that they are reworked from the underlying Dam Formation. The sandstones also contain rare reworked bivalve fragments and foraminifera that have been reworked and blown in. The foraminifera are tentatively identified as rotaliids ( cf. Ammonia), peneroplids, and elphidiids. These foraminifera live in very shallow water associated with algal mats and such like. Ammonia and Peneroplis spp . are known
from the Miocene to the present day. Unfortunately the foraminifera do not help to refine the stratigra phy or the environment of deposition because they are clearly reworked. Comparative studies of modern aeolian dunes in Abu Dhabi indicate that foraminif era occur in dune sands more than 80 km inland (Jonathan Pugh, personal communication, 199 5 ) . Sedimentary structures exposed in planform and cross-sectional outcrops indicate that the dunes were transverse to barchanoid in form, with a dominant palaeowind from north to south (Bristow and Hill, 1998) (plate 6.2, p. 57). Within the aeolian sand stones, sablma facies of only a few tens of metres in lateral extent have been identified. In one outcrop, sablma deposits onlap dune morphology, indicating that the sabkha rose around the dune. This contrasts with most of the other sablma deposits that are much more laterally extensive ( about 1 000 metres plus) and truncate dune morphology. The sablma facies that overly the aeolian sandstones on the west ern coast of Shuwaihat are a typical example. Here the aeolian sandstones are truncated by a sandy sablma facies overlain by a grey claystone with large selenitic gypsum crystals that cut across the bedding. The clay and gypsum bed forms a good marker hori zon along the western coast of Shuwaihat, which can be correlated with the eastern coastal section. The later-
SEDIMENTS OF THE SHUWAIHAT FORMATION
INTERPRETATION
Reworked calcrete conglomerate fluvial
F l uvial
Sabkha Playa lake
Aeolian
I nterdune sabkha
Aeolian Sabkha
(Base of section not exposed)
Figure 6 . 5 . Sedimentary log of the western side of Shuwaihat. For key to lithologies and sedimentary structures, see figure 6.2. ill�� restricted sabkhas that onlap dune morphology interpreted as interdune sabkhas whereas the :::1ore extensive sabkha facies with gypsum and clay asinal source that could be due to intrabasinal �ectonic uplift. Alternatively, the conglomerates could be derived from intrabasinal erosion due to fluvial incision caused by a fall in base level . In the Shuwaihat Formation this fall in base l evel could have resulted from desiccation of a lake or a fall in sea level . It is possible that the fluvial incision was produced during a sea-level lowstand and that accu::-:�c:
: .•.n
Formation with rivers flowing into an evaporitic
mulation occurred during the subsequent transgres sion . This model cannot be disproved, and although there is no clear evidence of marine sedi ments in the outcrops the presence of reworked foraminifera indicate that marine sediments were not totally absent from the basin. In this respect base-level changes could be invoked to account for some of the observed facies changes and stratigra phy of the Shuwaihat Formation. The sabkha facies are interpreted here as inland sabkhas, \Yith interdune ponds and plan lakes. The interdigitation of aeolian and sabkha sediments in the Shmnihat Formation and the onlap of an aeo lian dune b�· interdune sabkha f�Kies suggest that water levels within the evaporitic playa lakes often varied so that base-level changes can be explained without invoking sea-level change . The interdigita tion of aeolian, lacustrine, and fluvial facies is most likely to have been climatically driven with some autocyclic variations due to lateral migration of depositional systems. An intrabasinal tectonic con trol on base-level change cannot be ruled out, but it is clear that the climate fluctuated during the Miocene and it is possible that the observed strati graphic changes could be due entirely to climate change.
CONCLUSIONS The Shuwaihat Formation is an unconformity bounded sequence of fluvial, aeolian, and continen-
�
C. S. BRISTOW
tal sabkha sediments. The base of the unit is not properly exposed but is believed to be uncon formable on presumed middle Miocene carbonates of the Dam Formation . The top of the member is an erosion surface overlain by fluvial sediments of the Baynunah Formation, best exposed at Jebel Barakah. vVhere there is no clear exposure of an erosive base to the overlying Baynunah Formation the contact is taken at the top of the highest sabkha facies. The fluvial deposits in the Shuwaihat Forma tion show a dominant palaeoflow from west to east with a wide variance. Aeolian cross-stratification shows much more consistent palaeowind directions towards the south. The aeolian sandstones are usu ally interbedded with continental sabkha deposits . These include interdune sabkhas that onlap dune topography and more-widespread playa !alee sabkhas that truncate dune topography. The transition from an aeolian-dominated Shuwaihat Formation to a fluvial-dominated Baynunah Formation is best explained by a change in climate .
ACKNOWLEDGEMENTS This chapter forms part of the Birkbeck/University College London-Natural History Museum Global Change in the Biosphere research theme. I thank Peter Whybrow, Department of Palaeontology at The Natural History Museum, for inviting me to join this project; Nick Hill, Tim Goodall, and Jonathan Pugh for their help in the field; and the Abu Dhabi Company for Onshore Oil Operations (ADCO) for providing vehicles and financial support through their grant to Peter Whybrow, without which this work could not have been undertaken. I also thank John Whittaker (NHM) for identifYing the foraminifera.
REFERENCES Bristow, C. S . , and Hill, N. 1998 . Dune morphology and palaeowinds from aeolian sandstones in the
Miocene Shuwaihat Formation, Abu Dhabi, United Arab Emirates. In Quaternary Deserts and Climatic Change, pp. 5 5 3-64 ( ed. A. S. Alsharhan, K. W. Glennie, G . L. Whittle, and C. G . St. C. Kendall). A. A. Balkema, Rotterdam. Hailwood, E. A., and Whybrow, P. J. 1999. Palaeo magnetic correlation and dating of the Baymmah and Shuwaihat Formations, Emirate of Abu Dhabi, United Arab Emirates. Chap. 8 in Fossil Vertebrates of Arabia, pp. 75-87 (ed. P. J. Whybrow and A. Hill) . Yale University Press, New Haven. Van Wagoner, J. C., Posamentier, H. W. , Mitchum, R. M., Vail, P. R., Sarg, J. F., Loutit, T. S., and Hard enbol, J. 1988. An overview of sequence stratigraphy and key definitions. In Sea-level Changes: An Inte grated Approach ( ed. C. W. Wilgus). Society of Eco
nomic Paleontologists and Mineralogists Special Publi cation 42: 39--45 . Whybrow, P. J. 1989. New stratotype; the Baynunal1 Formation (late Miocene), United Arab Emirates: Lithology and palaeontology. Newsletters on Stratigra
phy 2 1 : 1-9 . Whybrow, P. J., Friend, P. F., Ditchfield, P. W. , and Bristow, C . S . 1999. Local stratigraphy of the Neo gene outcrops of d1e coastal area: Western Region, Emirate of Abu Dhabi, United Arab Emirates. Chap. 4 in Fossil Vertebrates ofArabia, pp. 28-37 ( ed. P. J. Whybrow and A. Hill) . Yale University Press, New Haven. Whybrow, P. J., Hill, A., Yasin al-Tilcriti, W. Y., and Hailwood, E. A. 1990. Late Miocene primate fauna, flora and initial palaeomagnetic data from the Emirate of Abu Dhabi, United Arab Emirates . Journal of
Human Evolution 19: 583-88 .
Diagenesis of the Baynunah, Sl1uwaihat, and Upper Dam Formation Sediments Exposed il1 the Western Re gion, Emirate of Abu Dhabi, United Arab Emirates PETER W. DITCHFIELD
this study samples from the various lithological .:::its of the Upper Miocene Baynunah Formation -- ae collected and analysed for their petrographic :.:�d geochemical variation to ( I ) help determine :::e depositional environments represented by these '� diments and ( 2) determine the degree and type , r· postdepositional alteration that has affected ::1em . In addition, samples from the underlying _'lliocene Shuwaihat and Dam Formations were also �ollected for comparison with those from the Bay :1Lmah Formation. The samples were collected from various out �rops between AI Mirfa in the east and As Sila in ::he west during the I 99 3 and I 994 field seasons fig. 7. I ) . All of the lithological units of the Bay nunah Formation were sampled (see fig. 7.2 for a schematic log through the Baynunah lithofacies) .
carbonate samples were obtained by inductively coupled plasma atomic emission spectrometry (ICP AES) and were calibrated against in-house multi element standards. Carbon and oxygen stable iso topic analyses were also carried out on carbonate samples. Samples for isotopic analysis were extracted using a tungsten carbide micro-drill, then routinely roasted at 400 oc under vacuum to remove any organic contaminants. Carbon dioxide ( C02 ) gas for analysis \vas obtained by reaction with I OO% phosphoric acid ( H3 P04) at 25 oc. Samples were analysed on a VG Isogas mass spectrometer. The results were corrected bY the method of Anderson and Arthur ( I 9 8 3 ) and calibrated to PDB (Peedee Formation belemnite standard) by repeated analysis of a carbonate standard (NBS I 9 ) . Analytical preci sion is better than 0 . I per mil ( %o) .
METHODS
RESULTS Petrography
= =:
The samples were analysed petrographically by standard thin-section techniques, which included cathodoluminescence ( CL) and staining for various carbonate phases ( Dickson, I 966). Some samples \Yere also analysed with a scanning electron micro scope (SEM ) . Mineralogies were confirmed by X ray diffi·action (XRD ) . Trace element analyses of
Copyright ©
Much of the sandstone within the Baynunah Forma tion is composed of fine- to very fine-grained sub quartzite with less than 5% feldspar grains (plate 7. Ia, p. 64) . This lithology is combined with varying amounts of probably soil-derived carbonate clasts, ranging in size from granules to cobbles, to give
1999 by Yale Universitv. All rights reserved. ISB;-,r 0-300-071 8 3 - 3
na_
. EEL
�
P. W. DITCHFIELD
53"
t
Arabian Gulf
N
0
24°
I
AI Ghuddah
\
bin
Jawabi
E m i rate of Abu Dhabi
Medinat Zayed
l<m
53"
Figure 7 . l . Outline map of the study area. the characteristic conglomerate lithology of tl1e Baynunah Formation (plate 7 . 1 b ) . More exotic extraformational clasts are seen only in a patchily developed, well-cemented conglomerate facies at the base of the Baynunah Formation (plate 7 . l c ) . Preservation o f porosity within tl1e remaining lithologies of the Baynunah Formation is good. Grain contacts are restricted to points only. The rare biotite grains within the sediment are generally " ell preserved with no petrographic signs of alter ation; these grains are also undeformed by com paction. These features suggest that the sediments of the Baynunah Formation have undergone only a minor amount of burial. Carbonate cements arc observed in samples only where there are locally abundant carbonate (originally aragonitic) bioclasts, in particular unionoid bivah·es or in some cases gastropods. These are largely preserved only as mouldic poros ity that occasionally contains pore-lining, equant, sparry, calcite cements (plate 7.2a, p. 6 5 ) . Sparse
circumgranular, equant, pore-lining or pore-filling, nonferroan calcite cements are present in the imme diate area of the bioclasts, which show a concentri cally zoned dull to moderate orange luminescence under CL (plate 7.2b ) . These cements occasionally show pendant and meniscate morphologies, the lat ter being especially common within the poorly cemented root casts in the cross-bedded sandstone facies of the upper part of the Baynunah Formation (plate 7.2c). These cement morphologies suggest that dissolution of bioclasts and precipitation of car bonate cements took place soon after deposition and above the permanent water table within the vadose zone. Similarly, the micritic limestones in the upper part of the Baynunah Formation show significant remobilisation of carbonate only in one location (Al Ghuddah) , where originally aragonitic ? cerithid gastropods were locally abundant; tl1ese are pre served only as mouldic porosity. At Al Ghuddah, fenestral cavities within the limestone show some
DL".GE>;ESIS OF SEDIMENTS
Displacive gypsum/anhydrite cap rock with tabular cherts Inter-bedded fine grained sandstones and micritic white limestones
r;-,:
THE WESTEfu'! REGION
PZIIJ �
Figure 7.2. Schematic log through the B aynunah, Shuwai hat, and upper part of the Dam Formations showing the charac teristic lithofacies in each forma tion.
Interbedded fine grained sandstones and clays
Large scale cross-bedded fine grained sandstone with abundant rootlets
I ntra-formational conglomerates and sandstones with large vertebrate remains
Conglomerate containing extra-formational clasts Mudstone with gypsum sheets
Dune bedded aeolian sandstone
Wavey laminated overbank deposits with abundant celestite rootlets Fine grained fluvial sandstone with climbing ripples, clay drapes and soft sediment deformation Dolomites with marine bivalves and abundant displacive anhydrite
pore-filling, nonferroan, dully luminescent cements. These micritic limestones from the upper part of the Baynunah Formation are preserved as nonfer roan, low-magnesium calcite and contrast strongly with the micritic limestones of the underlying Dam
Formation (as seen at As Sila) that have been per vasively dolomitised. Sulphate cements within the Baynunah Forma tion are largely restricted to a patchily developed basal conglomerate and to the "cap-rock" facies.
a
b
c
Plate 7.l.a, Photomicrograph of fine-grained subquartzite lithology characteristic of the fine-grained sands in the Baynunah Formation (field of view 2.5 mm). b, Outcrop view of the intraformational con glomerates typical of the coarse grained w1its within the Baynunah Formation. The fine-grained matri.x is composed of fine-grained sub quartzite and the larger clasts are dominated by reworked pedogenic carbonate (lens cap is 5 em in diam eter). c, Outcrop view of the well cemented extraformational con glomerate at the base of the Baynunah Formation as exposed in the small quarry on Shuwaihar Island. This figure shows the con glomerate disconformably overlaying d1e clay wid1 gypsum sheets of d1e Shuwaibat Formation .
Plate 7.2.a, Photomicrograph of bio mouldic porosity after dissolution of a unionoid bivalve, now partially filled by sparry dacite cement, Baynunah Formation (field ofview 5 mm). b, Cathodoluminescence photomicrograph of concentrically zoned equant pore-lining cement fi·om the Baynunah Formation (field of view 5 mm). c, Photomicrograph of intergranular mcniscate calcite cement from the upper Baynunah Formation (field of view 2.5 mm).
a
b
�
P. W. DITCHFIELD
Within the basal conglomerate (as seen in the small quarry on Shuwaihat Island) coarse-grained, often poikilotopic gypsum cements arc common (plate 7 .3a, p. 67 ) . These closely resemble the gyp sum cements in the aeolian sandstones of the underlying Shuwaihat Formation. The Baynunah Formation, however, does not contain the abun dant celestite cements ( replacing rootlets) tlut occur in tl1e overbank sediments of the Shuwaihat Formation. The gypsiferous "cap-rock" facies at the top of the succession contains abundant coarsely crystalline, displacive gvpsum cements with minor amounts of anhydrite . The classic chicken-wire texture associated with displacive gypsum growth in sabld1a environ ments, however, is only rarely seen. Gypsum crystals are usually arranged as parallel to subparallel equant laths up to 10 mm long. Large-scale displacive fea tures such as "Tee Pee" structures are also only occasionally seen in outcrop (plate 7.3b ) . The origin of this lithology is unclear, but the lack of character istic petrographic textures and the absence of dolomite from either tl1e cap-rock or tl1e underlying micritic limestones suggest that this lithofacies does not represent deposition in a classic coastal sabkha type environment. A more likely possibility is that it represents recent gypsum growth from the evapora tion of ground water on, or at, an erosion surface, the stratigraphic position of this surface being con trolled by the underlying tabular chert beds. Gypsum cements are volumetrically important in the dune-bedded aeolian sandstone facies of the underlying Shuwaihat Formation, where large, irregular, single gypsum crystals (up to 10 em long) form a coarse-grained poikilotopic cement. Silica cements 1vithin the micritic limestones at the top of the section are in the form of replaci1·e cryptocrystalline hYdrated cherts (plate 7 .4a, p. 68 ) , which often preserve the fine-scale algal laminations of the original limestone. Occasionally, however, as at the Hamra sites, silica cementation replaces large sub horizontal branching ( thalassinoides-type) bur row networks (plate 7.4b ) . These cemented bur rows often retain a circular cross-section, which contrasts witl1 tl1e slightly flattened adjacent nonce-
...._____ ...___
-
mented burrows and indicates silica cementation early in the diagenetic history prior to the small amount of compaction that has taken place. Wiiliin some of the larger bone fragments silica cements are also developed. These include a fine grained fibrous chalcedonic quartz phase and a later coarse-grained drusy quartz cement (plate 7.4c ) . The chalcedonic quartz cements are length-fast. No length-slow chalcedonic quartz-a possible indica tor of quartz precipitation in evaporative environ ments (Folk and Pittman, 1 971 )-was observed.
Geochemistry of Carbonate Phases The geochemistry and mineralogy of the analysed carbonate phases are shown in table 7. 1 . It is clear from these data that the dolomites of the Dam For mation are significantly different from the dolomites of the overlying Shuwaihat and Bay nunah Formations . In particular, the dolomites hom the Dam Formation are close to stoichiomet ric dolomites whereas those from the overlying units are enriched in calcium. The Dam Formation dolomites also contain considerably less iron than those from the overlying formations ( fig. 7. 3 ) . The most striking difference between these two sets of dolomites is in their stable isotopic compositions (fig. 7.4); the samples from the Dam Formation arc strongly enriched in 81 8 0 ( 3 .9 1-5 .63%o), indicating a probable evaporitic origin for the dolomitising fluids. In the dolomite samples fi·om tl1e overlying units, however, the 81 8 0 values range from -2 . 7 1 %o to -5. 54%o, ·which may indicate dolomitisation in a marine/meteoric mixing zone. Samples from the Shuwaihat Formation are more depleted in 8 1 3C (-2 . 04%o to -3.80%o) relative to PDB than the sam ples from the underlying Dam Formation ( -1 .08%o to -2 . 54%o ) . Although the mcygen isotopic data from the Dam Formation dolomites point strongly to an evaporitic origin for the dolomitising fluids, such dolomites are often calcium-rich and poorly ordered. The dolomites of the Dam Formation, however, are stoichiometric and well ordered. This may be the result of recrystallisation during a period of deeper burial before uplift and the forma-
DIAGENESIS OF SEDIMEKTS I� THE WESTERN REGION
a
b Plate 7.3.a, Photomicrograph of gypsum-cemented fine-grained subquartzite lithology from the Baynunah Formation (field of view 5 mm). b, Outcrop view of minor "teepee" structure within the gypsiferous cap-rock ofthe Baynunah Formation (hammer handle is 30 em long).
§J
- -- --
-
��-�-
�
- -
-----
Plate 7 . 4 . a, Hand specimen of replacive l aminated chert after lam inated micritic ( possibly ::d gal ) limestone, from the upper part of the Baynunah Formation . b, Out crop view of replacive chert infill ing a thallasinoides-type burrow network in the upper part of the Baynunah Formation. c, Photomi crograph of fine-grained chal cedonic quartz cements and later coarser-grained drusy quartz cements vvithin the porosity of a bone fi-agment from the elephant excavation site in the Baynunah Formation from Shuwaihat Island ( fi eld of view 5 rn m ) .
a
b
c
-
DIAGENESIS OF SEDIMENTS I:\ THE \VESTE&'.; REGION
tion of the unconformity separating the Dam For mation from the overlying units. The timing of such uplift is uncertain but could be related to the uplift of the Qatar arch (l\lsharhan, 1989 ) . Fenestral Micrites
Results of stable isotopic and trace element analy ses are shown in table 7 . 1 . The isotopic values for the fenestral micrites from the top of the sequence are shown in figme 7 . 5 . The 0180 val ues range from -3 .99%o to -5.29%o with a mean ,-alue of -4.40%o ( n 4 ) . The o13C values range trom -l .64%o to -4.68%o, with a mean value of -3. 22%o ( n 4 ) . The sparse faunal evid ence sug gests that these poorly consolidated micritic lime stones were deposited in a marine to slightly brackish-water environment (N. J. Morris, per sonal communication ) . The oxygen isotopic data are not incompatible with such an environment, but precipitation of the carbonate would have had to have taken place at temperatures in excess of 35°C, assuming an isotopic composition for the water of O%o ( relative to Standard Mean Ocean Water--8MOW). Modification of the isotopic composition of tl1e water by evaporation would reduce this temperatme estimate. The oxygen iso topic data wo uld also be compatible with diage netic recrystallisation vvitl1in a meteoric realm simi lar to that envisaged for the precipitation of the meteoric cements described below, but there is lit tle petrographic evidence for wholesale recrystalli sation of these limestone units. ==
Co/cite Cements
Isotopic results tor the rare, sparry calcite cements within the sandstone units of the Baynunah Forma tion are also shown in figure 7 . 5 . The 01 80 values range from -l . 5 1%o to -9 .44%o, \Vith a mean value of -4.99%o ( n = 7); o13C values range from - l .66%o to -10 .40%o, with a mean value of -4 .25%o (n 7 ) . There i s good petrographic evidence to suggest that at least some of these cements precipitated from meteoric water percolating down through the vadose zone (see "Petrography" section, above ) .
WJ
Modern-day precipitation in this area ranges in composition from 0%o to + 2%o (SMOW) (Yurtsever, 1 975 ) . The oxygen isotopic data are compatible with precipitation of the above cements from waters of such a composition. The moderately neg ative o13C values suggest the incorporation of some carbonate carbon from an organic source as well as trom the dissolution of carbonate bioclasts. Stable Isotopic Analysis of Baynunah Formation Bioclasts A variety of bioclasts from the Baynunah Formation were analysed p etrographically ( by SEM) for any textural evidence of recrystallisation. Of these, only the ratite eggshell san1ples that were originally cal citic preserve their primary texture and therefore possibly their original chemical composition . The carbon stable isotopic signature of ratite eggshell from the Neogene Siwalik sediments of southern Asia has been used to reconstruct a record of envi ronmental (vegetational) change (Stern et al. , 1994) . In this study the stable isotopic composition of 16 samples of ratite eggshell was analysed to try to deduce the isotopic composition of the dietary carbon ingested by the Baynunah avian fauna. In two of the samples (AUH 1 5 , locality Hamra 3, and i\B X2) there was evidence of significant amounts of recrystallisation of the original, main prismatic calcite shell layer. The remaining samples all showed well-preserved prismatic and layered tex tures characteristic of thick ratite shells (Silyn Roberts and Sharp, 1986). The preservation of these primary textmes was used by Stern et al. ( 1994) to suggest that no postdepositional recrys tallisation or significant isotopic exchange has taken place. Results
The stable isotopic results for the eggshell analyses are shown in table 7 . 1 and figure 7.5. The samples with no evidence of recrystallisation have o13C values ranging from -3 .05%o to -1 l .25%o, with a mean value of-7.48%o ( n = 14). The oxygen isotopic results, however, do not show as even a spread as the carbon isotopic data; the 0 18 0 values range from
fl!l
P. W. DJTCHFIELD
Table 7. 1 .
Analysis of mineralogical samples collected from the Emirate of Abu Dhabi
Sample
Type
Mineralogy
818 0
Total PDB
8 13C
PDB
oxide
Fe (ppm)
Mg (ppm)
Mn (ppm)
Sr (ppm)
Baynunah Formation AB 2 5/ 1
Fenestral
Calcite
-4. 3 3
-1 .64
54.54
178
2 614
334
380
Calcite
-4 . 0 1
-4 .45
5 5 .00
224
1 754
374
411
Calcite
-5.29
-4.68
54.44
3 02
2 314
544
2 353
Calcite
-3.99
-2 . 1 0
56.01
256
1 642
294
507
micrite AB 3 1/ 1
Fenestral micrite
AB 94.8
Fenestral micrite
AB 1 5
Fenestral micrite
AB 29
Concretion
Dolomite
-3.45
-2 . 8 7
50.85
3 997
1 1 6 006
3 431
205
A B 94. 1
Calcite-
Calcite
-1.51
-4.20
54.21
120
356
1 00
452
Calcite
-2 . 0 3
- 1 0 .40
5 3 .68
80
520
98
201
Calcite
-6 . 0 9
-3 . 54
52.27
1 07
427
50
328
Calcite
-3.47
- 3 . 66
5 3 .67
90
395
64
5 54
Calcite
-9.44
-1.66
54.88
1 46
561
66
352
420
cemented sandstone i\B 94.2
Calcitecemented sandstone
AB 1 4
Calcitecemented sandstone
AB 1 2
Calcitecemented sandstone
AUH 3 3 5/2
Calcitecemented sandstone
AUH 3 3 5 / 1
B ivalve
Calcite
-7.92
-2 .04
54.65
125
567
112
AB X 1
Bivalve
Calcite
-4.47
-4.25
54.69
1 34
2 499
515
739
AB X2
Eggshell
Calcite
-3 .70
- 1 0 .8 0
54.85
39
505
55
416
AB X3/1
Eggshell
Calcite
8 .2 8
-3.33
5 3 .47
100
657
36
571
A B X3/2
Eggshell
Calcite
8 . 04
-3 .05
53.10
150
453
48
631
AB X4/l
Eggshell
Calcite
7.82
-3.40
AUH 639 M3
Eggshell
Calcite
1 . 09
-11 .25
AUH 667 R3
Eggshell
Calcite
-0 .08
-9 .30
AUH 472 B2
Eggshell
Calcite
2.81
-7.76
AUH 3 1 5 K1
Eggshell
Calcite
2.13
- 1 0 .82 -9 .83
AGH 4 8 7 S4
Eggshel l
Calcite
-0 . 19
AGH 6 0 5 JD3
Eggshell
Calcite
2 . 99
-9.34
AUH 5 52 S6
Eggshell
Calcite
6.91
-7. 3 7
AUH 6 2 1 Ql
Eggshell
Calcite
1 . 64
-8 .43
AUH 446 H6
Eggshell
Calcite
2 .3 3
-5 .73
AUH 1 5 H 3
Eggshell
Calcite
-4 .98
3.15
AUH 6 1 2 Gl:2
Eggshell
Calcite
7.28
-6 .62
AUH 320 T H l
Eggshell
Calcite
8 .5 8
-8 .40
DIAGENESIS OF SEDIME:-ITS IN THE WESTERN REGION
Table 7.1.
( continued)
Sample
Type
Shuwaihat Formation AB Carbonate crust AB 7/2 Carbonate crust AB 8/1 Carbonate crust Carbonate AB 1 0/2 crust AB 1 0/2 Carbonate crust
Total oxide
Fe (ppm)
M g (ppm)
-2.04
49.32
2 401
1 16 006
112
120
-2.71
-2.27
50. 1 0
2 215
1 1 8 402
98
1 02
Dolomite
-5.54
-3.19
48.79
2 362
1 1 5 836
64
118
Dolomite
-5 .04
-3. 8 0
49.22
2 734
1 14 8 9 1
88
126
Dolomite
-4.82
-2.46
50.96
2 540
1 1 5 308
1 02
1 04
Dolomite Dolomite Dolomite Dolomite Dolomite
4.02 4.52 3 .9 1 5 .07 5 .63
- 1 .64 - 1 .98 -2.54 -1 .08 1 .45
5 1 .52 52.2 1 5 0 .92 5 0 . 74 5 1 .63
l 202
131 1 26 1 30 1 27 128
1 64 120 1 19 1 84 1 42
1 04 227 123 206 1 54
Mineralogy
8180 PDB
Dolomite
-3.5 1
Dolomite
813C PDB
Dam Formation AB AB AB AB AB
94. 1 2/ 1 94. 1 2/2 94 . 1 3/ 1 94. 13/2 94. 1 3/3
fJ]
Micrite Micrite Micrite Micrite Micrite
1 1 1 1
474 186 394 335
392 552 293 409 431
Mn (ppm)
Sr (ppm)
Note: The AUH reters to a fossil and the suffix follmving the number refers to the collection locality: B, Jebel Barakah; S, Shuwaihat; GU, Al Ghuddah; H, Hamra; JD, Jebel Dhanna; Q, Ras al Qa'la; R, Ras Dubay'ah; K, Kihal; TH, Thumayriyah; M, Jebel Mimiyah (Al Mirfa ) . The prefix AB refers to rock samples collected for this analysis.
Pi
P. w. DITCHl'lELD
5000
10 �-------,
4000
E Q. Q.
Lt
3000
2000
Ill c Q. 0
"
c 1!1
CD �
.. .
1000
1 1 0000
120000
Mg ppm
• •
1 30000
�
ple, Cande and Kent, 1 99 5 ) . Numerical ages may be
- --
..c:
.!? � .c ,$:! 22 .c Q.
16
observed within the sequence.
?
- _____ _: ___
E u.. . '"'
1.8
Lat (0N)
62.9 68.0 80.0 78.8 75.7 86.3 81 83 79.3 74.4 73.8
Long (0E)
F ormati on
231.2 242.8 247.9 248.8 222.0 199.5 297 210 309.2 285.5 241.3
Usfan basalts: Usfan basalts: Usfan basalts' Sarrat trachnes2 Sarrat trachnes2 Usfan basalts1 Jebel Khariz \·olcaci-.:s' Aden volcanics3 Madina Road basalt S:.��'�: �\�·lbi��� Madina Road basalt. S:.�:,�� -�:.=':::. Jordan basalts"
References: 1Yousif and Beckmann, 1981; 2Kellog and Reynolds, 1980; 3lrving and Tarling, 1961: Bakor, 1989; 5Sallomy and Krs, 1980. and the apw curve drawn through them describes a clockwise loop between about 40 and 2 million years (Ma) ago. The apw path for Mrica shows a similar broad clockwise loop for this period, but dis placed about 10° to the east. Reconstruction of the former positions of the Arabian and Nubian plates, by clockwise rotation of Arabia through an angle of 9° about a rotation pole at 29.3° N, 27.1o E to close the Red Sea (Hall, 1979), moves the Arabian and Mrican apw paths into closer conformity. A further shift of the Arabian Plate by 120 km in a SSE direc tion, parallel to the Red Sea axial trough, closes the Gulf of Aden and improves the fit of the two apw curves (Hussain and Balcor, 1989 ) . The apparent inconsistency of the Aden volcanics pole with other palaeomagnetic data for Arabia may reflect the fact that this pole is based on an early study by Irving and Tarling ( 1961 ) before modern methods for palaeomagnetic data analysis were available. Alterna tively, it could reflect the effects of local undetected tectonic rotations in the Aden area, or inaccuracies in the age of these volcanics (an age of less than 2 Ma would conform better with other palaeomag netic data for Arabia). This overall consistency between the Arabian and Mrican apw curves confirms the validity of the
.
�E�::::c::·
c, __ ,_
clockwise loop observed in both cun·c-s :·.: :· :;�� interval 40-2 Ma ago. Whether thi s loo�' ··�:��::: actual motion of the Mro-Arabian Plate ,-_.:·::�z this time, or includes an element of "e·:.: � �- _ :.::· wander (that is, departure of the gcom::.g:��::: dipole axis from the geographic axis t{y :. �:·:�: period), is uncertain. The existence of tic:, ::::c:·:c teristic feature in the Arabian apw curn:. :: �---ec·. greatly increases the potential value of tlc:' for palaeomagnetic dating applications. ··
,._._
c �--
. '
Palaeomagnetic Dating of the Shuwaihat and Baynunah Formations Constraints can be placed on the ages or ::-:e Shuwaihat and Baynunah Formations of _-\j�: Dhabi by comparing the positions of the palaeo magnetic poles derived from these formations \\·ith the age-calibrated apw curve for .-\rabia tig. 8.:The palaeomagnetic pole position for the Ba\· nunah Formation, based on the me an direction of ChRM in the Hamra 6 section, lies just south of poles 7 and 9 (which have ages of about 10 :\ia and 4 Ma, respectively) and north of pole 10 (with an age of about 3 Ma). Pole 7, from the Jebel I apertural view; B> adapcrtural view. Height 8.2 mm, diameter 4.2 mm.
24)
has warned with
respect to the inclusion of various bulirninid shells from Mghanistan in the genus
Subzebrinus:
"Inclu
sion . . . is a matter of convenience. Until they can be dissected, no meaningful generic reference is possible." If this is tl1e situation with recent shells, clearly one must be considerably more circumspect with a Miocene fossil cast. It is possible to state with some confidence that it does not belong to
B Figure 11.2. Drawing of fossil internal cast shown in figure 1 1 . 1 : A> apertural view; B> adapertural view. Scale bar 1 mm. =
bian Gulf area, but only one of a similar appear ance,
Imparietula jousseaumei (Smith),
occurs
locally, apparently being restricted to northern Oman (Mordan, 1986). Similarly, only one of the subulinid species,
Zootecus insularis (Ehrenberg),
is
close in shape to the fossil, and is a widely distrib uted coastal species found around the Arabian Gulf
Figure 11.3. Drawings of internal casts ofHolocene shells. A> ImpariettJ-la jousseaumei from Wadi Qatam, Saiq, Oman. Collected by M. D. Gallagher, 17.x.1984. B> Zootecus insularis from Bahrain. Col lected by S. Green, 1994. Scale bar = 1 mm.
--
-� --
---
-- -
--
· ----
-
-- -
-
-
-
om P. B. MORDA0i it is found at lower elevations only in synanthropic situations ( Mordan, 1 9 8 6 ) . It does not occur in the United Arab Emirates today. These snails spend the dry season in aestivation attached to rocks, trees, and shrubs well above ground level, where they can avoid the hottest conditions , and only become active during wet periods when feeding and repro duction take place. The presence of a pseudona paeine during the late Miocene suggests wetter, perhaps seasonal, conditions at that time . This notion compares well with other evidence of habi tats near to the fluvial system of the Baynunah For mation ( Friend, 1 9 9 9 ; Jeffery, 1 999-Chapters 5 and 1 0) . The Pseudonapaeinae, as a subfamily of the Buliminidae, are a group with Palaearctic and Ori ental affin ities. This is in contrast to the Cerastidae, a "southern" family that occurs in Africa, Madagas Figure 1 1 .4 . Dravv:ing of internal cast of Holocene Subzebrinus oxianus from Isfahan, Iran. Collected by E. S. Brown , 2 2 .xi . 1 9 5 8 . Scale bar = 1 mm.
car, the Seychelles, India, and Australia, and was previously included in the Buliminidae but is now thought to be distinct (Mordan, 1 984, 1 992 ) . The Holocene geographical distributions of these two families hardly overlap and there is a sharp line of
species or even genera currently occurring in the
demarcation in Oman, with the Buliminidae occur
immediate vicinity of the Emirate of Abu Dhabi. It
ring in the northeast and the Cerastidae occurring
almost certainly belongs to the family Buliminidae
in the mountains of Dhofar in the southwest. The
Kobelt, 1 880, but the genus remains uncertain;
presence in the Emirate of Abu Dhabi during the
Subzebrinus Westerlund,
Miocene of a putative species of Pseudonapaeinae
1 88 7 and
Pseudonapaeus
Westerlund, 1 8 87, both of which belong to the
does not conflict with present-day distribution pat
subfamily Pseudonapaeinae Schileyko, 1 978 (which,
terns at the family level.
incidentally, also includes
Imparietula Lindholm,
1 92 5 ) , appear to be the most likely genera.
REFERENCES BIOGEOGRAPHY AND ECOLOGY
Biggs, H. E. J. 1937. Mollusca of the Iranian plateau.
Journal of Conchology 20:
342-50.
The genera within the Pscudonapaeinae, and cer tainly those three mentioned above, are all charac
--- .
teristic of rather xeric habitats , but tend to occur in
Iraq.
1959. Some land Mollusca from northern
Journal of Conchology 24:
342--47.
situations where there is regular seasonal humidity, either in the form of precipitation at rather high
--- .
elevations, or lower down close to water at the base
Journal of Conchology 2 5 :
of wadis.
Imparietula fousseaumei is
1 962 . Mollusca of the Iranian plateau-H. 64-72 .
a case in point,
being restricted to sites above about 5 00 metres in
--- . 1 9 7 1 . Mollusca of the Iranian plateau-III.
the northern mountains of the Sultanate of Oman;
Journal of Conchology 27:
---- -----
2 1 1-2 0 .
pp. 1 2 +- 3 3
Ftiend, P. F. 1999 . Rivers of the Lower B aynunah
Studies on Non-marine Mollusca)
Formation, Emirate of Abu Dhabi, United Arab
Solem and A. C. van Bruggen ) . Btill. Leider .
Emirates. Chap. 5 in pp. 38-49 ( ed. P.
Fossil Vertebrates ofArabia) J. Whybrow and A. Hill ) . Yale
cd. _-\.
--- . 1986. A taxonomic revision of southern _-\ra
University Press, New Haven.
bian Enidae
Jeffery, P. A. 1999 . Late Miocene swan mussels from
ogy 50: 207-7 1 .
sensu lato) (Mollusca: Pulmonata . . Bul letin of the British Museum (Natural HistmJ)) Zool
the Baynunal1 Formation, Emirate of Abu Dhabi, United Arab Emirates. Chap . 10 in
ofArabia)
Fossil Vertebrates
pp. 1 1 1-1 5 (ed. P. J. Whybrow and A.
Hill) . Yale University Press, New Haven.
--- . 1992. The morphology and phylogeny of the Cerastinae (Pulmonata: Pupilloidea).
the British Museum (Natural History))
Bulletin of
Zoology 5 8 :
1-2 0 . Mordan, P. B . 1 980a. Land Mollusca of Dhofar. In
Journal of Oman Studies. Special Report no. 2: The Scientific Results of the Oman Flora and Fauna Survey) 1 97'7y pp. 1 0 3-1 1 . Ministry of lnformation and Cul
Shileyko, A. A. 1 9 8 4 . Mollusca: Terrestrial molluscs of
ture, Sultanate of Oman.
1 30 : 1 -399.
---
luscs.
. 1980b. Molluscs of Saudi Arabia-land mol
Fauna of Saudi Arabia 2:
---
3 59-67.
. 1984. Taxonomy and biogeography of the
sensu lato) (Pulmonata: World-wide Snails: Biogeographical
the suborder Pupillina of the famu of the USSR (Gastropoda, Pulmonata, Geophila) .
Fauna SSSR)
n.s.
Solem, A. 1979. Some mollusks from Mghanistan.
Fieldiana) Zoology no.
1 3 0 1 : 1-89.
Verdcourt, B. 1 9 6 3 . The non-marine Mollusca of
southern Arabian Enidae
Rusinga Island, Lalze Victoria and other localities in
Pupillacea). In
Kenya.
Palaeontographica
1 2 1A: l-37.
nss n . rr
nr
7
if
r::sr n
7
Late Miocene Fishes of the Emirate of Abu Dhabi, United Arab Emirates PETER L. F OREY AND SALLY V. T. YOUNG
Late Miocene fishes were reported from Abu Dhabi by Whybrow et al. ( 1990) and Whybrow ( 1 989). These reports were based on several collecting trips and we use these results, together with new collec tions made by The Natural History Museum/Yale University team from 1989, to present an up-to date summary of the Miocene ichthyofauna of Abu Dhabi and its significance . Specifically we address three areas of interest: ( 1 ) local palaeoecology of the fish-bearing sediments; ( 2) the significance of the finds in understanding the development of the Mrican freshwater fish fauna; and ( 3 ) the part, if any, that these Miocene fishes play in explaining the current fish fauna of the Arabian Peninsula.
MATERIAL AND PRESERVATION The fish remains were recovered from various local ities within the fluvial part of the Baynunah Forma tion (Whybrow, 1989; Whybrow et al. , 1 999Chapter 4), dated from the contained mammal fauna as Turolian (European Mammal Stage MN 1 3-14 ) . Deposition thus took place about 7-6 mil lion years ( Ma) ago-see Parts I and II of this vol ume (Whybrow and Hill, 1 999 ) . The geology of the sites is described in Chapters 4 and 5 (Why brow et al. , 1 999; Friend, 1 999 ) , and the localities are listed in Chapter 23 (Whybrow and Clements, 1999 ) . The fish remains are all fragmentary and are found on scree slopes beneath weathered sand stones and conglomerates within the Baynunah Formation. It is impossible, therefore, to relate fish occurrences to beds. At each of the sites the fish collections contain weathered, rounded bones as
Copyright ©
well as sharply fractured bones. The rounded bones are usually white whereas the sharply angled bones are pale brown. The rounded white bones are surely the result of recent erosion and weathering. Together, the size ranges and shapes of the bones are very variable and this implies that the original depositional environment was one of high energy, quick burial, and little water sorting. The fishes found in the Baynunah Formation are here referred to three taxa: two catfishes, Clar ias sp . and a new species of Bagrus) and a cyprinid, Barbus sp . We have decided to begin this report with general statements about the fishes: ecological preferences of modern relatives; biogeographic sig nificance of the Baynunah fishes; and some state ments about the modern fish fauna of the Arabian Peninsula. This is followed by systematic descrip tions.
MODERN D ISTRIBUTIONS AND EcoLOGY The modern distribution of Clarias is coextensive with that of the family Clariidae ( 1 00 species ) , which extends throughout Mrica, Syria, peninsular India, Southeast Asia, and the Philippines. The family is very rare in northern and Saharan Mrica. Of the 43 species, the majority ( 3 2 ) are found in Mrica, one in Syria, four in India, and seven in Asia ( C. batrachus is widespread in both India and Southeast Asia) . Where ecological data are available then most species prefer quiet, shallow waters; many prefer swamp conditions (Talwar and Jhin gran, 199 1 ) and can withstand intermittent periods of drought.
1999 by Yale University. All rights reserved. ISBN 0·300·07 1 8 3 · 3
rr · a
The Bagridae, as traditionally recognised, are a large group of primitive catfishes with a fossil record back to the Thanetian (White, 1926). This may be a misleading statement, however, because the bagrids may not be a monophyletic group . Mo 1 99 1 ) recently revised their systematics and split the family into three monophyletic families : a re stricted Bagridae, the Clareotidae, and the Aus troglanididae. The restricted Bagridae contains Bagrus) the only Mrican genus, with 1 0 species, and 1 5 Asiatic genera ( 1 25 species ) . Mo ( 1 99 1 ) considers that the Asiatic Aorichthys (with two species distributed in India and Burma) is the sis ter group of Bagrus. Species of Bagrus are widely distributed in a variety of freshwater habitats but most are bottom dwellers in relatively slow- moving waters . Barbus is a widespread genus with species in Africa, Asia, and Europe. Members inhabit a very wide variety of habitats and although individual species do show habitat preferences, no meaningful generalisations may be made about the ecology of the genus.
BIOGEOGRAPHIC IMPLICATIONS FROM THE FOSSIL RECORD The geographic position of Abu Dhabi at or near a possible migration route between Mrica and Asia is clearly relevant to the theories of historical bio geography that use dispersal (migration) as the causal explanation. Much has been written about the mammalian faunal exchange between Mrica and Asia during the Miocene ( for example, Bernor et a!. , 1987; Steininger et a!. , 1 9 8 5 ; Rogl and Steininger, 1984). The dating and direction of pos tulated migration depend in large part on the fossil occurrences, palaeoecological interpretations of tl1e putative corridors, and precise dating, which, in turn, may be correlated with theories of earth his tory derived from lithological comparisons, palaeo magnetism, and absolute dating. In the absence of the last categories of evi dence or where there is considerable dispute be tween models of earth history-see, for example,
earth histories proposed by Smith et al . ' 1 9 9 1 1 and Owen ( 1 983 )-palaeontological eYidence has to stand on its own. Of course there is some degree of reciprocal illumination in dating geological n·ents from fossils and then postulating dates of migra tions . In the case of the proposed connection be tween Asia and Mro-Arabia, however, there is con siderable and varied palaeontological evidence drawn t1·om continental sediments containing mam mals (Bernor et a!. , 1987) and intercalated marine sediments with foraminifera (Adams et a!. , 1 9 8 3 ) and echinoids (Ali, 1 9 8 3 ) . Opinions a s t o whether there was a single event leading to the contact between Asia and Mro-Arabia (Adams et a!. , 1 9 8 3 ) o r two events separated b y a marine interval ( Rogl and Steininger, 1 984; Thomas, 1 9 8 5 ) would seem peripheral to theories of the distribution of the fishes considered here. What is of more interest is the timing of the first occurrences of the fishes in Africa and Asia and what relevance this might have to the direction of migration. Menon ( 1 9 5 0 ) , when discussing the biogeo graphic history of clariid catfishes, adopted a strictly palaeontological centre-of-origin approach . At tlut time the earliest fossils of clariid fishes ( Clarias and Heterobranchus) were known from the middle Pliocene of the Siwalik Hills ( Lydekker, 1 8 8 6 ) , fol lowed by middle Pliocene occurrences in Egypt and Pleistocene occurrences in Java. He concluded, therefore, that Clarias originated in the Siwalik Hills and migrated west to North Africa with a sub sequent range extension to central, western, and southern Mrica during the Pleistocene. Menon's paper was written within the paradigm of fixed con tinents . He explained the absence of Clarias ( and Heterobranchus) from tl1e present Arabian Peninsula by suggesting that this part of the world had sub merged after the dispersal event from Asia to .-\frica sometime in the middle Pliocene. A similar scenario has been suggested for Barbus (Menon, 1 9 6-±: Gan:L 1982), which supposedly migrated from "\sia to Mrica as late as Plio-Pleistocene times . The severest test of Menon's methodologY came with subsequent finds of Clarias in the _\lio cene of the Arabian Peninsula and �orth _\ti."ica ( see table 1 2 . 1 ) . Accepting the fossil record at t�Ke
Table 1 2. 1 .
Neogene localities yielding
A
B
c
D
Clarias ( e ), Bagrus,
E
F
Holocene
Pleistocene
•
•
•
•
•
•
• •
• •
G
H
( + ), and
Barbus (•)
I
K
•
•
•
•
•
•
L
M
Middle Miocene Lower Miocene
•
p
Q
R
s
T
u
•
•
•
•
•
• •
•
0
•
• Upper Miocene
N
•
•
Pliocene
J
• •
•
•
•
•
•
•
•
•
•
•
•
•
A, Lake Eyasi, Tanzania, Upper Pleistocene ( Greenwood, 1 9 5 7 ) ; B , Olduvai, Tanzania, Lower Pleistocene (Greenwood and Todd, 1 970 ) ; C, Lake Omo, Kenya, Lower Pleistocene (Arambourg, 1 947); D, Lake Baringo, Kenya ( Bishop et al . , 1971 ) ; E, Lake Turkana ( Rudolf), Kenya, Upper Pleistocene (Thomson, 1966); F, Chad, Sahara, Upper Pleistocene to Lower Holocene (Joleaud, 1 9 3 5 ; Daget, 1 9 5 8 , 1 9 6 1 ) ; G, Lake Albert, Zaire, Upper Pliocene ( Greenwood, 1959); H, Lake Albert, Zaire, Lower Pleistocene ( Greenwood and Howes, 1 975 ) ; I, Lake Albert, Uganda, Upper Pliocene (White, 1926); J, Nubian Sudan, Pleistocene or Holocene ( Greenwood, 1968); K, Fayum, Egypt, Holocene (Stromer, 1904 ) ; L, Wadi Natrun, Egypt, Middle Pliocene ( Greenwood, 1 972 ) ; M, Chalouf� Egypt, Upper Miocene (Priem, 1 9 14 ) ; N, Sahabi, Libya, Lower Pliocene ( Gaudant, 1987); 0, Bled ed Dourah, Tunisia, Upper Miocene ( Greenwood, 1973); P, Huqf, S ultanate of Oman, Lower Miocene ( Roger et al. , 1 994); Q, Ad Dabtiyah, Eastern Province, Kingdom of Saudi Arabia, Lower Miocene ( Greenwood, 1987); R, As Sarrar, Eastern Province, Kingdom of Saudi Arabia, Lower Miocene (Thomas et al. , 1982); S , Emirate of Abu Dhabi, Upper Miocene (Forey and Young, 1998); T, Siwalik Hills, Middle Pliocene ( Lydekker, 1 886); U, Java, ?Pleis tocene ( Koumans, 1 949 ) .
FISHES FROM THE BAYNC.\IAH FoRMATION
value would place the centre of origin on the Arabian Plate/North i\£rica with migration to Asia. Discoveries of Barbtts in the late Miocene of Tunisia ( Greenwood, 1 97 3 ) and Abu Dhabi also might upset Menon's hypothesis for the history of Barbus. Table 1 2 . 1 is a plot of the Neogene fresh water fish-bearing localities containing Clarias, Bagrus, and Barbt>ts of Africa-Asia. It is an attempt to update that pmvided by Greenwood ( 1 974). Table 12 . 1 demonstrates j us t how incomplete, varied, and sparse is the sampling of Neogene fish history. Absence of a particular fossil takes on con siderable relevance in the "fossil record and centre of-origin approach". But absence is a difficult observation to square with reality. For instance, there are just three fishes currently known from the Baynunah Formation. Considering the relative rich ness of the remainder of the fauna, as described in this volume (v\lhybrow and Hill, 1 99 8 ) , the paucity of the fishes may reflect collecting effort ( unlikely) , taphonomic, or palaeoecological bias, or it may be a true reflection of tl1e faunal content. More particularly, the determination of the centre of origin using fossils is susceptible to new fossil discoveries, difficulties of dating, or the redat ing of already known localities. There may be in stances where the fossil record of a particular group of organisms can be assumed to be complete enough to tell a "true" story (for example, forami nifera-Adams et al. , 1 9 8 3 ) . The freshwater fish record of the Mrican/Asian �eogene is, however, so patchy ( and we do not know hmv incomplete it may be), that it would be umvise to base theories of origin and migration with any degree of confi dence. The fossil record does show that by late Miocene times Clarias was a widespread genus, the range of which extended from Tunisia in the west to the Arabian Peninsula in the east and to the Siwalik Hills on the Asian Plate in the middle Pliocene. Clarias was also present in Africa in the early Miocene. The genus Bagrus is limited to Mrica, where it is nearly always associated with Clafr·ias (table 1 2 . 1 ) This might suggest that the biogeographic histories of Clarias and Bagrus were congruent. .
IBl
Putting the stratigraphical distribution to one side then, in trying to account for the biogeo graphic history of the three fishes witl1in the Abu Dhabi fauna we meet three explanations based on three separate methodological premises. The distri b utional history of Clarias is explained ( Menon, 1 9 5 0 ) by assuming that because the greatest species diversity of tl1e genus is in Mrica this is the furthest place from the centre of origin ( Asia). For Barbus the scenario is the reverse, witl1 Europe/Asia being the assumed centre of origin because it has tl1e highest species diversity. Until we have species- level phylogenies of Barbus and Clarias it would appear futile to try and choose b etween these alternative historical explanations. For Ba,_rfru.;; accepting the phylogeny proposed by Mo ( 1 99 1 ) , the African Bagrus is the most derived member of a family that is othervvise distributed in Asia. Furthermore, the species of Bagrus described here shows some plesiomorphic similarities with the sister genus Aorichthys (see below). This suggests that there was single migration from Asia to Africa and might pro vide an example of Hennig's ( 1966) progression rule.
THE M ODERN FAUNA The modern freshwater fish fauna of the Arabian Peninsula is very depauperate compared with neigh bouring parts of the world. Banister and Clarke ( 1 977) record nine species for the entire penin sula. These represent eight cyprinids ( three species of Barbus, three species of Gara, two species of Cyp1·inion) and one cyprinodont, Aphanius dispar ( Ri.ippell) . The cyprinodont is a euryhaline species, tolerant of seawater, and has a ·widespread distribu tion from nortl1east i\£rica to northwest India. It is therefore of limited significance in understanding the biogeographic history of Arabia. No freshwater catfishes are present on the modern .Arabian Penin sula: this may well reflect the absence of slow-mov ing rivers . The modern fish fauna shows three different biogeographic affinities according to Banister and Clarke ( 1977 ) . Two species, now living in the west-
e
P. L. FoREY AND S. V. T. YOU:\G
ern part of the Arabian Peninsula, have affinities with relatives in the Horn of Mrica; one has a dis tribution spanning the Arabian Gulf; the remaining species (distributed along either side of the Arabian Peninsula) have affinities with species living in the Levant. It should be stressed that "affinities", as used by Banister and Clarke ( 1 977), does not mean cladistic affinities. Most of the resemblances that those authors show between species are phenetic similarities, which may or may not have phyloge netic significance. Even if we accept this phenetic resemblance as meaningful, the mixture of biogeographic relation ships suggests either that the modern distribution patterns arose at completely different times and were layered upon one another or, more likely, that the species were once widespread and that increas ing aridity has caused different but more restricted distributions . The absence of Clarias) the species of which are extremely hardy, might reinforce this idea. If the Arabian Peninsula were effectively "ster ilised" by harsh aridity then there may have been subsequent invasions from three separate directions: from the Ethiopian highlands, from the Levant, and from Asia. Because aridity was a relatively recent event ( Kassler, 1973) these invasions must have occurred within historic times. The idea of widespread extinction followed by reinvasion is also suggested for the North Mrican fish fauna. Here, the modern fish fauna consists mostly of southern European fishes such as tl1e cyprinids Phoxinellus and Barbus capita Pfeffer as well as the cyprinodont Aphanius fasciatus (Valenciennes) , all of which may well have migrated via the Iberian Peninsula. This complement is totally unlike the fishes from Miocene and Pliocene times. The exception to this scenario are the fishes of the Nile that appear to have had a relatively stable history throughout the Neogene. Parenthetically, it is worth noting that the mod ern and diverse herpetofauna of the Arabian Penin sula, which consists of some 1 3 5 species (Arnold, 1987), shows some interesting parallels with tl1e fish fauna. Some species of amphibians and reptiles of southwest Arabia have their closest relatives in the Horn of Mrica. According to Arnold the detailed phylogeny of semaphore geckos ( Pristurus) suggests
multiple movements between these regions. Some other species of amphibians and reptiles living in the northern Oman Highlands have conspecifics living in Iran across the Arabian Gulf; while others, living in the southwestern highlands, are most closely related to species in the eastern Mediterranean, Iraq, Iran, and even Pakistan and India. Arnold ( 1 987) suggests that the last category ofherpetofauna is absent from the central part of the Arabian Peninsula because of ecological conditions. All three of these patterns par allel those among the modern fishes. The reptile fauna has an additional component: the arid-adapted species that inhabit the central deserts of S audi Ara bia and have relatives in North Africa. The relatively moist Nile valley may provide an effective ecological barrier to these species .
SYSTEMATICS OF THE BAYNUNAH FISHES Superorder Ostariophysi Series Otophysi Order Siluriformes Family Clariidae Genus Clarias Scopoli
Material1 (figs 1 2. 1 - 1 2.5) Shuwaihat. Site S l , N 24° 06' 38.1", E 52° 26' 09.6": AUH 1 14j, cleithrum; AUH 1 1 7b, dermcthmoid; AUH 1 12, dermcth moid plus lateral ethmoids; AUH l l 4c, 1 17d, 280g, frontals; AUH l l 7c, 1 1 7m, 753, lateral ethmoids; AUH 1 1 7f, nasal; AUH 1 14d, l l7c, 280e, 752, supraoccipitals; AUH 1 17, cen trum; AUH l l 7j, pectoral fin spine. Site S2, N 24° 06' 4 1 .7", E 52° 26' 04.0": AUH 125d, dermethmoid; AUH l 25e, lat eral ethmoids. Site S4, N 24° 06' 44.7", E 52° 26' 1 2.7": AUH 732, cleithrum; AUH 728, dermethmoids; AUH 73 1 , fi·ontal; AUH 726, lateral ethmoid; AUH 730, post-temporal; AUH 729, quadrate; AUH 727, supraoccipitals; AUH 725 , centrum; AUH 733, coracoid. Site S6, N 24° 07' 06.6", E 52° 26' 32.7": AUH 493, frontal; AUH 493a, 6 1 7a, lateral ethmoids; AUH 493b, sphenotic; AUH 6 1 7, supraoccipital. Jebel Dhanna. Site JD3, N 24° 10' 3 1 .1", E 52o 34' 21 .0": AUH 1 34b, l 34c, 1 34e, 606a, cleithra; AUH 1 34e, frontal; AUH l 34c, 274b, lateral cthmoids; AUH 655, skull roof bones; AUH 1 34t; pterotic; AUH 134a, 274c, sphenotics; AUH 1 34d, 274d, snpraoccipitals. Site JD5, N 24° 10' 22.9", E 52° 34' 38.5": AUH l40i, 747, cleithra; AUH 745, denta1y; AUH 749, dermetlunoid; AUH 659b, dermopterotics; AUH 659c, 748, lateral ethmoids; AUH l40b, lateral ethmoid and pterotic; AUH 75 1 , post-temporal; AUH 750, sphenotic; AUH l40t; 140g, 659d, supraoccipitals; AUH 744, centrum; AUH 746, pectoral fin spine. Hamra. Site H3, N 23° 04' 28.7", E 52° 3 1' 37.3": AUH 276t; dermethmoid.
------- ��
FISHES fROtl
KihaL Site Kl, N 24° 07' 2 3.2", E 53° 00' 27.9": AUH 309,
frontal and lateral ethmoid; AUH 309a, supraoccipital. Jebel Barakah. Site B2, N 24° 00' 1 3.6", E 52° 19' 3 5 . 5": BMNH P62047, dermopterotic; AUH 469a, BMNH P62048, P62049, frontals; AUH 469c, post-temporal; AUH 469b, BMNH P62046, P62050, supraoccipitals; BMNH P6205 l , pectoral fin spine. Jebel Mimiyah (Al Mirfa). Site Ml, N 24° 04' 58 .2", E 53° 26' 07.6": AUH 632, cleithrum; AUH 668, sphenotic. Ras Dubay'al1. Site R2 (no co-ordinates available): AUH 645, bro ken skull bones; AUH 645b, cleithra; AUH 645c, dermeth moid; AUH 646a, R2, dermethmoid and lateral ethmoid; AUH 646b, lateral ethmoid; AUH 646, vomer.
The collections contain many isolated bones of this large catfish. The commonest recognisable ele ments are the supraoccipital, sphenotic, dermopte rotic, lateral ethmoid, dermethmoid, post-temporal, and orbitosphenoid amongst the head bones in addition to pectoral spines and cleithra. Some of these are illustrated in figures 1 2 . 1-1 2 . 5 , alongside a skull of the Holocene Clarias anguillaris ( Lin naeus) . The head bones are thick, coarsely orna mented with pronounced tubercles that rarely run together and show only faint regular patterning rel ative to ossification centres. The skull bones suggest that these remains belong to a species of the genus Clarias rather than the closely related Heterobranchus for the following reasons: the anterior suture between the supra occipital and fi-ontals is simple rather than complex and digitate; and the posterior end of the dermeth moid is a simple posteriorly directed V-shape rather than W-shaped. It is impossible to identifY these Clarias re mains with any living species because so few parts have been found. Equally, it would be unwise to erect a new species on such disarticulated pieces. But it is possible to add some comparative com ments. The Mrican Holocene species of Clarias have been revised by Tuegels ( 1986). He did not establish a phylogeny but did recognise several groups of species (his subgenera) based on the combination of shapes of the frontal and occipital foramina and the pattern of denticulation on the pectoral spine. In the Abu Dhabi Clarias the frontal fontanelle is long and narrow ( for example, AUH l l 2 ), the occipital foramen is located ante rior to the supraocciptal process and therefore lies in front of the posterior margin of the skull, and the serrations on the pectoral spine are restricted to
THE
K\Y:.:c:.:.\H fO�\L\TIO�
1m
the outer edge. This combination of teatures is seen in C. anguillaris, C. ngamensis Castelnau, and C. lamottei Daget and Planquette ( Roberts, 1 9 8 9 , questions the taxonomic status o f the last-men tioned species) . Clarias ngamensis is restricted to southern Mrica but C. anguillaris is one of the more widely distributed species, being found t1·om Senegal, along the southern edges of the Sahara to Ethiopia, and along the Nile to the Mediterranean . This corresponds to the Nilo-Sudan ichthyofaunal province of Roberts ( 1 975 ) . Other near-contempo raneous occurrences of Clarias in Mrica are in the late Miocene deposits of Bled ed-Douarah, Tunisia ( Greenwood, 1973), late Miocene of Chalouf, Egypt (Priem, 1 9 1 4 ) , and early Pliocene of Sahabi, Libya ( see table 1 2 . 1 ) . All these have been identi fied as Clarias sp . and they are closely comparable with the Clarias remains from Abu Dhabi. Mod ern species names have been applied only to fossil Clarias from the Pleistocene of Mrica ( Greenwood, 1974; Thomson, 1966). Clarias is also known as fossils from the middle Pliocene of the Siwalik Hills ( Lydeldcer, 1 8 8 6 ) and from the Pleistocene of Java ( Koumans, 1 949 ) .
Superorder Ostariophysi Series Otophysi Order Siluriformes Family Bagridae ( sensu Mo, 1 99 1 ) Genus Bagrus Bleeker, 1 8 5 8 Bagrus shuwaiensis sp . nov. (figs 1 2 .6-1 2 . 1 6 )
Diagnosis A species of Bagrus in which the ornament on the supraocciptal, pterotic, and sphenotic consists of coarse, irregular honeycomb rugosities; the supra occipital spine is relatively broad ( transverse width at base is equal to 1 8-20% of the length) and equal to about 40% of the total length of the supraoccipitaL the lateral profile of the supraoccipital is flat and lies in continuity with the remainder of skull roof.
Holotype Emirate of Abu Dhabi AUH 678 ; rear half of a braincase from the posterior le\·el of the orbit to the occiput and showing the entire length of the
1 0 mm
20mm
1 0 mm
1 0 mm
Figures 12.1-12.5. 12.1, Skull of Holocene Clarias anguillaris, BMNH 1866.9.9.70. The letters refer to parts of the skull represented by fossil specimens illustrated in figures 12.2-12.4. 12.2-12.5: Clarias sp., Baynunah Formation; 12.2, dermethmoid, lateral ethmoids, and anterior portions of frontals, AUH 1 1 2 (portion A in fig. 12.1); 12.3, right lateral ethmoid, AUH 117m (portion B in fig. 12.1); 12.4, partial supraoccipital, AUH 140 (portion C in fig. 12.1); 12.5, left pectoral spine in dorsal view (note restriction of the denticles to the outer edge; cf. Bagrus fig. 12.6), AUH 117j.
Frontal frontal fontanelle foramen for
Y �
·
supraorbital sensory canal
Supraoccipital
Pterotic
Extrascapular
1 0 mm
'-----J
Parasphenoid
\
Exoccipital
Pterosphenoid
Prootic
Figures 12.6 and 12.7. Bagrus shuwaimsis sp. nov., braincase, holotype AUH 678: 12.6, dorsal 'iew; 12. -, wn tral view. Scale bar applies only to photographs.
••
B P. L.
FOREY AND S . v. T. YOUNG
posterior fontanelle but lacking the supraoccipital spine. Late Miocene, B aynunah Formation, west side of Shuwaihat, Abu Dhabi.
Etymology Named after a contraction of the name of find place of the holotype.
Material There are many isolated bones representing those of a bagrid catfish. Isolated supraoccipitals, sphe notics, and frontals can be direcdy compared with the holotype and thus referred to this species. All other isolated elements can be referred only on their association in the same deposits. There is, however, no reason to suspect that more dun one species is present. Besides the holotype there is one articulated piece of skull (AUH 1 0 8 ) , which con sists of part of the frontal, sphenotic, pterotic, and half of the supraoccipital of the left side of a skull considerably larger than the holotype. The many isolated bones are listed below and represent the following elements: supraoccipital, sphentoic, frontal, basioccipital, parasphenoid, anguloarticular, dentary, three fused occipital vertebrae , post-tem poral, cleithrum , second dorsal fin spine, pectoral fin spine, and centra . All come from various locali ties within the late Miocene Baynunah Formation of Abu Dhabi. Specimens Assigned to Bagrus shuwaiensis sp. nov. 1 (figs 1 2.6- 1 2 . 1 6) Shuwaihat. Site 5 1 , N 24° 06' 38. 1", E 52° 26' 09.6": AUH l l4h, basioccipital; AUH 1 08 , braincase; AUH 1 14a, 1 171, ll 7n, cleithra; AUH 33b, dentaries; AUH l l4g, 742, frontals; AUH l l7g, supraoccipital; AUH l l4b, l l4i, centra; AUH 1 14, fused occipital vertebrae; AUH 33, l l7h, 280f, second dorsal fin spines; AUH l l4e, median fin spine; AUH 280i, second median dorsal fin spines; AUH 33a, 1 14£; l l7i, l 1 7k, 280h, 4 16, 743, pectoral fin spines. Site 52, N 24° 06' 4 1 .7", E 52° 26' 04.0": AUH 678, braincase; AUH 125b, clcithra; AUH 1 2 5 , frontals; AUH l25c, median fin spine; AUH l 25a, pectoral fin spine. Site 54, N 24° 06' 44.7", E 52° 26' 1 2 . 7": AUH 723, clcithrum; AUH 7 1 9 , denta1ies; AUH 724, frontal; AUH 7 1 8 , hyomandibular; AUH 722, post-temporal; AUH 7 17, sphenotic; AUH 720, sphenotic and frontal; AUH 754, supraoccipital; AUH 721 , centra; AUH 7 16, second dorsal fin spine; AUH 7 1 5 , pectoral fin spines. Site 56, N 24° 07' 06.6", E 52° 6' 32.7": AUH 492c, anguloarticular; AUH 492, frontal; AUH 492d, post-temporal; AUH 488a, 489,
492b, 554a, median fin spines; AUH 488, 492a, 554, pec toral fin spines. Jebel Dhanna. Site JD3, N 24° 1 0' 3 1 .1 ", E 52° 34' 2 1 .0": AUH 1 34i, cleithra; AUH l 34j, dentary; AUH 606b, frontal; AUH 655a, 655b, supraoccipitals; AUH 606e, anterior fi.1sed centra; AUH 1 34, fused occipital vertebrae; AUH 1 34h, 606d, median fin spines; AUH 1 34g, 606c, pectoral fin spines. Site JDS , N 24° 1 0' 22 .9", E 52° 34' 38.5": AUH 140k, 737, anguloarticular; AUH 741 , basioccipitals; AUH l40d, 736, cleithra; AUH l40h, 140j, 659, dentaries; AUH 140, 659a, 735, fi.-ontals; AUH 140a, parasphenoid; AUH l40c, 740, sphenotic; AUH 738, supraoccipital; AUH l40e, 739, pec toral fin spines. Hamra. Site H3, N 23° 04' 28.7", E 52° 3 1' 37.3": AUH 276g, 276h, cleithra. Jebel Baralcah. Site 132, N 24° 00' 1 3.6", E 52° 1 9' 35 .5": AUH 471 , cleithrum; BMNH P6205 3 , dentary; BMNH P62052, supraoccipital; AUH 469, pectoral fin spine. Jebel Mimiyah (AI Mirfu). Site M 1 , N 24° 04' 58 .2", E 53o 26" 07.6": AUH 669, dorsal pterygiophore; AUH 634, pectoral fin spines. Harmiyah. Site Y2, N 24° 04' 38.4", E 53° 1 9' 29.0": AUH 347, median dorsal fu1 spine. Ras Dubay'all. Site R2 (no co-ordinates available): AUH 645a, pectoral fin spine.
Remarks In a revision of the bagrid genera Mo ( 1 99 1 ) iden tified a single synapomorphy for the genus BagruSy· namely, a dorsal fin consisting of eight soft fin rays . This cannot be checked in the material described here. Nevertheless there are several other features commonly found in Bagrus and the closely related Aorichthys. The skull is relatively flat and shallow beneath the orbit. The frontal fontanelle (fig. 1 2 .6 ) i s elongate and continues posteriorly a s a groove reaching close to the base of the supraoccipital spine. The groove forms a plane of weakness and many specimens, including the holotype, are bro ken along this line . The supraoccipital spine ends posteriorly without a notch, implying that the nu chal plate does not overlap the supraoccipital. The post-temporal fossa is relatively deep and the open ing is slit-like (fig. 1 2 . 9 ) . The frontal (fig. 1 2 .6) is narrow posteriorly where it is flanked by the sphe notic and it broadens above the orbit. The para sphenoid is poorly preserved in the holotype but, on the left side (fig. 1 2 . 8 ) , it appears to form part of the border of the trigeminal foramen ( this is a feature of bagrid genera other than Bagrus or Aorichthys) . The dentary ( fig. 1 2 . 1 1 ) is shallow throughout and lacks the pronounced coronoid
1 0 mm facet for Hyomandibula Supraoccipital
Sphenotic
Pterotic
1 0 mm
trigeminal foramen
Supraoccipital Pterotic
Exoccipital
Figures 12.8 and 12.9. Bagrus shmvaiensis sp. nov., braincase, holotype AUH 678: posterior view. Scale bars apply tO photographs.
post-temporal fossa 12.8)
left lateral view; 12.9)
1 0 mm
1 0 mm
Figures 12.10-12.16. Bagrus sht$Waimss i sp. nov.: 12.10> supraoccipital, AUH 655a; 12.11> anterior end of left dentary, lateral view, AUH l40j; 12.12> three fused occipital vertebrae in right lateral view, AUH 114; 12.13> anguJoarticular in right lateral view, AUH l 40k; 12. 14, right cleithrum, lateral view, AUH l l7n; 12.15> second dorsal fin spine, anterior view, AUH l l 7h; 12.16> left pectoral spine in dorsal view (note denticles along mesial e?ge; cf. Clarias> fig. 12.5), AUH l l7i.
fiSHES FROM THE BAYK1.!NAH fORlvl.ATION
1m
process present in most other bagrids. The post temporal is pierced by a foramen that receives the anterodorsal prong of the cleithrum (a feature also seen in Mystus)· Mo, 1 99 1 ) rather than a notch, which is more usual in bagrids. The proportions of the frontal relative to the
5
mm
sphenotic and pterotic are much closer to those of
Bagrus spp. than to Aorichthys spp . There is n o notch between the pterotic and the extrascapular, as
B
is seen in Aorichthys) and the post-temporal shows
. -
no inflated bulla, which in Aorichthys receives an anterior limb of the Svv-:imbladder. For these reasons we consider the Abu Dhabi bagrid to belong to the genus Bagrus. The proportions of the supraoccipital spine, however, and the spine's flat contour are unlike those in other species of Bagrus and more like those of Aorichthys spp . and other genera such as Mystus and Chrysichthys. This might suggest that
Bagrus shuwaiemis is a relatively primitive species of
Figure 1 2 . 1 7 . Barbtts. Above is a drawing of the right lower pharyngeal of Holocene Barbus bynni (Forsldl) . Below are some variants of the isolated teeth found in the Baynunah Formation .
the genus. There are several "large Barbuf' modern spe cies living today in East Mrica and along the Nile Superorder Ostariophysi Series Otophysi Order Cypriniformes Family Cyprinidae Genus Barbus
Material 1 (fig . 1 2. 1 7) Shmvaihat. Site Sl, N 24° 06' 38.1" E 52° 26' 09.6": AUH l l l , teeth. Site N 24" 06' 41 .7" E 26' 04.0'': AUH 123, teeth. Site S4, N 06' 44.7", E 26' 1 2 7": AUH 734, teeth,
The collection contains an assortment of iso lated pharyngeal teeth of Barbus. Some are mam
as far as the Nile Delta, in which the dentition con sists of markedly different teeth, including all the differently shaped teeth found in the Abu Dhabi collection. Barbus bymzi (Forskil ), B. intermedius Rii p pell, B. macrolepis Pfeffer, and B. mirabilis Pap penheim and Boulenger all show such dentitions and it is noticeable that where a large growth series has been studied ( for example, B. by;mi-see Banis ter, 1973: fig. 1 9 ) tl1e large teeth are worn into concave crowns, similar to some of the teeth in the Abu Dhabi collection. Both Greenwood ( 1 987) and Gayet ( 1982)
milliform, some molariform, and some elongate
have referred early Miocene pharyngeal teeth from
with recurved and contoured tips. Together they
Saudi Arabia to B. bymzi) a species that currently
represent the normal array of teeth expected
lives in the Nile drainage system. This is a possible
in a pharyngeal dentition of many of the large
identification for the Abu Dhabi teeth but, in our
Barbus species. The size of the teeth suggest that
opinion, all of the Miocene teeth agree vv-:ith teetl1 of
they belong to a species referable to the " large
all of the species mentioned above. As Banister
Barbus group" of Banister ( 1987), which are gen erally over 200 mm standard length and show par
( 1973 ) pointed out, there is considerable variation in tooth morphology within a species and during the
allel or converging striae on the scales. There is
life of an individuaL Tooth shapes may also be
no implication that these form a monophyletic
related to diet. Therefore species-level identification
group.
of individual Barbus teeth would seem Yen- difficult.
1m
P. L. FOREY AND
S. V. T. Yoc::-:rG
of Abu Dhabi; numbers prefixed by BMNH P
CONCLUSIONS 1.
are in the palaeontological collections of The
Only three kinds of fishes can be recognised in the late Miocene Abu Dhabi fauna; only one has been identified to species level.
2. Clarias and Bagrus are
genera whose modern
representatives prefer slow-moving waters .
3.
The nature of preservation of the fishes suggests that they may have been deposited at times of flooding.
4.
The occurrence of
Clarias in
Neogene deposits
of Mro-Arabia at numerous localities casts doubt on traditional theories that these catfishes migrated to Mrica from Asia in the Pliocene. The phylogeny of bagrid catfishes strongly sug gests that there was a dispersal from Asia to Mrica but the fossil record suggests that this took place before Mro-Arabia met Asia.
5.
Distributions of the modern fish fauna of the Arabian Peninsula may owe more to reinvasion after "sterilisation" than to continuity of histori cal events reaching back to the Miocene.
� opportunity to travel to the Emirate of
Hill for th
Abu Dhabi to participate in the First International Conference on the Fossil Vertebrates of Arabia,
and to examine the B aynunah Forma
tion at first hand. He also thanks the Abu Dhabi Company for Onshore Oil Operations
(ADCO)
and
the Ministry of Higher Education and Scientific Research, United Arab Emirates, for their invitation to attend the conference. We thank William Lind say, Palaeontological Laboratory, The Natural His tory Museum, London, for preparing the holotype specimen of Bagrus shuwaiensis. The photographs were taken by Phil Crabb ( Photo Studio, NHM ) .
J. 1983. Dating the terminal Tethyan events. In Re construction ofMarine Environments ( ed. J. Meu lenkamp ). Utrecht Micropaleontological Bulletins 30: 273-98.
Ali, Mohamed S . M . 1983. Tertiary echinoids and the time of collision between Africa and Eurasia. Neues ]ahrbuch fur Geologie und Paliiontologie, Monatshefte 1983: 2 1 3-27. Arambourg, C .
1947.
Contribution a !'etude geolo
gique et paleontologique du bassin du lac Rodolphe et de la basse Vallee de l'Omo, 2. Paleontologic. Mission scientifique de POmo, 1 932-1933, vol. l . Geologie-Anthropologie, fasc. 3 , pp. 231-5 62 .
In
1987.
Zoogeography of the reptiles
and amphibians of Arabia. In
Proceedings of the Sympo sium on the Fauna and Zoogeography of the Middle East, pp. 245-56 ( ed. F. Krupp, W. Schneider, and R. Kinzelbach) . Dr Ludwig Reichert Verlag, Wiesbaden. Banister, K. E.
1973 .
A revision of the large
Barbus
( Pisces, Cyprinidae) of East and Central Mrica. Stud ies of African Cyprinidae Part II.
Museum (Natural History), . 1987.
Bulletin of the British 26: 1-148.
Zoology
The
Barbus perince-Barbus neglectus Barbus species (Teleostei, Cypriniformes, Cyprinidae ) . Bul letin of the British Museum (Natural Histor)0, Zool ogy 5 3 : 1 1 5-38.
---
problem and a review of certain Nilotic small
B aynunah Formation, AUH numbers refer to
1977. The freshwa The Scientific Results of the Oman Flora and Fauna Survey, 1975, pp . 1 1 1-54. Ministry of information and Culture,
specimens that will be deposited in the Emirate
Sultanate of Oman.
NOTE 1.
Adams, C. G . , Gentry, A. vV. , and Whybrow, P.
Arnold, E. N.
Peter Forev thanks Peter Whybrow and Andrew
1995,
REFERENCES
Museum National d'Histoire Naturelle, Paris.
ACKNOWLEDGEMENTS
March
Natural History Museum, London.
In the list of identifiable fish material from the
Banister, K. E., and Clarke, M. A.
ter fishes of the Arabian Peninsula. In
3 �rnor, R. L., Brunet, M., Ginsburg, L., Mein, P. , �< .:ldord, M., Ri:igl, F., Sen, S . , Steininger, F., and -::- :� omas, H. 1987. A consideration of some major � � ::>ics concerning Old World Miocene mammalian :.�·onology, migrations and palaeoecology. Geobios 20: �3 1-39. :=::shop, W. W., Chapman, G. R., Hill, A., and Miller, - . _\. 197 1 . Succession of Cainozoic vertebrate assem :-:�cges from the northern Kenya Rift Valley. Nature � � : 389-94.
J. 1958. Sur la presence de Porcus cf. docmac ::-oisson Siluriformes) dans le gisement neolithique ' .C:urien de Faya. Bulletin de l,Institut Fundamental .; �-lfhque Noire, Dakar 20A: 1 379-86.
::__!J.get,
---
. 196 1 . Restes de poissons du Quaternaires tmezan, 1 89 0 ) from the Pliocene of Perpignan, MN 1 5 ; and (4) to . 10, of Kohfidisch Hohle (Aus· tria) in Bachmayer and Ml-ynarski (1983). It differs from the extant western adult fonn of the Mediter ranean Basin, which acquired a narrow cervical
M
gaudryi from the late l\.1iocene, MN MaJtremys leprosa,
longer than its width ( 19 adults out of 2 4 specimens examined the MNHN; cen'ical wider or of equal width in the five in juvenile specimens), a derived character already mens from Maghreb ( Pleistocene-Holocene) , Ain l:IOUCJler:lt, ria
collection), and Douldcala II in Morocco But, like M has a
M. caspica,
leprosa
the nuchal plate by marginals 1 ( 69-95% of the odvuuJudu
form diJfers also from the i\1iocene forms (when
cer-
covering on the nuchal bone is known), 1'.1. from Artenay (48-70%), "'.f. Pnrschke, 1885, from Hernals bei
pygolopha
sa.rmatica sophiae
\Vien, Austria, and lYI. from Sandelzhausen, Germany, MN of 6 (Schleich, 198 1 ) , which have a more primitive shorter the smaller cervical (although with width greater than length) nMt!:)llldi> 1 on the nuchal. So, in the Baynunah form, this plate is by its long and tall scutes, and is primitive by its wide cervi cal, as in Pliocene forms and in the extant Consistent with the two nuchal plates, the frrst peripheral,
Mauremys caspica.
AUH 8
13.4.2) indicates a form with the vertebral l much
wider nuchal bone, contacting the marginal scute 2, as in the late Miocene (from Sandelzhausen, MN 6, of Germany at least), Pliocene and extant forms. In early to middle i\1iocene forms (including tl1e vertebral 1 just covers the nuchal and we believe that enlargement is secondary in the later forms. The other AUH 276a (fig. 13.4.1), AUH 412, AUH 412a
M.
Estimated shell lengths: 1 5-22 em.
AUH 82 (fig. 1 3.4.8) also indicate that the scute t on the plates, as in the fossil and extant forms
Decoration
'"" '"w'��n
The fragments of shell are almost smooth, finely punctuated , some of them pitted by parasitism like the extant western ."11auremys. The scute sulci are d1in, straight, or slighdy sinuous.
Rl0973 (pleural 5 , buttress of the
Dorsal Shell
part of the from the peripheral 1 , AUH 8 ; peripher686; peripherals 8 ,
(longer on anterior plates in Pliocene to ell-"tant forms than in Oligocene to Miocene fmms). AUH 310, fiV-s growth annuli. BMNH shows t h e position o f the ingninal pleurals 5 and 6. The ventral includes the rib, is strongly raised, growing of the to the lateral extremity (the eleva\vidth of the plate); the ingninal suture to the lateral part of this elevation, bur
CHELONIANS FROM
1 b
1 a
5a 6a
·" r' {-1.•:·. . I�1;,. I. .,•. t'"· .< �
' • I; 7
I
•
0
:.J.
6b
3cm
0
3cm
9c
10 a .
'
10 b
. . ;. . . ""....';
� ... ' ...· ·
11
1cm
0
·-� . ..· 11 b
a
�
-- -
-
12
THE
BAYNUNAH fOfu\1ATION
11m
Figure 1 3 .4 . Mam·emys sp. 1, AUH 276a, second left peripheral; a, b, dor sal, ventral views. 2, AUH 8 , first left peripheral; a , b , dorsal, ventral views. 3, AUH 276b, left part of nuchal; dorsal view. 4, AUH 6 8 8 , right part of nuchal; a, b, dorsal, ven tral views. 5, AUH 274, neural l ; a , b, dorsal, ventral views. 6, AUH 276b, neural 4; a, b, dor sal, ventral views. 7, AUH 4 1 2a, left peripheral 9 ; dorsal view. 8, AUH 8 2 , l e ft peripheral 8 ; a , b , dor sal, ventral views. 9, BMNH Rl 0973, pleural 5 ; a, b, c, ventral, dorsal, posterior views. 10, AUH 2 8 0 , j uvenile neural 4 or 6; a, b, dorsal, ventral views. 1 1, AUH 276c, j uvenile right peripheral 5; a, b, ventral, dorsal views. 12, AUH 3 8 3 , left hypoplastron; dorsal view.
� in
F. DE LAPPARENT
DE
BROIN /\ND P. P.
V:"u"i
DI)K
fact it is not well developed med ially ( the suture of the buttress is
around 40%, see l3.4.9b, 9c); the lllllg chambers ought not to be not well develope d eith er. Specimen AUH 148b, a fragment of hypop lastral inguinal bu ttress ,
conforms to this rela tively short devdopm ent of the buttress under the highly raised ventral part of the plate. In extmt western forms-A£. ca.1pim caspica, Af. c.
rivulata, and "�1. l.eprosa-tl1e ing uinal buttress may be linke d only under pleural 5 or, as well, between pleurals 5 and 6 as exhib ited in the Baynunah specimen and the fossils from Arte nay and Sandclzhansen . ln extant forms, in the Holocene form from Doukkala II, and in the type of the Pliocene M. gaudryi, wh ich is a juvenile specimen-- -not known in the unp repared adults "Par alic!Jelys carinata" and " Clcmmys romam" (see Bergounioux ,
1935), and in the Italian Plio- Pleistocene forms from ToscmJa (Por tis, 1 890), and Piem onte ( S acco ,
1889)-the elevation of the plate
towards the buttress is variably pronounced. The width of this elevation and of the suture of the buttress on the is, however, is around 48%, much smaller and much more lateral: the the su tur e is 8-15% the widd1 of the plate in lvi. With its \\ide elevation and suture, the Baynllllah specimen consistent
wirh the .Miocene torms from Artenav (elevation m·ound 55-65%.
suture around 40% the width of the plate) and Sandelzhausen (n�t other J'vliocene forms). Wi de elevation and suture of also known in the foss il form from the Oligocene to early J\•1iocene of Europe, plesion Mattnmys, and in exam eastern Asiatic "\iaurem1•s forms , M. ivetsoni Pritch ard m1d McCord, 199 1 , M. mtttica, anlAnrmnemys annamertsis (included in Mau1·emys by Iverson and McCord, 1994; studied i n the Bou rret collection , MHJ'..1) as well as i n Ocadia, an excl u sive ly Asiatic similar in shell shape to Afauremys that is a Bataguri nae secondary palate and not a Geoemydinei (see Hi rayama, 1985 ) . Many fossil Afauremys and Palaeochelys from Eu rope have, until recently, been atnibuted to tl1is genus ( see S chleich, 198 1 , 1984; " Ocadia" sop!Jiae). 276b) are adult, witl1out Nenrals 1 (AUH 274) and 4 known in the buttresses are
any keel, whereas tl1e juvenile
4 or 6 and the neural 5 (AUH
280) bear
longitudinal keel, rather smooth and wide, as in the western i�faurem.ys, Here, the three neurals 4-6 are hexagonal, short-sided in fi·ont. In Mauremys, their s hape is individually vari able, regular hexagonal short-sided in front or behind, or irre gular,
the pectorals . The fragments of hyoplastra confirm tl1at the h umero-pectoral sulcus curs
cntopl astron
in its
1 3 . 5.4) has a posterior third. The specimen BMNH Rl0974 i n a weak "'ide sinuous h umero-pectoral sulcus ending notch, similarly pres ent in fossil and i n exant species of lvfaurem_}�. In the other, small, younger specin1ens (fig. 1 3 .5 .2a), the notch is
variably accentuated. Th e fragme nt of hyop lastron AUH 274a (fig. 1 3 .5 .2a) shows the medial obliq uity bel1ind the pectora- abdomina l s ul cus (pe ctor al s l onger medially ) . Th e fragmentar y hypoplastra ACH 62 (fig. 1 3 . 5 . 5 ) , AUH 1 3 .5.6), AUH 50 (fig. 1 3 . 5 . 1 ) , and AUH 148b have obliquity
behind the abdomino-femoral sulcus (shortening femorals). The length of the ab dominal scutc on tl1e hypoplas tron is variable in tl1e same proportion as in western fossil and extan t species of the genus. The dorsal covering of d1e hypoplastra by the femoral scu te is also consistent with the genus: moderate "''idtl1 a little variable "''itllin the indivi dual, and no bone elevation under this covering 1 3 .4 . 1 2 and width is also variThe length of the bone relative to plastron varying i n width witllin individuals (in general able, wider and flat in tl1e females); the obliquity of d1e suture of the hyp o pl astron with tl1e hyo plastron, variable from back to front, al so conforms to The su mre between h yoplastron and hypoplastron i s straigh t . This indicates that there is no lat· eral mcsoplastron , occurs in d1c Tertiary pleu rodi ran Pel omedusoides and not in the Ter tiary cryptodiran turtles. The xiphip las tron A U H 679 (fig. 1 3 .5.7) exactly contorms to fossil and e.'\tant western ,:vfauremy. nv'ith its femoro-m1al notch, its deep anal rounded notch (although tl1ese notches can be variable in deptl1), the anal longer than d1e part of d1e femoral scute on d1e pl a te , and the moderate lateral covering of the dorsal part of tl1e plate by rl1e scutes (fig. 1 3 . 5 . 7 ) . The inguinal buttress AUH 148c is strong a t its base and does n ot cover much of the pleural 'Width altho ugh it covers more than in the extant Mediterranean specie s A1auremys caspica m1d AI. leprosa. It is like the Miocene European species Af. pygolopha speci mens from Artenay, Frm1ee, MN 4 (Broin, 1977), and 111. sop!Jiae specimens from S andelzha1J sen , Germany, MN 6 ( S chleich , 198 1 ), as mention ed above. the
alternatively seven-sided m1d five-sided. But in Ocadia they are reg
h exagonal, and short-sided in front. As in extant wes tern adult neural 4 (fig. 1 3 .4.6) is rather short and wider than it is differe nce from Ocadia, which has neural s l-4 longer than , or as long as, the wi dtl1 in the studied ma terial
ular,
M��urem·ys, the
(J\1NHN; MHNT; BMNH). In western 1¥fauremys, tl1ese neurals may also be approximately as wide as they are long.
Humerus
The head of hu me rus AUH 363 (fig. 1 3 . 5 . 8 ) is eroded bm con� tonns to ."tfauremys in the angle between tl1e trochanters, the weal< inter-trochanteric fossa, light dorsal oblignity of the head, m1d the p rolo ngation in a lip.
of the small trochanter
part of the
he ad
Plastron Material : entoplastron , ACH 470b; hyoplastrons, AUH 274a, 276b, 470c, 685, BMNH Rl0974; hypopl astrons, AUH 50, 62, 148b, 383, 470, 470a, 687, 689; inguinal buttress, AUH l48c; xiphip lastron , ADH 679.
The entoplastron, ADH 470b (fig. I 3 . 5 . 3 ) , has two anterior
borders, ori ented at 90° to each other, and a rounded posterior border, like lvfmtremj(( ( al though there is individual variation in the proportions) but not Pelomedus oides where it is rhomboid. The
specimen sho\·vs that two meeting gcuars are sufficiently long to cover a part of the entoplastron and tl1at the pectorals cov er the posterior third of the entoplastron . Consequently, they are lon ge r than the hnmerals, whid1 are short mc:dially between the gulars and
Discussion Shells of the pleurodiran turtles, which constitute the northern Gondwana population (see Broin, 1 98 8) , are known from at least the early Cretaceous in Africa to the present, south to latitude 20° N, including the extremity of the southwestern Ara bian Plate ( Gasperetti et aL, 1 993 ) . Pleurodirans
are present in the late
and early middle
CHELOl\!A..'IS FRO�I THE
\ \ I \ \ \ I I I I I
\/
1 a
0
0
1cm
0
3c m
[1:6
Figure 1 3 . 5 . Jfa u ronys sp. 1, AUH 5 0 , left hypoplastron; a, b, dor sal, ventral Yie\\·s . l., AUH 274a, right hyoplastron; a, b, dorsal,
- -,
,- - 1 I I I I I I \ \ \
B.\\�l:-:.-\J-1 FOR.\L\TIO>:
ventral views. 3, AUH
1 b
470b, entoplasu·on; ,-en tral view. 4, BMNH Rl 0974, partial left hyoplastron; ventral Yie\Y. 5, AUH 62, right hypoplastron; ventral view. 6, AUH 3 8 3 , left hypoplastron; ventral view. 7, AUH 679, left xiphiplasu-on; a, b, dorsal, ventral views . 8, AUH 363, proximal part of a right humerus; a, b, c, external, dorsal, ventral views.
•
E DE LAPPARE:-..'T DE BROI::l &'!D P. P.
DIJK
Miocene of As Sarrar (Saudi Arabia) and Ghaba ( Oman) and are principally characterised by the sutured link of the shell with the three parts of the pelvis that is seen on the xiphiplastra and on pleu rals 7 and 8. The preserved xiphiplastron 1 3 .5 . 7a) shows a pelvis (pubis and ischion) free from this plate. Other pleurodiran characters are also lacking. The absence of lateral mesoplastra between the hyoplastra and hypoplastra, and the shape of the plates (particularly the elements of the plastron and plates 5-6 bearing the inguinal but tress) are not of a pleurodiran turtle and are consistent with the Geoemydinei itf.auremys. The entoplastron shape fully agrees with lvfauremys. }vfauremys is a Laurasiatic Geoemydinei, B atagurinae without a secondary palate, unlike Ocadia (see Hirayama, 198 5 ) , with a vertebral ! as wide as the nuchal ( as in Ocadia) or wider. The extant A�iatic Geoemydinei Annanemys (included in Maurn1'tys by Iverson and McCord, 1994 ) , Sacalia, Notochelys, and the western fossil forms, Palaeochelys s .s . , Oligocene-early Miocene, that are plesions of "�fauremys, also have a wide vertebral l. Matn�emys has a rather flat nuchal bone, slightly concave on the ventral side. We have seen the evolution of marginals l and cervical size on the nuchal. In Ocadia, the cervical remains small and short when the l ateral covering by marginals l is long. Like Sacalia and Notochelys, neurals of Afauremys are sometimes regular but not always (see above ). The Baynunah form conforms to these characters . In Mauremys, there is no exact link between each pleural and each peripheral from peripheral 4 and pleural 2 back ( a plesiom orphic character present in Ocadia and Palaeochelys). There is generally contact between costal 3 and marginal 6 as in Palaeochelys and many other Batagurinae, or near contact (contact between costal 3 and marginal 7 as in the more derived Geoemydinei, from Heosemys to Geoemyda, and as in Ocadia) ; the vertebral 5 overlaps the sup rapygal 2 and not the pygal, like tl1e other Geoemy dinei, Rhinoclemysand Notochelys excepted : these characters cannot be seen in the B aynunal1 form . As mentioned above, the plastron has a slight femoro-anal notch and a deep or rather deep anal notch (not Palaeochelys, Sacalia, and Notochelys) .
The entoplastron is as described above, the gulars are as long as they are \vide or longer and more or less overlapping the entoplastron . The epiplastrons have a wide and slightly protruding or not protrud ing gular part: in its preserved fragments; the Bay nunah form conforms to these characters. The preserved fragments are sufficient to attribute the Baynunah torm to Jvfauremys without any doubt. Mau1·nnys is known in Europe from Oligocene times and it is known to be present in Africa from the late Miocene of Wadi Natrun ( Egypt) (Dacque, 1 9 1 2 : " Ocadia nov. sp . ind.", poorly known, figured by the anterior lobe part only ) and possibly of Sahabi, Libya ( Erasmo, 1 9 34 : fig. 15 ). Furthermore, Maut·emys is now dis tributed all around the Mediterranean Basin \Vith M. leprosa from France to western Libya in the west, and ltfaufemys caspica from Yugoslavia to northern Arabia, along the Mediterranean coast and inland in other countries in the east ( see Iver son, 1992; Gasperetti et al., 1993 ) . The two extant western species of 1tfaut·emys have diverged, at least from Pleistocene times (l'vf. leprosa lineage in North Africa) or possibly before . Neontologists have given discriminant charac ters for the extant species but they are based principally on the decoration of the scutes and the colours, which are, of course, not preserved in fossil forms. They are also concerned witl1 related proportions that are based on complete specimens and calculated on the mean averages of populations ( B usack and Ernst, 1980; Iverson and McCord, 1994; Pritchard and McCord, 1 9 9 1 ) . With isolated fragments or few specimens, we can not use these characters . We can differentiate the specimens only by the bony shell, tile relative pro portions of the impressions of tile horny scutes on the bones (sulci), and the sutures of tl1e plates. But we do not have the full proportions of the dorsal marginal and ventral scutes that may be longer than the bones on the border of the shelL On the bony shell, it appears that, in the type species "�f. leprosa, the cervical scute on the nuchal bone of the adult is longer tl1an its \vidth in most of the specimens ( on all the 19 adult specimens examined), and the femorals are medially shorter
CHELONI&"lS FROM THE BAY�C\�lli FO�\ETIO�
lier-Perpignan (long covering, \Yeaker but
:han the pectorals in most of the specimens ( 1 4 out of
16
specimens examined, males or females ) .
tresses, wide vertebral 1 ) , and possibly at 6e same time or before the " Pontian" (late Valle
This is valid for the living specimens, still with Lheir scutes, as well as for the bony shells. These tiYO characters are apomorphic. In
M. caspica,
I!§
sian, MN 1 0 ) of Kohfid isch (long coYering,
as in
wide vertebral 1 , buttresses unknown) . Too
most of the fossil forms studied, the width of the
many fossil forms are insufficiently prepared or
cervical is still
figured to compare them with the B aynunah
than its length and the
temorals are longer than the pectorals, and much
.lvfauremys sp .
longer than the pectorals in most of M.
level. Another problem is the reduction in num
caspica
compared with
M. c. rivulata,
caspica
and to determine this to species
bers, followed by the extinction of fossil
11Iaut·e mys species in Europe at the end of the Miocene d e Broin, in (MN 6-8 to MN 1 3 ) press; Schleich, 1984), possibly due to aridity,
where the
temorals can be as long as the pectorals. vVe can not consider these characters i n the B aynunah For mation form, which is represented only by isolated fragments of the shell. But in studied fossils from
which impedes comparisons with localities of
western Europe
equivalent age.
and North
caspica
Italy, Germany, Austria)
the condition is that of tl-f.
from the Oligocene to Recent, except in:
In conclusion, although possibly new, the B ay nunah Formation
l.
M.
aff.
gattdryi from
the "Pontian" of
1 0 ) , Austria ( B ach 1 9 8 3 ) , known by one
3.
Kohfidisch Hohle ( MN
are still not differentiated and we do not know at what period in geological time the differentiation
juvenile specimen, which has a long cervical,
occurred. It is impossible to know if the Abu Dhabi
although its width is still
species is the ancestor of the extant population of
than its
M. por tisi Sacco, 1 889, a specimen with femorals shorter than pectorals ( and narrow vertebral 1 ) . M. gat,tdryi etrusca Portis, 1 890 (see Kotsakis, 1980 ) , a specimen with femorals equal to pec torals (Plio-Pleistocene) .
4.
remains inde
mayer and Ml-ynarski,
length, and femorals shorter than pectorals.
2.
1Vfauremys species
terminate. In the late Miocene, the extant species
M.
leprosa from the
lv.[.
caspica in
northern Arabia or if the extant popu
lation is the result of another invasion much later than the age of the Baynunah
from
eastern Europe-western Orient al. ,
199 3 ) .
Gasperetti et
Furthermore, the fossil form might have
arrived in the United Arab Emirates from the west,
Rharbian (late Soltanian or
due to the presence of lvfaz£rnr�ys in 'Vadi Natrun
Holocene ) of Doukkala II, Morocco, with a cer
(Egypt) and possibly in Libya ( Sahabi), instead of
vical longer than its width and femorals longer
from the eastern Mediterranean. "Te do not, how
than pectorals; one specimen only ( Gmira,
eYer, know the relati\·e age of these localities and
199 5 ), which agrees with some extant M. lep rosa. In fact, the sample is too poor in most of
the Miocene, between "'IN
the fossils to define the species vvith these tlvo
( in Germany) and "'iN
the end of
the re\·erse is just as
13
6
(in
characters. But if we consider the derived long
MN
covering of the nuchal by the" cervical and mar-
endured climatic change such as
ginals
1,
the
wide vertebral l , and the
1 4-1 5
to MN
8
Spain and Italy) to
southern France), western Europe aridity,
an evolutive state comparable ·with a hypotheti
Mauremys (like Trionyx from MN 6-9 and MN 1 0- 1 3 ) disappeared or was extremely reduced (verified presence in Kohfidisch only, late MN 1 0 ) until M N 1 3 . vVe find few turtles at these times,
cal late Miocene form, after the Miocene ( M N
only terrestrial ( giant and small) and semiterrestrial
primitive wide cervical and strong inguinal but tress, we can then assume that the species is of
6)
o f Sandelzhausen (short covering, strong
so that
small forms. Were these two genera still
m
buttress but already wide vertebral ! ) , surely
southern Europe and/or the eastern Mediter
before the Pliocene ( MN
ranean, from where they invaded the Arabian Plate
1 4-1 5 )
of Montpel-
•
F. DE LAPPARE!;'T DE BROIN AND P. P.
VAN
DI)K
and North Africa, and then returned north into Europe?
Infrafamily Testudininei Batsch, 1 7 8 8
Geochelone Fitzinger, 1 8 3 5 Centroche!ys Gray, 1 8 72 Geochdone ( Centroche�vs) afT. sulc:ata (Miller, Subgenus
1 779)
Material by Sites (fig. 1 3 . 1 ) There are six identified specimens and four u niden tified fragments, representing at least seven individ uals. H3: AUH 276, fragment of periphetaL JD3: A'lJFf 295, lateral fragment of pleural 1 , 2, or 4. Kl: AUH 388, left epiplastron. S l : AUH 1 07, right part of the nuchal bone associated ,,;ith the peripheral 1 and the medial part of the peripheral 2; AUH 284a, part of a right pleural 8 and AUH 284, part of an unidentified plate. S4: ACH 768, a juvenile peripheral; AlJH 769, 770, 7 7 1 , three unidentified fragments.
Description Dimensions
Estimated shell lengths: 52-58 em long. Some thick undefined spedmens in the locality indicate longer shells.
Decoration
Apparently smootl1; very finely punctiform , granulous v.rith vascular foramina, and small vascular sulci. Most of the scute sulci are straight with sharp borders, sometimes elevated on a crest (figs l 3.6. 1 b, 2b) bur not always (fig. 1 3 .6.2c). This deco ration is common to the terrestrial forms of Tesrudininei. The raised sulci are sometimes found in tl1e African Pelusio;; a Pclome dusidae.
Dorsal Shell
Material: part of the nuchal, peripheral 1 , and peripheral 2, AUH 1 07; fragment of peripheral, AUH 276; fragment of pleural 1 , 2, or 4, AUH 295; part of pleural 8, AUH 284a; juvenile peripheral, AUH 768; part of an unidentited plate, AUH 284; three unidenti fied fragments, AUH 769, 770, 771. The anterior border of tl1e shell, AGH 107, is a right part of the nuchal bone associated vilith right peripheral 1 and the medial part of right peripheral 2 ( fig. l3.6.la) The width of the fragment is 1 2.5 em. The specimens belong to a tall species of tortoise (terrestrial form) of the genus Geochelone due to the size (maximw11 here around 58 ern long), the slight anterior notch of the shell in tl1e animals provided "''itl1 their marginal sulcata in Villiers, 1958), wide vertebral 1 scutes, see
fig. (see the transverse sulcus bet>.veen marginal l and vertebral 13.6. 1b), which covers the nuchal (although in some the vertebral l may also be anteriorly narrower than the nuchal bone and widened posteriorly on pleural l ) , and the thickness of the ventral of the nuchal and of the peripherals (3 em) covered by tl1e ventral border of the marginal scutes (fig. 1 3 .6.1a). If complete, tl1e notched nuchal should be seen 'With out the presence of a cen�cal scute. TI1e plates are thick, without secondary tlllnning (the lateral parr of the pleural AUH 295 measures 2.1 em).
Plastron
Material: left epiplastron, Al..TH 388; unidentified part of a plate, AUH 284. TI1e epiplastron AUH 388 (length 10.4 em; thickness at the posterior end of the dorsal lip 3.4 em; fig. 13.6.2) is similar to that of a female of the extant Geochelone ( Centrochelys) sz.tlcata from the pre-Sahelian and Sahelian part of Africa (Senegal to Ethiopia) and southwestern Arabian Plate (see Gasperetti et al., 1993; Iverson, 1992; Roser et al., 1990) because of the shape of the gnlar area and of the dorsal lip. 1be dorsal lip is moderately bent back (fig. 1 3.6.2a, 2b), in this case not covering the eoto plastron; the ventral face of the gular area is rmmded l3.6.2c), and tl1e dorsal fuce of the lip is concave posteriorly narrow with a rounded posterior border. Comparison is possible witl1 a specimen of extant Geochelone su!cata 'Without its scutes fig ured in Roset et al. ( 1 99 0 : pl. 3, figs 1 and 3). It is not wide, not and lacks a double posterior convexity, tl-om the &ei1cht?lor.'e ( Stigmochelys) fi·om eastern and southern other tall tortoise from Africa, Jrom Ethiopia-Somalia to southern Africa (Iverson, 1992 ). Like many tortoises, including Geochelo�u s.L, the protruding gular lip is notched tor a partial subdivision of the gular. The gular covers ali the medial part of the epipla.�tron along the symphysis and extends only slightly over tl1e enroplas tron. We have to decide if the long covering of the epiplastron by the gular is a plesiomorphic or reversal character for Geochelone. The short gtliars are primitive in Testudinidac. The long coverin g is present in most Geochelone tor instance, the extant G. parda!is, where the covering is variable, sometimes extending over the entoplastron, as in G. ( S.) brachy;;u.laris from Laetoli (Tanzar•ia) (.Meylan and AuHenberg, 1987). In the six extant G. sulcata specimens stndied (MNHN, AC, P), without scutes, the gular is shorter than the epiplastral symphysis length, and does not cover all the epiplastron backwards, as in a Holocene 1 990). In tl1e Sahahi specispecin1en from Niger (Roset et men-an undefined species of Geochelone" (Wood, 1987: fig. 5), cleaxly a ymmg Geochel01u of the sr�lcata lineage and exactly the same as the Baynunall form-the gulars just touch the ento plasrron. If the long covering is derived, the Baynunah form is more derived than the extant and Holocene .;S FR02.1 THE BAY:\'l'\_\1-: ::'OR\lUIO:\'
ocean currents: it ex-plains the invasion of the oceanic islands and probably their entrance into .-\frica.
Freshwater Turtles The Baynunah freshwater turtles belong to the sub family T1ionychinac and infrafamily Geocmydininei . There are no Cydanorbinae, or Carettochelyidac, Laurasiatic representatives previously known in the Afro-·Arabian Plate from the early-middle .Miocene. �either arc there at B aynunah Pclomcdusoides ( Podocnemididac and Pclomedusidae) , the principal Afi:ican representatives from the northern Gondwan ian fauna (Broin, 1 9 8 8 ) , known from the early Cre taceous. This change of tua after the late early-early middle Miocene fauna seems to indicate a new link with Laurasia, probably with the eastern Mediter ranean Basin, later than the Hofuf Formation (Saudi Arabia) in age. In the case of Geochelone, it might indicate a link with Egypt or Libya or with both, if the Wadi Natrun and Sahabi localities are older than, or of a similar age to, the Baymmah Formation. This hypothesis does not seem j ustified because of the absence of any unquestionable remains of Trionyx s.s. of the T triungr&is lineage in the B aynunah Formation, when the few Libyan or Egyptian remains arc sufficient to indicate its presence. The presence of the Pelusios sinuatus group-so \videly distributed in northeast Africa from Miocene to present-day times (the northern distribution limit of the group is now reduced in extcnsion)-in Wadi Natrun but not in the Abu Dhabi localities does not favour this hypothesis either, unless this absence is due to an ecological barrier, because PelusioJ- is not found in Arabia at the earlier Miocene time witl1 the Podocnemidi dae, the Cyclanorbinae, and the Carettochelyidae . The remains of Aiauremys are not well enough represented in Sahabi and vVadi Natrun to com pare the evolutionary state of the species with the B aynunah form and to provide evidence of a close link between the two areas. The B aynunah Formation fauna is also different from those specimens from the .Miocene of the
�
Siwaliks that are sufficiently well k:nmYn . In Pakistan, in the Potwar Plateau ( Pilbeam et 1 9 79 1 , Nepal ( West e t al., 1 99 1 ) , and India site, turtles unpublished; see Thomas et there are always: -Cydanorbinac , undefined genera or aff Cvclo derma with separated hyoplastra and hypoplastra as in the Dam Formation and Ghaba and, later, Lisse mys (an Indian subcontinent endemic form); -Trionychinae, with Chitra, A;pideretes ( Indian subcontinent endemic forms) , or undefined forms; -Podocnemididae of the Schweboemys group ( in the Lmver series) as at As Sarrar and Ghaba and as in the Bugti Hills ( Palcistan ) (Wood, 1970); -Geoemydinei and Batagurinae, with Indian sub continent endemic forms; and -Testudininei either endemic ( Indomtudo) or undefined genera ( see above ) . A link (probably filtered) e-vidently existed between tl1e Afro-Arabian Plate and the Indian subcontinent during the early-middle Miocene, to allow the passage of Schweboen-tys, of the cycloder mid with separated hyo-hypoplastra, and of the hypothetical ancestor of the Indian Geochelone sub genus ( in common with the two Mrican subgen era) present in the younger Siwaliks. ( Naturally, the late early .Miocene species from the Arabian Plate are different from those of the Siwaliks; the latter are younger in age. ) The link is not at all evident during late Miocene-Pliocene times . At least we are sure that neither lvlaunmyJ nor CentrochelyJ is present in the Siwaliks, and that none of the numerous endemic forms fi·om India-for example, Kachuga, .2vlelanochelys (for merly given as Geoemyda in Pilbeam et a!. , 1 979 1 , and Chitra s o well represented all along the Pot war Plateau series, from the Late Chinji to the Late Nagri Formations-penetrated into Arabia. The only doubts are for the undefined Trionyx s . l. and Geochelone s . l . s o poorly preserved. B u t experi ence shows that each time a genus becomes recog nisable in the Later Siwaliks ( from middle-late Miocene), it is different from those of �-\rfo-_\rabia. The B aymmah turtle fauna is inte restin g be cause it indicates a continuity ·with earlier times
__...____________________________________________....,.,..... . ...., .. ..,.......,,... ... = .. � �'·· .......,_
m
F. DE LAPPARE'lT DE BROIN ,'\.."' D P. P.
VAN
DIJK
within the Mro-Arabian Plate for the terrestrial torm. If a tectonic or a climatic event hit the mid
7-8 and MN 13 ( B rain, 1977; Schleich, 1984, 1985), these two taxa may have been present in
dle Miocene fauna of turtles in the Arabian Gulf
eastern 1v1editerranean Basin, thus allowing them to
area either Geochelone remained in the area or
p enetrate into Arabia via the Mesopotamian Basin.
the
returned from Mrica after the event. But before the B aynunah Formation, at the late early-early middle Miocene, an aquatic system connected the southern Arabian Gulf area, northern Egypt, and Libya.
STRATIGRAPHY The Baynunah Formation turtle fauna cannot be
Then a permanent break was established between
identified accurately enough to give any precise
Africa and northeastern Arabia tor the freshwater
data. In the context of Arabian Plate environments
turtles: the aquatic niche of these groups disap
during the Mio- Pliocene, the B aynunah Formation
peared and the fluviatile network did not return.
is definitely younger in age than the early-middle
Trionychidae are good swimmers, with paddled
Miocene ( Burdigalian-Vindobonian) fauna of Egypt
limbs ( elongated metapodials and phalangae, digits
(Moghara, Wadi Faregh ) , Libya (Jebel Zelten) ,
within the skin, leaving only three free hand nail s ) .
Saudi Arabia (As Sarrar) , and Oman ( Ghaba ) .
Although variable in their ecological niches, Tri
The morphology o f the Baynunah Mauretnys i s
onyx generally prefer flowing wide systems, lakes, or rivers. By contrast, �f.auremys is less of a swimmer,
more derived than the M N
with just membranes between unelongated digits; it
latest Miocene) to extant forms. The morphology
occasionally lives in muddy waters. Note that Tri
of the terrestrial for m Geochelone aff. sulcata and of
onyx> if carried away in an estuary stream, will swim
Mauremys sp. clearly indicates older taxa than the
6
forms from Europe
but more primitive than the Pliocene ( an d may b e
in the sea, following a coastal current to access a
Pliocene-extant forms that appear in the Pliocene.
new river further along the coast. But �f.auremys is
In a first approach , we suggested a correlation with
a true little freshwater form , not known for this
northern Egypt (Wadi Natrun) and with Libya
ability. Both are principally carnivorous. Unlike the
(Sahabi ) , evidence for which may be the terrestrial
intertropical Cyclanorbinae and Pleurodira, T1·ionyx
form Geochelone aff. sulcata. B ut the absence or the
and Rafetus now live as far as about
complete changes in the links between the freshwa
dubious presence of a true Trionyx of the T triun guis lineage, present in Sahabi and vVadi Natrun, the possible presence of a Trionyx linked with Rafe ttts, and the absence of PelusioJ present at Wadi
ter systems (rivers, ponds and lakes) in the area of
Natrun might favour an eastern correlation with
Turkey and iVIauremyJ� lives as far as Yugoslavia ( Iverson,
1992 )
.
37° N in 46° N in
Their presence indicates
the Arabian Plate and northeastern Mrica. Mean
the eastern Mediterranean, Tigris-Euphrates B asin,
while in Africa s.s. (see summaries in Broin,
and might indicate a correlation with the Balchtyari
Lapparent de Broin and Gmira,
1 994 ) ,
1979;
the expan
sion of the Cyclanorbinae and Pelomedusoides occurs at the same time as the Trionychinae and
Formation of iraq. The complete absence of Cyclanorbinae, Caret tochelyidae, and Pelomedusoides (Bothremydidae,
J.f.auremys (these taxa only in northern Africa)
Podocnemididae, and Pelomedusidae) , which may
arrived from Europe. If evidence is found for the
be due to drastic changes in regional ecology, is
link of the B aynunah fatma 'with the lnjana fauna, it
also a new biostratigraphic factor that relates to the
will be possible to deduce that a southern extension
middle Miocene of Arabia.
of the Tigris-Euphrates Basin altered the previous hydrographic network that ran from Egypt to India, thereby modifYing the zoogeographic affini
ACKNOWLEDGEMENTS
ties of the region. During the climatic event ( dry
We are especially grateful to Peter Whybrow, The
ness? ) that caused lvf.auremys and Trionyx s.l. to dis
Natural History Museum, London, for making
appear or sharply decline in Europe between MN
the study of the turtle material possible; Diana
lma
F.
DE
LAPPARENT
DE
BRO!N AND P. P. VAN DI)K
pp. 2 1 1-3 3 (ed. N. Petit-Maire and J. Riser). Private
Chdoniens. Memoires de Ia Societe Geologique de
printing, Marseille.
France, Paleontologie 3 : 140-68 .
---
. 1 9 8 8 . Les tortues et le Gondwana. Examen
des rapports entre le fractionnement du Gondwana e t I a dispersion geographique des tortues plemodires a partir du Cretace. Studia Palaeocheloniologica. Studia
Geologica Salmanticensia 2 : 1 0 3--42 . Brunet, M., and Heirrtz, E. 1 9 8 3 . Interpretation paleontologique et relations biogeographiques de Ia faune de verte bres du Miocene superiem d'Injana, Iralc Palaeogeography, Palaeoclimatology, Palaeoecology 44: 283-9 3 . Brw1et, M . , Heirrtz, E., and Battail, B . 1984. Molayan (Afghanistan) and the Kham Siwaliks of Pakistan: An example of biogeographic isolation of Late Miocene M an1.111alian faru1as. Geologie en Mijnbouw 1 984: 3 1-3 8 .
Eiselt, J. 1976. Ergebnisse zoologischer San1.111elreisen in der Tiirkei. Bemerkenswerte Ftmde von Reptilien,
II. Annalen der Naturhistorische Museum, Wien 8 0 : 803-14. Erasmo, G . d'. 1934. Su alcwu avanzi di vertebrat.i terziari della Si.rtica. In Missione della Reale Accade
mia d;Italia a Cufra, Viaggi di Studio ed Esplorazioni, pp. 1-2 3 . Reale Accadernia d' Italia, Fondazione Alessandro Volta, Rome. Gasperetti,
J.
J.,
Stimson, A. F., Miller,
J. D.,
Ross,
P., and Gasperett.i, P. R. 1 9 9 3 . Turtles of Arabia.
In Fauna of Saudi Arabia, pp. 170-367 (ed. W. Buttiker and F. Krupp) . National Commission for Wildlife Conservation and D evelopment, Riyadh, Saudi Arabia.
Brru1et, M . , Heintz, E . , and Sen, S . 1 98 1 . Datations
Geraads,
paleontologiques et sequence biochronologique dans
superieur du Djebel Krechem El Artsouma (Tunisie
D.
1 989. Vertebres fossiles du Miocene
le Neogene continental d'Afghanistan . Comptes Ren
Centrale) . Comparaisons biostratigraphiques. Geobios
dus de l'Academie des Sciences, Paris 29 3 : 305-8.
22: 777-8 0 1 .
Busack, S. D . , and Ernst, C. H. 1 9 8 0 . Variation in
Gilmore, C. W. 1 9 34 . Fossil tmtles o f Mongolia: Sec
Mediterranean populations of Mauremys Gray 1 869
ond contribution. American Museum Novitates no.
( Reptilia) . Testud.ines, Emydidae. Annals of Carnegie
689: 1-14.
Museum 49: 2 5 1-64. Ckhikvadze, V. M. 1989. On the systematic position of Tertiary giant land tortoises from Asia, Europe and Africa. Bulletin ofthe Academy ofSciences ofthe Gem;gian
Socialist Soviet Republic [in Russian] 1 3 3 : 429-32 . Dacque, E. 1 9 1 2 . D i e fossilen Schildkroten Aegyptens.
Geologische und Palaontologische Abhandlungen 14: 275-3 37. Deperet, C. 1 8 8 5 . Description des vertebres fossiles du terrairr pliocene du Roussillon. In Description
geologique du Bassin tertiaire du Roussillon, pp.
Gm.ira, S. 1 99 5 . Etude des cheloniens fossiles du Maroc:
Anatomie-systematique-phylogenie. Caluers de Paleon tologic, 1-140. Centre National de Ia Recherche Sci entifique, Patis. Gray,
J.
E. 1 873a. Notes on Chinese mud-tortoises
(Ttionychidae) with the description of a new species sent to the B ritish Museum by Mr. Swinhoei, and observations on the male organ of tl1is family. Annals
and Magazine of Natural Histo1-y 1 2 : 1 5 6-64. ---. 1 873b. Notes on Mud-tortoises ( Trionyx
geoffroy) and on the skulls of the different kinds. Pro ceedings of the Scientific Meetings of the Zoological Soci
1-274. Editions G . Masson, Paris.
ety of London 1 87 3 : 38-72.
D eperet, C., and Donnezan, A. 1 890. Les an.imaux
Harris,
pliocenes du Roussillon. Classe des Reptiles, 0. des
Research Project, Vol. 1: The Fossil Hominids and an
J.
M. 1 978. Palaeontology. In Koobi Fora
Introduction to their Context, 1 968-1 974, 3, pp. 32-53 ( ed. M. G. Leakey and R. E. Leakey). Clarendon
Kuhn, 0. 1964. Testudines. In }ossilimtt Cntnlogzu. Animalia, pp. 1-299 ( ed. F. Westphal 1.
Press, Oxford.
Gravenhage.
0. P. 1908 . The fossil turdes of North America. Publications of the Carnq_qie Institution of Washington 7 5 : 1-568.
sexual dimorLambert, M. R. K 1 99 3 . O n phism, and d1e general ecology of the Afi-ican Spurred Tortoise, Geochelone sulcata, in Mali. Chelonian Con
servation and Biology Heintz, E., and Bmnet, M. 1982. Une barriere geo graphique entre le sous-continent indien et l'Eurasie occidentale pour les faunes continentales du Miocene superieur. Comptes Rendus de t>Acadbnie des Sciences,
Paris 294:
477-8 0.
Heude , F. M. 1 880. Memoire sur les Trionyx. In
Jfhnoires concernant Phistoire nattH'elle de l'Empire Chi noispar des Peres de la Compagnie de]i:suJ� vol. 1 , pp. Y-viii, 1-38 . 1mprimerie Mission Catholique, Shanghai. Hirayama, R. 198 5 . Cladistic analysis of batagurine nu·des (Batagurinae: Emydidae: Testudinoidea); a pre liminary result. Studia Palaeocheloniologica. Studia
Geologica Salmanticensia 1 :
1: 37-46.
Lapparent de Broin, F. de. in press. Les cheloniens . In
La faune tniocene de Sansan ( Gers) et son environ nement, pp. 1-46. Mi:rnoires dtt Museum National d'Histoire Naturelle, Paris. Lapparent de Broin, F. de, and Gmira, S. 1995. Les cheloniens dulo;:aquicoles du Rift occidental, Ouganda. In Geology and Palaeobiolo!:fY of the Alber
ti1�e Rift Valley, Uganda-Zaire, Volume II· Palaeo biology/Paleobiologie, pp. 1 57-86 ( ed. B. Senut and M. Pickford) . CIFEG Publication Occasionnelle 1 994/29. Centre International pour la Formation et les Echanges Geologiques, Orleans.
141-57.
Hummel, K. 1932. Trionychia fossilia. In Fossilium Catalogus. L Animalia, pp. 1-106 (ed. W. Quen stedt). W. Junk, Berlin. l1·erson, J. B . 1992. A RePised Checklist with Distribu tion Maps of the Turtles of the World. Private Printing, Richmond, Ind.
Liebus, A. 1930. Neue Schildkri:itenreste aus den Ter tiarenreste susswassertonen von Preschen bei Bilin in ohmen. Rospravy Stamiho Geologickeho Ustavu
�
Ceskoslovenski: Republiky 4:
1-57.
Loveridge, A, and Williams, E. E. 1 957. Revision of d1e African tortoises and turdes of the Suborder Cryptodira. Bulletin of the 1Vfuseum of Comparative
Zoology at Harvard College Iverson, J. B . , and McCord, W. P. 1994. Variation in East Asian turdes of the genus Mauremys (Bataguri dae; Testudines) . Journal of Herpetology 28: 1 78-87.
1 1 5 : 1 63-557.
Lydek.k:er, R. 1 88 5 . Indian Tertiary and post-Tertiar\· Vertebrata. Siwalik and Narbada Chelonia. �Memoirs of
the Geological Survey of India. Palaeontologia Indica Jimenez-Fuentes, E., and Martin De Jesus, S. 199 1 . Ejemplares-tipo de quelonios f6siles espafioles. RCPista
Espanola de Paleontologia 6 :
98-106.
Kotsakis, T. 1980. Revision des tortues (Emydidae, Testudinidae, Trionychidae) du Plio-Pleistocene de Valdarno superieur (Toscane, Italic) . QJtaternaria 22: l l-37.
I:
3:
1 5 5-208 . 1 8 89a. Catalogue of the Fossil Reptilirr in tile British Museum ( Natural History). Part III: CiJdfmirr.
Trustees of the British Museum, London.
---. 1 889b. On the land tortoises of the Si\\·alL\;:s. Record�� Geological Survey of Indirr 2 2 : 2 09-12 .
•
---
F. DE LAPPARE!'IT DE BROIN AND P. P.
. 1 891.
Vi\!'1
DUK
On a new species of Trionyx from the
Miocene of Malta and a chelonian scapula from the
Q;tarterly Journal ofthe Geological Soci ety of London 47: 37-40.
London Clay.
McDowell, S. B.
1964.
Partition of the genus
Clem
m.ys and related problems in the taxonomy of the
C. 1986.
Decouvertes recentes dans les sites
Comptes Rendus de FAcademie des Sciences, Paris 302: 681-86.
Proceedings of the Zoological Society of London 2: 239-79. 1 987.
Pickford, M., Senut, B . , Hadoto, D . , Musisi, J., and Kariira,
miocenes de Moroto ( Ouganda Oriental): Aspects biostratigraphiques.
aquatic Testudinidae.
Mallouli, N . , Creuzot, G., and Geraads, D.
Denkschriften der 111athematich-nat11wwissenschaftliche Classe dct· Kaiserlichen Akademie der Wissenschaften, Wien 29: 1 1 1-24.
Un
Pilbeam, D. R., Behrensmeyer, A. K, Barry, J. C.,
1 979. Miocene sediments and fau Postilla 1 79: l-45 .
nouveau gisement de vertebres dans le Neogene de
and Shah, S. M . I .
Comptes Rendus de l'Acadimie des Sciences, Paris 305: 1 1 21-24.
nas of Palcistan.
]\misie centrale.
1 890. I Rettili pliocenici del Valdarno supe riore e di alcu.ne altre localita pliocmiche di Toscana.
Portis, A.
1 990 ( 1 989). Updating of MN zones. In European Neogene 1Vfammal Cht·onology, pp. 73-90
Mein, P.
( ed. E. H. Lindsay, V. Fahlbusch, and P. Mein) .
Plenum Press, New York.
Le Monnier Successeur, Firenze. Pritchard, P. C. H . , and McCord, P.
1 99 1 . A new HerpetokJgica 47: 139-47.
emydid turtle from China. Meylan, P. A.
1 987.
The phylogenetic relationships of
Bulletin of the American Afuseu.m of Natu.ral History, New York 1 86: 1-lOI.
soft-shelled mrtles (Family Trionychidae ).
Meylan, P. A., and Auffenberg, W.
1 986.
1885. Clemrnys sannatica n. sp aus Denkschriften dM· Mathematich-naturwissemchaftliche Classe der Kaiser lichm Akademie der Wissenschaften, Wien 50: 1 8 5-92.
New land
tortoises (Testudines: Testudinidae) from the Miocene of Africa.
Purschke, C. A.
dem Tegel von Hernals bei vVien.
Zoological Journal of the Linnean Society 86:
279-307. ---. 1987. The chelonians of the Laetolil Beds. In Laetoli: A Pliocene Site in Nonhern Tanzania., chap. 3 . 1 , pp. 62-78 ( ed . M. D . Lealcey and J. M. Harris).
1903. Schildkrotenreste aus dem Abhandlungendm Senckenbet;ge Natmfurkunde Gesellshaft 29: 1-64.
Reinach, A. von.
Agyptischen Tert:iar.
Robinson, P., and Black, C. C.
1 974. Vertebrate fau Annals ofthe Geo logical Survey ofEgypt 4: 319-32 .
nas from the Neogene of Tunisia.
Clarendon Press, Oxford . Roger }., Pickford M . , Thomas H . , d e Lapparent de Meylan, P. A., Weig, B . S . , and Wood, R . C.
1990.
Fossil soft-shelled turtles (Family Trionychidae) of the Lake Turkana Basin, Africa.
Copeia 2: 508-28.
Broin F., Tassy P., Van Neer W. , Bourdillon-de- Gris sac, C., and AI- Busaidi S . ,
1994.
Huqf au Sultanat d'Oman. Peters, K. F.
1855.
Schildkrotenreste aus den osterre
Denkschriften de1· il1athernatich-natu.t-wissenschaftliche Classe der Kaiser lichen Akademie der Wissenschaftm, Wien 9: 1-22.
ichisten Tertiar-Ablagerungen.
Annales de Palimztologie
80: 253-73 . Roser, J. P. , Broi n , F. de, Faure, l'vl ., Gayet, M., Guerin, C., and Mouchet, F. Ouaffadene et d'Adrar Bous
-- -. 1 869.
Zur Kenntniss der Wirbelthiere aus
den Miocanschichten von Eibiswald in Steiermark.
Decouverte de
vertebres fossiles dans le Miocene de la region du
1 990. La taune de Tin 1 0 , demc gisements
archeologiques de !'Holocene ancien au Niger nord oriental.
Giodynamique 5 : 67-89 .
CHELONL'h'\S FRO,\!
F. 1 889. I Cheloni astiani del Piemonte. Mmt lrie della Reale Accademia delle Scienze di Torino II, 39: 1-37. Salih, M. S . 1965. The shell of the soft-shelled turtle euphraticus (Daudin). Bulletin of the Iraq
Satu1·al History Museum 3: 1-8 . Schleich, H. H. 198 1 . Jungtertiare Schildkroten Siid deutschlands unter besonderer Beriicksichtigung der FundsteUe Sandelzhausen. Courier Forschunginstitut
Smckenbet;g 48: 1-372. . 1984. Neue Reptilienfu.nde aus dem Tertiar Deutschlands. 1 . Schildkr6ten aus dem Jungtertiar Siiddeutschlands. Naturwissenschaftliche Zeitschriftfor
---
Siederbayern 30: 63-9 3 . . 1 98 5 . Neogene Testudines o f Germany: Their stratigraphic and ecological evaluation. Studia Palaeocheloniologica. Studia Geologica Salmanticensia
---
l:
249-67.
Sen, S., and Thomas, H. 1 979 . Decouverte de rongeurs dans le Miocene moyen de Ia Formation Hofuf (Pro Yince du Hasa, Arabie Saoudite ). Compte Rmdu Sam
main de la Societe Geologique de France 1 : 34-37. Siebenrock, F. 1902. Zur Systematick der Schild krotenfamilie Trionychidae Bell, nebst der Beschrei bung einer neuen Cyclanorbis-art. Sitzungsberichte der
Kaiserlichen Akademie der Wissenschaften, Mathema tisch-naturwissenschaftliche Classe, Wien 9 1 : 807-46. - -. 1913. Schildkroten aus Syrien und Mesopo tamien . Annalen des Kaiserlisch-koniglichen Naturhis torischen Hofmuseum�� Wien 27: 17 1-225 . Thomas, H., J. J . , and Sastry, M. V. A 1980. Geologie et paleontologic des Siwaliks inferieurs (Miocene moyen) a Ramnagar (J ammu et Cachemire, Inde ). Be Reunion Annuelle des Sciences de La Ter1·e, Marseille, 21-23 fevrier, 1980, p. 346. Thomas, H . , RDger, J . , Sen, S . , Bonrdillon-de-Grissac, C., and Al-Sulaimani, Z. 1989. Decouverte de vertebres
fossiles dans !'Oligocene inferieur du Dhotar Sultanat d'Oman) . Geobios 22: 101-20. Thomas, H., RDger, J. , Sen, S . , Dejax, J., Schuler. .\ L Al-Sulaimani, Bourdillon-de- Grissac, C ., Bretor:. G . , Brain, F. de, Camoin, G., Cappetta, H., CarrioL R. P. , Cavelier, C., Chaix, C., Crochet, J. G., Gayet, M., Gheerbrant, Lauriat-Rage, Noel, D . , Pickford, M . , Poignant, A. F. , Rage, J. C ., Roman, ]., Rouchy, J. M., Secretan, S., Sige, B., Tassy, P., and Wenz, S . 1991 . Essai de reconstructior. des milieux de sedimentation et de vie des Primates anthropoYdes de ! 'Oligocene de Taqah (Dhotar, Sul tanat d'Oman). Bulletin de La Societe Geologique de
France, Paris 1 62 : 7 1 3-24. Thomas, H., Sen, S . , Khan, M . , Battail, B . , and bue, G. 1982. The Lower Miocene fauna of AI-Sarrar (Eastern Province, Saudi Arabia) . .AILAL, The Journal
ofSaudi Arabian A�·chaeology 5 : 1 09-36. Thomas, H., Sen, S., and Ligabue, G. 1980. La faune miocene de l a Formation Agha Jari du Jebel Hamrin (lrak). Proceedings Koninklijke Nederlandse Akademie
van Wetenschappen 83: 269-87 . Thomas, H., Taquet, P. , Ligabue, G . , and Del'Agnola, C. 1978. Decouverte d'un gisement de vertebres dans les depots continentaux du Miocene moyen du Hasa (Arabie Saoudite) . Compte Rendu Sommaire de la
Societe Geologique de France, Paris 1978 : 69-72. Villiers, A. 1958. Tortues et crocodiles de !'Afrique Noire fran.;aise. Initiations Africaines 1 5 : 1-354. vVermutl1, H . , and Mertens, R. 196 1 . Schildkriiten Krokodile-Bruckenechsen. Gustav Fischer Verlag, Jena.
---. 1977. Liste der rezenten Amphibien und Rep tilien. Testudi.nes, Crocodylia; Rhynchocephalia. In Das Tierreich, Lief, pp. 1-174 ( ed. R. Mertens, W. Heming, and H. Wermuth). Walter de Gruyter, Berlin. West, R. M., Hutchison, J. H . , and J'viunthe, J. 199 1 . Miocene vertebrates from tl1e Siwalik group, western Nepal. Journal of Vertebrate Paleontology 1 1 : 108-29.
ll'IJ
F. DE LAPPARENT DE BROIN Al'lD P. P
VAN
DIIK
Whybrow, P. J. 1 9 89. New stratotype; the Baynunah
---
Formation ( Late Miocene), United Arab Emirates:
tion. In Neogene Palaeontology and Geology of Sahabi,
Lithology and palaeontology. Newsletters on Stratitfra
phy 2 1 : 1 -9 .
. 1 98 7 . Fossil turtles from the Sahabi Forma
pp. 1 07-22 (ed. N . T. Boaz, A. El-Arnuati, A. W.
Gaziry, J. de Heinzelin, and D . D. Boaz). Alan R. Liss, New York.
Whybrow, P. J . and Hill, A . eds. 1 999. J?ossil Verte
brates ofArabia. Yale University Press, New Haven. \Vhybrow, P. }., f-Iill , A., Yasin al-Tikriti, W., and
Yeh , 1-I.-k. 1965. New materials of fossil turtles of Inner Mongolia. Vertebrata PalAsiatica 9: 47-78 .
Hailwood, E. A. 1990. Late Miocene primate fuuna,
Zug, G. R. 197 1 . Buoyancy, locomotion, morphol
flora and initial palaeomagnetic data from the Emirate
ogy of the pelvic girdle and hindlimb, and systemat
of Abu Dhabi, United Arab Emirates. Journal of
ics of cryptodiran turtles. Afiscellaneous Publications)
Human Evolution 1 9 : 5 8 3-88 .
Museum of Zoology) UniJ'mity of Michitfan 1 42 : l-98.
Wood, R. C. 1 970. A review of the fossil Pelomedusi dae (Testudines, Pleurodira) of Asia. Breviora no. 357: 1-24.
Fossil Crocodilians from the Late Mioce11.e Baynunah Formation of the Emirate of Abu Dhabi, United Arab Emirates : Osteology and Palaeoecology MICHAEL R-A.UHE, E B ERHARD "DINO" FREY, D"Au�IEL S .
Since 1 9 79 scientific explorations b y The Natural History MnseumjYale University team and others of the Baynunah Formation of Abu Dhabi have vielded remains of late Miocene crocodilians (Whybrow, 1989; Whybrow et al. , 1990 ) . The material, the property of the Emirate of Abu Dhabi, is currently housed in The Natural History Museum, London, but was loaned to the Staat liches Museum fur Naturkunde, Karlsruhe ( Ger many) tor investigation. The aims of this chapter are ( l ) to morphological descriptions of the ta,\:Onomically rele vant crocodilian specimens to clarifY their taxo nomic and systematic status; and ( 2 ) to reconstruct the habitat-habit relationships of the Baynunah gavialid and crocodylid species.
TERMINOLOGY AND ABBREVIATIONS The compression index of the alveoli is the ratio mesiodistal length/linguolabial length; interalveolar space 1 is the distance between the first and second alveoli, interalveolar space 2 means the distance between the second and third alveoli, etc . ; inter alveolar grooves are the grooves in the lower and upper jaw that are located between two alveoli for the reception of the upper or lower teeth. The position can be lingual, labial, or median to the alveoli . Interalveolar groove 1 is situated between the first and second alveoli, interalveolar groove 2 between the second and third alveoli, etc.
Al\D ToRSTEN RosstvLA.N N
Abbreviations for .mentioned in this chapter are as follows: AUH Emirate of Abu Dhabi; BMNH = The Natural Museum, London-tormerly the British Museum (;...Jatural History) ; PlCM = Palaontologisches Institut der Universitat Mainz; SMF Senckenberg Museum Frankfurt/Main; SMNK Staatliches Museum fur Naturkunde Karlsruhe. =
GEOLOGICAL BACKGROUND The Baynunah Formation outcrops in the coastal part of the Western Region of the Emirate of Abu Dhabi in the area lmmvn as the Baynunah plateau, ' "Which is of the vast oil-bearing, Rub al Khali Basin (\Vhybrow et aL, 1990). The area has been lit tle studied even by oil geologists (Whybrow et a!. , 1 9 9 0 ) . Lower Baynunah Formation faunas (particu larly the vertebrates) indicate an age of between 8 and 6 million years (Ma) during the Vallesian and/or Turonian, or late Tortonian to early Messin ian of the late :Miocene (Friend, 1 999-Chapter 5 ) . As Friend describes in Chapter 5 , the lower part of the Baynunah Formation consists mainly of fluvial sands, gravel beds, and some distinct mud units. The fossil vertebrate and inver tebrate faunas, root marks, and soils indicate that the unit \:vas fi:mned predominantly by river deposi tion . Sedimentary structures characteristic of river channel deposition indicate a braided river network, consisting of channels 2-1 0 metres deep and up to
Copyright © 1999 by Yale University. All rights reserved. ISBN 0, 300-071 8 3 , 3
-
•
M. RAt:HE ET AL.
1 00 metres wide, which probably occurred in belts separated by bars and vegetated islands. Direction of flow of the river system appears to have been to the east-southeast, and Friend speculates that the river of the lower part of the Baynunah Formation may have been ancestral to the present-day Tigris Euphrates system.
TAXONOMY, SYSTEMATICS, DESCRIPTIONS, AND DISTRIBUTION OF EAST AND NORTH AFRICAN AND ARABIAN CROCODILIANS Historical Review and Current Knowledge Andrews ( 1 90 1 ) was the first to mention fossil Mrican crocodilians from the Fayum (Egypt) and Libya in the literature. A short description of Croc odylus articeps ( Lower Oligocene ), C. megarhinttS ( Lower Oligocene), To1rtistorna gavialoides (Upper Eocene-Lower Oligocene), T africana ( Middle Eocene-Upper Eocene ), and T kenteme ( Middle Eocene) were published by Andrews in 1 9 0 5 . 1n 1 906 Andrews redescribed the Mrican crocodilian fauna in a catalogue. A full and detailed analysis was later supplied by Miiller ( 1927). Tchernov ( 1986) gives an extensive review of the taxonomy and systematics of African crocodilians and the fol lowing list completes the numbers of described African crocodilian species up until the work of Tchernov ( 1986: 1 3-14; 5 3 , fig. 24). 1.
Crocodylus Laurenti,
1768: C. lloydi Fourteau, 1920 C. lloidi [sic] Tchernov, 1 986)-uppermost Oligocene/lowermost Miocene to upper most Pleistocene/lowermost Holocene ( Libya, Kenya, Sudan, Ethiopia, Egypt, Tanzania, Uganda, and ?Tunisia); C. p igotti Tchernov and Van Couvering, 1 978- uppermost Oligocene/ lowermost Miocene ( Kenya); C. checchiai Maccagno, 1 947-Upper Miocene-lowermost Pliocene (Libya). 2 . Tornistoma Muller, 1 846: T dmvsoni Fourteau, 192 0-Lower Miocene ( Libya and Egypt) ;
T cairense Muller, 1927-Lower Oligocene ( Egypt); T tenuirostris Muller, 1927--Lower Oligocene ( Egypt) . 3 . Euthecodon Fourteau, 1920: E. bntmpti Joleaud, 1920-Lower Miocene-Middle Pleistocene (Kenya, Tanzania, and Ethiopia) ; E. arambour gia Ginsburg and Buffetaut, 1978-Lower Miocene ( Libya); E. nitriae Fourteau, 1 920Upper Miocene ( Libya and Egypt) .
More recently Aoki ( 1992 : pl. 1-4) reported fossil crocodilian remains from the Sinda Formation, Haute Zaire, eastern Zaire ( uppermost Miocene/ lowermost Pliocene) . He identified the fragmentary material as C. aff. nilotints> C. aff. lloydi> Mecistops ( Crocodylus) aff. cataphractus> Euthecodon sp. , and Osteolae;nus ati. osborni. Boulenger ( 1920: 9 1 3 ) referred jaw remains from the Plio-Pleistocene of Omo ( Ethiopia), which were identified by Joleaud ( 1 92 0 ) as Tomistoma> to Gavialis Oppel, 1 8 1 1 . Muller ( 1927: 87) agreed with the determination ofBoulenger ( 1 920) and mentioned also an indeter minate Gaviali:; species from the Burdigalian of Moghara, Egypt, described first by Fourteau ( 1920: 26, fig. 19). Of uncertain determination are also Tomi:;tmna africana and Tomi:;to ma gavialoides. The former species is referred to Gavialis african us by Hecht and Malone ( 1972), whereas Buffetaut ( 1982b) ascribed both species to a new genus: Eogavialis as E. gavialoides and E. africamts. Accord ing to Tchernov ( 1986: 1 3 , 46) both these Tomis tom.a species do not belong to Gavialis but to Euthe codon. As long as it remains unclear exactly which characters define the Gavialidae the systematic status of these species must be considered as being dubi ous ( see Steel, 1989; and this chapter) . With the exception of Crocodyltts lloydi the tem poral and geographical distribution of the African Crocodyltts species is limited. Crocodylus lloydi ranges from the uppermost Oligocene/lowermost Miocene to the uppermost Pleistocene, possibly into the lowermost Holocene throughout north and east Mrica. The extant C. niloticus can be proven with certainty to have existed since the Middle Pliocene (Tchernov, 1986: 5 3 , fig . 24), and C. cataphractus Cuvier, 1 824 since the Lower Pliocene. Both of =
CROCODILIA.i'IS l'ROl\1
these species co-existed during the Plio- Pleistocene
Euthecodon
in the Kenyan Turkana Basin with
brumpti and C. lloydi (Arambourg,
1 947; Tchernov
and Van Couvering, 1 9 7 8 ; Tchernov, 1 9 8 6 : 5 5-5 6 ) . Only and
C. checchiai, C. llo.ydi, E. brumpti,
E. nitriae are
reported from the Upper
Miocene of Mrica. Thomas et al. ( 1982) were the first to mention the existence of
otti
C. lloydi and C. pig
outside Africa from the Lower Miocene of As
Tchernov ( 1986: 1 6 ) dismissed the \·alue of the crocodilian postcranial skeleton as being useful tor drawing systematic conclusions. According to him ( 1 9 8 6 : 1 9 ) the length of the premaxillary-mamlan· suture in
Crocodylus is
"extremely stable" and is
therefore of high taxonomic value at species level. He also considered the following features to be of taxonomic importance: relative symphysis length ( alveolus to which posteriormost part of the symph
Sarrar, Saudi Arabia (see Tchernov et al., 1 9 8 7 :
ysis extends ) , symphysis ratios, and the number
3 0 1-2 ) .
of teeth in the maxilla and mandible ( Tchernov,
Buffetaut ( 1982a) briefly mentioned fossil croc odilian remains outside Africa from Saudi Arabia
1 9 8 6 : 2 0 , table 1 ; 22-23, table 2 ) . According to Kalin ( 193 3 ) and Wermuth ( 19 5 3 ) most of these
( 1 40 km northwest of Dhahran, Dam Formation,
features show only little differences within the
Lower Miocene, Burdigalian-Vindobonian) . Buffe
recent genus
taut ( 1 9 84: 5 1 6 figs lA-C ) published a precise
for mandible characters where the overlap between
description with a figure of a mandibular fragment
several species-for example,
from this locality, which he referred to
C. pigotti.
It is worth noting that no representative of the Tomistominae sensu Kalin, 1 933
Euthecodon)
( Tomistoma and
have been recorded from Saudi Arabia,
only from East and North Africa. We compared the crocodilians from the lower
Crocodylus. This
sus) C. sia�-nensis,
and
holds true especially
C. palustris) C. poro C. acutus-is far-reaching. We
believe that such characters on their own are not reliable enough to define taxa at the species level. For Tchernov ( 19 8 6 : 1 6 ) cranial indices are impor tant for species separation. The study by Tchernov and Van Couvering ( 1 978 : table 2) in which they
part of the Baynunah Formation of the Emirate
compared 1 1 different cranial indices between
of Abu Dhabi with fossil crocodilians from Mrica,
checchiai, C. pigotti)
and
C. lloydi
C.
does not confirm
India, South America, and the Pacific area (Solomon
the taxonomic importance they postulate for species
Islands ) as well as with modern specimens. The com
diagnoses. On the contrary, the skull proportions of
parison of the Abu Dhabi crocodilians with the fos
the quoted species are very similar to one another
sil taxa
Ikanogavialis Sill, 1970, Charactosuchus Langston, 1965 , and Euthecodon is based on litera
and lie, according to the experience of Kalin ( 1 9 3 3 :
ture research only.
8 34-77, tables
within the range of varia-
3 2 ) , two fragmentary mandibles (AUH 5 6 and
Crocodylus species (the same holds true for the several species of Diplocynodon and Allognathosuchus; Rauhe, in preparation ) . The skull indices of different African Crocodylus species from
6 1 6 ) , a fragmentary maxillary (AUH 2 8 5 ) , and a
the Lower Miocene to the present day overlap so
fragmentary jugal (AUH 2 8 1 ) . These specimens will
much in their intraspecific variability that they can
The determination of the Baynunah crocodilians is based on two fragmentary skulls (AUH 3 34 and
tion of a single
be described and figured in detail ( figs 1 4 . 1-1 4 . 4 ) .
not be used for the identification of species . Kalin
The postcranial specimens available for comparison
( 19 3 3 ) has pointed out how great the intraspecific
are taxonomically less significant or insignificant
variability of both characters can be in skulls of sim
because in previous studies the diagnostic charac
ilar-sized crocodiles coming from the same geo
ters for fossil crocodilians from Africa have been
graphical region. It is noted here that there are
restricted to the cranium. For the same reason we
great differences in the skull proportions of
could not take into account osteoderms and verte
lloydi documented
brae, despite their diagnostic relevance (Frey et a! .,
fig. 1 2 ) .
1987; Jesus et al . , 1 9 8 7 ; Laemmert, 1990, 1993; Rauhe, 1 99 3 ; Rauhe and Rossmann, 1 99 5 ) .
C.
by Tchernov himself ( 1986: 34,
I t i s also noted that Tchernov's ( 1 986) taxo nomical results concerning
C. pigotti
(KNM-RU
1m
M. RAmiE ET AL.
2 5 79 and B MNH R7729) (Tchernov and Van
struction share many similar characters in skul l
Couvering, 1 97 8 : 8 5 8 , pl. 99, figs 3 and 4) and
and mandibular dentition, and C . niloticzts and C. palustris are especially similar in these aspects. It is
Maccagno's concerning C. checchiai ( M accagno, 1947) are based on only two specimens. In the
unnecessary to compare the Abu Dhabi material
case of C. lloydi we have also only a few speci
with the skulls of all extant Crocodylus species, due
mens; some are very fragmentary and cannot be
to the distribution history of Crocodylus in Africa.
reliably assigned to this species. For the same rea
Until now only two fossil Crocodylus species are
son it is impossible to make reliable statements
with certainty known from northern and eastern
about the intraspecific variability of skull and
Africa that reach into modern times-C. niloticus
mandible featnres ( and therefore species dit1er
and C. cataphractus (Tchernov, 1 986 )-and we
ences) for these species based on our present state of knowledge. Although C. pigotti sensu Tchernov and Van Couvering, 1 9 7 8 is documented from several iso lated incomplete mandibles-KNM-RU 2 5 8 0 , 2 5 8 2 , and 2 5 8 3 (Tchernov and Van Couvering, 1 9 7 8 : 8 5 8 ) -this material has not been fully described and remains unfigured. The reference of a mandible to C. pigotti (AS 979) by ButTetaut ( 1 9 8 4 ) is presumably based on the brief and taxo nomically insufficient description of the mandible of
C. pigotti (KNM-RU 2 5 8 0 , 2 5 8 2 , and 2 5 8 3 ) by Tchernov and Van Couvering ( 1 978 : 8 6 1 ) . The reason tor the determination of the mandibular fragment by Buffetaut ( 1 984) is unclear and the coincidence in stratigraphical age with material from Kenya is insufficient for reliable species identi
therefore have no problem in restricting compari
son of the Abu Dhabi material to these two species.
Specimen AUH 285-Edentulous Left Maxilla (table 1 4. 1 and fig. 1 4. 1 ) AUH 285 is a fragment 91 mm long consisting of the first eight and part of the ninth alveolus . A vestige of the premaxillary-maxi!� lary suture on the palate (median surure of the premaxillaries; Tch� ernov, 1986: 19) is preserved. The posterior point of this suture is at the level of the anterior border of the second maxillary alveolus. The dorsolateral process of the premaxillar y--maxillary surure extends posteriorly to the level of the anterior border of the second maxiUaty alveolus. The arrangement, shape, <md dimension of the alveoli are as follows: 1.
The alveoli increase somewhat irregularly i n size from alveoli 1 to 5; the fifth alveolus is by tar the largest.
2. 3.
Alveolus 6 is about equal to the fourth in size. Alveoli 2, 3, 7, and 8 are nearly equal in size t o each other. Posterior to alveoli 3 , 4, and 6 the interalveolar grooves are linguolabially elongated and distinctly v,i_der than long. The interalveolar groove between the fourth and fifth alveolus is
4.
fication. Thus, the association of the Saudi Arabian material with C. pigotti is questionable.
much larger (mesiodistal diameter 6 . 3 mm ) and deeper than the interalveoloar groove 3 (mesiodistal diameter 4.5 m m ) . 5.
OSTEOLOGY AN D COMPARATIVE ANATOMY OF THE TAXONOMICALLY RELEVANT SPECIMENS- CROCOOYLUS SP. 11\JDET. , CROCODYLUS CF. NILOTJCUS, AND CROCODYLI DAE GEN . ET SP. INDET. Crocodylus articeps and C. rttegarhinus are excluded from the comparison with the B aynunah crocodil ians, because both these taxa in Africa extend only
6.
7. 8.
Posterior to alveoli 8 and 9 the interalveolar grooves are cir cular, rather deep, and almost as large as the neighbouring alveoli. Interalveolar spaces 6 and 7 are the largest; interalveolar space 8 is the next largest. I.nteralveolar grooves 6 and 7 are distinctly shallower than interalveolar groove 8 . Alveoli 2 a n d 7 are a compressed oval i n cross�section; all other alveoli arc almost circular. Neither the lingual nor the l abial borders of the alveoli are supported by a stout buttress.
Comparison ofC. niloticus (SMF 47244, PlUM S 7 7, Tchernov, 7 986: pl. 3, fig. 7; 20, table 1) with AUH 285 (table 1 4. 1 and fig . 1 4. 1 )
into the early Oligocene. The descriptions and fig ures of C. checchiai ( Maccagno, 1 947) do not allow
Similarities
a reliable taxonomic comparison with the material
Interalveolar space 6 is the second -Interalveolar space 7 is Interalveolar spaces and 4 are larger than l and 2. Inter� alveolar spaces 3 and 4 are equal in size to each other.
from Abu Dhabi. According to K11in ( 19 3 3 ) the recent Crocodylus species with a massive jaw con-
-Alveol i 2 and 3 arc of about the same size as 7 and 8.
CROCODILI.-1.1'1 5
Figure
14.1.
B.\\1\l'K\H fOR,\L-\TIOC:
ll'IJ
Crocodylus sp . indet. Fragment of an edentulous left maxilla, AUH 2 8 5 ; palatinal 20 m m .
·.ie\\'. Scale bar
=
-The most posterior extension of the palatinal part o f the premax illary-maxillary suture reaches the centre of the second ma'ciJlar y
ah·eolus
I'RO�I THE
(PlUM Sl l ) .
-There is a steep wa.U between the alveoli row and pal ate
(BMNH
R7729) : the palate is sunk below the level of the alveoli row so that
there is a near vertical wall posteriorly ( maxillary teeth 10-7) and a
7-l ).
-S mal l, linguolabially elongated interalveolar grooves are situated
steeply sloping wall anteriorly (ma'cillary teeth
::1edian between alveoli 3-5, large circular ones between alveoli 7-9.
-Interalveolar spaces
Differences between C. niloticus and AUH 285
Comparison of C. lloydi (BMNH R5893, R5893, R8333, and R4697, Tchernov, 1 986: 28-30, pl. 4, fig. 1; 20, table 1) with AUH 285 (table 1 4. 1 and fig . 14. 1 )
"The dorsolate ral processus of the prema�illa.ry-maxillary suture in C. uiloti.cus extends posteriorly to the fourth-fifth maxillary alveo :1!5 ( Tchernov, 1986: 20, table 1 ) ; in AUH 28 5 it reaches only the ' �cond. Acco rding to Kalin ( 1933: 557), who contrary ro Tcher•
O\" ( 1 986) studied recent individuals of
C. niloticus from
different
;.::ographical areas, the length of the premaxillary-maxillary suture
:S highly variable between the second and fourth tooth; in PIUM _ l l the dorsolateral processus of the prema'cillary-maxillary s u ture extends posteriorly to the anterior border of the third rooth.
7 and
8 are nearly equal i n size.
Similarities
-Alveoli 2 and 8 are nearly equal in size.
-1mcralveolar space 7 is the largest; interalveolar spaces 6 and 8
arc the second largest and nearly equal in s ize; intcralveolar space 5
is small.
Comparison of C. pigotti Tchernov and
-There are large in teralveolar grooves 6 (linguolabially elongated)
Van Couvering, 7 978 (BMNH R7729, Tcher
-Interalveolar spaces
nov, 7 986: pl. 7, figs 2, 4; Tchernov and Van Couvering, 7 978: 858-64, pl. 99, figs 7 -4) with AUH 285 (table 1 4. 1 and fig . 1 4. 1 ) Similarities -lnteralveolar space 7 is the largest. Alveoli 5-6 are separated by a
sma.U interalveolar space. Interalveolar space 6 is much smaller than 7.
-The most posterior extension of the palatinal part of the prem;u.: illary-maxillary suture reaches the second maullary alveolus. -The dorsolateral processus of the premaxillary-maxillary sunue extends posteriorly to the second ma'cill a.ry alveolus.
Differences of C. pigotti Tchernov and Van Couvering, 7 978 from AUH 285 pigotti alveoli 2-5 (and possibly 1) have
-In C.
a stout bumess
labially and l ingually, and the teeth therefore appear to emerge from long protruding collars.
and
7 (ci rcula.r).
1, 2, and 5 are smaller than the others.
-The most posterior extension of the pala tinal part of the prema-.:illary-ma.'cillary suwrc reaches the middJe of the second maxillar·1·
ah-colus (BM.1'\IH R8 3 3 3 ) ; according to Tchernm· ( 1 9 8 6 : 2 0 , tablc
1 ) : " l st
to the mid lst".
Differences of C. 1/oydi (BMNH R8333) from
AUH 285
3 and -± in C. l!ow!i are muc h br,;:er than 7 and S . trig;na.l in shape , and located far lin guall,·; in ,-\.t'H 285 i.nterah eol.lr !';:s 3 and -t ,u·e located directlv between the ah-eo li, the1· J.re :' "�guo!Jbialh· elon gated, ar1d rela tively l arge r than those in C. !!v:.-. and Crocodyhts acutus (Arch Creek dose to the estuary into the Biscayne Bay, southeast coast of Florida) . I n all these cases the different crocodile forms live in different ecological habitats: whereas Croco dylus porosus is a typical inhabitant of coastal brackish waters, Crocodylus jolmsoni lives in inland freshwa ters; whereas Alligator mississippiensis inhabits nearly all freshwaters in Florida, Crocodylus aczttus lives only in the coastal region of South Florida. In some recent ecosystems it is knmvn that such encounters between two different crocodilian forms can result in the breakdown of the population of the physically weaker form (Cott, 1 9 6 1 ; Jelden, 1 9 8 5 ; Johnson, 1973; Medem, 1 9 7 1 ; Otte, 1 978; Webb and Mano lis, 1989; Webb and Messel, 1978 ; Webb et al. , 1 9 8 3 ; Vanzolini and Gomez, 1 9 79). Geographically only four extant crocodilian species co-exist: in the Amazonian Basin, the largest crocodilian ecosystem (7 million square kilometres, one-fifth of the world's freshwater sources) . It is important to repeat, however, that such a co-exis tence can be stable only when the live in different habitats and/or have different breeding seasons or nesting sites ( Cott, 1 9 6 1 ; Jelden, 1 9 8 5 ; Johnson, 1 9 7 3 ; Neill, 1 97 1 ; Webb and Manolis, 1989; Webb and Messel, 1978; Webb et al., 1 9 8 3 ; \rVhitaker, 1 980; Vanzolina and Gomez, 1979). The social and, especially, the territorial behaviour may be useful for avoiding fatal intraspecific interactions. For the Miocene Baynunah river system the co existence of Gavialis and Crocodylzts ecotypes could be explained ( 1 ) by habitat splitting and territorial time-condensing of the fossilised control or remains by reworking. The co-occurrence of the two crocodilian ecotypes in the Baynunah river sys tem remains a problem. It might be explained ( 1 )
BAY:c;L:.::AH FOJUHTrON
lmJ
by a territorial behaviour not existing in recent crocodilian communities or (2) by time-averaging mixing two distinct assemblages. Taphonomic data suggest little reworking and support the co-exis tence of the crocodilians . \Ve therefore prefer the former explanation.
ACKNOWLEDGEMENTS We present our sincere thanks to Peter \rVhybrow (The Natural History Museum, London) for loaning us the crocodile material from Abu Dhabi, for references to the newest literature on the geology of the Baynunah Formation, and for manuscript. vVe the improvements to our also thank Alexandra Anders (Staatliches Museum fiir Naturkw1de, Karlsruhe) for her preparation work and Colin M . Wight for translation of the Ikanogavialis paper. The photographs were taken by D . F. and D. P. , >vith developing and printing by Volker Griener (SMNK) .
REFERENCES Alderton, D. 1 99 1 . Crocodiles and Alligators of the World. Blandford, London. Andrews, C. W. 190 1 . Preliminary note on some discovered extinct vertebrates fi:om Egypt. Geological Magazine, London 8 : 436-44.
---. 1905. Notes on some new crocodilia from the Eocene of Egypt. Geological lvfagazine, L<Jndon 1 1 : 48 1-84. ---. 1906. A Descriptive Catalogue of the Tertiar)' Vertebrates of the Faxum (Egypt). Trustees of the British Musewn (Natural History), London. Aok:i, R. 1 992. Fossil crocodilians from the Late Ter tiary strata in Sinda Basin, Eastern Zaire. African Study Monographs, suppl. 17: 67-85 . Arambourg, C. 1 947. Contribution a l'etude geologique et pakontologique du bassin du lac
e
M. RAUHE ET
M,.
Rodolphe et de la basse vallee de l'Omo. 2, Paleon tologic. In Mission scientijiqz�e de FOrno, 1 932-1 933, voL l . Geologie-Anth1'opologie, fasc. 3, pp. 5-562. Museum National d'Histoire Naturelle, Paris. Basu, D. 1979. Indien kampft urn den Ganges-Gavial. Siebnanns Tierwelt 3 : 4-1 3 . Berg. D . E . 1966. Die Krokodile, insbesondere Asiato sudJus und af[ Sebecus aus dem Eozan von Messel bei Darmstadt. Abhandhmgen des Hessischen Landes Amtesfiir Bodenforschtmg 52: 1-1 05. . 1969. Charactosuchus kugleri, eine neue Krolcodilart aus dem Eozan von Jamaica. Eclogae Geo logicae Helvetiae 62: 3 1-73 5 .
---
Boulenger, G. 1 920. Sur le gavial fossile d e l'Omo. Comptes Rendus de l'Academie des Sciences, Paris 170: l -914. Buffetaut, E. 1 982a. Crocodylia. In The Lower Miocene fMma ofAl-Sarntr (Eastern Province, Saudi Arabia) . AILAL, The Journal of Saudi Arabian Archaeology 5 : 1 09-36. 1 982b. Systematique, origine et evolution des Gavialidae Sud-Americains. Geobios, Memoirs Speciale 6: 127-40.
---
. 1984. On the occurrence of Crocodylus pig otti in the Miocene of Saudi Arabia, with remarks on the origin of the Nile crocodile. Ne�tes Jahrbuch jz:ir Geologie und Paliiontologie, Monatshefte 1 984: 5 1 3-20.
---
Cott, H. B . 1961. Scientific results of an enquiry into the ecology and economic status of the Nile crocodile ( Crocot{vlus niloticus) in Uganda and Northern Rhodesia. Transactions of the Zoological Society of L{}n don 29: 2 1 1 -356. D<miel, J . C. 1983. The Book ofindian Reptiles. Bom bay. Forey, P. L., and Young, S. V. T. 1 999. Late Miocene fishes of the Emirate of Abu Dhabi, United Arab Emirates. Chap . 12 in Fossil Vertebrates ofAt·abia, pp. 1 20-35 ( ed. P. J. vVhybrow and A. Hill). Yale Univer sity Press, New Haven . Fourteau, R. 1 920. Cont1·ibution it l'titude des vertebres miocenes de l'Egypte. Geological Survey of Egypt, Cairo. Frey, E. 1982. Der Bau des Bewegungsapparates der Krokodile und seine Funktion bei der aquatischen Fortbewegung. Diploma thesis, Fakultat fur Biologie der Universitat Ti.ibingen. ---. l 98 8a. Anatomic des K6rperstarnmes von Alligator mississippiensis Daudin. Stuttgarter Beitriige zur Naturkunde, ser. A, 424: 1-106. ---. 1988b. Das Tragesystem der Krokodile-cine biomechanische tmd phylogenetische Analyse. Stutt garter Beitriige zur Naturhmde, ser. A, 426: 1-60.
Bustard, H. R. 1980. A note on the nesting behav iour in the Indian Gharial, Gavialis gangeticus. Jou1r nal ofthe Bombay Natural History Society 77: 5 1 4-1 5 .
Frey, E., Laemmert, A., and Riess, J. 1987. Bary phracta deponiae n.g. n.sp . (Reptilia, Crocodylia), ein neues Krolcodil aus der Grube Messel bei Darmstadt (Hessen, Bundesrepublik Deutschland). Neues Jahrbuch fiir Geologie und Palaontologie, Monatshefte l : 1 5-26.
Bustard, H. R. , and Singh, L A. K 1 977. Studies on the Indian gharial Gavialisgangeticus (Gmelin) (Rep tilia, Crocodylia): Change in terrestrial locomotory pattern vvith age. Journal of the Bombay Natural His tory Society 74: 5 34-3 5.
Friend, P. F. 1 999. Rivers of the Lower Baynunah Formation, Emirate of Abu Dhabi, United Arab Emi rates. Chap . 5 in Fossil Vertebrates ofA�·abia, pp. 38-49 (ed. P. J. vVhybrow and A. Hill). Yale University Press, New Haven.
Galdikas, B. M. F., and Yaeger, C. P. 1 984. Croco macaque in Borneo. dile predation on a crab
Johnson, C. R. 1 973. Behaviour o f the Australian crocodiles, Crocodylus johnsoni and Crocod;'ltts porosw.
American Journal ofPrimatology 6: 49-5 1 .
Journal ofthe Linnean SociePy 52: 3 1 5-36.
Ginsburg, L., and Buffetaut, E . 1 978. Et1-thecodon arambout;gi n. sp. et !'evolution du genre Ettthecodon, crocodilien du Neogene d'Afrique. Geologique
Joleaud, L. 1920. Sur Ia presence d'un Gavialidae du genre Tomistoma dans le Pliocene de l'eau douce de l 'Ethiopie. Comptes Rmdus de fYAcademie des Sciences,
J1editerrane 5 : 29 1-302.
Paris 70: 8 1 6-18 .
Gorzula, S. G. 1978. An ecological study of Caiman crocodilus crocodilus inhabiting Lagoons in the Venezuelan Guayana. Oecologica 3 5 : 21-34 .
IGlin, J. A. 1933. Beitrage zur vergleichenden Oste ologic des Crocodilidenschadels. Zoologische
Grenard, S. 1 99 1 . Handbook ofAlligators and Croco diles. Krieger, Malabar, Fla. Giirich, G. 1 9 1 2 . Gryposuchus jessei, ein neues schmalschnauziges Krokodil aus den jiingeren _-\blagerungen des oberen Amazonas-Gebietes.
Jahrbuch der Hambu1;gischen Wissenschaftlichen Anstalten 29: 59-7 1 . Guggisberg, C. A. W. 1972. Crocodiles: Their Nat ural History, Folklore and Conservation. David and Charles, New1:on Abbot.
Jahrbucher (Abteilung Anatmnie und Ontogenie) 57: 535-714. Kraus, R. 1 996. Pisegavialis jugaliperforatus n. gen. n. sp. ein neuer Gavialide aus der Pisco-Formation ( oberes Miozan, Peru). Diploma thesis, University of Mainz. Laemmert, A. 1990. Korperpanzerung, die ersten drei Hals\virbel unci die ersten beiden Halsrippen der Mes seler Diplocynodonten ( Crocodylia, Eozan, Messel). Diploma thesis, Institut fur Geologie unci PalaontoloTiibingen.
Hecht, T. M. K, and Malone, B. 1 972. On the early history of the gavialid crocodilians. Herpetologica 28:
Laemmert, A. 1993. Dorsal and ,-ent:ral armour and various positions of embedding in Diplocwzodou ( Crocodilia) . Kaupia (Da nwriidttT Beitriige sur
28 1-84.
Natut;geschichte) 3: 35-40.
Iordansky, N. N. 1 973. The skull of the Crocodilia. Biology of the Reptilia, vol. 4, pp. 20 1--62 ( ed. C. Gans and T. S. Pearson). Academic Press,London.
Langston, \V. 1965 . Fossil crocodilians from Colom bia and the Cenozoic history of the Crocodilia in South America. Publications of the University of Cali
In
fornia 52: 1-1 57. Jelden, D. 1985. Brutbiologie und Okologie von Crocodylus porosus und Crocodylns n. novaeguineae am mittleren sepik (Papua Neuguinea). Stuttgarter Beitrage zu1' Naturkunde, ser. A, 378 : 1 -32. Jesus, de M . , Fuentes, E. J., Fincias, B., del Prado, J. M., and Alanso, E. M. 1 987. Los crocodylia del Eoceno y Oliogoceno del la cuenca del Duero. Dientes y osteoderms. Revista Espanola de Paleon
tologia 2 : 95-108.
Maccagno, A. M . 1947. Descrizione di una num·a specie eli Crocodilus del giacimento di Sahabi (Sirtica).
Atti della Reale Accadetnia ( Nazionale) dei Lincei. Mernorie; Roma 1 : 6 1-9 5 . Maskey, T. M . , and Mishra, H. R. 1 98 1 . ConseiTa tion of gharial ( Gavialis gangeticus) in Nepal. In Wild is Beatttiful, pp. 1 8 5-196 (ed. T. Ch Majupuria). Bangok.
eM
. RAUHE .
ET
AL.
Maskey, T. M., and Schleich, H. H . 1 992. Unter suchungen und Schutzma�nahmen zum Ganges gavial in Sudnepal. Nt:ttur tmd Museum 1 22 : 258-67. Medem, F. 1 97 1 . Biological isolation of syrnpatric species of South American Crocodilia. IUCN Publica tiom; n.s.) Suppl. Pape1· 32: 1 52-5 8 . Molnar, R. E. 1 9 8 2 . A longirostrine crocodilian from Murua (Woodlark), Solomon Sea. L'tiemoirs of Q;tems land Museum 20: 675-85 . Muller, L . 1927. Ergebnisse der Forschungsreisen Prof. E. Stromer in den Wusten )\gyptens V. Tertiare Wirbeltiere. 1 . Beitrage zur Kenntnis der Krokodilier des agyptischen Tertiars. Abhandltmgen der Bay
erischen Akademie der "Wissmschaften (Naturwis senschaftliche Abteilung) 3 1 : 1-96.
Rauhe, M., and Rossmann, T. 1995. News about fos sil crocodiles of the Middle Eocene from Messel and Geiseltal. Hallesche Jahrbuch fiir Gemvissenschaften B 1 7 : 81-92. Ross, C. A., Garnett, S., and Pyrzalcowski, T. 1989. Crocodiles and Alligators. Merehurst Press, London. Sill, W D. 1970. :Nota preliminar sobre un nuevo Gavial dal Plioceno de Venezuela y Lma discusion de los Gaviales Sudamericanos. Arnet;ghiniana 7: 1 5 1-59. Singh, L. A K., and Bustard, H . R. 1977. Locomo tory behaviour during basking and spoor formation in the gharial ( Gavialisgangetints) . British Journal of Herpetology 5 : 673-76. Steel, R. 1 973. Part 16: Crocodilia. In Handbuch der Paliioherpetologie (ed. 0 . Kuhn). Gustav Fischer, Stuttgart.
Neill, W T. 1 97 1 . 1J1e Last Ruling Reptiles, Alliga tm�s) Crocodiles and their Kin. Colwnbia University Press, New York.
---
Otte, K C. 1 978. Untersuchungen zur Biologic des
Taylor, E. H. 1 970. The turtles and crocodiles of Thailand and adjacent waters. Bulleti1� of the Univer sity of Kansas 49: 87� 1 79.
Mohrmkaimans (Melanosuchus niger Spix 1825) aus dem Nationalpark con Manu (Peru). Munich. Penny, M. 1 99 1 . Alligators tmd Crocodiles. Crescent Books, New York. Rauhe, M. 1990. Habitat-Habitus-Wechselbeziehung von Allognathosuchus gaudryi Stefano 1905 (=.A.llog nathosuchus haupti Weitzel 1935 ). Geologisches Jahrbttch Hessen 1 18: 5 3-6 1 . ---. 1993. Postkranialskelett und Taxonomie des Alligatoriden Allognathosuchus haupti (Mitteleozan von Messel, Darmstadt) unter Berucksichtigung der Anatomie und Altersvariationen von Allognathosuchus cf haupti. PhD thesis, University of Mainz.
---. 199 5 . Die Lebensweise tmd Okologie der Geiseltal-Krokodilier-Absehied von traditionellen Lehrmeinungen. Hallesche Jahrbuch fiir Geowis senschaften B 17 : 65-80.
1989. Crocodiles. Christopher Helm, London.
Tchernov, E. 1 986. Evolution of the Crocodiles in East and North .Afhca. Edition du Centre National de Ia Recherche Scientifique, Paris . Tchernov, E., and Van Couvering, J. A. H. 1 978. Crocodiles from the Early Miocene of Kenya. Journal of Paleontolo,_rry 2 1 : 857-67. Tchernov, E., Ginsburg, Tassy, P., and Goldsmith, N. F. 1987. 1-1iocene mammals of the Negev (Israel ). Journal of Vertebrate Paleontolog_y 7: 284-3 1 0. Thomas, H . , S., Khan, M., Battail, B., and bue, G. 1 982. The Lower Miocene fauna of Al-Sarrar (Eastern Province, Saudi Arabia) . AIL4I�) The Journal of Saudi Arabian Archaeology 5: 1 09-36. Thorbjarnarson, J. B. 1990. Notes on feeding behav ior of the gharial ( Gavialisgangeticus) under semi natural conditions. Journal of Herpetology 24: 99- 1 00.
CROCODILIA:-lS FRO,\!
Trutnau, L. 1 994.
Krokodile-Alli.J:Jatoren, Kaimane, Echte Krokodile tmd Gal'iale, :Keue Brehm Bucherei 593. Westarp Wissenschaften, Magdeburg.
Vanzolini, P.
and Gomez, N. 1 979. Notes on the
ecology and grm:vth of Amazonian caimans.
Papeis
Avulsos de Zologie 32: 205-16.
Wermuth, H . 1 95 3 . Systematik der rezenten
Mitteilungen atts dem Zoologi.w:bm Jfuseum der Humboldt- Univmitdt in Berlin 29: 3 75-5 1-l: .
Krokodile.
\:Vmtaker, R. 1980.
Status and Distribution of Croco diles in Papua New Guinea. Wildlife Division, ment of Lands and Environment, Papua New
FAO/UNDP field document, Port Moresby. Waitkuwait, E. 1 98 1 . Untersuchungen zur Brutbiolo des Panzerkrokodils
( Crocodyhts cataphractus)
---. 1987.
The management of crocodilians in
im Tai-:Kationalpark in der Republik Eltenbeinkiiste.
India. In
Diploma thesis, University of Heidelberg.
gators.
Webb, G. J., and Manolis, S. C. 1989.
VVhitaker, R, and Basu, D. 1983. The gharial
Australia.
Crocodiles of
Reed Books, :Kew South Wales, Australia.
Wildlife Management-crocodiles and Alli
Chipping Norton, N ew South Wales, Australia.
( Gavi alisgangeticus) : a review. Journal of the Bombay Nat ural History Society 79: 5 3 1 -48.
Webb, G. J., Manolis, S. C., and Buck:worth, R.
1 982. Crocodylus johnstoni in the McKinlay River area,
N. T. I. Variation in the diet and a new method of assessing the relative importance of prey.
Australian
Journal of Zoology 30: 877-90. --� .
I 989.
New stratotype; the Baynunah
Formation (Late Miocene) , United Arab Emirates: Lithology and palaeontology.
Newsletters on Stratigra
phy 2 1 : 1-9.
1983. Crocodylus johnstoni in the McKinlay
River area, N. T. II. Dry season habitat selection and an estimate of the total population size.
Whybrow, P. J.
Australian
Journal ofZoology 1 0 : 373-82.
Whybrow, P. J., Hill, A., Yasin al-Tikriti, W., and Hailwood, E. A. 1 990. Late Miocene p1imate £1una, flora and initial palaeomagnetic data from the Emirate of Abu Dhabi, United Arab Emirates. journal
Human Evolution 1 9 : 583-88. Webb, G. J., and Messel, H. 1978. Movement and
Crocodylus porosus in some rivers Australian Wildlife Research 5: 263-8 3 .
dispersal pattern of
of i\rnhem Land, Northern Australia.
of
A Late Miocene Insectivore and Rodent Fauna from the Baynunah Formation, Emirate of Abu Dhabi, United Arab Emirates HANS
The small collection of rodent remains from the B aynunah Formation (Whybrow, 1 989) made in 1 992 (de Bruijn and Whybrow, 1 9 94) was, with few exceptions, obtained by dry-screening sediment from the surface of blow-holes . Because of the technique used duri ng the reconnaissance for small mammals the exact provenance of the specimens is unknown and the enamel is often corroded. The opportunity presented to revisit the Shuwaihat localities after participating in the First Interna tional Conference on the Fossil Vertebrates of Ara bia in Abu Dhabi, March 1995 was therefore wel comed. During the second collecting expedition we restricted ourselves to the S 1 and S4 sites at Shuwaihat, which presumably represent about the same lithostratigraphical horizon. After cleaning a section near the northern limit of the S 1 outcrop two conglomerate beds separated by an erosional surface could be seen. Both these beds show cross stratification and seem to have accumulated in river bars. The clasts of the lower bed are small caliche pebbles; those of the overlying bed are on average larger caliche pebbles and day-balls. Three different lithologies-the lmver conglomerate, the upper conglomerate minus the clay-balls, and the clay balls-were collected separately and subsequently wet-screened after drying. A trench dug at about 30 metres to the south of this small outcrop revealed the presence of only one, much thinner conglomerate bed, which was also sampled. These four samples from site 5 1-named 5 1 -low, Sl-high, S l - balls, and 5 1 -trench-all appeared to contain some rodent remains and fragments of fish bones,
DE
B RUIJN
crocodile teeth, eggshell, internal casts of gas t:ropods, and oogonia of Characaea. A trench dug in the small escarpment of site S4 revealed the presence of one narrow fossiliterous conglomerate bed. The clasts of this bed are caliche pebbles and small indurated day-balls. The amount of sediment processed from S l low, S l -high, S l -trench, and S 4 i s of the order of magnitude of about 200-2 5 0 kg per site; that from 5 1-balls is about 5 0 kg. Exact data are not avail able, because the sediment was dry-screened on a 0 . 5 mm mesh at the locality b efore wet-screening.
M ETHODS The small-mammal material described below com prises the specimens collected in 1 992 and 1 995 . Although the various sites are not strictly contempo raneous, the rodent material is treated as a sample from one community because there is no evidence that the fluvial part of the Baynunah Formation comprises a geologically long time interval. 1\1easurements of the teeth were made using a Leitz Ortholux measuring microscope with mechan ical stage. All meas urements are given in 0 . 1 mm units. The teeth figured are all x40 and illustrated as if they are fi·om the left side.
SYSTEMATICS Thryonomyidae Pocock, 1 922 (cane rats) Thryonomyidae gen . et sp. indet. 15.1)
Copyright © 1999 by Yale University. All rights reserved. ISBN 0-300-07 1 8 3-3
SMALL
.VIAMMALS l'RO�l THE B.-1\'-l.''.vith
crest as in most cricetids . The upper incisor has one
a dental pattern that is intermediate between unques
longitudinal groove .
tionable Myocricetodontinae and Gerbillinae. This last objection is no longer valid and it seems pro bable that the root of the Gerbillinae is in the Myocricetodontinae. I therefore retain the three subfamilies of Chaline et al. ( 1 977) with the obser vation that the Gerbillinae and Taterillinae cannot be differentiated on the basis of fossils of which the ear region is unknown. The contents of the subfamilies :Myocriceto dontinae and Gerbillinae plus Taterillinae differ considerably among authors. Whereas Tong ( 19 8 9 )
(1\1ascaramys excepted)
places a l l fossil genera
in the
Myocricetodontinae, de Bruijn et a! . ( 1 9 7 0 ) , Sen ( 1 977, 1 9 8 3 ) , Jaeger ( 1977), and others allocate the genera
Pseudomeriones and Protatera to
the
Gerbillinae s.l. Although the classification of fossils that are known only by teeth remains necessarily speculative, I am inclined to include those species
Differential Diagnosis Abudhabia is presence in
tines, but that lack
(I)
the enterocone in the upper
and ( 3 ) have more or less opposing cusps in the cheek teeth,
as
gerbils (
=
Gerbillinae and Taterilli
((Myocricetodon» ;nag ((Myocricetodon)) ultimus
nae ) . That decision refers to
nus Jaeger,
1 9 7 7 and
Jaeger, 1 9 7 7 . Gerbillinae s . l .
Abudhabia
d e Bruijn a n d Whybrow, 1 994
A.budhabia
of a central posterior cusp in
Abudhabia
is among fossils most similar to the
((Myocricetodon )) mag ( Morocco) , ((Myocricetodon))
insufficiently known species
nus from Pataniak ultimus from Kendek-el-Quaich ( Morocco) and ((Protatera )) sp. nov. (Munthe, 1 9 82 ) from Sahabi, Libya. It is not impossible that some or all of these
Abttdhabia as «Protatera)) kabulense Sen, 1 9 8 3 .
species vvill eventually be referable to is the case with
Abudhabia baynunensis d e
Bruijn an d Whybrow,
1994 ( figs 1 5.2-1 5 . 5 )
Material and Measurements (tables 1 5. 1 and 1 5.2) M\ AUH 572, 591 (damaged) (site S 1 ) , AUH 573 (site :Nt2 , AUH 574, 576, 592, 593, 594 (damaged) (site S 1 ) ; M3, 575 M1), AUH 596; 1M\ AUH 779 (site S1); M�, 567, 781 (fragment) (site S4), AUH 566 holotype , 597, 598, 599 (damaged) (site S 1 ); M 2 , AUH 568, 600, 755, 756, 765 ( damaged) (site S 1 ), AUH 782 (site S4); M3, AUH 569, 570, 766, 767, 775 (site S1), AUH 783 (site S4)
Remarks With the exception of the M 1 ( fi g . 1 5 .2 ) the mate
Remarks The additional material collected in 1 9 9 5 shows that the original diagnosis and differential diagnosis of Abudhabia need slight changes.
Emended Diagnosis Abudhabia is
Taterillus among the
the M 1 and of remnants of the anterior cingulum in 2 the M and M2 .
that were previously described as myocricetodon molars; ( 2 ) h ave very reduced longitudinal ridges;
most similar to
extant Gerbillinae. These two genera differ by the
a medium-sized rodent with dental
rial of A.
baynunensis collected in 1 99 5
is similar to
the specimens described ( de B rnijn and Whybrow, 1 9 94 ) . The stage of wear of the complete M 1 from site S 1 -low is somewhat more advanced than in the specimens figured on plate 1 , figures 7 and 8 of de Bruijn and Whybrow ( 1 994 ), which may be the rea son that it does not show a posterior cingulum and
characters that are intermediate between the Gerbil
has an incipient connection between the anterocone
!inae and the Myocricetodontinae. The M 1 always
and the protocone. In both these respects this tooth
3
2
� r _Ai,.
....
I
r �
'
.��
�
I
'"
.
'l
"' '
I 11/JA . :. .I ! ' �
�:'!
i
'·
-
\
\
..r
I
'.Jt'•
..
:•
\
�
I
•.
·
.
"\:
. -. �--�,
· >
4
�-' 5
2 Figures 1 5 .2-1 5 . 5 . Abudhabia baynunensis. 1 5.2, M 1 , AUH 573; 15.3, M , AUH 593 ; 15. 4, Ml> AUH 597; 15. 5, M2, AUH 568. A1l x40 and figured as left teeth regardless of which side the tooth really is. Figures 1 5 .4 and 1 5 . 5 are from the right side.
- H.
DE BRU!JN
Table 1 5 . 1 . Summary of the small mammals collected from the Baynunah Formation by site and taxon; the numbers refer to almost complete cheek teeth from rodents and to molars and incisors for the Soricidae
Small mammal taxa
Slundivided
51low
51high
S lclay-balls
51trench
Zapodinae gen. et sp. indet. Soricidae gen. et sp . indet. Total from sites
54trench
Ml
K1
l
Thryonomyidae gen. et sp . indet.
Abudhabia baynztnensis Parapelomys cf. charkensis Dendromus aff. me/anotus Dend1·omus sp. .�-Myocricetodon sp.
54undivided
5 2
7
4
3
1
1 1
2
l
2 1
2 l
1 2
1 2
22 6 7
1 1 1
1
l
3 5
1
1
l 12
13
Total
5
2
4
4
2
4
2
3
2
48
Sites: 51 Shuwaihat on the western flank of the eastern jebel; 54 Shuwaihat on the eastern flank of the jebel known as Jebel Mershed (sec Whybrow and Clements, 1999-Chapter 2 3 ) , about l metre above site 56, the proboscidcan excavation (sec Tassy, 1999-Chapters 1 8 and 24; Andrews, 1 999); Ml Jebel Mimiyah (Al Mirfa); K1 Kihal. =
=
=
is different from the specimens collected earlier. The
development of longitudinal connections at advanced stages of wear is common among Gerbillinae, but the presence of a posteromedial cusp in the M 1 is unex pected because this feature is lost at an early stage of evolution in the Myocricetodontinac, the alleged
ancestral group of the Gcrbillinae. The material avail able, however, is insufficient to be able to decide whether the posteromedial cusp is present in unworn M 1 and lost due to abrasion or whether the M1 figured on plate 1 , figure 8 of de Bruijn and Why brow ( 1994) has an aberrant morphology.
Table l 5 .2. Dental measurements of Abudhabia baymmensis de Bruijn and Whybrow, 1 994 (units
=
0 .I mm)
Width Min.
Max.
Mean
n
Mean
Min.
Max.
15.5 8 .2 22.2 1 5 .0 7.6
1 5 .9 9.7 24.6 1 6.7 9.8
24.7 1 5 .8 8.8 2 3 .4 1 5 .9 8.7
2 4 3 4 3 6
17.5 1 6.0 1 0 .2 1 5 .2 1 5 .2 10.5
1 5 .4 9.5 14.5 1 3 .7 9.5
17.5 1 1.0 1 5 .9 1 6 .0 12.0
Ml
l'vrZ
M3 M1 M2 M3
n
=
number o f teeth measured.
SMALL MAMMALS
Abudhabia baynunensis is of particular interest because its dental characteristics are intermediate
FRO�! THE B.\\o;l-�.'.H FOR..\L\TI Ol'
161
tine, whereas th e true gerbils seem to have origi nated in Mrica.
between the Myocricetodontinae and the Gerbilli nae. Structurally Abudhabia makes a perfect ances
Family Muridae Gray,
tor for Tateril!us. It is peculiar that Abudhabia bay
nunensis is in many respects more primitive than
Murinae Gray,
the much older ''Myocricetodon >> magmts from the Upper Aragonian of Morocco. This suggests that
Parapelomys cf. charkensis Brandy, ( figs
tition from the Myocricetodon-type of dentition endorsed by the "Protatera>> material that has be come known from the Upper Turolian of south ern Spain. These specimens, from strata that are roughly of the same age as the B aynunah Forma tion, are different in having strong longitudinal ridges and prismatic cusps and obviously belong to a different group of Gerbillinae. The polyphyly of the Gerbillinae s . l . is further enhanced by including the genus Pseudomeriones>
1821 1978
Parapelomys Jacobs,
the evolutionary development of the gerbil den has taken place several times. This conclusion is
1821
(Old World rats and mice )
15.6
and
1979
1 5 .7)
Material and Measurements (table 1 5. l ) One M 1 (23.3 x 1), AUH 776 (damaged) (site Sl); 1 M2 (16.8 x 1 5 .9), AUH 5 8 1 (site M 1 ) plus three other specimens: AUH 757 (site 51), AUH 582 (damaged) (site M1), and AUH 583 (dam aged) (site 54); 1 M2 ( 1 5.7 x 1 5 . 1 ), AUH 579 (plus AUH 758, too badly damaged to measure) (site 51) and 1 M3 ( 1 5 .2 x 13.0), AUH 580 (plus AUH 777, worn) (site S l )
Remarks The specimens from the Baynunah Formation re
because the fossil record of this genus strongly sug
ferred to the Murinae represent one species. Their
gests that it derives from an Asiatic myocricetodon-
simple dental pattern, the large difference in size
7
6 2 Figures 1 5 .6 and 1 5 . 7. Parapelomys cf. char!zensis. 15. 6, M , AUH ured as left teeth although both are from the right side.
5 8 1 ; 15. 7,
M2 , AUH
579 .
Both x-4:0 and fig
lflJ
H. DE
BRUIJN
between the t1 and t3 of the M 2 , the absence of a posteroloph in the .rve, and the inflated t2, t5 , and t8 of the M 1 clearly show that these specimens rep resent a small rat. Although the material is poor, its similarity in size and dental pattern to P. charkensis is striking. I therefore list these specimens as Para pelomys cf. charkensis. Dendromurinae Alston, 1 8 76 (Mrican forest mice) Dendromus Smith, 1 829 Dendromus aff. melanotus Smith, 1 8 34 (figs 1 5 .8-1 5 . 1 1 ) Synonym : Dendromus sp . 1 ( de Bruijn and Whybrow, 1994)
Material and measurements (tables 1 5. 1 and 1 5.3) M 1 , AUH 584, 587 (site Sl ) ; M2, AUH 759 (sire Sl); Ml> AUH 760 (site S l ); M2, AUH 585, 588, 761 (sire Sl)
Remarks These teeth, which represent a small species of Dendromus, have about the same size and relative dimension as the extant species D. melanotus. The fossil cheek teeth show minor differences in details of the dental pattern. The anterocentral cusp in front of the anterocone of the M 1 is smaller and the paracone-metacone connection of the M 1 is absent. Similarly the endolophid between the meta conid and the entoconid is absent in the M1 and M 2
9
8
, ·
,;, ·
.
. . . ,
�
.
l
. \"'- j -,.
10
.
..
�.
11
Figures 1 5 .8-1 5 . 1 1 . Dendromus aff. melanotus. 15.8, M 1 , AUH 5 87; 15. 9, M2 , AUH 759; 15. 1 0, M l > AUH 760; 1 5. 1 1, M2, AUH 76 1 . All x40 and figured as left teeth regardless of which side the tooth really is. Figures 1 5 . 8 and 1 5 . 1 0 are from the right side.
SMALL MAMMALS FROM THE BA\c;'l.ll'\,\H
fO&\ lAT!ON
�
Table 1 5 . 3 . Dental measurements of Dendromus aff. melanotus; synonym Dendromus sp. 1 ( d e Bruijn and Whybrow, 1 994) (units = 0.1 mm)
Width
Length
Ml
M2 M1 i\1
2
Miri .
Max.
Mean
n
Mean
Min.
Max.
1 5 .7
16.8
9.2 8.3 8.1 8 .4
9.3
1 0.0
2 1 1 3
9.0
9.5
16.3 9.4 1 3 .2 9.8
7.9
9.0
11 = number of teeth measured.
and the labial cingulum of the M1 is much weaker in the material from the Baynunah Formation than in the extant material that I have seen . The material is very limited, however, so I can not verify whether or not the differences from the extant species are consistent and of sufficient impor tance to define a new species. I prefer tentatively to assign the fossils from the Baynunah Formation to the extant species D. melanotus because of their sur prising similarity. True dendromw-ids apparently developed much earlier than hitherto thought. Dendrormts
sp . ( figs 1 5 . 1 2 and 1 5 . 1 3 )
Synonym : Dendromus sp. 2 ( de Bruijn and \Vhybrow, 1 994)
12
Material and Measurements (table 1 5. 1 ) 2 One M ( 1 1 .4-1 1 .7 x 1 0.3-10.8), AUH 586 (site Sl); ( 12.9 x 1 1 .5), AUH 762, 772 (site S l ) .
two
M2
Remarks The two teeth referred to Dendromus sp . differ sufficiently in size and morphology from D. aff. 11ulrwotw to establish the presence of a second species of Dmd1"omus in the B aynunah Formation. The central sinus of the M 2 and the sinusid of the M2 are more transverse, the metaloph of the M2 is not connected to the metacone, and the posterior cingulum is developed as a cusp. These differences relative to Dmd1·omw aff. 111elrmotw and their about 25% larger size are far beyond the range of
13
Figures 1 5 . 1 2 and 1 5 . 1 3 . Dendromus sp . 15. 12, M2 , AUH 5 8 6 ; 15. 13, M2 , AUH 772 . Both x40 and figured as left teeth regardless of which side the tooth really is. Figure 5 . 1 3 is from the right side.
E
H.
DE
BRUI)N
variation expected in one population even when we
Baynunah Formation seem to document a new
allow for some minor age difference between these
species of Myocricetodon. The available material is,
fossils. The material is insufficient for identification
however, too scant to serve as a basis for a new name.
at species level. Dipodidae Fischer, 1 8 1 7 (jerboas and jumping mice) Myocricetodontinae Lavocat, 1 9 6 1
Zapodinae Coues, 1 8 75
Myocricetodon Lavocat, 1 9 5 2
Zapodinae gen. et sp. indet. ( fig. 1 5 . 1 7)
Myocricetodon sp . nov. (figs 1 5 . 14-1 5 . 1 6 )
Material and Measurements (table 1 5. 1 ) Two M1 ( 14.0 x 8.3, ± 13.8 x 1 0 . 1 ) , AUH 773 (site $ 1 ) , AUH 589 (damaged) (site Ml), AUH 778 (roo badly damaged to measure) (site Sl); rwo M2 (9.5 x 9 .0, 9.2 x 9.0), AUH 763, 780 (site S l ) .
Material and Measurements (table 1 5. 1 ) One M1 ( l l . S x 6.9), AUH 774 (site Sl).
Remarks This small, damaged tooth from site S 1 - high is inter
Remarks
preted as a right first lower molar of a zapodine. The
1 The three M s show that the Myocricetodon from the B aynunah Formation had a weakly split anterocone, an enterocone that is either very weak or absent, and a short posteriorly directed protoloph. The Bay nunah specimens are small and within the range of
M. parvus Lavocat, 1 9 6 1 from Beni Mella! ( Morocco) but differ from that species in not having the protocone and paracone aligned and confluent. The two M2s both have a strong anterolabial cingu lum and, for a myocricetodontine, a distinctly bun
peculiar position of the metaconid in front of the protoconid is, in fact, more similar to the configura tion seen in Allactaginae than in Zapodinae, which usually have a more symmetrical pattern of the ante rior part of the M 1 . The ectomesolophid and the hypoconulid of the M1 from the Baynunah Forma tion are strong. The tooth has these characteristics in common with many Zapodinae and Allactaginae. The anteroconid of the M1 is absent. The entoconid is broken but was probably high.
odont dental pattern. The posterior cingulum is
Soricidae Gray, 1 8 2 1 ( shrews )
developed as an individual cusp, a configuration that
Soricidae gen. et sp. indet.
is also seen in M. trerki and M. ouedi from the Valle
(see de Bruijn and Whybrow, 1 994)
sian of Oued Zra (M orocco ) . The teeth from the
15
14
M\_
16
Figures 1 5 . 14-1 5 . 1 6 . Myocricetodon sp . 15. 1 4,. AUH 773 ; 15. 1 5, M 2 , AUH 763; 15. 1 6, M2 , AUH 7 8 0 . All � 40 and figured as left teeth regardless of whtch stde the tooth really is. Figures 1 5 . 1 5 and 1 5 . 1 6 are from the nght stde.
SMALL MAMMALS
17 Figure 1 5 . 1 7 . Zapodinae gen. et sp . indet. M1, 774, figured as a left tooth x40 although it is rrom the right side.
:\UH
FROM THE BAY �L'>.".-\l-1
FO!t.\L\TION
lfE
The zapodine M1 is among the comparative material available to me most similar to specimens from the early Miocene of Turkey. The zapodine su bfamily has many representatives that have con servative dental characteristics, which makes them inappropriate for biostratigraphy. In conclusion, the presence of a small rat and of two species of Dendromus suggests that the Bay nunah Formation cannot be older than late Turo lian. An early Ruscinian age cannot be excluded.
B IOGEOGRAPHY Materia l (table 1 5. 1 ) Three incisors: AUH 577, 578 (site Kl), AUH 764 (site Sl).
B IOSTRATIGRAPHY The age of the small association of rodents from the Baynunah Formation is difficult to assess at pre sent because the faunal succession of the Arabian Neogene is virtually unknown (Whybrow et al. , 1 982 ) . The presence o f the small rat Parapelomys cf. charkhensis suggests that the fauna is not older than late Turolian. A late Turolian age is also sug gested by the stage of evolution of Abudhabia bay1l1tnensis because the dental pattern of this gerbil is close to A. kabulense from the Lower Ruscinian of Pul-e Charl2, ffith a single small mental foramen. A short diastema separates the dP2 from the partially pre· served alveolus of the deciduous canine. The crown of the dP2 is slender and noticeably ffider across the talonid than at the front due to a slight lingual expansion of the talonid. The principal cusp is low and has anterior and posterior longitudinal crests. A large antetior accessory cusp is positioned slightly lingually at the rront of the tooth, while a small posterior accessory cusp is situated on the rear of the principal cusp. The fragment of the dP3 indicates it had a large, well-separated ante1ior accessory cusp as well .
Figure 1 7 . 3 . Plesiogulo praecocidem, right mandible fragment, AUH 702, lateral view.
having a relatively larger and more separated ante rior accessory cusp. The size of the fossil. hO\\'cTer, precludes reference to Crocuta crocJJtn or. in de ed . almost all known hyaenids. Juvenile mandibles of Plioc1rocuta perrieri1 as described and ill ustrated by Viret ( 1 954), have a generally si milar dP2 morphol ogy, although they appear to be narro,,·er J.ll ha,·e l ess distinct cusps. These mandibles also appear to be significantly smaller, although Viret does n o t give measurements. ·
Genus indeterminate Medium-sized indeterminate species
Table 1 7.2. Measurements (in em ) of dP2 in AUH 294 and Crocuta crocutn
Comparisons The morphology of dP2 is not particularly distinc tive among hyaenas . The fossil's form compares closely to the dP 2 of Crocuta crocuta1 except for
AUH 294 MCZ 25403 MCZ 2 5404
Length
Width
1 . 62 0.86 0.83
0 .90 0 .42 0 . 39
f1!m
f.
C. BARRY
cuta with fully erupted dentition i n the Museum of Comparative Zoology, Harvard collections (MCZ 1 45 5 8 ) . In the fossil the diameter of the greater sigmoid notch is slightly smaller, but the preserved portion of the shaft is longer and more gracile than the corresponding part of the ulna of MCZ 1 4 5 5 8 . Although not determinable to genus o r species, the specimen clearly indicates the presence of a second hyaenid species in the Abu Dhabi faunas.
Felidae Machairodontinae Genus and species indeterminate
Material AUH 24 1 , a complete ti ght calcaneum, and A U H 202, a last lum
bar vertebra. Sires: Kihal, Kl (AUH 24 1 ) and Ras Dubay'ah, R2
(AUH 202).
Description
2 cm
Plate 1 7 . 1 . Hyaenidae gen. indet. , very large inde terminate species, fragment of left mandi ble, AUH 294; lateral view.
The specimen belongs to a large species, about the size of Panthcm
leo or Panthera tig1'is. It is broad for its length (fig. 17.4 ). The
antetior and posterior parts of the inferior sustentacular fucet are
continuous, ri ot separated as in many extant felids. Both facets are separated /Tom the superior sustentacular facet by a wide sul cus.
The mediaUy directed process supporting the posterior part of the
Material AUH 370, an incomplete right u.lna. Site: Jebel Mimiya.h
Mirfu), M l .
inferior sustentacular facet is reduced in size, as is the area of the
(A.l
Description
inferior susentacular facet. The supctior sustentacular fucet is lower
a.t1d flarrer th an in Panthera leo. There is a large, distinct navicular f:'lcet, with is set at a 45° angle to the cuboid fucet.
AUH 202 comprises the centrum and dorsal arch, with the
bases of the spinal and transverse processes. It is wide relative to its
length, with the distance between the posterior zygapophyses being
AUH 370 comptiscs a tight ulna, lacking d1e olecranon process
greater than that between the ante1ior zygapophyses. Bod1 features
shows evidence of postmortem abrasion i.n the rounding of the
tion of d1e base of the n·a.t1sverse process indicate that it is a fetid,
and about one- fifth of the distal end of d1e shaft. The specimen
suggest it is a last lumbar vertebra. The construction and Oiienta
broken ends and polished surfuce, and perhaps gnawing by a carni
not a large hyaenid. It is of a size to fit the calcaneum.
Cmcttta crowta. The radial tubercle is much reduced, while the coronoid process is mediolarerally compressed and extends anteri orlv onto the dorsal surf:'lce of the shaft as a supporting buttress.
Compa risons
I'Orc. The ulna is gracile, suggesting an animal slightly smaller than
The articular surface for the head of d1e radius is a shallow arc,
indicating that supination of d1e radius was resnicted. The greater
sigmoid notch co mp rises a narrow trochlear articulation and a more c"ensi,·e capitula.t· articulation. There is a large, deep fossa just .m tctior ro the radial rubcrcle. Distally the shaft is broken just prox im.11 rn rhc expansion for rhc styloid process .
Co m pa risons hyaenid in the con nf the coronoid process and radial tuber compares i n size to a subadu lt O'ocuta. C1'0 -
T h � s p eci m e n is de monstrably for m a t i o n
cle.
It
The generally heavy construction of the calcaneum and the presence of a distinct navicular facet indi cate that this large felid is a machairodont. The specimen compares in size to Smilodon ( Merriam and Stock, 1932), but differs in morphology.
DISCUSSION OF THE ABU DHABI TAXA Four carnivore taxa are identified i n the collections from Abu Dhabi . They include Plesiogulo praecoci dens; a very l arge, ind eterminate species of hyaenid;
CARJ'IIVORA FRO,\ I THE B.wxc:::-;.".H Fofu\lATION
Bl!lJ
Figure 1 7 .4 Machairodontinae gen. et sp . indet., complete right calcaneum, AUH 241 .
2 cm
a second medium-sized, indeterminate hyaenid; and
markedly different from the taxon at Abu Dhabi.
an indeterminate machairodontine felid. With low
They include Plesiogulo monspess-vtlanus from Europe
population densities and moderately high diversity
and P mafor from China. The oldest Eurasian record
within communities, individual carnivore species
I
tend to be represented in fossil assemblages by one
from Pakistan that is c.
or at most a few specimens. A fossil assemblage
records a possible occurrence at Pikermi in Greece
know for the genus is an unpublished occurrence
9
Ma, while Solounias
(198 1 )
may therefore have a seemingly disproportionate
that may be of similar age. The only Mrican record
number of carnivore species given the paucity of
to date is P monspessulanus from South Mrica at
remains overall. The Abu Dhabi carnivores illustrate
about
this double principle, with four taxa for six speci
4
Ma (Hendey,
1978 ) .
As noted, Plesiogulo praecocidens i s based on
mens. It is also noteworthy that the known taxa are
two specimens, both from Baode in China. Flynn
of medium to large size. Presumably many smaller
et al .
taxa were also present, but were not preserved in
nal analysis that most of the Baode localities are
the fossil assemblages due to chance and biases
late Miocene in age, and are probably older than
against the preservation of small animals.
6 . 5 Ma. The closely similar species Plesiogulo crassa
The documented Abu Dhabi carnivore species
( 1 995 )
have suggested on the basis of a fau
also occurs in China in the Baode and younger lev
are typical of the late Miocene and Pliocene of
els, and is documented in Pakistan as coming from
Eurasia and Mrica. Medium-sized hyaenas and
sediments lying in the lower normal zone of chron
sabre-tooth cats are widely distributed by the late
C3A of the magnetic time scale, that is between
Miocene, about
and 6 . 3 Ma.
9
or
1 0· million
years ( Ma) ago.
6.6
Large species of both groups, however, more typi
The Abu Dhabi carnivores as a whole do not
cally are not present until the end of the Miocene
suggest any particular habitat. Plesiogulo is similar
or even within the Pliocene-that is, in the time
to, and possi bly an ancestor of Gulo, which has usu
range between
ally been considered as indicative o f boreal forest or
7
and
4
Ma.
The appearance of Plesiogulo near
6
Ma ago is
woodland. Harrison ( 1 9 8 1 ) , however, has pointed
one of several events used to define the base of
out that wolverines are also animals of open tundra,
the late Hemphillian Land Mammal Age in North
and that North American species of Plesiogulo are
America, where the genus has a restricted time range
most abundant in fossil assemblages with hypsodont
( Harrison,
19 8 1 ;
Tedford et al. ,
1987) .
In the Old
World, however, the genus is both younger and older, with an approximate time range from
9
to
and cursorial ungulates. She interprets such assem blages as indicating grassy and open habitats, which
4
Ma ago. The early Pliocene species are larger and
might as well characterise the habitats of the Abu Dhabi fossils.
litml
J, C. BARRY
ACKNOWLEDGEMENTS
1932. The Felidae of Publications of the Carnegie Institu tion of Washington 422: 1-2 3 1 .
Merriam, J. C . , and Stock, C. Rancho La Brea.
I thank Phil Crabb ( Photographic Studio, The Nat ural History Museum, London) and AI Coleman
198 1 .
(Peabody Museum, Harvard University) for pho
Solounias, N .
tographs of the specimens.
Pikermi: Resurrection of a classic Turolian fauna.
Mammalian fossils o f Samos and
A nnals qf Carnegie Museum 50: 231-69.
REFERENCES
Tedford, R. H., Skilmer, M . F. , Fields, R W., Rens
Flynn, L. J., Qiu, Z., Opdyke, N., and Tedford, R H.
1 995. Ages ofkey fossil assemblages i n the Late Neo gene terrestrial record of northern China. In Geo chronology, Time Scales, and Stratigraphic Correlation. SEPM Special Publication no. 54: 365-73. Society of Economic Paleontologists and Mineralogists, Tulsa. Harrison, J. A.
1981.
A review of the extinct wolver
America. Smithsonian Contributions to Paleobiology
46: 1-27. 1 978.
Late Tertiary Mustelidae (Mam
malia, Carnivora) from Langebaanweg, South Africa. Annals of the South African Museum, Cape Town
76:
329-57. Kurten, B .
1 970.
The Neogene wolverine Plesiogulo
and the origin of Gulo ( Carnivora, Mammalia). Acta Zoologica Fennica
1 3 1 : 1-22.
1 987.
through Hemphillian interval (Late Oligocene through
Ceno zoic Mammals ofNorth America, pp. 1 53-2 1 0 ( ed. earliest Pliocene Epochs) in North America. In
M. 0. Woodburne) . University of California Press,
Berkeley.
ine, Plesiogulo ( Carnivora: Mustelidae ) , from North
Hendey, Q. B.
berger, J, M., vVhistler, D . P., and Galusha, B . E.
Faunal succession and biochronology of the Arikareean
Viret. J.
1 954. Le l oess
a banes durcis de Saint-Vallier
(Drome) et sa fatme de mammiferes villafranchiens.
Nouvelles Archives du Museum d 'Histoire Naturelle de Lyon 4: 1 -200. Zdansky, 0. 1 924. Jungtertiare Carnivoren Palaeontologia Sinica C, 2: l-1 55.
Chinas.
Miocene Elephantids (Mammalia) fi�om the Emirate of Abu Dhabi, United Arab Emirates : Palaeobiogeographic Implications PASCAL TASSY
The first
of an elephantoid in the Neo-
Other localities have yielded elephantoid re
gene of Abu Dhabi was an isolated tooth found at
mains. One edentulous juvenile mandible (AUH
Jebel Barakah by Glennie and
( 1 968 ) and
475 ) can be allocated to
identified by Madden et al. ( 1 9 8 2 ) as
Stegotetrabe
isolated teeth ( except one fragment, which is a
lodon grandincisiV7tm
v\'hybrmv, 1 9 8 9 ) . Later,
Peter Whybrow and A.ndrew Hill visited Abu Dhabi and collected fossils in the B aynunah
at Jebel
B aralVith the palatine plane is about 5 0-60 ° ) .
be measured with precision because its posterior face is crushed ( it is estimated to be 3 7 0 mm). The dorsal face of the mandible, especially the mid
The measurements o f the cranium (in m m ) are as follows:
observed during this study.
to the choanae
300
L:ngth of basicranium fi·om the choanae to the foramen magnum
Width of the basicraniu m taken at the lateral borders of the glenoid fossae
Maximum \\'idth of the choanae
External maximum width of the palate
Internal \\�dth taken between the first loph of the M2s
Internal width taken between the fourth loph of the M\
L:ngth of the basicranium from the condyles to the area of the pterygoid process
its plaster jacket (including the occlusal surface of the molars ) , and therefore this face was not
Palatal length fi·om the anterior border of M 1
Maximum \\�dth taken at the zygomatic arches
portion, is heavily wead1ered and still embedded in
300
c. 820 c.
820 90 265 58 75
c. 350
The upper tusks are long, straight, oval in cross-section , and devoid of an enamel band ( figs 1 8 . 5 and 1 8 . 6 ) . The right tusk is the best preserved and its tip is nearly complete. The overall length is 1020 mm. The cross-section is 8 5 . 6 x 70 . 3 mm at 270 mm from the tip . The right lower tusk is pre served. It is long, protruding out of the symphysis. The visible length of the tusk is 5 3 0 mm, longer than the symphysis length, which is estimated at
ELEPHANTIDS !'ROM THE BAYNUNAH fOR.\L\TION
PJ:B]
Figure 1 8 .4. Stegotetrabelodon syrticus, manclible, AUH 503 (same inclividual as AUH 502); ventral view. Scale 20 em. =
c . 370 mm. The cross-section i s oval with a longi
show a primitive plate- like pattern typical of prim
tudinal dorsal sulcus
itive elephantids-that is, a mosaic of primitive
(
=
subpiriform cross-section)
( figs 1 8 . 7 and 1 8 . 8 ) . The dentine of the lower tusk
( " gomphotheriid" traits) and derived (elephantid)
has a concentric la mellar structure . The intermediate molars ( M 1 and M 2 ) belong
traits.
to the tetralophodont grade ( fig. 1 8 .9, tables 1 8 . 1 and 1 8 . 2 ) . M 1 is tetralophodont sensu stricto. M 2 is
cusps forming the lophs are more or less of equal
nearly pentalophodont; its postcingulum is strongly
De rived elephantid traits are as follows: the size, including the mesoconelets ( conelets that are close to the meclian sulcus); the cusps are trans
inflated, forming a narrow and incipient fifth loph
versely aligned-when marked, the wear of the pre
but this loph is much lower than the fourth and is
trite and posttrite parts is nearly equal so that it
only followed by an enamel bump at the cervix (on the right M2 only ) . In conclusion, the fifth loph of M 2 is not complete and is a perfect intermecliate
photheres, a typical trefoiled one ( especially clear
between tetralophodont and pentalophodont
forms a plate-like wear figure, not, as for gom on the worn M 1 s ) ; cement is plentiful. Primitive "gomphotheriid" traits are as follows: the cusps are rather low; the persistence of a median
molars. Five lophs of the left M3 are visible on the labial side but the posterior part of the molar lies in the bone so that it is impossible to ascertain
sulcus; the lophs are made of few cusps-four cusps
of M 2 ( two on each half-lop h ) , except the fourth of right M 2 , which shows five apical digitations ( three
the complete ridge formula (fig. 1 8 . 1 0 and tables
on the pretrite half-loph ) , five to six on M3 ( three
1 8 . 1 and 1 8 . 3 ) . The lophs of the upper molars
on the pretrite half and two on the posttrite, except
20 cm
Figure 1 8 . 5 . Stegotetrabelodon syrticus, right upper tusk of cranium AUH 502; lat eral view.
e
P. TASSY
Figure 1 8 . 6 . Stegotetrabelodon S)Wticus, cranium A UH 5 0 2 . Cross-sections of the right upper tusk taken at 70 mm from the tip (A) and 270 mm from the tip (B); rear views. Scale = 2 em .
3 cm
Figure 1 8 . 7 .
Stegotet1'abelodo11 syrticus,
taken at 5 0 0 mm from the tip .
20 cm
Plate 1 8 . 1 . labial view
Stegotetrabelodon. S)Wticus,
right M_,,
(front is righ t ) . Scale bar = 20 Cin .
mandible
AUH 5 0 3 . Cross-section of the right lower tusk
AUH 456: top, occlusal view; bottom,
Figure 1 8 . 8 . A> Cross-section of the lower tusk of Stegotetrabelodon syrticus from Shuwaihat, AUH 5 0 3 , com pared with (B) cf. Stegotetrabelodon from Jebel Semene; C> Elephantidae indet. ( cf. Stegotetrabelodon seu Pri'frzelephas), Lukeino; D> Tetralophodon longirostris> Grossweissendorf; E> "Mastodon» grandincisivus> Maragheh holotype); F, Pestszentlorincz; G> Kertch; H> Sahabi. Rear views are (B) from B ergounioux and Crouzel 1956), (D) from Steininger ( 19 6 5 ) , (E) redrawn from Schlesinger ( 19 1 7), (F) redra\vn from Schlesinger 1 92 2 ), (G) redrawn from Pavlow ( 19 0 3 ) , and (H) from Gaziry ( 1 987). Scale 5 em.
Table 1 8 . 1 . Measurements (in mm) of the molars of the Shuwaihat cranium
rt M 1 1 lt M 2 rt M It It
p
L
w
4 4 4-5 4�5 5+
l l 9.0 l l6 . 7 1 74 .4 1 76.9
7 3 . 6( 3) 72 . 1 (4) c . 1 0 1 (4J 1 0 1 .4 (4 ) c . 1 0 3 (4 )
H
c .46 (4) 50.2 5 9 . 7( 4 )
LF
ET
3 .9 3.8 3.0 2.9 2 .9
4.0
HI
c .4 5 . 5 49. 5 60.3
Table 18.2. Comparative measurements (in mm) of M2 ----·--- ···---
Shuwaihat 5. syrticus S.
orbus
p
L
w
H
LF
4�5 4-5 5
1 74 .4-1 76.9 1 5 8 .0-172 . 0 1 4 5 .0-174.0
1 0 1 .4 96.0 7 5 .8-9 3 . 1
5 0.2 49.0 60.9
2 .9-3 . 0 3 .2-4.0
ET
HI
-19.5 5 . 0-6 .-1
/0 . 0
For key t o abbreviations see Table 1 8 . 1 . Som-ces of data: S. syriticus from Sahabi ( Gaziry, 1 9 8 7 ) ; S . orbus from Lothagam and �\du-.\sa .\ laglio, 1 9/3; Kalb and Mebrate, 1 99 3 ) .
�
P. TASSY
for the fourth loph which has three on each half lop h ) ; the persistence of the pretrite conules, although they are weakly developed, seen on M
2
(on
d1e posterior face of the second loph and, as a bump, on ilie anterior face of d1ird loph) and on M3 ( one on the posterior face of the first loph and two on the posterior face of d1e second loph ) ; and the shape of the lophs convex/ convex ( according to the criteria of l M2 , and partial right M 2 , AUH 475 ; left, occlusal view; right, lateral view.
of the tooth. True pentalophodont M 2s ( that is
I n any case, the shape and height of the unworn
with a postcingulum separated from the fifth loph )
upper M3 of ilie skull from Shuwaihat (estimated
are not present in primitive tetralophodont gom
height index is only slightly lower ilian in S. syrticus
photheres, and are common at Sahabi. But an iso
fro m Sahabi) , and the morphology of ilie isolated
lated M2 from Sahabi with exactly the same formula as that of the M 2 of the skull from Shuwaihat is
any geological age. Although molars of S. syrticus
described by Gaziry ( 1 9 8 7 ) .
wiili loph(ids) showing numerous apical digitations
Two traits combined on the Shuwaihat molars,
M3 are clearly outside the range of Stegolophodon of
(up to six or seven cusps) are known at Sahabi,
probably primitive, are reminiscent of Stegolopho
simpler molars wiili four cusps on each loph are
don, a genus of tetralophodont grade known in Asia
also found in this locality ( three M2s described by
from the early Miocene up to the Pliocene . The
Gaziry, 1987: 1 94 ) .
molars are relatively low and the lophs are made of few big cusps ( four on tl1e M2s ). Intraspecific varia
The premol ars found in the Baynunah Forma tion seem more primitive ilian iliose described at
tion was already suspected as "Stegolophodon sahabi
Sahabi by Gaziry ( 1987) and at Loiliagam and
anus sp . nov." (one M3 only) was described by Petrocchi ( 1954) at Sahabi and later considered a synonym of Stegotetrabelodon syrticus ( M aglio, 1 9 7 3 ) . The height index is much lower at Abu Dhabi than i n S. orbus from Lothagam (table 1 8 .2 ) .
Mpesida ( Kenya) by Maglio and Ricca
( 1 977)
and
Tassy ( 19 8 6 ) . The posterior loph o f P4 is not plate like, and this tooili could be allocated to any other tetralophodont genus, such as Tetralophodon. The intraspecific variation of ilie premolars of trilopho-
110 mm . '
• 2
90
• 3 A 4
"'
A 5
0 6 8 7
70 H
*
50
H�
30
.j
10
...
o•
• o oo o ' 30
� .:
�. .
� X
0 0
6.
• •
fiE]
Abhandlungen der Bayerischen Akademie der Wis senschaften, Mathenzatisch-Naturwissemchaftliche 2 : 1-22 1 .
P. , and Darlu, P. 1986. "-\nalyse cladistique nu merique et analyse de parcimonie; l'exemple des Ele phantidae. Geobios 19: 587-600.
Smith, A. G . , Smith, D . G., and Funnell, B . M . 1994. Atlas ofMesozoic and Cenozoic Coastlines. Cambridge
---. 1987. Les Elephantidae : nom·eau regard sur les analyses de parcimonie. Geobios 20: 487-9-±.
lJniversity Press, Cambtidge. Steininger, F. 1965. Ein bemerkenwester Fund von Mastodon ( Bunolophodon) lmgirostris Kaup, 1 832 (Proboscidea, Mammalia) aus dem Unterpliozan (Pannon) des Hausruck-Kobernausserwald-Gebietes in Osterreich. ]ahrbuch der Geologischen Bundesanstalt
108: 195-212.
Tassy, P. 1983. Les Elephantoidea miocenes du Plateau de Potwar, Groupe de Siwalik, Pakistan. Annates de
Paleontologic 69: 99-1 36, 235-97, 3 1 7-54.
. 198 5 . La place des mastodontes miocenes de l'Ancien Monde dans Ia phylogenie des proboscidiens ( Mammalia) : hypotheses et conjectures. These Dr. es Sci. Memoires des Sciences de la Te1n, Universite Curie,
---
Paris 85-34: 1-862.
Tassy, P., Anupandhanat, P. , Ginsburg, L., Mein, P. , Ratanastien, B . , and Sutteethorn, V. 1992. A ne\Y Ste golophodmz ( Proboscidea, Mamma lia) in the Miocene of northern Thailand. Geobios 25: 51 1-23. Tobien, H. 1978. O n the evolution o f mastodonts (Proboscidea, Mamma lia), Part 2: The bunodont tetralophodont groups. Geologisches ]ahrbuch Hessm
196: 1 59-208. . 1980. A note on the mastodont taxa (Pro boscidea, Mammalia) of the "Dinothetiensande" ( Up per Miocene, Rheinhessen, Federal Republic of Ger many). lvfainzer Geowissenschaftm Mitteilungen 9 :
---
1 87-201.
1986. Noz1-veattx Elephantoidea (Mammalia) dans le Miod:ne d$t Kenya. Cahiers de Paleontologic.
Tobien, H., Chen, G., and Li, Y. 1988. Mastodonts ( Proboscidea, Mammalia) from the Late Neogene and Early Pleistocene of the People's Republic of China. Part 2 : The genera Tetralophodon, A1tancus, Stegote
Editions du Centre National de la Recherche Scien tifique, Pads.
some generalities on the Chinese Mastodonts. lvfainzer
---
. 1990. Phylogenie et classification des Pro boscidea ( Mamma lia) : histotique et actualire. Annates de Paleontologic 76: 1 59-224.
trabelodon, Zygolophodon, Afa;mmtt, Stegolophodon; Geowissenschaften Mitteilungen 1 7: 9 5-220.
---
. 199 5 . Les proboscidiens (Mammalia) fossiles du Rift Occidental, Ouganda. 1n Geology and Palaeo
--
biology of the Albertine Rift Valley Uganda-Zaire. Vol ume II: Pateobiologie/Palaeobiology, pp. 2 1 5-55 ( ed. B . Senut and M. Pickford). CIFEG Publicational Occasionnelle 1994/29. Centre International pour la Formation et les Echanges Geologiques, Orleans.
. 1996. Who is who among the Proboscidea? In 1JJe Proboscidea: Evolution and Palaeoecology of Ele phwnts and their Relatives, pp . 39-48 (ed. J. Shoshani and P. Tassy). Oxford University Press, Oxford. ---
Whybrow, P. J . 1 9 89 . New stratotype; the Baynunah Formation Miocene) , United Arab Emirates: Lithology and palaeontology. Newsletters on Stratigra
phy 2 1 : l-9. VVhybrow, P. J., Hill, A., Yasin al-Tikriti, w: , and Hailwood, E. A. 1990. Late Miocene primate fauna, flora and initial palaeomagnetic data from the Emirate of Abu Dhabi, United Arab Emirates. ]ozmwl of
Human Evohttion 19: 583-88. Zhou, M., and Zhang, Y. 1983. Occurrence of the proboscidean genus Stegotetrabelodou in China. l 'erte
brata PalAsiatica 2 1 : 52-5 8.
Hipparions from the Late Miocene Bay11unah Formation, Emirate of Abu Dhabi, United Arab Emirates VERA EISEN�L'\N N AND PETER J. WHYEROW
The n ames Baynunah Formation (Whybrow, 1 9 8 9 ) , and the Shuwaihat Formation (VIlhybrow et al. ,
istration numbers of the Emirate of Abu Dhabi
1 999; B ristow, 1 999-Chapters 4 a n d 6) are given
in the i\merican Museum of Natural History,
(AGH ) . Comparative hipparion material is housed
to deposits of mainly clastic, late :Miocene sedi
New York (A...\1NH) , Museum National d'Histoire
mentary rocks outcropping in an area of about
Naturelle, Paris (MNHN) , and Ungarische Geolo
1 6 000 km2 in th e Western Region of the Emirate
gische Reichsanstalt ( UGR) .
of Abu Dhabi, Gnited Arab Emirates. It is likely that the fossil-bearing B aynunah Formation is coeval with parts of the Agha Formation of Iraq (Thomas et al. , 1 9 8 0 ) and Iran (James and Wynd , 1 9 6 5 ) . From the most western tossil site, Jebel
METHODS AND MATERIAL Bone scraps can be found o n the slopes of most
B arakah (fig. 1 9. 1 ) to the most eastern (Tarif), a
of the Miocene exposures. The Natural Histo1y
distance of about 1 5 0 km, and further east to Abu
Museum/Yale University team has attempted to
Dhabi city, the regional dip is merely 1 o east. All
be nonselective in their collecting to reduce sam
sedimentary units are horizontally-bedded; there is
pling problems ( Hill, 1 98 7 ) . All hipparion fossils
no regional folding and therefore no compli cated
were collected by close examination of outcrop
stratigraphy. The sedimentary associations are,
surfaces. The genus Hipparion is represented
however, complex and tremendously variable
among the Abu Dhabi fossils by 41 specimens,
(Whybrow et a!. , 1 999; Friend, 1 999; Bristow,
including 1 3 upper cheek teeth, 14 lower cheek
1 999-Chapters 4-6 ) .
teeth, and 1 4 limb bones, nearly all fragmentary,
Except tor a small collection ( 1 982-8 3 ) made
collected at six localities (fig. 1 9 . 1 ). The sizes of
by Al Ain Museum and German archaeologists
bones and teeth show that two species are repre
(Vogt et al., 1 9 8 9 ) now in Al Ain Museum, Emi
sented: one small or middle-sized, and one larger.
rate of Abu Dhabi , tossil collections from the \.Vest
What makes the sample most interesting is the
ern Region made since 1 98 6 by The Natural His
presence of two mandibular fragments belonging to
tory Museum/Yale University team are temporarily
the smaller species. In size and proportions they
housed in the Department of Palaeontology, The
differ fro m other Hipparion mandibles and warrant
Natural History Museum, London : material col
the description of a nevv species.
lected pre- 1 98 4 has BMNH numbers. A collection
Mandibles and other bones and teeth were
made by Andrew Hill and Walid Yasin in 1 9 8 4 is
measured according to the recommendations of the
stored in AI Ain Museum. All palaeontological
New York International Hipparion Conference of
material collected by the NHMjYal e team has reg-
1 98 1 ( Eisenmann et al. , 1 9 8 8 ) . Two new mandibular
Copyright
ISBN
0-300 - 0 7 1 8 3 - 3
HIPPARJO:-;s FRml
52"
t
N
24°
Emirate of Abu Dhabi
km I
0 52°
40
53°
Figure 1 9 . 1 . Map of the Western Region of the Emirate of Abu Dhabi, showing the locations (Shuwaihat, Hamra, Jebel B aralcah, Jebel Dharma, Ras Dubay'ah, and Kihal ) where specimens of Hipparion have been tound.
measurements were introduced to evaluate the length of a fragmentary muzzle. They are: _\
distance between infradentale
( the point between
the alveoli of the first incisors) and the symphy seal constriction (where the width of the sym physis is smallest) ; and B = the distance between the symphyseal constric tion and the second premolar. The sum of A plus B is the muzzle length. A third measurement comes from the system used by Gro mova ( 19 5 2 ) : G3
=
the distance between the mandibular angle and the front of the second premolar, which gives an idea of size independent of the muz zle length
length plus G3 is the max
imal anteroposterior length of the mandible ) .
Simpson's ratio diagrams (Simpson, 1 94 1 ) were used to compare the sizes and proportions of the Abu Dhabi
Hipparion
mandibles to others.
Unworn or little worn cheek teeth were sectioned at mid-crown to provide evidence of the enamel p attern. The ages of
HippaTion
localities given here are
according to the updated European Land J\lammal Zones ( MN) of Mein ( 1990) . Comparisons \\ ere
made chiefly witl1 hipparions for which the sample s are good enough to give points of comparison
( mandibles, cheek serie s ) , and where the ,1ge is at least approximately known. B ecause of tcl.e geo graphical position of the Emirate of .\bu
Dhabi
in the Arabian Peninsula comparisons " ere also attempted with some African, .\ liddk Eastern. and Asiatic hipparions.
�
V. E!SENMA
Libycochoents rnassai LibycochoentS ldJinzilubirus
Source: after Pickford ( 1 98 8 ) .
basis o f fossils from Lothagam, Kenya ( Cooke and Ewer, 1 972 ) . Cooke later determined that the Lothagam suids were indistinguishable from Siva choerus syrticus Leonardi, 1 952 described from Sahabi in Libya, but more correctly placed witl1in Nyanzachoerus ( Cooke, 1 9 8 7). Nyanzachoerus syrticus is a large snid of the subfamily Tetraconodontinae. In addition to the Lothagam and Sahabi samples, it has been identi fied in Kenya from Ekora, Kanam, and the Lukeino Formation of the Tugen Hills, all of which date to the late Miocene (Harris and \'Vhite, 1 979; Hill et al. , 1 9 8 5 , 1986, 1 992). The species is not cur rently knmvn south of Kenya, and its last well dated appearance is at 5 .6 million years (Ma) from the Tugen Hills ( Hill et al. , 1 992) though White ( 1995) points out that the Sahabi occurrence could
be slightly younger. A smaller and more primitive taxon, Nyanzachoerus devauxi) was described from deposits at Oued al Hamman, bourg, 1 96 8 ) . This provenance is now in doubt, and the holotype is tl1ought to come from Dublineau (Thomas et al., 1 982). Nyrmzachoerus devauxi is similarly distributed in space and time, and is kno,Yn trom Sahabi, the Tugen Hills, and Lothagam ' Cooke. 1987; Hill et al. , 1 992). Most late Miocene African localities presen e maximum of two ta.x a, both tetraconodo::1ts. :-...-. devauxi and N. syrticus. The sole is Sahabi, where the more deri,-ed X. usually found in later deposits in c..lstern _-\iii.::a. also recovered with the t\\'O more forms ( Kotsakis and Ingino, 1 9 79 ; Cooke. 1 9 8 - Since this co-occurrence is unique \Yithin any geo.
�
L BISHOP
A.i'-:D
A. HJLL
logical stratum in Africa, it may imply that the
Owen ,
Sahabi fauna averages a greater period of time or
These pigs, which have complex cusp patterns, are
1 840-4 5 )
and Pilgrim
( 1 926),
respectively.
that ecological conditions were unique at the site.
unlike any of the specimens recovered from Abu
Another possibility is that the Sahabi fauna pre
Dhabi or in any other Old World deposits of similar
serves a stage of tetraconodont evolution unsam
age.
pled in the remainder of the African record .
Europe and the Eastern Mediterranean
Asia
In the middle Miocene, European suid diversity is
Late Miocene deposits of the Indian subcontinent
quite high, with numerous taxa of listriodont,
preserve a variety of suids. The
hyotheriine, and tetraconodont pigs represented
Sivachoerus,
tetraconodont
described by Pilgrim
( 1 926),
is com
mon in sediments of the upper part of the Dhok
(van der Made and Moya-Sola,
1 9 89-90),
1 989; van der Made,
but by the later Miocene there are fewer
Pathan Formation, and apparently occurs in later
species representing iewer subfamilies. In recent
sediments of the Siwaliks ( Pickford,
work, taxonomy has been revised, leading to a
however, been hypothesised by
ture of less suid diversity in the late Miocene
1988 ) . It has, Pickford ( 1 9 8 8 )
that its ancestors are not in the tetraconodont and lophodont forms from the earlier part of the Siwa lik sequence, such as
choerzts)
Tetraconodon
and
Lopho
but rather among the earlier tetracon
odonts of A±hca.
Sivachoe1'US has
heavily crenellated
cingulation of the
premolar enamel and P 4. The third and fourth
of
Sivachoerus
are even larger, relative to the anterior premolars in either
and molars, than are
choerus or
in the Abu Dhabi
Nyanza-
The third molar
( approximating MN zones
l l- 1 3 ) (van der Made 1989; van der Made, 1 989-90; Fortelius et al. , 1 996). Microstonyx, a suine, i s the dominant late
and Moya-Sola,
Miocene genus, present at numerous Eurasian localities. The genus
1 971 ,
Korynochoeru.s Schmidt-Kittler,
also common in the late Miocene and
Pliocene, has been sunk into
Propotamochoerus)
originally named for Siwalik specimens
has an inflated appearance most apparent in the lin
et al. ,
1996). One etruscus) is a highly
gual side of the protoconid/metaconid junction.
form tound only in Tuscany
These characters are apparently absent in the Abu
MN
1 6a
suine pig,
Eumaiochoerus
derived and endemic island
1 9 82 ).
( middle Pliocene),
Dhabi sample. The mandible is also much more
is further diminished with most specimens sub
massive in
sumed into
Sivachoerus than
in
Nyanzachoerus.
Pigs of the subfamily Suinae are also common in deposits of the Indian subcontinent.
choerus hysudricus,
described by Stehlin
common in the
Propotamo ( 1 899), is
and Dhok Pathan Forma
tions and is also known trom Perim Island (now called Piram Island; Pickford,
1988).
This species is
not known from Europe (van der Made and Moya Sola,
1989).
The large suine pig
described by Lydeldcer
( 1 8 77 )
Hippopotamodon
is also known from
(Azzaroli,
By
suid diversity
Sus) a genus of presumably Asian origin 1 975; van der Made and Moya-Sola,
1 989 ) .
MATERIAL FROM THE BAYNUNAH FORMATION So far,
17
sites from
specimens have been recovered a t
6
localities (table
20.2).
10
B oth dental and
India in the Nagri and Dhok Pathan Formations,
fragmentary postcranial remains are known. There
and from sediments that post-date those in the
are at least two taxa represented in the sample. The
Dhok Pathan type area ( Pickford,
1988) .
Later
larger and more common of the two is provisionally
Siwalil but i s larger and better preserved, missing only the distobuccal portion of the hypoconid. The dentine lalzes are roughly star-shaped. There appears to be a basin separating the metaconid and entoconid, and another, cuspule-lined one between the protoconid and the hypoconid . The anterior and posterior cin gula are well developed, but somewhat obliterated during eruption. M3 is a triangular tooth with a small, somewhat waisted talonid. The anterior cingulum appears distinct and well developed. The four main trigonid cusps are strong, and there one major pillar in the talonid, located distolabially. The pillar on the trigonid-talonid junction is also well developed. There are numerous small accessorv cuspules in d1e talonid, and d1e anterior and posterior portions of • the trigonid are separated by labial and lingual indentations and cuspule-lined enamel basins between the major cusps.
aff. Nyanzac!Joe rus syrticus
Specimen AUH 238: Thumayriyah, Site TH 1 Juvenile left mandibular corpus with !3, dP2, dP3, Mb and unerupted C and M2 in the crypt (fig. 20.2 . l a, lb ) . specimen also has a fragment of the ascending ramus, including a partial mandibular angle. It has been damaged considerably during the
taphonomic process, and reconsolidated
hardeners and glue
during recovery. The preserved morphology
compatible with an assignment to Nvanzacboerus S)7'ticus but the morphology is rmdi agnostic because there is no comparative juvenile sample assigned to the species. The morphology of the developing permanent denti tion is tetraconodont in character. Only d1c tip of the canine is sent, and i t exhibits a D-shaped cross-section. The I3 is robust a pronounced lingual crest running down the long axis. The dP 1 is missing throngl1 apparent breakage. dPrdP4 have long mots, exposed through damage, and relatively thin enameL dP2 is long and narrow, with a pronounced central cusp. At the cenix, there is a marked constriction and raised enan1el level betw·een the mesial and distal roots. The a-own of dP 3 is severely damaged but appears to have had the same overall basal shape as dP2. dP4 is well worn and damaged. The enamel thick. There are three pillar pairs M 1 is in full occlusion and exhibits and the tooth broadens a tettaconodont pattern, with strong simple cusps indented by grooves and strong anterior and posterior cingula. There is a small basin joining the protoconid and hypoconid, but morphologv is difficult to discern due to weathering and brealcage. An unerupted M2 is exposed in the The crown had not fully formed. but visible morphology similar to M 1 and lilce f\!,;ar.c?.arhl•en'"
Specimen AUH 328: Thumayriyah, Site TH 1 Large right hemirrcandible in bad condition. Broken tooth �rowns
and roots are apparently present for PrM3. The rr:andibic is ken, very weathered and cracked, with cortical bone bdh d2:::: ;ol!c,i on all surfaces. Some tooth is present, although �:c'le enc..-::e: C:.cs � been exfoliated from aU exposed portions of the ::oo:c': ;::ov.-c:s Tie crowns are still covered in matrix and furL':e:· ·:c�cc.:-":'c::: might reveal more occlusal morphology remains beneath. The mandible is on:� -3 ·:::-:: tl1e region of the ascending ramus. broken :ee::."' we:-e ently massive <md bunodont . The P" anJ absolutely The size of the teeth anJ premolar ate compati ble
-
L. BISHOP
AND
I
A. HILL
la
�----2 em
lb
Figure 2 0 . 2 . Aff. Nyanzachoerus syrticus, AUH 2 3 8 , juvenile left mandible with unerupted C, I3, dP , dP3, dP4 , and M 1 in occlusion and M in the crypt: la, occlusal 2 2 view; l b, labial view.
particularly Nyanzachoerus. Further preparation of this specimen might prove useful for a secure taxonomic assessment.
Specimen AUH 557: Shuwaihat, Site S6
Specimen AUH 784: Ras Dubay' ah, Site R2
labial cusp with thick enamel and relatively deep grooves on the cusp. Cusp relief is not high, suggesting a btmodont tooth. There are preserved fragments of a relatively high cingulum. The morphol ogy and size of the specimen are compatible with Nyanzachoe1'Us. The tooth is not worn, but the enamel is pined and weathered.
This specimen is an unerupted crown fragment, most probably from an M3 according to size, morphology, and enamel thickness. The cusps are relatively low, pyramidal , and have relatively deep grooves. This cusp morphology, as well as the squarish relative posi tion of the cusps, is similar to that of Nyanzachocrus.
Specimen AUH
1 28:
Shuwaihat, Site S2
Lower tight canine fragment, split longitudinally; this is an infero lateral portion. The canine is large. No wear is apparent on the pre served fragment and the dentine is absent so the tooth might be unerupted. The preserved portion of the canine suggests that it is suboval in cross-section. The enamel is slightly rugose and the tooth has a longitudinal ridge, similar to that seen on some speci mens of Nyanz.achoerus.
P4 fragment. This tooth is very fragmentary, but appears to be the
Specimen AUH 690: Shuwaihat, Site S4 Left P3 fragment consisting of the mesial portion of the rootl1
crown. The tooth is broken distal to tl1e indented portion of the cervix. The fragment appears to be relatively narrow, and because it is only a mesial fragment the breadth reported in table 20.3 must be considered a minimum value. The anterior cingulum is pro nounced with a very jagged appearance on the cranial border. The ridges on the mesial surface of the tooth, which connects the ante rior cingulum with the apex of the protocone, is raised and marked, separated off from the buccal face of the tooth by a depressed line .
Scrvs
in the enameL Several observations suggest that the tooth might have been unerupted. The central portion of the tooth is missing and there is a large cavity in the dentine. No roots are preserved. There is no apparent wear on the preserved portion, and the ante rior cingulum bears no contact facet.
Subfamily S uinae Gray, 1 8 2 1 Genus Species
Propotamochoerus Pilgrim, 1925 Propotamochoerus hysud1·icus (Stehlin, 1 899)
Specimen BMNH M49433: Jebel Barakah, Site B2 BMNH M49433 is a right mandibular corpus bearing PcM3 (figs 20.3 . l a, lb). The inferior border of the corpus is not preserved. The bone is missing anterior and inferior to the distal roots of the P4 , which are exposed in this specimen. The corpus is also broken posterior to the M3, preserving only the initial rise of the lateral
(buccal) portion of the ascending ramus. The specimen has consid erable postmortem damage to the bone at the cranial border of the alveoli; this exposes considerable areas of the tooth roots. Much of the enamel has been lost through wear during life combined with postmortem damage.
The individual was a mature adult, with M3 in full occlusion, and all the teeth are well worn. The teeth appear to be thick-enam elled, bunodont, and of relatively simple occlusal morphology. There
is no indication oflophodonty or other dental specialisation. The mesial portion of P.1 is missing, but the distal portion has a typical suine pattern of morphology and, apparently, of size relative to the molar row. P4 has a large and pronmmced posterior cingulum set off from the central conid by deep, V-shaped lingual and buccal furcula. Unfortunately, its morphology cannot be assessed completely due to brealcage, which contributes to some level of uncertainty in attribut ing the specimen to Propotamochoe-rus hysudricus.
The lv11 appears relatively small, although it is worn and bro ken to below the cervix. All occlusal morphology is obliterated. M2 is in a sirrjlar, although less state. Brealcage has exfoliated much of the enamel. The
that remains suggests it was rela-
tively thick, and that there was a well-developed basin joining the protoconid and the hypoconid on the buccal side of the tooth. M3 is the best preserved of the teeth , although it is missing a large por tion of the enamel from the mesiobuccal aspect of the protoconid. It has a small anterior cingulum, and the entoconid is worn to a mushroom-shaped dentin e !alee. The cusps of the tooth are conical and isolated, and enamel is thick.. A cuspule-bordered basin joins the hypoconid and the protoconid, apparently flaring lingually. Cusp morphology seems simple, without deep grooves dividing the surface of the individual cusps. Because the tooth is worn and abraded, these features might have been lost through wear or post mortem abrasion. The talonid is strong, with two very well-devel oped cusps and a strong cusp on the trigonid-talonid junction. For dental measurements of this and otl1er specimens assigned to Propotamochoe1'US hysudricus, see table 20.4.
Specimen AUH 55: Shuwaihat, Site S 1 Tlus specimen is an incompletely erupted left M3 (figs 20.3.2a, It is missing the antetior cingulum from just lingual to the midline to the labial border, and the labial aspect of tl1e paracone. The
FRmi
tooth roots are present but fragmenta.rT_ and extensive. The overall shape of the too':!': the simple talon comprised of one flanked by two rows of labial cusplets. and simple in morphology; the enamel is
The para..:one and
metacone are well separated, with a smaU basin bet\' cen On the lingual side of the tooth, the hnJoconid is a\ntted ::1'":21 pp.
ACKNOWLEDGEMENTS
2 1 -64 (ed.
R.
L. Ciochon and
R.
S. Corruccini).
Plenum Press, New York.
Many indhiduals and organisations have assisted with the fieldwork that has lead to the retrieval of
---. 1984. A zoogeographic theater and biochrono
these specimens . In particular, Andrew Hill thanks
logic play: The time/biofaces phenomenon ofEurasian
the Undersecretary of the Department of Antiqui
and Mrican Miocene mammal provinces.
ties and Tourism, Al Ain, His Excellency Saif Ali
Continentale 14:
Paleobiologic
1 21-42.
Dhab'a a! D armalci for the help provided by his department, especially from Walid Yasin. Subse quent \York has been largely funded by grants to
---. 1 9 8 6. Mammalian biostratigraphy, geochronol ogy, and zoogeographic relationships of the Late
the project from the Abu Dhabi National Oil
Miocene Maragheh futma, Iran.
Company UI.D:-.:oc) and the Abu Dhabi Company
Paleontology 6:
tor Onshore Oil Operations
(ADCO ) .
Andrew Hill
76-9 5 .
Journal of Vertebrate
Colbert, E. H. 1 9 3 5 .
Palaeochoerus perimensis.
Siwalik
--
. 1 999. Late l\;liocene sub-Saharan African verte
mammals in the American Museum of Natural His
brates, and tl1eir relation to the Baynunal1 fauna, Emi
Transactions of the American Philosophical Society, Philadelphia, n.s. : 1-40 l .
rate of Abu Dhabi, United Arab Emirates. Chap. 29 in
tory.
Fossil Vertebrates ofArabia, pp. 420-29 ( ed. P. J. \Vh�· brow and A.
Hill) . Yale University Press, );"ew HaYen.
Cooke, H . B . S . 1987. Fossil Snidae from Sahabi Libya. In
Neogene Paleontology and Geology of Sahabi,
pp. 25 5-66 (ed. N. T. Boaz, A. El-Arnauti, A. W.
Gaziry, J. de Heinzelin, and D. D. Boaz). Alan R.
Hill, A., and Ward, S. 1988. Origin of the Hominidae: The record of African large hominoid evolution between 14 My and 4 My.
ical Anthropolq_IJY 3 1 :
Liss, :;\lew Yorlc.
Yearbook of Pbys
49-8 3 .
Cooke, H. B. S., and Ewer, R. F. 1972. Fossil Suidae
Hill , A . , Curtis, G., and Dralce, R. 1 9 8 6 . Sedimentary
from Kanapoi and Lothagam, Northwestern Kenya.
stratigraphy of the Tugen Hills, Baringo, Kenya. In
Bulletin ofthe Museum of Comparative Zoology 143: 1 49-296.
Sedimentation in the African Rifts,
pp. 285-95 ( ed.
L. E. Frostick, R. Renaut, and J. J. Tiercelin). Geo logical Society of London Special Publication no. 2 5 ,
Falconer,
and Cantley, P. T. 1 846 ( 1 845-49).
Fauna Antiqua Sivalensis. Smith,
Elder and Co., Lon
don. Fortelius, M.,
van
Blackwell, Oxford . Hill , A., Dralce, R., Tauxe, L., Monaghan, f..i ., Barry,
J. C., Behrensmeyer, A. K, Curtis, G . , Jacobs, B . F., der Made, J., and Bernor, R. L.
Jacobs, L., Johnson, N . , and Pilbeam, D. 1 9 8 5 .
1996. l\;liddle and Late lVliocene Suoidea of Central
Neogene palaeontology and geochronology o f the
Europe and tl1e Eastern Mediterranean: Evolution,
Baringo Basin, Kenya.
T11e Evolution of \Vt:stern Eurasian Later Neogene Faunas, pp. 348-73
14: 759-73 .
( ed. R. L. Bernor, V. Fallibusch, and H .-W. Mittmann) .
Hill, A., VVard, S., and Brown, B. 1992. Anatomy and
Columbia University Press, New York.
age of the Lothagam mandible.
biogeography and paleoecology. In
Evolution 22:
Journal ofHuman Evolution
Journal ofHuman
439-5 1 .
Golpe-Posse, J. M. 1979. Suides des gisements du Bled Dourah (W de Gafsa, Tunisi e ) .
Geok�gique, 'I�misie 44:
Notes du Service
75-107.
Hiirzeler, J. 1982 . Sur l e suide du lignite d e Monte
Comptes Rendus de FAcademie des Science�; Paris 295: 697-701 .
bamboli (prov. Grosseto, Italic).
Hanlllton, W. R., Whybrow, P. J., and McClure, H. A. 1978. Fauna of fossil mammals from tl1e Nliocene of
Saudi Arabia. Nauwe 274: 248.
Kotsakis, T. , and Ingino, S . 1979. Osservazioni sui
Nyanzachoerus (Snidae,
Artiodactyla) del Terziario
Superiore di Sahabi ( Cirenaica, Libia) . Harris, J. M., and White, T. D. 1 979 . Evolution of the Plio-Pleistocene African Snidae.
American Philosophical Society 69:
Se1·vizio Geologico d;ltalia
Transactions of the
1-128 .
Lealcey, L. S. B . 1958. Some east African Pleistocene
Hill, A . 1995. Faunal and environmental change in
Snidae.
Fossil Mammals ofAfrica no.
Hills Sequence, Baringo District, Kenya. ln
Paleocli
pp. 1 78-93 ( ed. E. S . Vrba, G.
H. Denton, T. C. Partridge, and L. H . Burlde ). Yale University Press, New Haven.
14. British
Museum (Natural History), London.
the Neogene of East Africa: Evidence from the Tugen
mate and Evol�ttion,
Bollettino del
C: 39 1-408 .
Leonardi, P. 1 952. Resti fossili di
ccsirnchot'i'lfo�:· dei Set;:entri
giacimento di Sahabi in Cirenaica r .-\frica
Atti ddln Acmde111ia Nazionale dei Lincei, Rendiconti, ser, 8m, 13: 1 66--69,
onale). Notizie preliminari.
fJliJ
L. BISHOP Ai'iD A. HILL
Lydekker, R. 1 8 77. Notices of new and rare mamm als from the Siwaliks.
India 9 : ---
Records ofthe Geological Servey of
Memoirs of the Geological Society of India. Palaeontologia Indica 8 :
---. 1926. Fossil Suidae ofinclia. 1-65 .
1 54.
. 1 884. Indian Tertiary and post-Tertiary Ver
Mem oirs of the Geological Survey of India. Palaeontologia Indica 3 : 35-104.
tebrata. Siwalik and Narbada bunodont Suina.
Schmidt- Kittler, N. 1971 . Die obermiozane Fossilager statte Sandelzhausen. 3. Suidae (Artiodactyla, Mam malia). Mitteilungen der Bayerischen Staatssamrnlung fiir Paldontologie und hi.>torische Geologie 1 1: 1 29-70.
Nalcaya, H., Pickford, M . , Nakano, Y., and Ishida, H.
Stehlin, II. G. 1899. Ueber die Geschichte des
1 984. The Late Miocene large mammal fauna from
Suiden-Gebisses.
the Namurungule Formation, San1buru El.ills, North ern Kenya.
African Study Monographs,
suppl. i ssue 2 : Thomas, H . , Bernor,
87-1 3 1 . Owen,
Abhandlungen der Schweizerischen Paliiontologischen Gesellschaft 26-27: 1-527.
R.
R., and Jaeger,
J.-J. 1982. Orig
ines du peuplement mamm alien en Afrique du Nord
1840-45.
Odontogt·aphy.
H. Bailliere,
durant le .Miocene terminal.
Geobios 1 5 :
2 83-97.
London. van der Made, J. 1989-1990. A range-chart for Euro
A Revision ofthe Miocene Suidae and Tayassuidae, (Artiodactyla, M.ammalia) ofAfrica.
Pickford, M. 1986.
pean Snidae and Tayassuidae .
Paleontologia i Evolucio
2 3 : 99-104.
Tertiary Research Special Paper no. 7. E. J. Brill , Lei van der Made, J., and Moya-Sola, S. 1989. European
den.
Suinae (Artiodactyla) from the Late Miocene onwards. ---
. 1987 . .Miocene Suidae fi·om Ad Dabtiyah,
eastern Saudi Arabia.
(Natural History),
Bulletin of the British Museum
Bollettino della Societit Paleontologica Italiana
28:
329-39.
Geology 41 : 441-46. White, T� 1995. African omnivores: Global climatic
--
. 1988. Revision of the Miocene Suidae of the
Indian subcontinent.
Miinchner Geowissenschaftliche
Abhandlungen A12:
1-92.
-·-.
T. C. Partridge, and L. H. Budde) . Yale University
1993. Old World Suoid systematics, phy
logeny, biogeography and biostratigraphy.
gia i Evolucio 26-27:
change and Plio- Pleistocene hominids and suids. In
Paleoclimate and Evolution, with Emphasis on Hwman Origins, pp . 369-84 (ed. E. S. Vrba, G. H. Denton,
Press, New Haven.
Paleontolo Whybrow, P. J. 1987. l'vliocene geology and palaeontol
237-69.
Bulletin ofthe British Mztseum (NaturalHi:;tory), Geology 41 : 367-457.
ogy ofAd Dabtiyah, Saudi Arabia. Pickford, M., Nalcaya, H., Ishida, H., and Nalcano, Y. 1 984. The biostratigraphic analyses of the faunas of the Nachola area and Sarnburu Hills, Norrl1ern Kenya.
African Study Monographs, suppl.
issue 2 : 67-72 .
Pilgrim, G. E. 1925. Presidential Address to the Geo
logical gress.
:_7; ·,·.'·'·
Section of the Twelfth Indian Science Con
Proceedings of the T·welfth Indian Science Con � 06-10.
v·V11ybrow, P. J,, Hill, A., Yasin al-Tikriti, W., and Hailwood, E. A. 1990. Late :Miocene primate fauna, flora and initial palaeomagnetic data from the Emirate of Abu Dhabi, United Arab Emirates.
Hzmzan Evolution
19: 5 8 3-8 8 .
Journal of
�t\ Fossil Hippopotamus from the Emirate of Abu Dhabi, United Arab Emirates
The Baynunah Formation of the 'Western Region, Emirate of Abu Dhabi (Whybrow, 1 9 8 9 ; Whybrow ct al., 1999-Chapter 4 ) , is predominantly com posed of poorly consolidated, horizontally bedded sands, and land vertebrates and freshwater molluscs ha,·e been recorded and collected from the Bay nunah intermittently since 1 949 . Collecting expedi tions of l984- 1 996 by a Natural History Museum/ Yale University team have explored numerous local ities, particularly Jebel B arakah (stratotype locality) , Hamra, Jebel Dhanna, and the islands of Thu mariyah and Shmvaihat. The recovery of hipparion ine horses has suggested that its age is no earlier than late Miocene. Among the mammalian fussils are remains of hippopotamuses to be described below. In the toll owing descriptions comparisons are made with
Formation, Siwaliks, Pliocene Formation, East Turkana, Kenya,
and Flynn, 1990); Koobi and early Pleis
l'lei:stocene; Lothagam , tocene ( Harris, 199 1 ); Kuguta, Kenya, late Kenya, late Miocene (Leakey ct al., 1996); Miocene 1978a); Lusso Beds, Semliki, Pliocene et al., 1 992); Maboko, Kenya, middle Miocene (Feibd Brovm, 1991); Mpesida, late Miocene don, 1978a); Nachnkui Formation, West and Pleistocene (Harris et al., 1 9 88 ); Nakali, Miocene (Aguirre and Leakey, 1974}; Ngcringcrowa,
Miocene ( Pieltford, 1 9 8 1 ; Hill, 1 995), c. 9.0 Ma; Nkondo tion, Ug;mda , Pliocene (Faure, 1995); Pinjor ( or Kansal) Forma tion, Siwaliks, Pliocene spanning 2.5-1.5 Ma (Barry and Flynn,
1990}; Sahabi, Libya, late M iocene (Boaz et al., 1987; Geraads, 1989); Saint Arnaud, Algeria, Pliocene (i\rambourg, 1970); Shun
gura Formation, Omo, Ethiopia, Plio-Pleistocene (Heinzelin, 1983); Siwaliks Group, Pakistan and India, 1\liocene-Piiocene Pliocene and Flynn, 1990 ); Tatrot Formation, Natrun, ''P""ll,.u,6 5 . 0-2 . 5 Ma (Barry and Flynn, 1990); Miocene (Howell, 1 980); Warwire Fomution, Uganda, Pliocene (Faure, 1995); vVembere-Manonga Fonnarion, Tanzania, latest Miocene-early Pliocene (Hanison, 1997).
tossils in the Palaeontology Department of The Xatural History Museum, London (institutional abbreviation BMNH ) . Localities and deposits men tioned in the text other than those of the Baynunah Formation are:
PREVIOUS STUDIES OF FOSSIL H IPPOPOTAMUSES The extant common hippopotamus, Hippopotamw
Baringo Kenya, middle and late Miocene and mLwuiu15 r:he Ngorora Formation (Hill et al., 1 985; Hill, 1995); Beglia marion, Tunisia, late Miocene (Piekford, 1990); Bone, Algeria, of uncertain age ( Howell, 1 9 8 0 ); Dhok Pathan, Siwaliks, late Miocene Barry and Ft�1m, 1990); Fort Te rnan, Kenya, dated to 14.0 million years (Ma) (Ceding el al., 199 1 ) ; Irrawaddy Group, 1983); Ka.iso Formation, l:ganda, Burma, Pliocene Pliocene (Cooke and 1970; Pickford et a!., 1988); et al., 1996); Kansal (= Pin-
Copyright
amphibius Linnaeus,
1 75 8 , l ived during the his tori.:
period in sub-Saharan Africa and northwards in �he
Nile Val ley. It also occurred in other parts of nor::.h
ern Africa until the Holocene wheneYer -:limat; c environmental conditions allowed. At ir:rc:-,::..� 3 the P l eistocene H.
amphibiw:
ranged across southern and \\TStcrn
1999 by Yale University. All rights reserved. ISBN 0 � 300-071 8 2
may be less raised, the P
1
the sagittal crest
is single-rooted (Aram
bourg, 1 947: 3 1 8 ) while that of Hex.
sivalensis
looks as if it has two roots, and the p4 comes to have only one cusp. P3 may be shorter relative to than in
Hex. sivalensis.
Harris ( 19 9 1 : 45 ) suggested or implied that
protamphibius evolved into
the extinct diprotodont
contact the nasal , P more gracile than
Hex. sivalensis or protamphibius
Hippopotamuses from Mpesida and Lukeino ( Coryndon, 1 978a) may be earlier than
vardi from
Hex. har
Lothagam, and are more certainly ear
lier than the first
Hexaprotodon in the
Siwaliks.
Teeth from Mpesida are the size of Hex.
harvardi.
Upper cani nes have a deep posterior groove, lower canines have smooth enamel, premolars are large and pustulate with marked cingula.
At Lukeino there are many remains including a 4 hexaprotodont mandible and a P with two sub-
equal ( b uccal and lingual) cones surrounded
a
strong cingulum. A smaller species vYith lower crovvned teeth may also be present.
Hippopotamus kaisensis Hopwood , 1 926 Hip. kaisensis is
a name for hippopotamus te eth
from the Kaiso Formation. The molars resemble those of
Hex. sivalensis
and
protn mp!Jibiw,
but
- A.
w. GE:X1RY
isolated teeth are not easily distinguishable from Hip. amphibius. Moreover, some fossils can be very rge:-tor example, mandibular pieces of the O'Brien collection from Kaiso Village ( Cooke and Coryndon, 1970: 196, table 36 ) . Tetraprotodonty can b e demonstrated from a symphysis BMNH M2527l (Co oke and Coryndon, 1970: pl. 14B), upper canines can have a shallow groove posteriorly (Cooke and Coryndon, 1 970: 1 92 ), and a left P4 is a robust tooth 5-1 0% wider than in Hex. sivalensis or protamphibius ( Cooke and Coryndon, 1970: 1 9 5 ; my own measurement of BMNH 36722; Harris, 199 1 : table 2 .2 ) . Hitherto the age o f the Kaiso F ormation has been established only by faunal correlation; its upper t'l una may be a later Pliocene equivalent to Shun gura Formation Members F-G, and its lower fauna earlier in the Pliocene. Thus Hip. kaisensis could be conspecific with Hex. protamphibius) or possibly with d1e hippopotamuses that begin to appear higher in the Shungura Fo rmation, such as Hex. karumensis and Hip. gorgops. Pavlakis ( 1990), workon hippopotamuses of the Lusso Beds, con cluded that Hip. kaisensis was not distinguishable from Hip. amphibius. Faure ( 1995) used the name Hip. kaisensis for hippopotamus fossils of the War wire and Nkondo Formations in Uganda, going back perhaps to 5 . 0 Ma.
Hexaprotodon sahabiensis Gaziry, 1 987 This Sahabi species appears to be slightly smaller than Hex. sivalensis and the mandible less deep. It is hexaprotodont according to Gaziry ( 1987) and on the testimony of Petrocchi ( 1 9 5 2 : 27), who was describing a skull that can no longer be found. The canine teeth have smooth enamel, the cheek teeth are low crowned, P4 has avo cusps, and Gaziry thought the premolar and molar rows were of the same length. The cusps on the molars are only poorly tretoliate, as in Hex. harvardi. The P4 has avo main cusps side by side cen( metaconid and protoconid) and t\Vo sub sidi an· ones side by side posteriorly. These latter cusps appear not to have survived in later species like Hex. sim.lensis and p1totamphibius) nor in Hex.
harvardi) although a syntype P4 (BMNH Ml262 l ) o f the dwarf species Hex. itnagunculus has pre served into late wear a single wide posterior cusp. The P 3 of Hex. sahabiensis has rather clear ante rior and posterior lobes with a prominent constric tion between them ( Gaziry, 1 987: fig. 6 ) . A similar morphology appears to be present in one or two Hex. sivalensis but is absent in a syntype Hex. ima gtmculus ( B MN H M l 2 6 1 9 ) and in other African fossil hippopotamuses. The of Hex. saha biensis also has a small posterolingual cusp rather than sim ply a cingular outgrowth as in more advanced hip popotamuses. Although litde is lmmvn of Hex. sahabiensis, i t looks a s if the and P4 characters d o differentiate it from the more advanced African hippopotamuses described above. The anterior and posterior lobes of P 3 may indicate a link to Hex. sivalensis in so far as one could posmlate a longer surviving primitive characteristic in that persistently hexaprotodont lineage.
Hexaprotodon hipponensis
(Gaudry, 1 876) Hex. hipponensis was founded for some hippopota mus teeth ( Gaudry, 1 8 76: pl. 1 8 ) of uncertain geo logical age from Pont de Duvivier, south of Bone, Algeria (Joleaud, 1920: 16): they were redescribed by Arambourg ( 1 94 5 ) . Parts of six more or less equally sized incisors, two canines, two premolars, and half a molar all seemed to have come from the same mandible . Hence the species must have been hexaprotodont, like Hex. sivalensiJ� and it was the first record of an African hexaprotodont. The of ilie enamel. canines lacked any strong Pomel ( 1 896) assigned a tetraprotodont fossil from Saint-Arnaud to Hex. hipponensis, but Aram bourg ( 1947: 327; 1 970: 19) held that it was not conspecific with th e Pont de Duvivier fossils. Hippopotamus remains from Wadi Natrun were attributed by Andrews ( 1902 ) to Hex. hipponensis. Stromer ( 1907: 1 10 ; 1 9 14 : 5 ) affirmed that the ani mal was tetraprotodont with t\Vo more or less equally sized incisors on each side of its lower jaw, but he did not illustrate this feature. Arambourg
A HIPPOl'OTA.l>,HcS FRmi cHF
( 1 947: 328 ) , rejecting a tetraprotodont as conspe cific with He:x:. hipponensis) referred the ·wadi Natrnn form to a new subspecies, andrewsi) of �Hex. protam phibizts. IfWadi Natrun is indeed early Pliocene in age ( Geraads, 1 987: 22 ) , this would be an excep tionally early tetraprotodont species according to Geze's ( 1 9 8 5 : 94) estimate that tetraprotodonty appears in East Africa at around 3 . 0 Ma ago. \Vadi Natrun hippopotamus teeth and casts thereof in London are small and low crowned. The upper molars have strong cingula, and the cusp shape is triangular rather than trefoliate. According to Stromer's ( 1 9 1 4 ) illustrations the upper canine has a narrow but shallow posterior groove (an advanced 3 character), the P does not have clear-cut front and back halves, and P 4 is without posterior cusps. These characters accentuate the differences between the Wadi Natrun and Sahabi hippopotamuses.
Hexaprotodon crusafonti Aguirre, 1 963 Hex. siculus ( Hooijer, 1946) and Hex. puntanellii (Joleaud, 1920) are possible senior synonyms of this species, which occurs in the latest Miocene and possibly the early Pliocene of southern Europe. No evidence of the number of incisors was found by Aguirre ( 1 96 3 : 2 1 8 ) . Pantanelli ( 1 879: pl. 4, fig. 5 ) illustrated a hexaprotodont premaJci !la o f Hex. pan tanellii from Italy, whereas Lacomba et al. ( 1986: 1 77, pl. 1) interpreted a Spanish mandible of MN 13 age assigned to Hex. crusafim ti as tetraprotodont. This claim tor tetraprotodonty in another early hip popotamus, in addition to the Wadi Natrun one, is noteworthy. Moreover, this mandible appears from the illustration to have a very narrow symphysis in comparison with Hex. harvardi) Hex. protamphibiz1-s) Hex. sivalensis) and modern hippopotamuses. We see here, therefore, a combination of a primitive narrow symphysis with precocious, if occasional, tetraprotodonty. The back of the symphysis in Lacomba et al. ( 1 986: pl. 1 , fig. 1 ) is level with the front of P3. The P4 ( Aguirre et al ., 1973: fig. 6) has two posterior cusps like Hex. saht�biensis. (The reported by Aguirre in 1 963 was said to be a by Lacomba et al., 1 986. ) The lower molars in Lacomba et al. ( 1986: pl. 1 ) are low crowned.
BAlc-:'L'XAH FO&'v!ATIO�
-
Lower molars of the specimen figured by Cuvier ( 1 8 04: pl. 2, fig. 1 ) as "grand hippopotame fossile" (other specimens of which were Pleistocene hip popotamuses) show narrow elongated cusps in occlusal view with wide transverse valleYs in between them. The enamel is very roughened \'ertically and the hypoconulid of M3 is narrow and isolated. Faure and Meon ( 1 984) two M1-M3 lengths as 1 2 0 m m and 124 mm, seemingly slightly smaller than in 3 Hex. sivalensis. The P ( Lacomba et al., 1986: fig. 8 ) does not have very clear front and back halves , nor does it have a posterolingual cusp. Gaziry ( 1987: 3 1 0 ) thought that Hex. crusa fonti differed from Hex. sahabiensis in the vvidth of upper canines, larger molars and premolars with less prominent cusps, the M3 hypoconulid larger and more elongated, and with a shallower horizontal ramus of the mandible. These differences do not appear marked; the shallow ramus of the specimen in Aguirre ( 1963: pl. 3, fig. 3 ) could arise from its immaturity, and in any case Gaziry had shown from measurements that the ramus of Hex. sahabiensis was not itself very deep. Coryndon ( 1978b) thought that Hex. crusa fonti differed from the hippopotamus at Mpesida and Lulceino ( previously mentioned here under Hex. harvt�rdi) by being smaller, v;ith higher crowned molars and proportionately smaller premo lars. Tllis is surprising because the premolars of crusafonti appear ro be relatively large and the molars low crowned. Coryndon took large premo lars , low-crowned molars, a deep posterior groove on the upper canines, and smooth lower canines as primitive, so that Hex. crztsafonti appeared to be more advanced . She thought that the Sahabi and Wadi Natrun hippopotamuses were closer to crwa fonti than to those of sub-Saharan Mrica.
Hexaprotodon iravaticus Falconer and
Cautley, 1 847
comes from the Irra,,·addy Group. Burma (Bender, 1 9 8 3 : 1 0 0 ) . The lecron·pe is a hexaprotodont mandibular s\·mphysis_ R\I::;H 1 4771 (Falconer and Cautley, 1 8-±:-: 01 , 10 and l l ) , o n which t h e cro,Yns o f all t..ceth ha,-e Hex. iravaticus
fl'm
A, w. GE:\'TRY
been destroyed. Other pieces in London are a part of a left mandible with broken M3 and much dam aged M2 (M10525 ) and part of a left juvenile sym physis (M10526). Colbert ( 1938: 4 1 9 , fig. 61 ) described a partial skull, 20037, in the American Museum of Natural History, New York. Hex. iravaticus is smaller than Hex. sivalensis and its most obvious feature is that the symphysis is narrow, as in the Hex. crusafonti mentioned above. The back of the symphysis looks as if it would be level vvith the back of P3• I2 is set slightly above I1 and I3. The orbits are little elevated, but perhaps more so than in Hex. ha1·vardi. The front of the orbit appears to lie above the back of M2 . The front of the lachrymal looks as if it has only a point contact with the nasal. Upper molars retain wider transverse valleys between the front and back pairs of cusps according to Colbert ( 1938: 420 ) . An upper canine showed a less-deep posterior groove (less-expanded internal lobe or pillar) and less heavy longimdinal ridging on its enamel than in Hex. sivalensis or Hip. amphibius. Irrawaddy fossils have been collected since 1826-27. The fauna of the lower Irrawaddy beds is equivalent to Dhok Pathan (late Miocene) faunas, but Stamp ( 1 92 2 ) and then Colbert ( 1938, 1943) and Bender ( 19 8 3 ) showed the hippopotamus as coming from the upper Irrawaddy beds, taken as equivalent to the Tatrot or even the Pinjor zones of tl1e Siwaliks. Although Colbert ( 1 9 3 8 : 276, 4 2 3 ) noted that the provenances o f specimens from the t\vo faunal levels of the Irrawaddy Group are not always clear, he strongly supported (p.42 2 ) a late age for Hex. iravaticus. This would make the nar row-muzzled Hex. iravaticus a contemporary of more advanced hippopotamuses in Africa and of Hex. sivalensis in India. It is also evidently contem porary with Elephas hysudricus (see Colbert, 1935: 4 1 5 ) , a species that Hussain et al. ( 1992 ) date back only to 2.7 Ma. Perhaps Hex. iravaticus is a long-surviving relict of a primitive narrow-muzzled grade of hippopotamus evolution, with some adnnced characters evolved in parallel with other bppopotamuses. From the illustrations in Colbert 1938 1 and Coryndon ( 1 978b: fig. 2 3 . 5 ) it looks adnnced than Hex. harvardi in more ele. ::.:o: -:i orbits. lachrymal contacting nasals, and tooth
row positioned slightly more anteriorly relative to orbits.
Dwarf Hippopotamuses Coryndon ( 1977) decided that Chon·opsis Leidy, 1 8 5 3 , for the extant West Mrican pygmy hippo potamus C. libe1·iensis ( Morton, 1 849 ) , was a junior synonym of Hexaprotodon Falconer and Cantley, 1 8 36 . Hex. liberiensis differs from Hip. amphibius mainly in the following characters: I.
Smaller size.
2. Only two lower incisors ( diprotodonty) although the upper incisors remain at four. 3 . The upper and lower incisors occlude more or less horizontally at their tips and not intersti tially. 4. The upper canines have a deeper posterior groove. 5 . The enamel of the lower canines is smooth and lacks ridges anterolabially. 6. Shorter premolar rows. 7. Shorter diastemata. 8 . P4 has two cusps. 9 . A simpler occlusal pattern of molar cusps. 1 0 . The anterior ends of the premaxillae remain fused medianly and do not diverge. 1 1 . The orbits are set low. 12. The face is relatively shorter and the orbit is further forwards relative to the tooth row. 1 3 . The lachrymal is not in contact with nasal. 14. The back of the braincase remains level or even curves downwards rather than passing upwards in profile. 1 5 . The limbs are more gracile. Of these 1 5 characters I, 3-5 , 7-13 , and 1 5 are probably primitive in liberiensis and shared with fos sil Hexaprotodon species and unlike Hippopotamus) while 2, 6, and 14 are advances in liberiensis. Harri-
A HIPPOPOTA!>IUS FROM THE BAYNUNAH FORMATION
son ( 1 997: 1 7 7 ) believes that
liberiensis is
a "some
what specialized derivative of the sister taxon of all
-
conid and it does have a posterior subsidiary cusp . Pickford ( 19 8 3 : fig . 1 8 ) dated the disap
other hippopotamuses" and should not be included
pearance of
in
hippopotamuses in East Mrica to around 8-7 Ma.
Hexaprotodon. Fossil dwarf hippopotamuses are known in addi
tion to those of Madagascar and the Mediterranean islands.
Hex. imagunculus ( H opwood ,
teeth from the Kaiso Formation, and the name has
Pickford ( 1989, 1 9 9 0 ) discusses
THE ABU D HABI H IPPOPOTAMUS
been used for teeth from other African localities. It
A hexaprotodont mandibular symphysis in the Kaiso Formation ( "Hip popotamussp." of Cooke and Coryndon, 1970: pl. may be small enough to be part of the Hex. imagunculus lineage. Its I2 is slightly than the
line from I 1 to I 3 , and Cooke and Coryndon ( 1 970: 1 8 8 ) note that I2 is the smallest incisor. Faure ( I 99 5 ) has recorded it back to 5 .0 Ma.
Order Artiodactyla Owen, 1 848
and Coppens,
tetraprotodont. It occurs
in the Shungura Formation Members E-L and in the Koobi Fora and Nachukui Formations. It shows additional differences from Kaiso Formation cranial remains of small hippopotamuses, BMNH M l 48 0 I and M26 3 3 6 , i n more elevated orbits and occiput. It could be a descendant of Hex.
imagunculus.
The
of M l 4801 has only one main cusp, unlike the t\vo cusps on one of the imagunculus syntypes, but Hex. aethiopicus P4 also may have either one or
the
1 99 1 : 48 ) .
t\vo cusps
Kenyapotamus for
early
hippopotamids from the Miocene of the Baringo deposits in
doni,
The type species is
K. coryn-
and the holotype a right lve + M
3
from
)\[geringerowa. The Nakali teeth are higher crowned and more robust than those from Ngeringerowa. An earlier species comes from Fort Ternan and possibly Maboko . The teeth are smaller than in
popotamus
Kenyapotamus He.'!Caprotodon or Hip
and lower crowned than in most or all
of them, the
has a posterolingual cusp in addi
tion to its main cusp ( similar t o
Genus
Hexaprotodon Falconer and Cantley, 1 8 36 Hexaprotodon aff. sahabiensis Gaziry, 1987
The holotype of Hex.
sahabiensis is
a left mandible
with PcM3 ( Gaziry, 1987: fig. 2 ) , housed in the Department of Geology of the Garyounis U niver Libya.
Material from Abu Dhabi AUH 457, Hamra, H l : lower jaw. Parts of left 11 and 12 and right cheek teeth survive. AUH 48 1 , S4 : immanw: lower jaw \vith right ramus. Surviving crowns of teeth and front of left are right I1-I3, erupting P2 + worn dP4; left I2�3, canine, erupting Pc-P3, much of worn Occlusal length right d P4, 43.4 mm. Figure 2 1 . 1 . AUH 235, Shuwaihat, S4: pieces o f a mandible wid1 incisors. with right and left B.IVL'!H M49464, Jebel Barakah, B2: lower tooth rows and horizontal rami. crovms of teeth are right P3 + P4, M3; left I1, I3, M3 in early middle wear, M2 in
wear, and M1 in late
middle wear. Occlusal \V.idth (mesiodistal) 13, 1 3 . 6 mm. Occlusal length P2-4, c. 127.0 occlusal length M1_3, c. 1'3, 39.2 x 23.0 mm; 139.0 mm. Occlusal lengths and
The Earliest Hippopotamuses Pickford ( 1 9 8 3 ) founded
Family Hippopotamidae Gray, 1 82 1
sity, Benghazi, Libya. The type locality is Sahabi,
1 9 7 5 ) is a more completely known species, as small
Hex. imaguncuJus and
Kenyapotamus
further and records it at Beglia.
was probably hexaprotodont.
as
and appearance of later
1926) was
originally described on some isolated small cheek
Hex. aethiopicus ( Coryndon
Kenyapotwmus
Hex. sahabimsis),
P4 has no metaconid cusp alongside the proto-
P4, 34.0 x 23.8 mm; M1, 3 5 .0 X 2 5 . 8 (posterior); M2, 45.5 x 3 l . l ,-. . .-...." , M 3 , c. 64.0 x 3 1 .8 m m
26.6 mm
AUH 3 6 , Jebel Dh;uma, unworn length and >vidth 42.9 x 44.6 mm (posterior). Figure 2 1 .4 .
AUH 99, Shuwaihat, S L base o f a left M3 i n a AUH 1 1 0, Shuwaihat, S l : parts lower mohu·.
i\.UH 262, Kihal, Kl : partial !eft , tmworn. Occlusal cc§:':::. ;:. 27.9 mm. BMNH M49465, Jebel Baralcah, B2 :
Occlusal length and width 38.7 (posterior). AUH 446, Jebel Barakah, B2:
X
i\.UH 29, Shuwaihat, Sl: canine IT,l_;rrn,·:--• AUH 359, Harmiyah, Y l : canine cnarr.e: . fi-agment. AUH 57, Shuwaihat, S l : incisor
-
A. W. GENTRY
Figure 2 1 . 1 . Hexaprotodon aff. sahabiensis, partial immature, lower j aw, AUH 48 1 ; lat eral view. Scale bar 10 em. =
AUH 2, Hamra, H l : incisor piece incl uding the occlusal surface. AUH 292, Jebel Dhanna, JD3: upper incisor piece. AUH 37, Jebel Dhanna, J D 5 : ?incisor piece, possibly hippopotamid.
AUH 3 1 2, Kihal, Kl: ?incisor piece, possibly hippopotamid.
AUH 247, Shuwaihat, S4: left metatarsal 3 . Maximwn length 1 1 9 . 1 mm, width of distal condyles 28.2 mm.
AUH 420, Jebel Dh
ann
a, JD4: proximal left metatarsal 4.
AUH 49, Hamra, H5: stem of a right scapula. AUH 1 1 8, Shuwaihat, S l : stem of a 1ight scapula. Figure 2 1 .7.
AUH 60, Shuwaihat, S 1 : tooth fi·agments.
AUH 443, Jebel Barakah , B2: part of a right scapula.
AUH 92, S huwaihat, Sl: tooth fiagments.
AUH 497, Jebel Dhanna, JD3: right humerus in four pieces.
AUH 3 1 , Shuwaihat, S 1 : tooth fi-agmenrs.
Length from the top of lateral tuberosity to the base of lateral
AUH 42 1, Jebel Dharma, JD4: tooth fragment.
groove distally c. 350 mm; minimum transverse width of shaft
AUH 103, Shuwaihat, S1 : right ilium with part of the acetabular surface. AUH 1 33, Shuwaihat, S2: part of a left ischium. AUH 59, Shuwaihat, S 1 : distal left femur, much damaged. AUH 498, Jebel Dhanna, JD3: damaged pieces of a proximal right
a left humerus. AUH 253, Shuwaihat, S4: most of a proximal right radius. Figure 2 1 .7.
tibia.
AUH 98, Shuwaihat, Sl: distal right radius, damaged posterolater-
AUH 498.
AUH 170, Hamra, H 1 : right lunate. Figure 2 1 .7.
AUH 499, Jebel Dhanna, J D 3 : probably part of the same tibia as AUH 1 50, Hamra, H 3 : distal right tibia. Ma.ximum distal width 77.5 mm, ma.ximum anteroposterior width 59.7 mm. Figure 21.5. AUH 288, Jebel Dharma, J D 3 : distal right fibula. Figure 2 1 . 5 . AUH 5 , Hamra, H 1 : left astragalus, medial surface missing, lateral and posterior surfaces much damaged. Lateral height c. 84.0 m m . AUH 3 6 8 , Jebd Barakah, B 1 : l e ft asmgalus. Lateral height
75.2 mm, medial height 68.7 mm, ma.'<s.
The ilium fragment AUH 103 has little or no sign of the origins
of the recms femoris muscle; it could be of a hippopotamus accord
1---i 1 em
Figure 21.4. Hexaprotodon aff. sahabiensis, unworn tight upper molar, AUH 36; occlusal view, anterior side is to the right.
ing to the evident great area and backward-facing rather than lat
eral-facing plane of the front part of the acetabular articular surface. The distal femur is smaller than in
Hex. sivalensis. Little of it
survives other than the fossa on the shaft, a part of the medial
condy le and a smaller part of the patellar fossa. The fossa is placed
too anteriorly on the lateral surface for the bone to be giraffid.
A HIPPOPOTA�\ICS
THE RW:\UNAH FO!Uv!ATION
m:l
B, Figure 2 1 . 5 . Hind limb bones and \'ertebra. D-F, Hexaprotodon aff. sahabiemis; C� Hex. sivatensis. A, Distal right tibia, Al."H 150; anterior view. B) C) Distal articular surfaces of .--\CH 150 and of right tibia BMNH 170 18a; anterior side towards the top. D) Distal right fibula ArH 2 8 8; medial view, anterior side towards the left. E, Right astragalus, AUH 44; anterior vie\,-. � Tho racic vertebra, AUH 1 0 5 ; anterior view. Cross h atching matrix. Scale bar 50 mm. =
=
E
The articular surfaces of the proximal tibia ALJH 498 are
more concave than in Hex. sivalmsis.
The distal tibia is the postcranial bone that least resembles other hippopotamuses. T11e astragalus facets are narrow and antero
articulation is centrally placed and not closer to the lateral edge. Also there is a longitudinally running ho!Jow on the posterior sur :fuce medial to the articular facet for the calcaneum. The metapodials of the Abu Dhabi hippopotamus appear
1 3% shorter and 28% more slender than in Hip.
posteriorly long and the flange behind the fibula is much enlarged. The proportions of the astragalus facets are pig-like, but the bone
to be about
would be large for
from Madagascan hippopotamuses. It seems to be general in hip popotamuses that metacarpals are longer than metatarsals, that dis
pig. Even the large Siwaliks }.fiocene pig
HippopotM;wdon sivalense has proximal astragalus breadths of only
26.2 mm, 2 8 .2 mm, and 30.5 mm (Pickford, 1988: 76), about half the size needed to fit the Abu Dhabi tibia. Moreover, the large flange behind the fibula is not a pig character. The lateral :fucet for the astragalus is less expanded anterolaterally than in Madagascan hippopotamuses. The bone is a bit smaller than in Hex. sivalensis.
amphibius.
are also longer than metapodials of similar girth
tal ho!Jows on anterior surfaces are better developed in metatarsals, and that metatarsals 2 and 5 are more reduced proximally than metacarpals 2 and 5. distal ho!Jow anteriorly than in The metatarsal 3 has a Hex. and the proximal for metatarsal 4 is weak, small
On the distal fibula the facet for the lateral side of the astra galus is arrved downwards anteriorly or at any rate it has a ventral
compared with its co ndition in Hex. sivalensis. or not very The scapulae are a bit smaller than in Hex. siJ>alensis. The
edge that is concave downwards. The astragali are smaller and narrower than in Hex. sivalensis. They are not of pigs because the ridge on the ventral condylar
and distingnishes them from giraffids or bovids. The tuber is nar· rower in lateral view through being less expanded anterodorsally,
A
spine is centrally placed between the anterior and posterior edges
Figure 2 1 .6 . Front limb bones. A-D, Hexaprotodon af£ sababiensis; E) Hex. sivatensis. A) Right scapula, AUH 1 18 ; lateral view. B, Right scapula, AUH 1 18 ; glenoid view, lateral side towards the top. C, Right radius, AUH 253; proximal articular sur fuce, anterior side towards the base. D, E) Right radii, AUH 98 and BM:KH 16478 ; distal articular surfaces, anterior sides towards the top. Key: a, indented anterior margin of scaphoid and lunate facets; b, localised tubercle anteromedia11y to scaphoid facet; c, ridge on medial part of anterior surface; d, concave profile of posterior of lunate facet. Scale= 50 mm .
� A. W. GENTRY
and the articular fu.cet for the hwnerus is rounded in ventral view
rather than being narrower u-ansversely. A hippopotamus scapu.la
from exposure 3 of the Kazinga Channel South in the Kaiso For mation, BMNH M25176, looks intermediate in size and in tuber morphology between the Abu Dhabi examples and Hex. sivalensis;
its humeral articulation is damaged so that its original shape cannot be determined. Compared with Madagascan fossil hippopotamuses,
the Abu Dhabi scapulae have narrower and longer stems below the blade and the blade widens more gradually above the stem. The humerus AUH 497 has lost the medial tuberosity, the posterior eminence behind the lateral tuberosity, the surface of the
bicipital groove, and most of the distal end below the coronoid
fossa. It has, however, preserved most of the distal lateral groove. It
may be slighdy smaller and is certainly more gracile than in He...:. sivalensis. Compared with proximal ends BMNH 16467, 16468, and 16670 of Hex. sivalensis, the Jebel Dhanna humerus has an
infraspinatus insertion set more anteriorly on the lateral tuberosity and standing more proud of the surrOLmding bone surface. Also the back of d1e articular head projects further behind the stem or shaft of the humerus in lateral view.
Figure 21.7. Hexaprotodon aff. sahabiensis. A, Right metatarsal III, AUH 243; proximal articular sur face; anterior view, anterior side tow ards the base of drawing. B, Right metacarpal III, AUH 96 and 97; same view as A. C, Right metacarpal V, AUH 154; same view as A. D, Right cuneiform, AUH 252; anterior and medial views. E, Right lunate, AUH 170; lateral view. F, Proximal phalanx, AUH 83; dorsal view. G, Distal phalanx, AUH 478; dorsal view. Scale 50 mm. =
The proximal radius AUH 253 is smaller than Hex. sivalensis,
having linear dimensions no bigger than 80% of those of the Siwa liks species at best. What remains of the top of d1e shaft appears considerably more gracile than Hex. sivalcnsis. The back of d1e lat eral facet is slanted from posteromedial to anterolateral, there is a broad and shallow groove between lateral and medial fu.cets, and the medial facet has a considerable anteroposterior dimension.
These features arc more like hippopotamuses than modern pigs.
Modern pigs arc too small in size to fit this radius, but a sufficiently
large fossil species
might have a
more sinlliar morphology. For the
present this bone is accepted as being of a hippopotamus.
The distal radius AUH 98 is not liom a pig because of the
indented anterior edge between the scaphoid and ltmatc fu.cets, and
Figure 21.8. Hexa protodon aff. sahabi ensis, right humerus in four pieces, AUH 497; lateral view. Scale bar 20 em.
A HIPPOPO'L\.\!CS
because this indentation precludes there being room for a promi nent concavity in the anterolateral part of the scaphoid facet. It \muld also be rather large lor a pig, and the overall width of the lunate and scaphoid facets together is perhaps too gteat. It is a little smaller than several examples of Hex. sivalensi-"1.X \H fORMATION
-
The p reservation of the anterior surtace on the na\iculocuboid
AUH 460 leaves open the possibilitY that it is really a na,icular bone and thus not the united na,iculocuboid of � ruminant. If it
were a ruminant it would fit this s pecies b1· size. On the distal metatarsal AUH 249 the thnzes on either side of the ventral end of the central vertical gulh- on -the ::mrerior sur face have been destroyed. The bone agrees in size "ith Pa!n but the shaft narrows as it tises abm·e the distJ.! end. The distal end is wider than in living Boselaplms t;·ro._qoc:? ;;:d!!.'. Irs maximum \vidth across the condyles is about 15% greater than in Tragoportax acrae ( Genu·y, 1974: fig. 2 1 ) . Figure 22. 1 4 . A> B> Right horn-core base of Gazella aff. lydekkeri> AUH 3 8 9 ; lateral and anterior views. C) Cross-section of AUH 389 just above its base, lateral side to the left, ante rior to the foot of the page. D> Left horn-core base of Pachyportax latidens> AUH 1 06 ; ante rior view. Key : f, dorsal surface of frontal; p , side surface o f pedicel; o, surface o f orbit. Scale 20 mm. =
Discussion The low level of the frontals relative to the dorsal orbital rim on AUH 1 06 is unlike post-Vallesian Tragoportax and agrees with the Siwaliks genera Selenoportax and Pachyportax. The horn core also matches these genera in its probable large size and wide insertion. It agrees with Pachyportax latidens rather than Selenoportax vexitlarius in a very low inclination of the horn core in side view, and less divergence of the horn cores. It could also be
Figure 22 . 1 5 . Pachyportax latidens> right mandible AUH 266; lateral vie"·· Scak bar
20
em.
m:l
A.
W. GENTRY
from a boselaphine bovid but too large to be accommodated in Tragoportax cyrenaicus. It is of appropriate size to match the holotype upper molar of Pachyportax latidens (BMNH M34567, cast) and is slightly larger than mandibles of Tragoportax acrae of Langebaanweg. Parabos cordieri appears to have had the lower premolar row with an occlusal length 77% that of the molar row (my own mea surements of a specimen in Lyon); it was also slightly smaller than the Arabian fossil . The mandible is too large to match the upper dentitions on two Pikermi skulls of Palaeoryx pal lasi (Wagner, 1 8 5 7) (BMNH M l 0 8 3 1 , M 1 08 3 2 ) . Pachytragus ligabue Thomas ( 19 8 3 ) from the Hofuf Formation is about the size of AUH 226 but has the premolar/molar row ratio at c. 5 5 % .
Genus indet. Bovidae, sp . indet. Figure 22 . 1 6 . Pachyportax latidens, distal metatarsal AUH 249; anterior view. Scale bar 5 em. =
Material Shuwaihat, Sl: medial side of right astragalus. Height mm. AUH 407, Ha.rmiyah, Yl : left and right astragali. Much abraded. Lateral height ofd1e left one 50.5 mm, medial hcight 47.1 mm, distal widili 28.9 mm. Lateral height ofilie right one c. 48.0 mm, medial height c. 45.0 mm, distal widili c. 28.0 mm. AUH 39 1 , Harmiyah, Yl : partial left naviculocuboid.
AUH 63,
48.2
These tarsal bones are rather smaU to be con specific with the other postcranial bones assigned to Tragoportax cyrenaicus. They have a size appropri ate for a late Miocene bovid such as one of tl1e similar to the left horn core of a large boselaphine or bovine from Piram Island , BMNH M2402a, referred by Pilgrim ( 1 9 3 7 : 746; 1 93 9 : 1 7 5 ) to
Pachytragus species of Samos or perhaps a species of about the size of the Sahabi ? Hippotragus sp. ( Lehmann and Thomas, 1 9 8 7 : 32 8 ) .
Selenoportax lydekkeri ( Pilgrim, 1 9 1 0 ) . The horn cores of Parabos cordieri from the start of the
Tribe Antilopini Gray, 1 8 2 1
French Pliocene (Gromolard, 1 9 8 0 ) were similar to those of Pachyportax. The mandible differs from a giraffid mandible
Remarks
like the Pikermi Palaeotragus 1'oueni BMNH
Spiral-horned antilopines of a variety of genera are
M 8 3 67 in tlut the teeth are too hypsodont, their
especially characteristic of the Graeco- Iranian Turo
basal pillars are too large, and the central fossettes
lian. Knowledge of them has been much extended
on the occlusal surface are isolated from the exte
in tl1e past 15 years by tl1e work of Bouvrain (see
rior even in the middle wear of M3. It is definitely
Bouvrain, 1 992 ) , and further revision is likely. The
type species of Prostrepsiceros comes from lower and middle Maragheh, around 9-7 Ma in age. Lehmann and Thomas ( 1 987) assigned to Pro:;trep siceros a large and late species, P. libycus> from Sahabi. The ancestry of Antilope cervicapra> the liv ing blackbuck of India, lies within this group.
Genus Pr·ostrepsiceros Major, 1 89 1 Type species Prostrepsiceros houtumschindleri (Rodler and Weithofer, 1 890) Prostrepsiceros afi libycus Lehmann and Thomas, 1 9 87
Material AlJH 237, Shuwaihat, S4: two pieces o f a left horn core, Figure 22.17. AUH 236, Shuwaihat, S4: of a horn core. AUH 4 1 3, Harmi.yah, Yl : core possibly spiralled. AUH 258, Kihal, Kl: left astragalus. Lateral height 43.6 rnm, medial height 39.4 mm, distal widd1 24.6 . Figure 2 2 . 17. AUH 191, Ras Dubay'ah, R2: left astragalus, much abraded. Medial height 35.9 mm, distal width 23.0 mm. in two pieces. AUH 400, Shuwaihat, S2: much of a left Medial height 37.8 mm, distal widd1 mm. AUH 41 1 , Hamra, H3: much of a left ectocuneiform . AUH 1 7 3 , Hamra, H 1 : part o f a proximal left humerus. AUH 1 32, Shuwaihat, S2: distal left humenLs. Width across articu lar condyles c. 38.0 mm. AUH 392, Harmiyal1, Yl : proximal left radius. Maxinmm width 45.8mm, vvidth across articular surface 42.7 mm. Figure 22.17. ACH 223a Ras Dubay'ah, R2: light unciform, damaged anterome. diaUv AUH 32 S Thumayriyah TH1: light lunate.
mm
:
Description The pieces of d1e horn core AUH 237 are much damaged. One piece, about 90 mm long, could be from its because there core on one of appears to be a boundary bet\veen the pedicel its surfaces. The anteroposterior and transverse basal diameters would be c. 40.0 x 30.6 mm, giving an index of compression at 76.5%. The presumed lateral surface is less rounded than the medial surface, but most of the medial surface is missing. The l evel of max imum transverse thickness lies centrally. There is an approach to a posterior keeL There is torsion. A distal piece of horn core, about 60 rnm long, shows a deep longitudinal groove. The pieces of horn core AUH 236 are more poorly than AUH 237, but are probably conspecific. AUH 4 1 3 doubtfully referred t o this Th� postcranial bones above are of a suitable size to be conspecific with the horn cores, but otherwise there is no associa tion with the horn cores. The distal humerus AUH 1 32 has rolled edges . It is slightly smaller than the Samos partial humerus BMNH M4309, which is of a size to go with the Samos Pachytragus radius BMNH M43 1 5 .
c
t
H
Figure 2 2 . 1 7 . A> Supposed base of left horn core of Prostrepsiceros aff. libycus, AUH 2 37; lateral view. C> Cross-section of AUH 237 just above its base, lateral side to the right, anterior to the foot of the page. B> A more distal piece of ( ?the same) horn core, AUH 2 37 . D> Cross-section of the more distal horn-core piece. E> Basal cross-section of a cast, BMNH M32982, of the paratype left horn core of Prostrepsiceros libycus (Lehmann and Thomas, 1987: fig. 7 C, D ) ; lateral side to the right, anterior to the foot of the page. F, Distal cross-section of BMNH M32982. G1 Medial view of left astragalus AUH 2 5 8 ; a = hollowing for medial malleolus of tibia, b absence of indenta tion for naviculocuboid. H> Proximal articular sur face of left radius, AUH 392, assigned by size to Prostrepsiceros aff. libycus> anterior side to foot of page. Scale = 20 mm for A-11 and 1 3 . 3 mm for G and H. =
The condyles are upright and the hollow for the lateral humero radial ligament is not very deep as preserved. The lack of steepness of the sides of the ridge between the medial and lateral condyles suggests bovid rather than cenid affinities. The proximal radius AUH 392 differs from those of Trago portax cyrenaints by the posteromedial part of the medial facet not projecting posreriorly. The medial snrfuce of the astragalus AUH 258 shows large holloVi-ing for the medial malleolus of the tibia and almost no indentation of the back edge for the naviculocuboid 22 . l ;-a. b), both characters being unlike modern Boselaphw. 191 looks similar t o AUH 258. Thev are o f a suitable size t o mat�!: � Sahabi metatarsal (BMNH cast lv132984) of P. libyws Le'o:n 2..:'.::, and Thomas, 1987: pL 3, fig. 3).
Discussion The horn core AUH 2 37 difters trom Pi'o!Trcp�·ict'!'os houtumschi-ndleri of Jmyer and midcLe :\Iaragheh by its greater compression. we��er keels. and
�
"�- VV. GENTRY
less-developed torsion. Among lmown species of Prostrepsiceros it appears closest to P. libycus from
Sahabi. It is similar to this species by its large size, the level of maximum transverse thickness of the
horn core lying centrally, some sign of a posterior keel, weakness of the torsion, and, on the distal
piece of horn core, the presence of a deep longitu dinal groove. In P. libycus this groove lies on the
anterior p lane or edge. AUH 2 3 7 differs, however,
in its greater degree of compression; Lehmann and
Thomas ( 19 8 7 : table 6 ) give an average index for Sahabi P. libycus of 8 3 .6%. The medial surface of
AUH 2 3 7 is too poorly preserved for it to be
apparent whether there would have been, as in
P. libycus, well-marked anteromedial and postero
medial surfaces on either side of a narrowly zoned
level of ma.ximum transverse thickness of the horn core.
Hitherto Prostrepsiceros libycus has been lmmvn
only from its type site at Sahabi. The closest species among the Prostrepsiceros of the Graeco-Iranian
Turolian faunas would be P. fraasi (Andree, 1926),
which is also large and has more compression in the distal part of its horn cores, but which has the
compression oriented anteroposteriorly rather than
BMNH M5071 3, Jebel Barakah, B2: proximal left radius. Maxi mum vvidth 39.1 mm, width across the articular surface 3 5 . 3 mm. Figure 22.19. AUH 321, Thumayriyah, TH1 : epiphysis of a juvenile distal left radius. VVidth across the articular surfaces 34.8 mm. 39.8 mrn, AUH 343, Jebel Dharma, JD3: first phalanx. proximal width 10.8 mm.
Description The horn core AU:H 441 has anteroposterior and transverse basal diameters of29.6 x 20.9 rnm, giving it an index of compression of 71%. The lateral surface is flatter than the medial surface, there is an anterior keel and an approach to a posterior one, it is inserted above the orbit, slightly inclined in side view, not divergent, curving backwards in side view, �'ith slight torsion, there is a shallow postcornual fossa. The radius BMNH M507l3 is smaller than those considered The latunder Tragoportax cyrenaicus or Prostrepsiceros aff. eral fucet is not drawn out posteriorly and the mbercle is larger than in tl'le others. The drop in kvd of the medial relative to the lateral facet in anterior view is much more pronounced than a Samos radius, BA-:INH M4315, possibly belonging to Pachytragtts. The tarsal bones, being larger than Pikermi Gazella capt·icornis (Waguer, 1 848 ), are thereby of an appropriate size to go into bostrepsiceros aff. vinayaki. The naviculocuboid AUH 322 is rather small but not here as a smaller species. It, too, is larger than in G. cat,ricornis The metatarsal AUH 364 is slightly smaller than the cast, BM:KH M32984, of a Sahabi metatarsal assigned by Lehmann and Thomas ( 1987: pl. 3, to P. libycus. It has well-marked hollows above th.e condvlar anteriorly to receive phalanges in
mediolaterally, no posterior keel, an approach to an
anterior keel descending anteromedially rather than anteriorly, no deep longimdinal groove anteriorly,
horn cores more uprightly inserted and more diver gent, and a higher level of the frontals between the
horn cores relative to the orbital rims.
Prostrepsiceros aff vinayaki (Pilgrim , 1 9 3 9 )
Material AUH 441, Hamra, H6: right horn core. Figure 22.1 8. AUH 76, Shmvaihat, Sl: lateral side of a left astragalus. Height 33.4 mm. AUH 279, Jebel Dhanna, JD6: right astragalus. Medial height 30.5 mm. _\1_.1-I 415, Shuwaihat, S l : medial side of a astragalus. Height 29.6 mm. _\1_"H 322, Thumayriyah, Till: leii: naviculocuboid. Al.." H 286, Hamra; H5: part of a right calcaneum. .\1_l-1 +. Hamra, Hl: part of a left calcanemn. .\c.:-I 36-±. Jebel l\llimiyah, M1 : distal metatarsaL Width across 25.0 mm, maximum anteroposterior diameter of 18.7 mm. Figure 22.19.
Figure 22 . 1 8 . A, B; Right horn-core base of Prostrep siceros aff. vinayaki; AUH 44 1 ; anterior and l ateral views. C, Cross-section of AUH 441 just above its base; lateral side to the left, anterior to the foot of the page. D) E; Cast of holotype horn-core base of Prostrepsiceros vin.ayaki) BMNH M42957; anterior and lateral views, drawn as if of the right side. F; Cross section of BMNH M42957; lateral side to the left, anterior to the foot of the page. Scale 20 mm. ==
BAYNUNAH FORMATION
Nisidorcas
ing an index of compression of 76%.
planicomis (Pilglim, 1 9 39 )
of Piram Island and the
Graeco-Turkish Turolian r Bounain,
1987: 199) A
-
1979;
Kohler,
is similarly sized, and compressed to
about the same extent. but has
a
posterior keel
strengthening distalh- and no anterior keel . The drop in ki-d o f
the
m edi al
relati1•e to
the lateral facet in antelior \ ie"� on the radius
M 5 07 1 3 Figure 2 2 . 1 9 . A, B, Distal metatarsal AUH 3 64 and proximal left radius BMNH M 5 07 l 3 , both o f a size to fit Prostrepsiceros aff vinayaki,· anterior views. C, Right horn core AUH 3 32, a possible female of Gazella afi. lydekkeri; l ateral view. Scale = 20 mm. extreme extension and also well-marked hollows on the sides of the condyles for ligaments. The aspect of the bone is of an open-cmmtty or fleet-footed not a skulker in tbickets.
could suggest ..: er,i�i affiniti e s according
to the criteria of Heintz
however, can occur ir: 2'-"' i.is-ror examp l e , n,-o
,-: ; · : ��::::, J c,·p!m lopfm.r among 1 6 sylvic11ltor 1 9 6 1 . 8 . 9 . 8 Ci . C; of Zooltraensis 1 9 65 .3 .2 :i . .2 '� :::c ogy, BMKH. I \\ Ot::,�
for the presence
o�
�Lls s:� C c�'·
Peninsnla. Thi s rad':.:s _ :. : :,, :::;:c: a no::- b o s e -
Prostrepsiceros aff. libycus just
..
Gw::.c ::
having a fairly narrow zone of maximum
anteromedial and posteromedial surfaces, in inclina tion and divergence, and in frontals being at a low
1 75 8 ) .:· �"�,-::-;: ,�; ·i Pilgrim, 1 9 37
Material l:orn core. Anteroposterior and = c . 22.4 x c. 1 8 .2 mm. Figure
AUH
level between the horn cores. It is very s imilar to a
BMNH M429 57, of the holotype horn core vinayaki from the Dhok Pathan of the Siwa liks (see Pilgrim, 1 9 3 9 : 42, pl . 1 , fig. 10; also see Thomas, 1984: 4 1 , pl. 3, fig. 7 for additional mat
cast,
"';;::::� : horn core. Anteroposterior and
of P.
erial) . The resemblance is shown in less compres
libycus,
the level of maximum
transverse thickness being
narrow and thus
.:.:c.-:ccters
The hom core
_-\CH 389
half Its basal diameters
index of 8 pressed
maintains that the posterior keel is stronger than
is also smaller than a cast,
G.
diameters of M42957 being
Jebel Hamrin named as
19.5
mm,
a compresston
latter species in its wealcer backward cur nture than in
x
is quite badly preserved
Gazella capricornis of Pikermi. It BMKH M 1 5 760, of Pilgrim, 1 9 3 7 (figs 35-39), but resem-
larger, the anteroposterior and transverse basal
25.5
12.7 mm. Figure 22. 19.
It is smaller and a little more com-
anterior keel and a slight posterior keel (Thomas the anterior one) , insertion above the orbit, in
x
with the surtace missing over much of its anterior
separating anteromedial and posteromedial compo-
inclination and divergence. It is, however, slightly
1 2 .7
Description and Discussion
nents of the medial
the presence of an
1816
Type spe ci e s G_; :::-�#·��.:. -�·- :·.�;;J Linnaeus,
marked anterior keel. It agrees with the Sahabi P.
transverse thickness and hence some indication of
Sa:nos Prrcbytrngtts
Ge:-:;_;s G.:.:,��:.:
considered, and it also
differs in being more compressed and with a better
sion than in P.
.
.. c
-��
radius B.MXH :\1-±2 1 � .
The horn core is from a smaller species than the
libycus in
E�scx. _-\ dror in
deer radii from
laphine bm·id sraa!.: c·
Discussion
This drop is in
l 9 -0
fact more pronoun..:e�i than in British Pleistocene
G. capricornis. A horn core from Gazella cf. deperdita by
BI!nJ A. W. GENTRY
Thomas, Sen, and Ligabue (1980: pl. 1, fig. 7) is
This is a respectable ruminant fauna, nine
more backwardly curved than in AUH 389 or the
species strong, fi'om which it is possible to derive an
G. lydekkeri cast.
indication of the age of the B aynunah Formation.
The small horn core AGH 332 is about the
Sivailieres are not known from betore the late
same size as a rather large example of the African
Miocene unless one adopts the doubtful proposi
steinbok, Raphicertts campestris. It shows no com
tion that the tour horns of the middle Miocene
pression. It has a preserved length of about
Giraffokeryx indicate sivatheriine affinity. If ilie
35.0 mm and when complete might have been a
? Bramatherium sp. is a sivathere, it agrees better
little over 50.0 mm long. The lateral surface is flat
with late Miocene sivatheres than with those of the
and parts of tl1e otl1er surfaces are probably miss
Pliocene or later.
ing, thereby giving the horn core an irregularly
The postcranial bones assigned to Girafftdae
shaped cross-section. It shows a slight backward
sp. indet. reach a larger size than in middle
curvature. A slight indication of the midfrontal
Miocene Giraffokeryx> and so are more likely to
plane on the medial side of the horn pedicel sug
belong to a late Miocene species. Furtl1ennore,
gests that the insertion was probably fairly upright.
the distal metatarsal of ? Palaeo tragus sp. is more
There was a postcornual fossa. This horn core could be a female or immature specimen of Gazella. Gazella lydekkeri in the Dhok Pathan of the Siwaliks had hornless females (Pil
robustly built than in middle Miocene Girajjoke1·yx> and thus fits better with a late Miocene age.
Tragopor tax cyrenaictts at its type locality is a morphologically advanced species of the very late
grim, 1937: 800), but possible Gazella female horn
Miocene (Geraads, 1 989b: 790 ) . An allied species
cores are present in the Samos collection of the
is found in the presumed early Pliocene of Lange
BMNH (Pilgrim and Hopwood, 1928: 12, pl. 1,
baanweg.
fig. 1). The other alternative tor AUH 332 is that
Pachypor tax occurs in the Nagri and Dhok
it is an early neotragine or a relative of the Nagri
Pailian zones of the Siwaliks, and Proamphibos
(Siwaliks) supposed small boselaphine Elachistoceras
replaces it soon after ilie start of the Tatrot. Hence
Thomas ( 1 977 ) . The latter, however, like many
around 7.0 Ma, or the latest Miocene (see Barry et
neotragines, shows a slightly concave anterior edge
al., 1 991 for the dating of the Siwaliks deposits) ,
of its horn core in lateral profile.
would b e a likely date for a fauna >vith Pachyportax
tatidens. The French boselaphine Parabos cordieri is
DISCUSSION OF THE BAYNUNAH FORMATION GIRAFFIDAE AND BOVIDAE Age and Correlation The ruminant species identified from the B aynunah Formation are as tallows:
also like Pachyportax and occurs in European Mam mal Zone MN l4 (Mein, 1990: 79 ), which equals early Pliocene; hence it is younger than most
Pachyportax. Prostrepsiceros aff. libyars is at least close to the species at Sahabi, while tl1e smaller Prostrepsiceros aff.
vinayaki is close to a late Miocene Siwalilcs species, for which Thomas (1984: 4 2 ) accepts a date of around 7.0-7.5 Ma. Together, the two Abu Dhabi
Family Giraffidae
? Palaeotragus sp. ?Bramathr:rium sp. Giraffidae sp. indet.
Family Bovidae Tribe Boselaphini
Tragoporrax cyrenaict�s Pachyportax la.tidens
Tribe indet. Tribe Antilopini
Bovidae sp. indet. Prostrepsiceros aff. tibycus
Prostrepsiceros aft: vinayaki
Gazella aff. lydekkeri
species suggest a date in the latest Miocene. Gazella
lydekkeri> which the limited morphological knowl e dge ofilie gazelle horn core AUH 389 suggests as a possible near relative, is again a late Miocene species. The Baynunah ruminants hence support an age for the formation of late in the Miocene, perhaps equivalent to MN 1 3 in terms of the European Mein zone scale and around 6.0 Ma ago in terms of dating.
PECORANS FROM THE BAYNUNAH FoRMATION
Zoogeogra phy and Palaeoecology
-
dance of spiral-horned i\ntilopini (all the listed �.A,ntilopini except
Gazella)
is also noteworthy; they
The Baynunah ruminants can be compared with
make up 38% of the listed species. Caprini and
u'lose of late Miocene faunas in adjacent continental regions. Making comparisons at regional level
archaic and specialised 0Yibovini are present here 1 but absent at Abu Dhabi. Cervidae occur sparingly
rather than with particular localities increases the
in the Graeco- Iranian Turolian and there are even
deficiencies of temporal resolution. The beginning
rare records of Traguhdae, but as yet there are no
of the late Miocene at l 0 Ma ago is twice as old as
records of these families from the B aynunah For
its end, so if the B aynunah fossils are indeed close
mation. The presence o f a redundne in the Graeco
to the end of the Miocene in age, we shall be com
Iranian list is based onlY on a
paring them with some faunas that are twice as old.
Turkey ( Kohler, 1987: 2 14 ) , and the tribe was not
from MN 1 3 in
Hence it becomes more difficult to separate effects
a normal component of MX I 1 and ,\1::\ 1 2 faunas
of geography from those of time.
in this region.
Below tribal level the B aynunah pecorans show
( "Redunca"
erem opolitana Erdbrink,
1982 from the Maragheh area is based on a Pachy
little resemblance to the rich Turolian assemblage
tragus or some later caprine like Norbenia.) The
known from the Graeco-Iranian region. The fol
differences at species level in Tragoportax and
lowing list of giraffids and bovids for tlut region is
Prostrepsiceros suggest that the B aynunah Formation
based mainly on Solounias ( 19 8 l a), Bernor ( 1986),
is of later date.
Kohler ( 1987), and Kohler et aL ( 1995 ) .
At Langebaanweg the following species are pres ent ( Harris, 1 9 76; Gentry, 1 9 8 0 ) :
Family Giraffidae
Family Bovidae
Tribe Boselaphini Tribe ?Bovini
Tribe Reduncini
Tribe Antilopini
Tribe 0\�bovini
Tribe Caprini
Palaeotragus, 2 spp. Samotherium boissieri Decennathe;•ium macedoniae He!ladothe;·iwm duvernoyi Bohlinia attica Tragoportax, 3 spp., not incl. cyrmaims Samokeros minotaurus ?Redunca sp. Prostrepsiceros, 4 spp., not incl. /:ibycus or vinayaki Protragelaphus, 2 spp. Palaeoreas, 3 spp. Nisidorcas planicomis ( ?Hispanod01·cas) rodleri Oioceros, 3 spp . Samodorcas kuhlmanni Gazell , 2 spp. not incl. lydekkeri Urrniatherittm, 2 spp. Plesiaddax inundatus Criothcrittm m;galioides Palaem·�x pallasi1 � Protm·yx carolinael Pachytragus, 2 spp 1 Norbertia hel!enica Pseudotragtts, 2 spp 1
a
Family Giraffidae Family Bo�dae
Tribe Tragelaphini
Tribe Boselaphini
Tribe Bovin i
Tribe Reduncini
Tribe Alcelaphini
Tribe Neotragini Tribe Antilopini
Tribe 0\ibovini
Girajja sp. Sivatherium hmdeyi Tragelaphus, ? 2 spp. TragopM·tax acrae Simathet·ium demissum Kobus, 2 spp. Damalacra, 2 spp. Raphice1·us paralius Gazella sp. Gen. et sp. indet.
This is a fauna >vith an early bovine and the African tribes Tragelaphini, Redtmcini, Alcelaphini, and Neotragini. There are also a boselaphine and ovibovine looking like Miocene immigrants from far to the north. Most of the species show primi tive characters and are thereby differentiated from later species of East and South Africa. Only the boselaphine has any resemblance to a species in the Baynunah Formation and this is likely to reflect the fact that both the species concerned are late and advanced representatives of Tmgoportrr.x.
These faunas contrast with that of Abu Dhabi in the definite presence of the distinctive mixed
The presence of African tribes of antelopes is more likely to show a zoogeographical than a temporal
feeder or grazing giraffe Samatheriu.m baissieri
difference, but the bovine could indicate a slightlY
(Solounias et aL, 1 9 8 8 ) . The much greater abun-
later date for LangebaamYeg.
11m A. W. GENTRY
In East Africa, at Mpesida, Lukeino, and the Manonga Valley (Thomas, 1980; Gentry, 1 997), the as: combined giraffid and bovid tauna can be Family Giraffidae
Tragelaphini Tribe Bovini Tribe Cephalophini
I1·agelaphus sp. U gandax cf. gatttieri Cephalophus sp.
Tribe Reduncini
Kobus aff. subdoltts Kobus aff. pot"'?cticornis
Tribe Hippotragini
Praedamali.r sp.
Tribe i\lcelaphini
Damalacra sp. Alcelaphini, 2 more spp.
Tribe Aepycerotini
Aepycems sp. lvfadoqua sp.
Tribe Neotragini Tribe Antilopini
? Gazella sp.
This assemblage has some resemblance to Langebaanweg, but Cephalophini, Hippotragini, and Aepyceros are additional African constituents. The only possible resemblance to the Bavnunah Formation pecorans seems to be the pre�ence of a sivatheriine giraffid at a comparable stage of evolutionary development. This ? Sivatherium sp . could as easily belong to Bramatheriunt by what little is known of its primitive charac ters. There are no B oselaphini or spiral-horned Antilopini, although boselaphines are present ear lier in tl1e East African Miocene and Tragoportax at 1. Also Nakaya et al. ( 1984: 1 09, pl. 9, 5 ) ) recorded a possible spiral-horned antilopine Palaeoreas sp. ) in the earlier Namurungule and Smart ( 1976) listed A-ntilope sp. for Lothagam l . The North African pecorans, a s known from Bled ed-Douarah, Sahabi, and Wadi Natrun ( Ger aads, 1989b; 1 987; Lehmann and Thomas, 1 9 8 7; also see comments in Gentry, 1 98 0 ) , are: Family GiraHidae Family Bovidae Tribe Boselaphini Tribe Bovini
Giraffidae, sp. indet. SiJ>atherium ati. hendeyi Tragoportax cyrenaicus "Leptobos" syrticm
Tribe Reduncini
Kobus sp.
Tribe Hippotragini Tribe i\lcelaphini Tribe Neot:ragini
Damalacra sp.
Tribe Antilopini
This list is quite close to the Abu Dhabi fauna with a conspecific boselaphine and spiral-horned antilopine, and two giraffids, but it ditters in the presence of several African tribes. It needs to be remembered ( Geraads, 1989b) that several strati graphic levels are represented at these localities, probably ranging up into the Pliocene, and some of the _;\frican tribes of antelopes may be present only in higher levels. Towards the end of the Miocene in the Siwa liks, in the Dhok Pathan, one finds the following giraffids and bovids ( Pilgri m, 1 937, 1939; Barry et aL , 1982, 1 99 1 ; Barry, 1 99 5 ) : Family Giraffidae
Boselaphini
Giraffa punJabiensis Bramatherium megacephalum Tragoportax salmontanus Tragopm'tax browni Pachyportax latidens
Tribe Reduncini Tribe
NrnlOIOllll
Kobus porrecticornis Prostrepsiceros vinayaki Gazelltt l;,dekkeri
This late Miocene fauna is very like that of Abu Dhabi in the Pachyportax and spiral-horned antilop ine, but the Tragoportax are ditierent at species level. If the really is a GirajJ'a, the metapodials would be too long to agree with the Bay nunah Formation Giraffidae, sp. indet. It is likely that smaller sivatheriines are present in addition to Bramathe�·ium megacephahnn, and this could be a resemblance to the Baynunah Formation 1 Bra matherium sp . , if the latter should also be a mod estly sized species. The presence of late Miocene Reduncini shows both that this tribe had evolved before the end of the Miocene, and that the Siwa liks and, by implication, Arabia and North Mrica h ad a zoological connection with tl1e whole of Africa at this period. Mter th e end of the Miocene, Cervidae, the bovine Proamphibos, and possibly Hippotragini appear in the Siwaliks while boselaphines become rare or temporarily absent. This fauna is much less similar to that of the Baynunah Formation and sug gests that the latter cannot be as late as the Pliocene. One fossil fauna with pecorans that deserves mention in comparison with the Baynunah Forma tion is that of Piram (formerly Perim) Island, Gulf
PECOR.\:-.rs
THE BAYl'D.NAH
FORMATION
D
M 3 6 8 3 , the holotype of Cambayella 1vatsoni
of Cambay, India ( Prasad, 1 974) . From the collec
Pilgrim, 1 9 39, may be the same species.
tion in the B MNH and from the older literature, 6.
especially Pilgrim ( 1 93 9 ) , the pecorans appear to be:
Prasad ( 1974: 1 5 , pl. 5 , fi g. 7) referred a small Piram Island horn core to Rttticeros compressa)
l.
2.
3.
Giraffidae, smaller sp . : a left ma.xilla with
sp . nov. The genus Ruticeros is otherwise
P3-M3, B MNH 37259, about the size of
known by horn cores of R. pugio Pilgrim ,
Samotheritun boissieri (see Matthew, 1 929 :
1 9 3 9 , supposedly of late Miocene age. They
551).
are compressed, with a flat lateral surface, slight
Bramatherit1 pp. 3 1 72 1 (ed. N. T. Boaz, A. El-Arnauti, A. W. Gaziry, J. de Heinzelin, and D. D. Boaz). Alan R. Liss, New York. Harrison, T. J. ed. 1 997.
Neogene Paleontology ofthe
Manonga Valley, Tanzania: A Wzndow into East
African Evolution. Topics Plenum Press, New York.
in
Geobiology vol. 14.
Heintz, E. 1970. Les cervides Villafranchiens de France et d'Espagne. Mhnoires du Musl:unz National d 'Histoire Naturelle, Paris 22 : 1-303. Heintz, Brunet, M., and Sen, S. 198 1 . Un nou veau Gi.raffide du Jvliocene superieur d'Irak: Injana therium hazirni n. g., n. sp . Cornpte Rendu Hebdo madaire des Seances de t>Academie des Sciences, Paris
292: 423-26. Hill, A. 1987. Causes of perceived taunal change in the later Neogene of East Afnca. Journal of Human Evolution 16: 583-96. Hill, A. 1995 . Faunal and environmental change in the Neogene of East Africa: Evidence from the Tugen Hills Sequence, Baringo District, Kenya. In Paleocti lnate and Evolution, with Erttphasis on Human Ori gins,
pp. 1 78-93 (ed. E. S. Vrba, G. H. Denton, T. C. Parttidge, and L. H. Burlde). Yale University Press, New Haven. Hill , A., Drake, R. , Tauxe, Monaghan, M., Barry, J. C., Behrensmeyer, A. K., Curtis, G., Jacobs, B. F.,
Jacobs, L., Johnson, N., and Pilbeam, D . 1985. Neogene palaeontology and geochronology of the Baringo Basin, Kenya. Journal ofHuman Evolution 14: 759-73. Kohler, M. 1 987. Boviden des turkischen Miozans (Kanozoikum und Braunkohlen der Turkei 28). Pale ontologia y Evolucio 2 1 : 1 3 3-246. Kohler, M., Moya-Sola, S., and Morales, J. 1995. The vertebrate locality Maramena (Macedonia, Greece) at the Turolian-Ruscinian boundary (Neogene) . 1 5 . Bovidae and Giraffidae (Artiodactyla, Mammalia). i14iinclmer Geowissenschaftliche Abhandlungen A 28: 167-80. Leal�.ey, M. G., Feibel, C. S., Bernor, R. L., Harris, J . M., Cerling, T. E., Stewart, K. M., Storrs, G. \¥. ,
Walker, A. , Werdelin, L., and Winkler, A. J. 1996.
licacione.• de Geologia, Universidad Aitro:wnw df
Lothagam: A record of faunal change in the late
Barcelona 1 8 : 1-236.
Miocene of East Africa. ]ottrna! of Vertebrate Palaeon
tology 16: 5 56-70 .
Nakaya, H., Pickford, M., Nakano, 1. and Ishida. H. .
1984. The late Miocene large mammal tinma from Lehmann, U., and Thomas, H. 1987. Fossil Bovidae fi· om the Mio-Pliocene of Sahabi, (Libya). In Neogene
Paleontology and Geology of Sahabi, pp. 323-35 (ed.
X T. Boaz, A. El-Arnauti, A. W. Gaziry, J. de Heinzelin, and D. D. Boaz). Alan R. Iiss, New York.
the Nanmrungule Formation, Samburu Hills. north em Kenya. African Study Monographs, suppl . issue 2 : 87-1 3 1 . Pickford, M. 1988. The age(s) of the Bugti faunat s : , Paldstan. In The Palaeoenvironment of East A1ia .fi ·o m
Lydeldcer, R. 1 87 8 . Crania of ruminants from the Indian Tertiaries. Memoirs of the Geological Survey of
India. Palaeontologia Indica 1 : 88-1 7 1 .
---
. 1 8 8 5 . Catalogue o f the Fossil Mammalia in
the Bt·itish Museum (Natural History), II- Artio dactyla, Boridae. British Museum (Natural History), London.
---
. 1 886. Siwalik Mammalia-supplement l.
1'Wemoirs of the Geological Strrvey of India. Palaeontolo gia Indica 4: 1-2 1 . Matthew, W D . 1929. Critical observations upon Siwalik mammals. Bulletin of the American Musuon of
Natu1·a! HiJ1:ory, Ne111
Yin·k
56: 4 3 7-560.
Mecquenem, R. de. 1924-2 5 . Contribution a !'etude du gisement des vertebres de Maragha et de ses envi rons. Annates de Paleontologie, Paris 1 3 : 1 3 5-60; 1 4 : 1-36.
( 1 989).
lJpdating of M)l' zones. In
European Neogene Afarnmal Chronology, pp. 7 3-90
H.
Kong. Pilgrim, G. E. 1 9 3 7 . Siwalik antelopes and oxen in the American Museum of Natural History. Bulletin of
the Ame1·ican }..J.uJeum of Natural Hi�1:ory, New York 72: 729-874. ---. 1939. The fossil Bovidae of lndia. Memoirs of
the Geological Surrey of India. Palaeontologia Indica, n.s. 26: 1-356. Pilgrim, G. E., and Hopwood, A. T. 1 928. Catalogue
of the Pontian Bovidae of Europe. Trustees of the British Museum (Natural History), London. Prasad, K. N. I 9 74. The vertebrate fauna from Piram Island, Gujarat, India. Memoirs of the Geological Sur
vey of India. Palaeontologia Indica, n.s. 4 1 : 1-2 3 . Schlosser, M. 1904. Die fossilen Cavicornier Yon
Mein, P. 1 990 (ed. E.
the Mid-Tertiary, pp. 9 3 7-5 5 (ed. P. Whyte) . Centre
of Asian Studies; lJniversity of Hong Kong, Hong
Lindsay, V. Fahlbusch, and P. Mein).
Plenum Press, New York. Morales, J . , Soria, D., and Thomas, H. 1 9 8 7 . Les
Samos. Beitriige zur Palaontologie zmd Geologie Orter
reich-Ungarns und dn Orients 1 7 : 28-1 1 8 . Smart, C . 1 976. The Lothagam 1 fauna: Its netic, ecological, and biogeographic significa.:Ke. In
Ea1·liest Man and Environments in t!Je Lake
Giraffidae (Artiodactyla, Mammalia) d'Al Jadidah du
G. Ll. Isaac, and R. E. F. Leake\· i .
.Miocene Moyen de la Formation Hofu:f (Province du
Chicago Press, Chicago.
Rasa, Arabie Saoudite). Geobios 20: 44I -67.
Rurfo(f
Basin, pp. 361-69 ( ed. Y. Coppens. F. C. HowelL of
Solounias, N. 1981a. 11arnmalian tossils of Samos a.:1d Moya-Sola, S. 1 9 8 3 . Los Boselaphini (Bovidae
Pikermi. 2. Res urrection of a classic T:1rolian tauna.
Mammalia) del Neogeno de la peninsula Iberica. Pub-
A nnalJ Gf Camegie Jfztsw;n. Pitr.cim;p/J 5 0 : 23 1-69 .
D
A. w. GENTRY
. 1981b. The Turolian fauna from the island of Samos, Greece. ContributiortS to Vertebrate Evolution
---
--- . 1984. Les bovides ante-hipparions des Siwa liks inferieurs (plateau du Potwar), Paldstan. Mhnoires
6: l-232.
de !a Societe Geologique de France, Paris
Solounias, N., Teaford, M., and Walker, A. 1988. Interpreting the diet of extinct ruminants: The case of a non�browsing giraffid. Paleobiology 14: 287-300.
Thomas, H., Sen, S., Khan, M., Battail, and Ligabue, G. 1982. The Lower l'vliocene Fauna of A.l-Sar rar (Eastern Province, Saudi Arabia) . AnAL, The Jour nal of Saudi A rabian Archaeology
Smart, A. J. 1982. Pleistocen.e Vertebrates in the British Isles. Longman, London. Thomas, H . 1979. Miotragocerus cyrenaicus sp. nov. (Bovidae, Artiodactyla, Mammalia) du Miocene superieur de Sahabi (Libye) et ses rapports avec les autres Miotragocerus. Geobios 12: 267-8 1 .
. 1980. Les bovides du Miocene superieur des couches de Mpesida et de Ia formation de Lukeino ( District de Baringo, Kenya). In Proceedings of the 8th
---
Panafricart Congress of Prehistory and Q;tatemary
Studies, Nairobi, 5 tol O September pp. 82-91 (ed. R. E. F. Leakey and B. A. Ogot). The Interna tional Louis Lcalcey Memorial Instimte for Afiican Prehistory, Nairobi.
---. 1983. Lcs BoYidae (.l\rtiodactyla, Manunalia) du Miocene moyen de la formation Hofuf (PruFince du liasa, Arabie Saoudite). Palaeovertebrata 1 3 :
157-206.
145 : l-68.
5: 109-36.
Thomas, H., Sen, S., and Ligabue, G. 1980. La faline Miocene de l a Formation Agha Jari du Jebel Hamrin (ll'ak). Proceedings Koninkli;J<e Nederlandse Akadetrtie van Wetenschappen B
83: 269-87.
Whybrow, P. J. , Friend, P. F., Ditchfield, P. W. , and Bristow, C. S. 1999. Local stratigraphy of the Neo gene outcrops of the coastal area: Western Region, Emirate of Abu Dhabi, United Arab Emirates. Chap. 4 in Fossil Vertebrates ofArabia, pp. 28-37 ( ed. P. J. vVhybrow and A. Hill) . Yale University Press, New Haven.
Late Miocer1e Baynunah Formation, Emirate of Abu Dhabi, United Arab Emirates : Fauna, Flora, and Localities PETER J.
The jebels scattered along the coastal plain of Abu Dhabi's Western Region are the only vertebrate bearing late Miocene rocks known from Arabia. .\ lost of the tossils collected by The Natural His tory Museum/Yale University team described in this volume come from outcrops in the Baynunah Formation located between the road from As Sila to Abu Dhabi and the coast (Whybrow, 1 98 9; \\11ybrow et al., 1 990; see also Whybrow et al. , 1 999-Chapter 4) . Here, fluvial clastics deposited by a major river system (Friend, 1 999-Chapter 5 ) contain remains of invertebrates, reptiles, birds, and mammals as well as poorly fossilised plant materiaL Few vertebrate fossils have been found at localities south of the main road. The j ebels in this area lack the gravels and coarse sandstone lithologies seen in the lower part of the Baynunah Formation at the coastal exposures; sediments forming these south ern j ebels might be higher in the Miocene sequence 1 the "upper" part of the Baynunah in Friend, 1 999-Chapter 5 ) . The most eastern fossiliferous locality is south of Tarif. Here a magnificent partial skull of Tragoportax cyrenaicus (Gentry, 1 999b Chapter 2 2 ) was the only fossil found in homoge neous red-coloured sandstone (a road-cut expo sure) by staff of the United States Geological Survey ( Hill et al. , 1 999-Chapter 3 ) . Excepting the sea cliff outcrops of the nunah Formation, "clean" exposures of ll,1iocene sediments are rare. The region is hyperarid and during the 1 5 years or so of work by the NHM/ Yale team torrential rain h as occurred just five times. Detritus, a mixture of weathered Miocene
Copyright © 1999 by Yale University.
WHYBROW
AND D IANA
CLElvl E:\TS
sediments v-rith modern wind-blown sand, on the slopes of most jebels can be 20 em deep before unweathered rock is reached. Although the Baynunah Formation crops out in an area of about 1 8 00 km2 , the fossils identified so far (more than 900) come from a 560 km2 area. Except for the disarticulated, scattered bones of a proboscidean found at Shuwaihat (Tassy, 1 999; Andrews, 1 999-Chapters 1 8 and 24), almost all vertebrate fossils are found as isolated elements, often fragmented by extremes of temperature and rehydration of microcrystalline gypsum during pre cipitation of \\inter fog and dew. During the period that the NHM/Yale team, and others, have been finding fossils from the Bay nunall Formation, we have collected all material from each locality that might be identifiable so that a possible collecting bias is minimised (Hill, 1 987). Some bias has been unavoidable, however, because further collecting at some localities h as been impos sible, especially from 1 992 . The development of the Western Region of Abu Dhabi proceeds rapidly. Access to some localities that were relatively easy to visit in the early 1 980s is now prohibited. vVe haYe suggested to the Abu Dhabi authorities that some localities should be conserved for the scientific itage of the Emirate as well as for internatimul science. This suggestion has been ( 1 99 ;) received and i t is likely that Shuwaihat \\ J[ 2c recognised as a locality of scientific Here we list and describe the localir:cs c::�-� sites ( vvithin the localities) , trom \l·esr . ,fauna and flora, and present a
All rights reserved. ISBN 0-300-0- 1 3 3 - �
11m
P. J. WHYBRO\V AND D. CLEME:-:Ts
j ..
--
---
., .
r---
�
t �-
.
_,
.L- -
-·
�
. i>IJTil" - -----
J
Plate 2 3 . 1 . Jebel Barakah viewed from the west. The gravel bed, top left, is at the base of the lower part of the Bavnunah Forma rio� ; the remaining outcrop, below, is the Shuwaihat For mation .
· ....-.:... 't i ii o!!l � -�
·
Plate 2 3 .2 . Shuwai hat, site 56. Gillian Comerford (The Natural History Museu m ) begins the excavation of the proboscidean (Tassy, 1999-Chapter 1 8 ) . The lower jaw is seen at bottom right and the "planed off' cranium is to the left (Andrews, 1999-Chapter 24 ) .
--
h l " l l i .OL\I .ITII:� I� l"Hb. B.W N L1NAH FORMATION
-
Plat..: 2 3 . 3 . Sou th r:K i n g escarpment of the H;lmra localitv
H 5 ) . From the rhnhmic da,· la v . e rs , bottom ri g h t pc Africa : Paleosol carbon isotopic eYicknce.
Human Evolution 27: 25-45 . Sillen, A. 1986. Biogenic and d:agc:netic Plio-Pleistocene fossils of th e Omo tion. Paleobiolo'-rry 1 2 : 3 1 1-2 3.
forma-
&m
J. D. KINGSTON
Singh, J . S . , and Gupta, S . R.
1977.
Plant decomposi
Tiezsen, L. L. , Tesdahl, K G., Boutton, T. W., and
tion and soil respiration in terrestrial ecosystems.
Slade, N. A.
Botanical Review 43: 449-528.
ble carbon isotopes in animal tissues: Implications for
1983.
Fractionation and mrnover of sta
the 13C analysis of diet.
Smith, B . N . , and Epstein, S. 13Cj12C ratios for higher
1971. Two categories of plants. Plant Physiology 47:
380-84. K.,
Teaford, M . , and "Walker, A.
1988.
Interpreting the diet o f extinct ruminants: The case of a non-browsing giraffid.
Paleobiology 14: 287-300.
Sternberg, L. S. L., Mulkey, S. S., and Wright, J. S.
Ecological interpretations of leaf carbon isotope
ratios: Influence of respired carbon dioxide.
Ecology
70: 1 3 1 7-24.
and Nowak, R. S.
1989.
Oecologia 80:
1 1-18. van der Merwe, N. J . , and Medina, E.
1989.
Photo
synthesis and 13C/12C ratios in i\mazonian rain
forests.
Geochimica et Cos·mochimica Acta 53: 1 091-94.
van der Merwe, N. J., and Vogel, J. C.
1978.
13C
content of human collagen as a measure of prehistoric
Sullivan, C. H . , and Krueger, H. W.
1981 .
Carbon
isotope analysis of separate chemical phases in modern and fossil bone.
---. 1983.
Nature 292: 333-35.
1999.
diet in woodland North America.
Nature 2 76:
815-16. Vogel, J . C., and van der Merwe, N. J .
Carbon isotope ratios of bone apatite
and animal diet reconstruction. Tassy, P.
Anderson, J .
\.Yater use efficiency and carbon isotopic composition of plants in a cold desert environment.
Solounias,
19 89.
Toft, N.
Oecologia 37: 337-50.
Nattwe 301: 177-78.
Miocene elephantids (Mammalia)
1977.
Isotopic
evidence for early maize cultivation in New York State.
Amer·ican Antiquity 42: 238-42.
vVang, Y. , Ceding, T. E . , Quade, J., Bowman, J . R.,
from the Emirate of Abu Dhabi, United Arab Emi
Sinith, G. A. , and Lindsay, E. H.
rates: Palaeobiogeographic implications. Chap.
topes of paleosols and fossil teeth as paleocology and
1 8 in
�Fossil Vertebrates ofArabia) pp. 209-33 (ed. P. J . Why
brow and A. Hill). Yale University Press, New Haven. Thackeray, J. F. , van der Merwe, N. J., Lee-Thorp ,
1993.
Stable iso
paleoclimate indicators: Ari example from the St.
David Formation, Arizona. Climate change in conti
nental isotopic records.
Geophysical Monograph 78:
241-48 .
J . A., Sillen, A. , Lanham, J. L . , Smith, R. , Keyser, A., and Monteiro, P. M. S.
1990.
Changes in carbon iso
Whybrow, P. J . , and McClure, H. A.
1981. Fossil
tope ratios in the late Permian recorded in therapsid
mangrove roots and palaeoenvironments of the
tooth apatite.
Natur·e 347: 75 1-53.
Palaeogeog r·aph» Palaeoclimatology) Palaeoecology 32: 2 1 3-25 .
Tieszen, L. L.
199 1 .
Nanrral variations in the carbon
isotope values of plants: Implications for archaeology, ecology, and paleoecology.
Journal of Archaeological
Sciertce 18: 227-48. Tieszen, L. L., Senyimba, M. M., Imbamba, S. K , and Troughton, J. H.
and
1979.
The distribution of C3
grasses and carbon isotope discrimination
Miocene of rl1e eastern Arabian peninsula.
vVhybrow, P. ]., Hill, A., Yasin al-Tiktiti, W., and Hailwood, E. A.
1990.
Late Miocene primate fauna,
flora and initial palaeomagnetic data from d1e Emirate of Abu Dhabi, United Arab Emirates. Journal of
Human Evolution 19: 583-88. Woodward, F. I.
1990.
Global Change: translating
along an altitudinal and moisture gradient in Kenya.
plant ecophysiological responses to ecosystems.
Oecologia 37: 337-50.
in Ecology and Evolution 5: 308-10.
Trends
Earliest Stone Tools from the Emirate of Abu Dhabi, United Arab Emirates SALLY 1\IcBRBRrr
�-\rtificially fractured fragments of chert signal the
level no doubt vastly accelerated the erosion of
first entry of the human genus Homo into Abu
Baynunah Formation sediments, which must for
Dhabi. At most outcrops of the Baynunah Forma
merly have covered an area of many tens if not
tion, the sequence is capped by a thick layer of 1 resistant tabular chert. The horizontal disposition
hundreds of thousands of square kilometres
of this unit lends the characteristic flat-topped
( Friend, 1999-Chapter 5 ) . The Arabian Gulf today is quite shallow, with a
appearance to the jebels of Abu Dhabi's Western
mean depth of only about 35 metres, and thus
Region. Weathering of superficial Baynunah Forma
recurring drops in sea level caused its episodic
tion sandstones, probably under conditions moister than those at present, resulted in the solution of
retreat or even disappearance . The Gulf's current maximum depth of about 165 metres lies near the
quartz grains and the redeposition of silica in the
Iranian shore . While relative shifts in plate loca
form of silcrete (Ditchfield, 1999-Chapter 7 ) . It is
tions, local uplift, downwarping, and sedimentation
unclear whether this process occurred subaerially or
may be expected to have affected local topography,
at a subterranean bedding plane, but it may have
a marine remnant or embayment probably persisted
required as much as a million years to complete
here during glacial periods, except when sea levels
I C. S . Bristow, personal communication. ) . Tllis
were at their lowest. During the last glacial maxi
siliceous diagenetic product provided the raw mate
mum, however, the Arabian Gulf is lmown to have
rial for implement manufacture by Abu Dhabi's earliest toolmakers.
retreated beyond the Straits of Hormuz. At such
When humans first occupied the Emirate, Bay
times western Abu Dhabi would have been sepa rated from Iran by a sandy plain. The combined
nunah Formation rocks were already very ancient.
discharge of the Tigris and Euphrates may have
The landscape has apparently been attritional for a
emptied into the Gulf of Oman, though it is not
considerable period of time, as there is no remain
certain that the volume of water would have been
ing trace of any local sedimentation subsequent to that represented by the late Miocene Baynunah
sufficient to maintain flow ( Kassler, 1973; Rice,
Formation. The location of the shoreline of tl1e �-\rabian Gulf fluctuated dramatically during the
Occasional periods o f increased precipitation in the Late Pleistocene and early Holocene are
Pleistocene. During the most recent glacial maxi
demonstrated by widespread shallow lacustrine
mum at 1 8 000 years ago, global sea levels are
deposits in the interior of the Rub' al Khali, ,,-ith
1994 ) .
known to have fallen by about 1 1 5 metres, and
associated vertebrate fauna and traces of human
drops of similar magnitude occurred periodically
habitation ( Zenner, 1954; Field, 1 9 5 8 , 1 9 6 0a..b:
throughout the Pleistocene at intervals of roughly
Masry, 1974; McClure, 1 976, 1988
1 0 0 000 years (Imbrie and Imbrie, 1 9 8 0 ; B erger et
ditions may have prevailed intermittmtlY at earlier
and Straight, 1991 ). This dramatic lowering of base
Peninsula in the Early or :\liddle Ple istocene is
al, 1 984; Ruddiman and Wright, 1987; Johnson
'·
Similar con
times, and hominid occupation of the _-\.rabia.n
Copyright © 1999 by Yale University. All rights reserved. I S B� 0-300-0:-lS3-3
mJ s. McBREARTY
attested by the occurrence of heavily patinated han
asl. The raw material at Shuwaihat is a silicified
daxes and other stone tools with a pronounced
limestone, usually light yellow, but sometimes
Middle Palaeolithic aspect at a number of sites in
weathering to a blackish colour. While its quality is
the Kingdom of Saudi Arabia ( Masry, 1 974; Zarins
poor, the edges of most artifacts are quite sharp .
et al., 1 98 0 , 1 9 8 1 ) . Sea levels in the current inter
There are no formal retouched tools in the col
glacial are near their apparent maximum, but even a
lection from Shuwaihat. Ratl1er, all artifacts consist
small rise in sea level, such as the 2-metre rise at
of cores and the resulting debitage, though a few
about 7000 years ago, resulted in large-scale flood
elongate high-backed cores on slabs might be clas
ing in this region of low relief. At such times some
sified by some as "pushplanes" or heavy-duty scrap
resistant outcrops of coastal Baynunah Formation
ers (fig. 2 6 . l a) . Cores are rather large, with a mean
rocks may have been isolated as islands. The famil
size of l l 6 mm ( range 7 5- 1 6 3 mm,
iar sabkhas of Abu Dhabi are thought to be fairly
of these ( 43 . 8 % ) are radial forms, and include disc,
recent in origin, the result of a marine transgression
subradial, and high-backed types (fig. 26. l b,c) .
perhaps dating to no more than 4000 years ago
Raw material at Shuwaihat may seem unpromising
( McClure, 1 97 6 ) .
rt =
1 6 ) . Seven
for blade manufacture, but two of the Shuwaihat cores ( 1 2 . 5%) are bidirectional blade cores (fig. 2 6 . 1 d) , and one is a crescent-shaped blade core
ABU DHABI 's STONE ARTIFACT OCCURRENCES
( 19 8 3 ) from the Pre- Pottery l'Jeolithic B (PPNB)
Stone tools made fi·om siliceous cap-rocks are
levels at Jericho, Palestine. Technologically they
found draped over several Baynunah Formation outcrops ( McBrear ty, 1 99 3 ) . Isolated artifacts have been observed at other localities in the area, but the largest numbers were encountered at Shuwai hat, Jebel B arakah, Hamra, and R.:1.s al Aysh; discus sion here will be confined to these four localities. All artifacts were found on the surface of Baynunah or Shuwaihat Formation o utcrops or in very shal low superficial deposits produced by their erosion. Primary artifact collection was uncontrolled, in that objects were selected fro m the total site area to provide an idea of the range of material present. Frequencies of artifact categories may therefore not be strictly representative, but at all four sites addi tional small controlled collections were made to determine artifact density and size distribution.
Shuwaihat
(fig. 2 6 . 1 f). These distinctive cores are identical to "naviform" cores described by Crowfoot-Payne
resemble the bifacially prepared plaquettes on tile flint described by Inizan ( 1 9 80b, 1 9 8 8 ) for Qatar. The Shuwaihat cores were radially prepared around their circumference and then blades were removed fro m an axis perpendicular to that of the radial striking platform. The first blade removed from such a prepared core has a d orsal "crest" of inter secting flake scars, the remains of the radial striking ( c[ fig. 2 6 . 3 n ) . Blades removed subsequently show parallel dorsal scars. Upon close examination, three additional Shuwaihat cores (fig. 2 6 . l g,i) and two bifacially trimmed slabs (fig. 2 6 . 1 h) exhibit attempted blade removals and four show breakage that probably resulted from failed attempts at blade removal. O nly one true blade was observed. Rather, Shuwai hat debitage consists of flakes vvith breadth/length ( B/L) ratios of approximately 1 .0 , and a simple pattern of dorsal scars.2 Although most artifacts at
Artifacts occur here in the vicinity of the Shuwai
Shuwaihat have a some\vhat crude appearance, their
hat, site S2, Miocene fossil-collecting area ( N
lack of elegance reflects primarily the poor quality
24° 0 6' 4 1 .7", E 5 2 " 2 6' 04.0 " ) . They are scattered
of the lithic raw material. Indeed, multiple bulbs
oYer an area of about 30 0 0 0 m2 on the sea cliffs
of percussion show that several blows were some
and wave-cut platform on the south and west faces
times required to detach flakes from the core (fig.
of the jebel below an elevation of about 40 metres
2 6 . 1 j ,k ) .
Figure 26. 1 . Lithic artifacts from S huwaihat: a , elongate high- backed core or " p ushplan e " ; b, radial core; c, high - backed radial core; d, bidirectional blade core ; c, sin gl e platfo r m core; f, crescent-shaped or "naviform" blade core; g, radial core with attempted blade remoYal; /;, bifacial ly tri mmed slab with attempted blade removal; i, single-platform core with attempted blade remo,·al; ;; fND A. HlLL
to extant species (Hill, 1995), a series of significant
straddles the subtropical high-pressure belt and
synchronous first and last appearances in the fossil
much of the area is today climatically arid or semi
record of the Siwalik succession in Paldstan ( Barry
arid, dominated by subtropical deserts. Several
et al. , 1990; Barry, 1995), the spread of grassland
major climatic regimes intertace in this region and
adapted fauna (Gabw1ia and Chochieva, 1982;
the climate and weatl1er of the Arabian Peninsula
MacFadden and Ceding, 1994), and the origin of
are influenced by
the human lineage ( Hill and Ward, 1988; Hill,
combinations of high- and low-pressure systems
1994). By late Miocene Baynunah times, palinspastic (palaeogeographic) reconstructions of the region
a
complex variety of seasonal
superimposed on annual s olar variations. These include the year-round equatorial lows, ridges of the Azores High and seasonal anticyclones, the
place the Arabian Plate roughly in its m odern con
regional highs lying over the Armenian Plateau in
figuration relative to the Mrican and Eurasian
the cool half of tl1e year, summer depressions over
Plates (Briggs, 1995). C onnections between the
Paldstan and the Arabian Gult� winter lows from
Eastern Paratethys and the Indo-Pacific had been
the Mediterranean Sea, and depressions from the
severed permanently by the Afro-Arabian Plate
Sudan in the transition seasons ( Hastenrath, 1985;
impinging onto the Eurasian continent 12�14 mil
Roberts and Wright, 1993; Schneider, 1996). The
lion years (Ma) ago ( Lyberis et a!., 1992; Rogl,
Yemen highlands in the extreme southern portion
1999-Chapter 35). Development of the modern
of the Arabian Peninsula m anage to intercept some
day extension of the Red Sea, i\rabian Gulf, and Gulf of Aden was incomplete during the mid- to late Miocene and did not form the geographic bar riers they do today along the margins of the Ara
of the moisture borne by the southwesterly South
Asian monsoon summer winds. An important factor mediating the interplay between these circulation
systems is the relief of the region, characterised by a
bian Peninsula (Coleman, 199 3). ::-J o consensus
northeastward inclination with considerable altitude
exists regarding the initial stages of rifting in the
along the western and southern margins. The mer
Red Sea but evidence for extension and widespread
idional mountains of Lebanon effectively block the
volcanism extends back to at least the early Miocene,
influence of the Mediterranean and tl1e westerly cir
at which time the structural shape of the Red Sea
culation in the winter, most of which travels well
depositional basin was defined (Coleman, 1993).
north of the Arabian Peninsula anyway, while the
Although there is evidence that the Red Sea trough
Kurdistan and Zagros Mountains check the south
occasionally contained deep-water sediments during
ward flow of winter air.
the Miocene (Crossley et al. , 1992), extensive evap
The northern portion of tl1e peninsula receives
oritic sequences throughout this interval suggest
most of its precipitation during the winter half of
episodic batch filling of the basins followed by
the year in association with middle- to high-latitude
evaporitic draw-down and shallow-water e vapora
westerly depressions whose tracks are steered by the
tion associated with sabkhas.
subtropical
stream (Wigley and Farmer, 1982).
Most of the precipitation falls to tl1e north but
PRESENT CLIMATE
occasionally moist air masses penetrate to the inte
Considering that the regional geographic and tec
over the western part of the Arabian Peninsula
tonic setting of the Arabian Peninsula has remained
from October to April but are usually too shallow
consistent over the past 15 �1a, the modern climate
to bring p recipitation. Occasionally, depressions
rior of the peninsula. Central African depressions lie
represents a valuable model for attempts to under
from northern Egypt reach the Arabian Peninsula
stand the palaeoclimatic patterns that may have
where they affect the weather along the Red Sea as
influenced the region in the past. Presently lying
far as the K.amaran Island and possibly even as far
between 12° and 38° N, the Arabian Peninsula
as Bahrain. In the interior of the Arabian Peninsula,
.MlOCE"'E P.1\LAEOE�1RONME:-.!TS IN ARABLit
m:J
rainfall is typically no greater than 50-:100 mm and
Aeolian sands coYer 770 000 lun2 or almost 90%
in the Rub' al Khali and along the Gulf of Aden it
of the peninsula's land surface (Whitney et al.,
is even less. Exposed to the minimal influence of
1983).
the Indian monsoon, the Arabian Sea coast receives little more than 50 mm of rain£11l while the Yemen :\lountains get 200-400 mm. The interior averages less than 10 rain days a year while the monsoonal of the Arabian Peninsula coast exceeds 25 days ..Mean surface temperatures over the central
.\rabian Peninsula average 15 oc during the winter \Yith variation associated vvith elevation (Schneider, 1996). The thermal low centred over the Arabian
PAST CLIMATES As the ;\Iiocene terresnial palaeoem·ironmental record remains poorly knmYn, interpretations of climatic trends in tropical
lm1·-Iatitude) terrestrial
regions during the past 20 :\Ia haYe dram1 heavily
on global eYents recm·ded in tae ma1ine record,
and to a limited extent from terrestrial sequences else\Yhere. This period of time incorporates seYeral
Gulf effectively influences the weather and its con
major flucmations in world\\ide climate and the
stancy in summer. The beginnings of this low are
onset of Milankovitch mid-latitude northern hemi
relt in April and it is at its deepest in July and
sphere glaciation ( deMenocal and Rind, 1993).
_\ugust. During the summer, the rather dry nor
All of these probably had profound etfects on the
therly and northwesterly etesian winds dominate
evolution of the Arabian fauna and flora. In the
this region. Thermal stratification in the tropical
long-term evolution of global climate, :-Jeogene
trade-wind circulation zone does not promote
climatic conditions appear to reflect a continua
cloud formation between 1\1ay and November.
tion of the general trend documented for the past
\"ery low evaporation also accounts for the incon
100 Ma, characterised by a shift from the mid
siderable cloud cover and rainfall. Over 80-90% of
Cretaceous thermal maximum to a world domi
days in Iraq are clear and clouds appear on only
nated by bipolar ice sheets. Antarctic and South
35-40 days in winter. Around the Gulf of Oman
ern Ocean cryospheric development occurred
and Arabian Gulf, the monsoon increases the rela
throughout the Cenozoic while northern hemi
ti\·e humidity to over 50-60%. While the mean
sphere glaciation developed in the latest Neogene
surface temperature during the summer is 34 °C,
(Miller et al., 1987). This sequential cooling and
the mean maximum surface temperature can reach
cryospheric development did not occur uniformly
45
oc.
Dust storms are characteristic of the arid and
semi-arid zones. They are associated both \vith
but rather as a series of abrupt shifts representing threshold events (Kennett, 1995). Accompanying this cooling trend was a presumed increase in
strong convection along a cold front and with
aridity in low latitudes (Shackleton and Kennett,
strong, constant winds that transport dust and
1975a). Explanations for the trend are incomplete
sand. In the Arabian Peninsula, these storms can
but research so far suggests that several processes
affect immense areas, from the Syrian Desert to
are involved in this long-term evolution of cli
the Rub' al Khali. They pass over tl:e eastern,
mate. They include shifting orbital parameters,
less-elevated half of the peninsula, directed from
changes in continent-ocean distribution, ocean
the west by mountains over 1000 metres high. In the south their progress is restricted by the moun tains near the Arabian Sea, and in summer by the presence of the intertropical convergence zone and the southwest monsoons associated with it. A prominent physiographic feature of the peninsula is sandy deserts referred to as sand seas or ergs.
heat transport, orography, and atmospheric C02
kvels (Crowley and North, 1991; Prell and Kutz bach, 1992). The following discussion prm ides an overview of global and continental :\liocene cli
matic trends and changes that may be rd::Yant tor
interpreting the evolution of landscapes in _\rabia during this period.
1m
J. D. lns a
the cycle of enporation must haw been repeated
significant increase in aridity documented by a n
about 40 times in the latest ). liocene. Climatic
increase in aeolian deposition throughout the late
deterioration manifested as increasing aridity in low
Cenozoic (Rea et aL, 1 9 8 5 ) , high-latitude shifts in
latitudes may ha,;e had a significant influence on
vegetation to more seasonal and arid-adapted flora
the African and Arabian flora as the Red Sea Basin
(Wolfe, 1 9 8 5 ) , and a hypothesised general transi
was also dry during this inten·al ( -:an Zinderen
tion from forested environments to habitats with
B akker and Mercer, 1 9 8 6 ) .
abundant grasses ( Potts and Behrensmeyer, 1 99 2 ; Williams, 1994 ) . Such shifts have been documented
in both western North America (Axelrod and Raven, 1 9 8 5 ) and Australia (Tedford, 1 98 5 ) . At 7 .4-7.0 Ma,
Quade et aL ( 1989) detected a dramatic shift from
vegetation dominated by c3 plants (forest/grass
land) to one dominated by c4 ( grassland) plants in
MIOCENE PAlAEOENVIRON MENTS OF THE ARABIAN PENINSUlA Early to Middle M iocene
the Siwalik sediments of Pakistan, possibly corre
Known empirical evidence of terrestrial environ
lated >vith inception or strengthening of monsoonal
ments on the Arabian Peninsula during the early
conditions due to uplift of the Tibetan Plateau or
to middle Miocene is limited to continental sedi
to declining atmospheric pC02 ( Ceding et
ments and associated fossil fauna and flora exposed
1 99 3 ) . In India, humid forest taxa rapidly retreated
in four areas of eastern Saudi Arabia ( Powers et al.,
eastwards to areas of moister climate during this
1 966; Hamilton et al., 1 978; Thomas et al., 1978;
period ( Prakash, 1 9 72 ) . Fossil macroflora from the
Thomas, 1982; Whybrow et a! . , 1 98 2 ; Whybrow,
southwestern Cape of South Africa indicate replace
1 984; Whybrow, 1 987; Whybrow et al. , 1990 ) and
ment of subtropical rainforest by the present
the western part of the Emirate of Abu Dhabi
"f)m
bos" or macchia ( Coetzee, 1 9 7 8 ) and the transition
(Whybrow et al . , 1 990; Whybrow et aL, 1999-
to a Mediterranean t:yve of climate. Evidence of
Chapter 4; Bristow, 1 999-Chapter 6 ) These
vegetation in East Africa during this period does
deposits have been divided into the late early Mio
not support the widespread replacement of forests
cene Hadrukh Formation (c. 1 9- 1 7 Ma), the over
by grasslands but rather a persisting heterogeneous
lying late early Miocene Dam Formation ( c . l7 -1 5
landscape ( Ceding, 1992; Kingston et al. , 1 994) .
Ma), and the middle Miocene Hofuf Formati on in
The terminal Miocene t o early Pliocene ( 6.55. 5 Ma) is characterised by extensive climatic vari ation, which resulted in significant changes in
.
Saudi Arabia and the ?middle Miocene Shuwaihat Formation in Abu Dhabi. Table 2 7 . 1 presents a compilation of palaeoen
the size of the polar ice sheet. At the Miocene
vironmental data derived from lithofacies studies
Pliocene boundary ( 5 . 5 Ma) the Antarctic lee Sheet
and analyses of fossil material recovered from these
may have exceeded its glacial maximum extent by
sequences. In general, the data indicate that this
as much as 50% ( Shackleton and Kennett, 1975b;
p ortion of the early to middle Miocene Arabian
Denton, 1 9 8 5 ), resulting in an appreciable drop in
Peninsula was dominated by more open em·iron
global sea level of up to 5 0 metres. The drop in sea
ments than during the Eocene ( �\s-Saruri et a! .,
level coupled v;,ith the closure of the Straits of
1 999-Chapter 3 1 ) and Oligocene ! Thomas et a ! . ,
Gibraltar, due to tectonic impingement of the
1 999-Chapter 3 0 ) . Interpretations o f the fauna
Table 27. 1 .
Palaeoenvironmental reconstructions of pre-Baynunah time i n the Arabian Peninsula
Basis of reconstruction
Formation
Site
Age
Reference As-Saruri et al., 1999 (this vol., Chapter 31)
Eocene of Wadi Rayan, and Fayum of Egypt
Fossil fruits and seeds,family ?Nymphaeaceae
Kaninah
Kaninah, southern Yemen
Middle Eocene
Tropical lowland evergreen forests
Family Anonaceae
Kaninah
Kaninah, southern Yemen
Middle Eocene
Strongly seasonal climate with marked rainy season
Sedimentary lithofacies (lack of evaporites, the presence of siliciclastics, and of broad permanent lakes with carbonate sedimentation)
Ashawq (Shizar Member)
Taqah, Oman
c. 33 Ma
Thomas er al. , 1999 (this vol., Chapter 30)
Semi-arid climate
Erycine snake and embrithopod A rsinoitherium
Ashawq (Shizar Member)
Taqah, Oman
c. 33 Ma
Thomas et al., 1991
Open savannah
Bovids
Hadruld1
Eastern Province of Saudi Arabia
Late early Miocene
Whybrow et al., 1982
Dry rather than arid (freshwater environment containing dissolved solutes probably concentrated by evaporation)
Fossil fruits ( MidraMlra arabica of the family Potamogetoneaepondweeds and ditch grassesand sediments)
Hadrukh
Jabal Midra ashShamali, Eastern Province of Saudi Arabia
Late early Miocene
Palm wood
Fossil vvood
Hadrnld1/Dam
Ad Dabtiyah, Saudi Arabia
19-17 Ma
Thomas et al., 1978 (VVhybrow et al., 1987)
Woodland habitat
Rhinoceroses
Hadrukh/Dam
Ad Dabtiyah, Saudi Arabia
19-17 Ma
Gentry, 1987a
As-Saruri et al. , 1999 (this vol., Chapter 31)
Whybrow et al . , 1982; Collinson, 1982
Dorcatherium
forest
Hadrukh/ l )am
A d Da htiy�1 h , Sa u d i
1 9 1 7 ,vta
l ; l' l l l l'y, 1 9 8 / h
19-17 Ma
Gentry, 1987b
19-17 .Ma
Gentry, 1987b
Lower Miocene
Thomas et al. , 1982b
Arabia Forest if okapi is used
Canthztmeryx
Hadrukh/Dam
as modern
Ad Dabtiyah, Saudi Arabia
analogue, savannah if giraffe is used Tropical to subtropical near-shore envi-
Vertebrates
Hadrukh/Dam
( general)
Ad D abtiyah, Saudi Arabia
ronment, tidal flats, and a large estuarine system More open type of
Microfauna
Dam
woodland and
As S arrar, S audi Arabia
17-15 Ma
bushland More forested
Dominance of
Dam
browsing
As Sarrar, Saudi Arabia
Lower Miocene
Thomas et al., 1982b
17-15 Ma
vertebrate fauna Mangroves-tropical
Fossil roots
Dam
climate
Dawmat al 'Awdah, Saudi Arabia
Palm wood
Fossil wood
Dam
Lower Miocene 17-15 Ma
South of Jabal
Lower Miocene
Dawmat
17-15 Ma
vVhybrow and McClure, 1981 Hamilton et a!., 1978
al 'Awdah, S audi Arabia Ope n milieu
Ve rtebrate fauna
Hofuf
Vertebrate fauna
t\ h nn;ll i ng, i n t er d u ne
Sedimentary
Hofuf
wit l 1 marked aridity
s;1 hk h as,
1 Lmsvcrse
.liHI h;Jrc h a noid
d u nes, ,I I H i river
systems; cvid cnn: of marine SL' d i lllCIJls ha si 1 1
wi t h i n t h e
c. 14 Ma
Thomas et a l . , 1 978
Al J adidah, Saudi
c. l 4 Ma
Sen and Thomas, 1 9 79
Arabia
( )pen environment
ami ro;Jst a l t·ypc
AI Jadidah, Saudi
Arabia lithofacies
S huwaithat
Abu D habi, UAE
Middle? Miocene ( unconformabl y
underl y i n g
the B ay n u n ah Forma t ion )
Bristow, l 999 (this vol . ( : h a p t cr
6)
lim
J . D. KINGSTON
Al"•lide
B arry, T. C. 1999. Late Miocene Carnivora from the
spread semi-arid to hyperarid conditions that domi nate the Arabian Peninsula today, the environment during the late Miocene may have been more var
Emirate of Abu Dhabi, United Arab Emirates. Chap.
17 in Fossil Vertebrates of Arabia, pp. 203-08 ( ed.
ied and supported a number of ditierent types of
P. J. \Vhybrow and A.
habitats. In this case it would be unreasonable to
New Haven.
extrapolate the interpreted environments for west ern Abu Dhabi to a regional scale. It remains difficult to link global climatic trends
Hill). Yale
University
Barry, T. C., Flynn, L. J,, and Pilbeam, D. R. 1990. Faunal diversity and turnover in a Miocene terrestrial pp. 381-421 ( ed. R. Ross and W. All
Causes ofEvolution: A Paleontological
and events documented in the marine record w�th
sequence. In
the evolution of terrestrial communities, primarily
Perspective)
because of a lack of sites and resolution in conti
mon). University of Chicago Press, Chicago.
nental sediments. Although it is tempting to corre late climatic shifts with faunal interchange, local
Barry, J. C., Johnson, N. M., Raza, S. M., and
and regional effects of global shifts on terrestrial
Jacobs, L. L. 1985. Neogene faunal change in south
ecosystems are intricate and basically unknown.
ern Asia: Correlations with climate, tectonic and
Late Cenozoic cooling at high latitudes clearly had
eustatic events.
an aridifYing effect on low latitudes but the timing and
of these changes on the continents
Dc;ruur,
Geology 13: 637-40.
R. L., Brunet, M., Ginsburg, L., Mein, P. ,
varied greatly depending on buffering by regional
Pickford, M . , Rogl,
and local atmospheric circulation patterns.
Thomas, H. 1987. A consideration of some major
Sen, S . , Steininger, F., and
topics concerning Old World Miocene mmalian chronology, migrations and palaeoecology.
Geobios 20:
431-39.
REFERENCES
Andrews, P. T., and Van Couvering, J. H. 1975.
Bonis, L.
Palaeoenvironments in the East African Miocene. In
G. 1992. Diversity and paleoecology of Greek late
Approaches to Primate Paleobiology,
Miocene mammalian faunas.
F. S. Szalay). Karger, Basel.
pp. 62-103 (ed.
Bouvrain, G . , Geraads, D., and Koufos,
Palaeogeography; Palaeo climatology, Palaeoecolog_y 91: 99-121.
LATE MIOCENE PALAEOEN\1RON�1ENTS IN ARi\BL->
Bonnefille , R. 1984. Cenozoic vegetation and envi ronments of early hominids in East Africa. In The Evolution of the East Asian Environment. II. Palaeo botany, Palaeozoology and Palaeoanthropology, pp. 5 79-612 (ed. R. 0. Why1:e). University of Hong Kong, Centre of Asian Studies, Hong Kong. J. C. 1995 . Global Biogeograph_y. Elsevier, :\.msterdam. Bristow, C. S. 1999. Aeolian and sabkha sediments in the Miocene Shuwaihat Formation, Emirate of Abu Dhabi, United Arab Emirates. Chap . 6 in Fossil Verte brates of Arabia, pp. 5 0-60 ( ed. P. J. vVhybrow and A. Hill) . Yale University Press, New Haven. Brown, G. F. 1960. Geomorphology oHVestern Cen tral Saudi Arabia. In Proceedings of the 21st Interna tional Geological Cortgress, Copenhagen, pp. 1 5 0-59. Gerling, T. E. 1992. Development of grasslands and savannas in East Africa during the Neogene. Palaeo geography, Palaeoclimatology, Palaeoecology (Global and Planetary Change Section) 97: 241 -47. Ceding, T. E., Wang, Y., and Quade, J. 1993. Expansion of ecosystems as an indicator of global ecological change in the late Miocene. Nature 361: 344-45. Coetzee, J. A. 1978. Climate and biological changes in south-western Africa during the late Cainozoic. In Palaeoecology ofAfrica, pp. 1 3-29 ( ed. J . A. Coetzee and Z. van Zinderen Baldcer). Balkema, Rotterdam. Coleman, R. G. 199 3 . Geologic Evolution of the Red Sea. Oxford University Press, New York. Collinson, M. E. 1982. Reassessment of fossil Pota mogetoneae fruits with description of new material from Saudi Arabia. Tertiar.Y Research 4: 83-104. Crossley, T., vVatkins, C., Raven, M., Cripps, D . , Car nell, A., and vVilliams, D. 1992. The sedimentary evo lution of the Red Sea and the Gulf of Aden. Journal of Petroleum Geology 15: 157-72.
m:J
Crowley, T. J., and North, G. R. 1991. Paleoclimatol ogy. Oxford UniYersity Press, New York. deMenocal, P. and Rind, D. 1993. Sensitivity of Asian and �-\frican climate to nriations in seasonal ice CO\Tr. sea-surface temperature, insolation, and �\sian orograph\·. Journal of Geophysical Research 98: 72 6 5-8 7. Denton, G. H. 1985. Did the �-illtan::ti c Ice Sheet influence late Cainozoic clin1ate and e,·olution in the Jouma! southern hernJsphere? South 8 1 : 224-29 . Ditchfield, P. W. 1999. Diagenesis of the Baynunah, Shuwaihat, and Upper Dam Formation sediments exposed in the \.Vestern Region, Emirate of Abu Dhabi, United Arab Emirates. Chap. 7 in Fossil Verte brates ofArabia) pp. 61-74 (ed. P. J . vVhybrow and A. Hill) . Yale University Press, New Haven. Eisenmann, V., and \Vhybrow, P. ]. 1999. Hipparions from the late Miocene Baynunah Formation, Emirate of Abu Dhabi, United Arab Emirates. Chap. 19 in Fos sil Vertebrates ofArabia, pp. 2 34-53 ( ed. P. J. \\"'by brow and A. Hill) . Yale University Press, New Haven. Fleagle, J. G. 1988. Primate Adaptation and Evolu tion. Academic Press, San Diego. Flower, B. P., and Kennett, J. P. 1994. The Middle ivliocene climatic transition: East Antarctica ice sheet development, deep ocean circulation and global car bon cycling. Palaeogeography, Palaeoclimatology, Palaeoecology 108: 5 37-5 5 . Flynn, L. J . , Tedford, R. H., and ZhaiL,iang, Q. 199 1 . Enrichment and stability in the Pliocene mam malian fauna of North China. Paleobiology 1 7: 246-65. Forey, P. L., and Young, S. V. T . 1999. Late }, liocene fishes of the Emirate of Abu Dhabi. C nited Arab pp. Emirates. Chap. l 2 in Fom·! Te·;·td;;-etc·.c 120-35 (ed. P. J . \.VhybrmY <md A Hill . Yale CniYer sity Press, London and �ew Ha,·en.
l!!:im J,
D. KINGSTON AND A. HILL
Friend, P. F.
1999.
Rivers of the Lower Baynunah
Formation, Emirate of Abu Dhabi, United Arab Emi rates. Chap.
38-49
5 in Fossil Vertebrates Q{Arabia)
( ed. P. J. v\-nybrow and A.
sity Press, Kew Haven. Gahunia, L. lution of the Paratethys
pp.
Hill ) . Yale L'niver
K,
and Chochieva, K. 1. 1982. Co-evo Hipparion fauna and vegetation in the Evolutionary Theory 6: 1-1 3 .
Past: Paleoanthropolq_rrical Advances in Honor ofF.
(ed. R. S. Corruccini and
R. L. Ciochon) . Prentice-Hall, Englewood CliffS; X J.
Clark Howell,
pp.
--............ 1995.
123-45
Faunal and environmental change in
the Neogene of East Africa: Evidence from the Tugen
Paleocli mate and Evolution, with Emphasis on Human Ori gins, pp. 178-93 ( ed. E. S. Vrba, G. H. Denton, T. Hills Sequence, Baringo District, Kenya. In
C. Partridge, and L. H . Burkle) . Yale University Press,
1987a. Rhinoceroses from the Miocene of Saudi Arabia. Bulletin of the British Museu1n (Nat ural History), Geology 4 1 : 409-32 .
Kew Haven.
1987b. Ruminants from the Miocene of Saudi Arabia. Bulletin of the British Museum (Nat ural Hi.-tory), Geology 41 : 433-39.
rate ofAbu Dhabi, United Arab Emirates. Chap. 29 in
Gentry, A . W.
---. 1999 . Late .Nliocene sub-Saharan Africar1 verte brates and their relation to tl1e Baynunah fauna, Emi
Gentry, A. W.
Glennie, K W., and Evamy, B . D.
1968.
Dikalca:
Plants and plant-root structures associated with aeo
Palaeq_qeography, Palaeoclimatology) Palaeoe colog_v 4: 77-87.
lian sand.
Grobe, H . , Futterer, D . K , <md Spiess, V.
1990.
Fossil Vertebrates ofArabia1 pp. 420-29 ( ed.
Hill ,
and Ward, S .
1988.
Origin of tl1e Homin-
between I 4 .My and
idae: The record of African
hominoid evolution
4 My. Yearbook of Physical Amh1'opology 3 1 : 49-83. Hill, A . , and vvnybrow, P. J.
continental margin. In
P. J. Why
brow and A. Hill). Yale University Press, New Haven.
1999.
Summary ar1d
Oligocene to Quaternary processes on the Antarctic
overview of the Baynunah fauna, and its context.
Proceedirtgs of the Ocean Drilling Program, Scientific Results, pp. 1 21-31
P. J. V\ll1ybrow and A. Chap.
(ed.
2
in
Fo.Lril Vertem'ates ofArabia>
7-14
pp.
(ed.
Hill) . Yale University Press,
R. F. Barker and J . P. Kennett). College Station, Tex.
New Haven.
Hamilton, W. R., Whybrow, P. J., and McClure, H .
Hodell, D. A., Elmstrom, K. .M., and Kennett, J.
A.
1 978.
Fauna o f fossil man1mals from the rvliocene
of Saudi Arabia.
Nature 274: 248-49 .
1986.
Latest rvliocene benthic o180 changes, global
ice volume, sea level and the 'Messinian salinity crisis'.
Nature 320: 4 1 1-14. Haq, B . U.
1980.
Biogeographic history of Miocene
calcareous nannoplankton and paleoceanography of the Atlantic Ocean.
Micropaleontology 26: 4 14-43.
Hastenrath, S. I 985.
Holm, D. A.
1960. Desert morphology in Science 1 32: 1 369-79.
the i\ra
biar1 Peninsula.
TI'opics.
Reidel, Dordrecht.
Hill, A.
Erickson, A. , Garrison, R. E . , Kidd, R. B . , .Milieres,
1987.
Mediterranean Salinity Crisis.
Climate and Circulation of the
Hsu, K. J., Montadert, L., Bernoulli, D., Cita, M. B . , F., Mueller,
Causes of perceived faunal change in
the later ::'-Jeogene of East Afi'ica.
Joumal ofHuman
Late �liocene and early Pliocene
hominoids from Africa. In
1977. History of the Nature 267: 399-403 .
Hut, P., Alvarez, W. , Elder, W. P. , Hansen, T., Kauf. man, E. G., Keller, G. R. , Showmaker, E . .M . , and
Evolution 1 6: 5 8 3-96 .
---. 1994.
and Wright, R.
Integrative Paths to the
vVeissman, P. T.
1987. Comet showers as Nature 329: 1 1 1 8-26.
mass extinctions.
a cause of
LATE MIOCE:-JE PAJ�'iliOENVIRONMENTS IN
Imbrie, J., and Imbrie, J.
Z. I 980.
Modeling the cli
matic response to orbital variations.
Science 207:
943-5 3. Janis, C. M.
1993.
Tertiary mammal evolution in the
context of changing climates, vegetation, and tectonic
events. Annual Review ofEcology and System ati cs 24: 467-500.
ARABIA -
Kroon, D . , Steens, T., and Troelstra, S. R.
199 1 .
Onset of monsoonal related upwelling in the western
Arabian Sea as reYealed by planktonic foraminifers. Proceedings Ocean Drilling Program) Scientific Remit' 117: 2 3 7-6 3.
Kutzbach, J. E" Guetter, P. L Behling, P. }., and R. 1993. S imulate d clli'lutic changes; results of
1999.
Late Miocene swan mussels from
the B aynunah Formation, Emirate of Abu Dhabi, United Arab Emirates. Chap.
ofArabia,
Hill) .
pp.
l l i- I 5 ( ed.
10
in
Fossil Vertebrates
P. J. Whybrow and A.
1 982. Marine Geology.
Climrr.tes since the Lrrst Gla cirtl Jfnximum, pp. 24-93 H. E . } . iY1ight, J. E. -r. L \Yebb,
T. I. Webb, W. F. Ruddiman, F. ,\. Su·eet-Perrott, and P. J. Bartlein) . GniYersity oL \ linnesota Press
,
1--linneapolis.
Yale University Press, New Haven.
Kennett, J. P.
Global
In
the C OIL\ L\P climate-mode� Jeffery, P. A.
Prentice-Hall,
I2pparent de Broin, F. de, and Dijk, P. P. van.
I 999.
Chelonia from the late Miocene Baynunah Forma
Englewood Cliffs, N .J .
tion, Emirate of Abu Dhabi, United Arab Emirates: -·--.
1995.
A review o f polar climatic evolution
during the Neogene, based on the marine sediment
Paleocli'frnate and Evolution, with Emphasis on H�tman Origins, pp. 49-64 ( ed. E . S . Vrba, G. H .
record. In
Denton, T. C . Partridge, and L . H . Burckle) . Yale University Press, New Haven.
Kennett, J. P., and Barker, P. F.
13
Palaeogeographic implications. Chap.
tebrates ofArabia, and A.
Hill ) .
pp.
136-62 ( ed.
in
Fossil Ver
P. J. Whybrow
Yale University Press, New Haven.
Leinen, M., and Heath, G. R.
1 981 .
Sedimentary
indicators of atmospheric activity in the northern
Palaeogeography, Palaeoclimatology> Palaeoecology 36: 1-22.
hemisphere during the Cenozoic.
I990.
Latest Creta
ceous to Cenozoic climate and oceanographic devel
I99 l .
Proceed ings of the Ocean Drillitltf Progmm, Scientific Results, pp. 937-60 (ed . R. F. Barker and J . P. Kennett) . Col
Le Loeuff, J.
lege Station, Tex.
graphical implications.
opments in the Weddell Sea, Antarctica. In
The Campano-Maastiichtian ver
tebrate fannas from southern Europe and their rela tionship \vith other faunas in the \Vorld; palaeobiogeo
Cretaceous Research 12:
93-ll 4. Kingston, J. D.
1 999.
Isotopes and environments of
the Baynunah Formation, Emirate of Abu Dhabi, United Arab Emirates. Chap.
ofArabia>
pp.
354-72 ( ed.
25
in
Fossil Vertebt·ates
P. J. Whybrow and A.
Lyberis, 'N., Yurur, T., Chorowicz, J., Kaspoglu, E., and Gnndogdu, N.
1992.
An oblique collision belt.
The east Anatolian Fault:
Tectonophysics 204: 1-I 3 .
Hill). Yale University Press, New Haven. MacFadden, B . J . , and Ceding, T. E .
Kingston, J . D., Marino, B . D. and Hill A.
1994.
Isotopic evidence tor Neogene hominid paleoenviron ments in the Kenya Rift Valley.
Fossil
Ecology and Evolution 9: 481-86.
Science 264: 955-59. Maley, J.
1980. Les
changements
de
Ia fin
du Tertiare en Afrique. In Ihe Snh:; ;·,? ::;u! rbc "\-i!c.
1972. New Perspectives on Ape and Human Evolution. Stichting voor Psychobiologic,
pp.
Amsterdam.
BallAcademie des Sciences, Paris 322: 5 01-5. Pickford, M., and Senut, B. 1994. Palaeobiology of the Albertine Rift Valley: General conclusions and synthesis. In Geology and Palaeobiology of the Albertine
Rift Valley Uganda-Zaire, Volume IL Palaeobiology/Paleobiologie, pp. 409-23 ( ed. B. Senut
and M. Pickford) . CIFEG Publication Occasionnelle 1994/29. Centre International pour Ia Formation et les Echanges Geologiques, Orleans. Pickfurd, M . , Ishida, H., Nalcano, Y., and Nakaya, H . 1 984a. Fossiliferous localities of the Nachola-Sarnburu Hills area, northern Kenya. African Study }dono graphs, suppl. 2: 45-56. Pickford, M., Nalcaya, Ishida, and Nalcano, Y. l 984b. The biostratigraphic analyses of the faunas of the Nachola area and Sarnburu Hills, Northern Kenya. African Study Monographs, suppl. 2: 67-72. Picld'ord, M . , Senut, B . , and Hadoto, D. 1993.
Geology and Palaeobiology of the Albertine Rift Valley, Uganda-Zaire, Volume I: Geology. CIFEG Publica
. 1 986. Geochronology of Miocene hominoids east of the Kenya Rift Valley. In Primate Evolution, pp. 35-45 (ed. ]. G. Else and P. C. Lee). Cambridge University Press, Cambridge.
tion Occasionnelle no. 24. Centre International pour Ia Formation et les Echanges Geologiques, Orleans.
Morgan, M. Kingston, J. D., and Marino, B. D. 1994. Carbon isotopic evidence for the emergence of C4 plants in the Neogene from Pakistan and Kenya. Nature 367: 162-6 5 .
Quade, ]., Cerling, T. and B owman, J. 1 989. Development of Asian monsoon revealed by marked ecological shift during the latest :Miocene in northern Palcistan. Nature 342: 163-66.
---
LATE MIOCENE SlJB-SAHARAN AFRICAc'\1 VERTEBRATES
1m
Senut, B . , and Pickford, M. eds. 1994. Geology and
Vertebrates ofArabia) pp. 209-33 ( ed. P. J. Whybrow
Palaeobiolq_IJY ofthe Albertine Rift Valley Uganda Zaire) Volume II: Palaeobiology/Pa!Cobiologie. CIFEG Publication Occasionnelle 1994/29. Centre Interna
and A.
tional pour Ia Formation et les Echanges Geologiques, Orleans. Sickenberg, 0 . , and Schonfeld, M. 1975. The Chorora Formation-lower Pliocene limnical sediments in the southern Mar (Ethiopia). In Afar Depr·ession of Ethiopia) pp. 2 77-84 ( ed. A. Pilger and A. Rosler). Schweizerbart'sche Verlagsbuchhandlung, Stuttgart.
Hill) .
Yale University Press, New Haven.
Thomas, H. 1979. Le role de barriere ecologique de la ceinture saharo-arabique au Miocene: Arguments paleontologiques. Bullt'tin du lvfushnn National
d )Histoire Naturelle) Paris 1 : 127-3 5 .
Thomas, Bernor, R. , and Jaeger, J.-J. 1982. Origines du peuplement mamrnalien en Afiique du Nord durant le Miocene terminal. Geobios 1 5 :
283-97.
Tassy, P. 1 986. Nouveatt.'x Elephantoidea rnalia) dans le Miocene du Kenya. Cahiers de Paleontologic. Centre National de la Recherche Sci enrifique, Paris.
Tiercelin, Micham::, J., and Bandet, Y. 1 979 . Le Miocene superieur du sud de Ia depression de l'Mar, Ethiopie: Sediments, fam1es, isotopiques. Bulletin
. 1999. Miocene elephantids ( Manunalia) from the Emirate of Abu Dhabi, United Arab Emirates: Palaeobiogeographic implications. Chap. 1 8 in Fossil
Whybrow, P. J . , and Hill, A. eds. 1998. Fossil Verte brates ofArabia. Yale University Press, New Haven.
---
de la Societe Geolq_JJiqzte de hance 7: 2 5 5-58 .
Oligocene and Miocene Terrestrial Vertebrates in the Southern Arabian Peninsula ( Sulta11ate of Oman) and Their Geodynamic a11d Palaeogeographic Setti11gs HERBERT TIIOlvLI\5,
JACK ROGER, S EVlZET SEN , MARTIN PICKFORD, EJ\1lv1Al'\UEL GHEERBRANT, ZAHER AL-SULAIJ\L'L'H, AND SALL\1 A.L-BUSAIDI
During the last four field seasons of the Franco Omani Palaeontology Expedition ( 19 8 6- 92 ) sev
The Miocene localities >vere found i n 1 99 1 , also within littoral marine strata that crop out in
eral Oligocene and Miocene vertebrate
the northern region of Huqf massif some 50 km
fossil sites were discovered in the Sultanate of
southeast of the Ghaba North Oil Field (Roger et
Oman. Whereas the Oligocene faunas came from
a! ., 1 994a) . Marine molluscs and foraminifera from
localities situated in the southern part of Dhofar,
the succession are extremely similar to those d e
the Miocene faunas were collected in the northern
scribed b y Powers et al. ( 1 966) from the type local
margin of Huqf (Central Oman ) near Ghaba (fig.
ity of the Dam Formation in Saudi Arabia and by
3 0 . 1 ). These discoveries have filled important gaps
Cavelier ( 1 97 5 ) from the same unit in Qatar, dated
in the knowledge of the distribution and the evo
as Burdigalian-Langhian. 1n the Ghaba sector, the
lution of terrestrial vertebrates during the Tertiary
Miocene series assigned to the Dam Formation ,
of the Arabian Peninsula. The faunal associations
only 50 metres thick, displays a mixed carbonate,
could, among other things, be placed into their
siliciclastic, and evaporitic sedimentation, whose
geodynamic and palaeogeographic context thanks
vertical arrangement comprises a complete cycle of
to the revision of the Tertiary units, carried out by
transgression and regression.
the Bureau de Recherches Geologigues et Minieres ( B RGM) during geological mapping of Oman at a scale of 1 :2 5 0 000 ( Le Metour et al . , 1 99 5 ) . T h e initial discovery of Oligocene terrestrial ver tebrates in the Sultanate of Oman dates t!·om
During the Tertiary, the Arabian Peninsula was affected by two major geodynamic events (fig. 3 0 . 2 ) d1at played a significant role in sedimentatary processes and in the distribution of marine and ter restrial taunas: the beginning of the opening of the
1 986-87 , when fossils v.rere found at two localities-
Gulf of Aden in the Rupelian and the Alpine tec
Thaytiniti and Taqah-in the coastal strip of Dhotar,
tonic phase that began in the Burdigalian (Le
within a littoral m arine u nit (Shizar Member at the
M etour et a! ., 1 99 5 ) The succession of transgres
base of the Ashawq Formation ) dated to the early
sive-regressive cycles that affect the very extensive
Oligocene on the basis of the nummulite
Arabian Platform determined the distribution of d1e
mulitesfichteli (Thomas et al.,
Nu1n
1 989a ) . The fossilifer
ous unit, about 1 00 metres thick, has a siliciclastic
.
deposits and, over the course of time, displaced the shoreline of the Arabian Sea by large distances.
and carbonate content and constitutes the transgres
B ecause of the large predominance of marine sedi
sive episode of the early Oligocene depositional
mentation in contrast with the poorly developed
cycle.
continental deposition during Paleogene-Miocene
Copyright © 1 999 by Yale University Press. All rights reserved. ISBN 0-300-07183-3
OLIGOCENE .�'-'D MTOCENE VERTEBRi\TES TN OMAN
42 '
' 56
50'
the palaeogeographic setting. Along the eastern flanks of Oman this period was characterised by an
IRAQ
JORDAN
!:m
important lmYering of sea level that led to severe ,..
erosion of the margins with the formation of tl1ick slope deposits. At tl1e same time in southern Dho far the opening of the proto-Gulf of Aden began. This nent is manit�sted by the opening of several grabens aligned in a step-like fashion ( en echelon ), characterised by strong subsidence and the accumu lation within them of mixed siliciclastic and carbon
200
ate sediments. It is probable that the terminal Bar
Taqah Thaytlniti J.t *
(notably Egypt) and the Dhofar region, especially
Arabian Sea
YEMEN
12'
tonian emergence of much of the plate facilitated communication between the eastern parts of Africa
L---===-----,-....l..---,---..,.--J 42 '
50 '
for terrestrial faunas, while the opening of the 12'
58'
Figure 30 . 1 . Map of the Arabian Peninsula shmving fossil vertebrate localities in the Sultanate of Oman.
ens, p artly invaded by the sea, led to the appear ance of palaeoenvironments favourable for a diversi fied terrestrial vertebrate fauna. Climatic modifications in the Oligocene led to greater rainfall; tl1is favoured the development of vegetation, the extent of which may have been lim
times, it i s unsurprising that terrestrial vertebrates were discovered mainly in littoral facies close to the sea shore where the depositional environment was propitious for concentration and preservation of bones and teetl1 .
ited to the southern grabens, which were partly inundated by fresh to brackish water. Only the transgressive basal horizons of the graben fillings, consisting of alternating marine and continental sediments, have yielded the remains of terrestrial vertebrates. These fossils appear to be concentrated mainly at the base of transgressive marine sequences
PALEOGENE EVOLUTION
that overlie more continental facies according to
During the Paleogene, the southeastern belt of the
plains the narrow association between the marine
fluctuations in the littoral zone. Such a model ex
Arabian Peninsula acted as an epeirogenic pladorm
and continental faunas. The model also accounts
over which were several extensive marine transgres
for the concentration of fossils and their fragmen
sions. Thick accumulations of marine carbonate
tary nature, caused by their immediate but localised
strata rich in foraminifera, molluscs, and algae were
reworking by the sea. Rapid subsidence and burial
deposited from the end of the Thanetian up to the
of the deposits ensured the preservation of tl1e
middle Eocene. Up to now, the oldest transgressive
fossils.
marine facies of coastal plains of Thanetian to Iler
dian age have yielded only marine vertebrates (sir
enians) , no terrestrial ones. The terminal Bartonian regression initiated a progressive retreat of the plat forms. Consequently, during the Priabonian, only
DHOFAR OLIGOCENE FOSSILIFEROUS SITES
the southeastern margin of the plate remained sub
Thaytiniti occurs in the broad AshaYI q Graben,
merged , the rest being subjected to erosion.
which constitutes the type localin- tor the Shizar
Reorganisation of the Mro-Arabian Plate, which began in the Oligocene, profoundly modified
Member (Roger et al . , l 994b J . The principal fossil iferous level (Thomas et al. , 1 9 8 8 , 1 992 ), ,,-hich is
1m
H. THOMAS
ET AL.
MAIN
PALEOGEOGRAPHIC
TERRESTRIAL
FACIES
and GEODYNAMIC
VERTEBRATE
TYPE
EVENTS
FOSSIL SITES
Alpine orogenic phase
()HOF AR MARGIN (rill
onset)
E a ter N. fabianii lineage that appears
manifest. The Fayum, it needs to be said, comprises
early Oligocene ( Rupelian) . The absence of the
the only important African Oligocene locality if one
foraminifer
excludes Malembe (An gola ), Zellah ( Libya), and
bed prompts us to refine the age to basal early
Gabal Bou Gobrine (Tunisia), of which the faunas
Oligocene, since
are little known.
in the series.
Lepidocyclina from the Thaytiniti fossil Lepidocyclina only appear higher
At the end of d1e
AGE OF THE TERRESTRIAL VERTEBRATES OF DHOFAR Among the Dhofari mammals, at the present stage
Oligocene, the geody-
namic conditions that had prevailed until then were interrupted. On the one hand, there was an accen tuation of the subsidence of the Dhofar margin. This caused the accumulation of calciturbidites, which continued until the early Miocene ( Roger et
of their study, only a few seem at first inspection to
a!. , l 994b). On the other hand, the continuous
be identical to those of the Fayum, although the
uplift of the interior of the Arabian Plate was
relationships between them are manifest. The stage
responsible for d1e formation of weathering profiles
of evolution of d1e rodents and proboscideans lead
(silicification) and absence of continental deposi
us to suggest, however, that the deposits of Thay
tion.
tiniti, at least, are slightly older than those of the Qatrani Formation or are more or less contempora neous with Fayum locality L 41. Correlations "'ith marine formations on the basis of characteristically early Oligocene nummulites
(N. jichteli), which
N EOGENE EVOLUTION The last major marine transgression that Hooded the central part of the Arabian Plate covered the Eastern
occurs in both of the D hofari fossiliferous levels,
Province of S audi Arabia, the Rub' al Khali and
combined >:vith palaeomagnetic data, indicate that
Central Oman during the Burdigalian-Langhian
the fossil horizon at Thaytiniti is dated around 33
( Cavelier et
million years ( Ma) ( chron C l 3n ) .
epicontinental sea was formed with lagoonal carbon
These nummulites, which are not present in the first
8
metres of the Shizar Member nor in the
underlying lacustrine Zalumah Formation, appear
al., 1993;
Le Metour et al., 1995 ) . An
ate facies ( Dam F ormation) bordered by a nst lacustrine belt in which sedimentation of carbonates also predominated. D eposits exhibit a clear trans
for the first time in the vertebrate fossiliferous level
gressive-regressive cycle marked by the passage from
at Thaytiniti. Unconformably underlying the
intertidal to subtidal bioclastic marl and carbonate
Zalumah Formation, the Aydim Formation has in contrast yielded abundant and varied foraminifera,
to restricted fa cies composed of gypsum-interbed
ded marl (fig. 30.4) . The proximity of the emergent
1m
H. THOMAS
ET AL.
SE
NW G h a ba
Central Oman
Huqf high
fossiliferous site
T
'j I 1'-'.--___.·l �1 lol==-
'·� � -" ·
�
HST : Highstand Systems Tract M F S : Maxi mum Flooding Surface
::---._ � *
TST : Transgressive Systems Tract
Facies lnterdidal to subtidal bioclastic marl and carbonate Marginal littoral mar1 and gypsum
with channeled sandstone Lacustrine limestone
Alluvial sandston·e and siltstone
Ante-Miocene substratum Vertebrate-fossiliferous bed
Figure 3 0 .4 . Southeastward evolution of the Miocene Dam Formation towards the Huqf high.
Huqf massif explains the intercalation, within the
The most striking facet o f the Ghaba fauna
marine succession, of the quartzo- lithic sandstone
concerns the minute size of the two proboscideans.
facies of fluvial origin. In the Ghaba area this has
Their dwarfing, although relative, could (if not a
yielded a predominantly terrestrial vertebrate fauna.
primitive character) be related to a period of insu
Whereas numerous fossil sites are !m own in Saudi
larity that affected the Ghaba region during the
Arabia, in Oman only the southeast margin of this
Bmdigalian-Langh.ian for a limited time. This
intracratonic sea, close to the Huqf high, has yielded
hypothesis is supported by palaeogeographic evi
terrestrial vertebrates ( Roger et a! . , 1 994a ) .
dence suggesting the existence of an isolated emer gent axis that comprised the Oman Mountains and their extension towards the south (regions of
GHABA EARLY-MIDDLE M IOCENE FOSSILIFEROUS SITE
Wahibah and Huqf) . The terrestrial fauna of Ghaba, the Mrican affinities of which are cl ear, show two taxa
Vertebrate remains from Ghaba show limited diver sity and are very fragmentary ( table 3 0 . 2 ) . Besides freshwater fishes such as
Clarias and Lates and
cer
tain marine forms (pristids and rays ) , they include brevirostral and longirostral crocodiles, l ittoral and terrestrial chelonians, and several mammals-notably two proboscideans (a bunodont mastodont and a deinothere ) , an anthracotheriid and a primitive giraffoid
(Afromeryx zelteni),
( Canthumeryx
sp. ) .
thumeryx and Afromeryx zelteni)
( Can
very common in
the Jebel Zelten fauna of Libya. This enables corre lation of the Ghaba sites with faunal zones Pilla and Pili b of Pickford ( 1 99 1 ) , the ages of which are
1 7 . 5-1 5 . 5 Ma.
Finally, from the middle Miocene until the pre sent day, the region became fully continental, with only rare transgressions along its margins due to fluctuations in sea level. Curiously enough, the
0LIGOCEl'E A>Univer site de Louvain 2 1 : 1 22-79 . ---. 1 966. Geology of the Arabian Peninsula; Yemen. United States Geological Survey Proftssional Paper 560B: 1 -2 3 .
---. 1966. Geology of the Arabian Peninsula,
Grolier, M . J . , and Overstreet, W. C . 1 978. Geologic Map of the Yemen Arab Republic (San 'a). United
El-Nakhal, Hamed A. 1 987. A lithostratigraphic sub division of Kholan Group in Yemen 1\rab Republic. Iraqi Journal ofScience 2 8 : 1 49-80.
Horner, J. R. 1 979. Upper Cretaceous dinosaurs from the Bearpaw Shale (marine) of south-central Montana with a checklist of Upper Cretaceous
eastern Aden Protectorate and part of Dhufar. United States Geological Survey Professional Paper 560H: 1-49.
States Geological Survey :Niiscellaneous Investigations Series. Govenm1ent Printer, Washington, D.C.
A DI::\OSAUR J'ROM YEME:>!
dinosaur remains from maline sediments in North "�merica. journal of Paleontology 53: Jacobs, L. VVinkler, D. A., Murry, P. A., and Manlice, J. M. 1 994. A nodosaulid scuteling fiom the Texas shore of the \Vestern Intelior Sea\vay. In Dinosaur Eggs and Babies, pp. 337-46 ( ed. K. Car penter, K. F. Hirsch, and J. lZ. Horner) . Cambridge University Press, Camblidge. Mateer, N. J., VVycisk, P. J . , Jacobs, L. L., Brunet, .Nl., Luger, Dina, A., Hendriks, \Veissbrod, Gvirtzman, G., "lalfJUS sp.
Phosphatosaurus sp. Dryosauridae gen. et sp . indet.
Middle Eocene Republic ofYemen: locality, I
pp.
290-3 16 (ed. P. J. Whybrow and A. Hill) . Yale Uni
versity Press, New Haven.
Gheerbrant, E . , and Thomas, H. 1992. The two first possible new adapids from the Arabian Penirlsula (Taqah), Early Oligocene of Sultanate of Oman. Paper presented at d1e 14th Congress of the Interna-
-
P. J. WHYBROW &'lD D. CLEME:..'TS
tional Society of Primatology, Strasbourg, 1 6-21
United Arab Emirates. Chap. 10 in
August, 1992. Abstracts p. 349.
ofArabia,
Gheerbrant, E., Thomas, H., Roger, J., Sen, S., and
Fossil Vertebrates
pp. 1 1 1-15 (ed. P. J . Whvbrow and A.
Hill) . Yale University Press, New H�ven.
Al-Sulaimani, Z. 1993. Deux nouveaux primates dans
Lapparent de Broin, F. de, and Dijk, P. P.
!'Oligocene inferieur de Taqah (Sultanat d'Oman ) :
1999. Chelonia from the late .\1iocene Bavnunal1
Premiers Adapitormes ( ?Anchomomyini ) d e I a penin
Formation, Emirate of Abu Dhabi, Unite
sule Arabique?
Palaeovertebrata 22:
141-96.
d Arab
Emirates: Palaeogeographic implic;tions. Chap. 1 3
in Sulairnani, Z. 1995. Nouveau primate Oligopithecinae (Simiiformes) de !'Oligocene inferieur de Taqah, Sul
Comptes Rendtts de PAcadhnie deJ· Sci ences, Paris 32 1 : 425-32.
tanat d'Oman.
K
cology 4 :
Palaeogeograpl�y, PalaeoclimatologJ; Palaeoe
77-87.
Greenwood, P. H. 1 9 8 3 . A new Palaeocene fish from Saudi Arabia. }ottmal of Natttral Histor_y 1 7 : 405-18 . Hamilton, W. R . , \¥hybrow, P. J . , and McClure, H .
A . 1 978. Fauna of tossil mammals from the Miocene of Saudi Arabia.
Hill,
Nature
274: 248-49.
mates, Cercopithecidae) from the late Miocene of Abu Dhabi, United Arab Emirates. Chap. 1 6 in pp. 1 98-203 (ed. P. J.
Whybrow and A. Hill). Yale University Press, New Haven.
Hill ,
A., Whybrow, P. J., and Yasin al-Tikriti, W.
1990. Late Miocene primate fuuna from the
Arabian Peninsula: Abu Dhabi, United Arab Emi rates.
A merican Journal of Physical Anthropology
81:
240-41 . Jeffery, P. A . 1999. Late .Miocene swan mussels from the Baynunah Formation, Emirate of Abu Dhabi,
New Haven.
Leakey, R. E . , and Lealcey, M. G. 1 986. A new �1iocene hominoid from Kenya.
Natttre 3 1 8 :
1 73-75.
deposits near Chinji Village. Part I. Megacricetodonti nae, !vlyocricetodontinae, and Dendromminae.
Palaeovertebrata
1 8 : 95-1 54.
Madden, C. T. 1982. Paleocene, l�vcnodont Fishes from Jabal Umm Himar; Harrat Hadan Area, A t Taif Region, Kingdom of Saudi Arabia. United States Geological Survey Preliminary Report, Jiddah. Madden, C. T., Naqvi, I. M., Whitmore, F. C . , Schmidt, J . D . , Langston, J . , and Wood, R. C.
and Gundling, T. 1999 . A monkey (Pri
Fossil Vertebrates of Arabia,
pp. 136-62 (ed.
Lindsay, E. H . 1988. Cricetid rodents from Siwalik
W., and Evamy, B. D . 1968. Dikaka:
Plant� and plant-root structures associated with aeo lian sand.
Fossil Vertebrates of Arabia,
P. J. Whybrow and A. Hill ) . Yale University Press,
Gheerbrant, E . , Thomas, H., Sen, S., and Al
Glennie,
van .
1979. Paleocene Ve1·tebrates frorn Coastal Deposits in Ha�·rat Hadan Area, At Taif Region) Kingdom of Saudi Arabia. United States Geological Survey, JiddalL Madden, C. T., Schmidt, D . L. , and \Vhitmore, F. C. 1983. Masrithetium (Artiodactyla, AJzthracotheriidae) from Wadi Sa!Jya> Southwestern Sattdi Arabia: An Eadiest Miocene Age for Continental Rift-valley Vol canic Deposits ofRed Sea Mat;gin. Open-file Report USGS-OF-03-6 1 . Ministry of Petroleum and Mineral Resources, Jiddah, Saudi Arabia. Mordan, P. B. 1999. A terrestrial pulmonate gastro pod from the l ate Miocene B aynunah Formation,
ARAll L'L'> TERTJAR\: FLORA, FAL':-A -: , AND
Emirate of Abu Dhabi, United Arab Emirates. Chap.
ll in Fossil Vertebrates ofArabia, pp. 1 1 6-19 ( ed. E J. Whybrow and A. Hill ) . Yale University Press,
ogy,
pp.
2 5 5-60
LOCALITIES -
( ed. B. Thierry, ]. R. Anderson, ]. J.
Roeder, and N. Herrenschmidt ) . UniYersite Louis Pasteur, Strasbourg.
:\'ew Haven.
Sige, B . , Thomas, H . , Sen, S . , Gheerbrant, E . , Roger,
1994. Sexual dimor phism i n Moeripithecus markgrafi from the early Oligocene of Taqah, Oman. In CtnTent Pritrtatology, pp. 261-64 (ed. B . Thierry, J. R. Anderson, J. J. Pickford, M., and Thomas, H.
]., and Al-Sulaimani, Z.
1994.
Les chiropteres de
Taqal1 (Oligocene inferieur, Sultanat d'Oman ) : Pre
1liiinchner Geowis senschaftliche Abhandlungen 26: 3 5-48.
mier inventaire systematique .
Roeder, and N. Herrenschmidt) . Universite Louis Tassy, P.
Pasteur, Strasbourg.
1999.
Miocene elephantids (Mammalia )
from the Emirate of Abu Dhabi, United Arab Emi rates : Palaeobiogeographic implications. Chap.
Pickford, M., Thomas, H . , Sen, S . , Roger, ]., and Al-Sulaimani, Z.
Gheerbrant,
1994.
Early
Oligocene Hyracoidea ( Mammalia) from Thayt:initi
Fossil Vertem·ates ofArabia, pp. 209-3 3 (ed.
18
in
P. ] .
\Vhybrow and A . Hil l ) . Yale University Press, New
and Taqah, Dhofar Province, Sultanate of Oman.
Haven.
Comptes Rendus de i>Acadbnie des Sciences, Pat·is 3 18 : 1395-400.
Thomas, H .
1982 .
The Lower and 1\!Iiddle Miocene
land connection of the Afro-Arabian plate and A�ia: A Rauhe, M . , Frey, E., Pemberton, D. S., and Ross
mann, T.
1999.
major event tor hominoid dispersal? In
Hard Evidence,
Fossil crocodili<ms tl·om the late
Miocene Baynunah Formation of the Emirate of Abu
Liss, New York.
Dhabi, United Arab Emirates: Osteology and palaeoe
---. 1983.
cology. Chap.
163-85
14
in
l:ossil Vertebrates ofArabia,
pp.
(ed. P. J. \Vhybrow and A. Hill). Yale Univer
sity Press, New Haven.
pp.
42-50
Ancestors: The
(ed. E. Delson ) . Alan R
Les Bovidae (Artiodactyla, Mammalia)
du Miocene moyen de Ia Formation Hofuf (Province
du Rasa, ltrabie Saoudite) .
Palaeove1·tebrata 1 3 :
1 57-206. Roger, J., Pickford, M., Thomas,
Lapparent de
Brain, F. de, Tassy, P. , Van Neer, vV., Bourdillon-de
Thomas, H . , Roger, J. S., and Al-Sulaimani, Z .
Yertebres fossiles dans Jc Miocene de la region du
du continent arabo-africain et d'un primate tarsi
Grissac, C., and i\1-Busaidi, S.
Huqf au Sultanat d'Oman.
1994.
Decouverte de
Annales de Palfontologie,
Pa1·is 80: 2 5 3-73 . S . , and Thomas, H .
1979 .
Decouverte de
Rongeurs dans le 1\!Iiocene moyen de Ia Formation
Compte Rendu Sommaire de la Societe Geologique de France 1 : 34-37.
Hofuf (Province du Hasa, Arabie Saoudite) .
Senut, B., and Thomas, H .
1994.
First discoveries of
anthropoid postcranial remains from Taqal1 ( Early Oligocene, Sultanate of Oman ) . 1n
Current H·imatol-
1988 .
Decouverte des plus anciens "anthropoYdes"
iforme dans !'Oligocene du Sultanat d'Oman.
Cmnptes Rendus de l'Acadenzie des Sciences, Paris 306: 823-29. Thomas, H . , Roger, ] . , Sen, S . , Bourdillon-de-Grissac, and }J-Sulaimani, Z.
1989a.
Decouverte de
vertebres fossiles dans !'Oligocene inte1ieur du Dhotar (Sultanat d'Oman).
Geobios 22: 101-20.
Thomas, H., Roger, ] . , Sen, S . , Deja.>:, ] . , Schuler. _\1.,
Al-Sulaimani,
Bourdillon-de-G1issac, C . . Breton,
-
P. J, WHYBROW A.c'iD D. CLEMENTS
G., Brain, F. de, Camoin, G., Cappetta, H., Carriol,
R. P. , Cavelier, C., Chaix, C., Crochet, J, Y., Farjanel,
G., Gayet, M., Gheerbrant, E., Lauriat-Rage, A., Noel, D., Pickford, M., Poignant, A. F., Rage, J, C., Roman, J, Rouchy, J. M., Secretan, S., Sige, B . , Tassy, P. , and \Venz, S . 1991a. Essai d e reconstruction des milieux de sedimentation et de vie des Primates anthropoi"des de l'Oligod�ne de Taqah (Dhofur, Sul tanat d'Oman). Bu-lletin de la Societe Geologique de France) Paris 162: 7 13-24. Thomas, H., Roger, J., Sen, S., Pickford, M., Gheer brant, E., Al-Sulaimani, Z., and Al-Busaidi, S. 1999. Oligocene and l\tiiocene terrestrial vertebrates in the southern Arabian peninsula (Sultanate of Oman) and their geodynamic and palaeogeographic settings. Chap. 30 in Fossil Vertebrates ofArabia) pp. 430-42 (ed. P. J. Whybrow and A. Hill). Yale University Press, New Haven. Thomas, H., Sen, S., Khan, M., Battail, B., and Liga bue, G. 1982. The Lmver Miocene Fauna of Al-Sar rar (Eastern Province, Saudi Arabia). ATIAL) The jour nal of Saudi Arabian Archaeology 5 : 1 09-36. Thomas, Sen, S., and Ligabue, G. 1980. La faune .Miocene de Ia Formation Agha Jari du Jebel Hanrrin (Irak). Proceedings Koninklijke Nederlandse Akademie van Wetmschappen B83: 269-87. Thomas, H., Sen, S., Roger, J., and Al-Sulaimani, Z. 1991b. The discovery of Moeripthecus markgrafi Schlosser (Propliopithecidae, Anthropoidea, Primates ) in the Ashawq Formation (Early Oligocene of Dhofar Province, Sultanate of Oman). Journal ofHuman Evolution 20: 33-49. Thomas, H., Taquet, P., Ligabue, G., and Dei'Ag nola, C. 1978 . Decouverte d'un gisement de vertebres dans les depots continentaux du l\1iocene moyen du Rasa (Arabie Saoudite). Compte Rendu Sommaire de la Societe Geologique de France 1978: 69-72 .
Thomas, H., Tassy, P., and Sen, S. 1989b. Paleo gene proboscidean remains fiom the Southern Dho far (Sultanate of Oman) . Paper presented at the Fifth International Theriological Congress (Rome, 22-29 August, 1989), sect. "Evolution and Paleoe cology of Proboscidea" . Abstracts, Papers, and Posters I: 166. Tleel, J. vV. 1973. Surface geology Damman Dome, Eastern Province, Saudi Arabia. Bulletin American Association ofPetroleum Geologists 57: 558-76. vVhitmore, F. C., and Madden, C. T. 1995 . Paleocene Vertebrates from jabal Urmn Himary Kingdom of Saudi Arabia. United States Geological Survey Bul letin no. 209 3 . Government Printing Office, Wash ington D.C. Whybrow, P. J. ed. 1987. l\1iocene geology and palaeontology of Ad Dabtiyah, Saudi Arabia. Bulletin of the British Museum (Natural History)) Geology 4 1 : 367-457. \iVhybrow, P. J. 1989. New stratotype; the Baynunah Formation l\1iocene ), United Arab Emirates: Lithology and palaeontology. Newsletters on Stratigra phy 2 l ; 1-9 .
---. 1992. Land movements and species dispersal. In I11e Cambridge Encyclopedia ofHt1man Evolution) pp. 169-73 (ed. S . Jones, R. Martin, and D. Pil beam). Cambridge University Press, Cambridge. Whybrow, P. J., and Clements, D. 1999 . Late l\1iocene Baynunah Formation, Emirate of Abu Dhabi, United Arab Emirates: Fauna, Flora, and localities. Chap. 23 in Fossil Vertebrates of Arabia> pp. 3 1 7-33 (ed. P. J. Whybrow and A. Hill). Yale University Press, New Haven. Whybrow, P. J ., and Hill, A. eds. 1999. Fossil Verte brates ofArabia. Yale University Press, New Haven.
ARABIAN TERTIARY FLOIL", FAUNA, Al'ID LOCALITIES
\'v'hybrow, P. J., and McClure, H. A. 1 98 1 . Fossil mangrove roots and palaeoenvironments of the :Miocene of the eastern Arabian peninsula. Palaeogeog raphy, PalaeoclimatologY, Palaeoecology 32: 2 1 3-25. \Vhybrow, P. }., Collinson, M. E., Daams, R., Gen try, A. W., and McClure, H A. 1982. Geology, fauna (Bovidae, Rodentia) and flora from the Early
-
Miocene of eastern Saudi Arabia. Tertiary Research 4: 105-2 0. 'v'Vhybrow, P. J., Hill, Yasin al-Tik:riti, W., and Hail\vood, E. A. 1990. Late Miocene primate fauna, flora and initial palaeomagnetic data from the Emirate of Abu Dhabi, United Arab Emirates. Journal of Human Evolution 19: 583-8 8.
Tertiary volcanics, Gulf of Aden, Republic of Yemen.
The Tethyan Arabia11 Gulf, the Mediterranean, and the World's Tertiary Oceans
PART
This part brings together three research projects that focus on the geochronological relationships between the oceans and landmasses. These projects provide an overview of events that relate to the geology of the marine rocks of the Arabian Penin sula during the Tertiary period, 65 to 2 million years ago. C. Geoffrey Adams, Deryck D. Bayliss, and John E. Whittaker ( Chapter 34) describe three modern hypotheses: those of Drooger, of Steininger and Rogl, and of Adams and colleagues. These hypotheses relate to the dating of the event that finally separated the proto- Mediterranean from the Indian Ocean, and they are analysed and dis cussed in the light of data that have become avail able during the past decade. Several problems of particular importance to the dating and correlation of marine Neogene sediments in the lvliddle East are also discussed. It is concluded that although the available palaeontological and l ithostratigraphical evidence now points most strongly to an early Miocene (Aquitanian) disconnection, and suggests that a later reconnection in middle Miocene times is highly unlikely, further studies of certain sequences in Iran and Iraq will have to be under tal�:en before a general consensus can emerge. Fred Rogl ( Chapter 3 5 ) provides up-to-date information on palaeogeographic reconstructions of the area around the Mediterranean from the Oligocene to the Pliocene . The palaeogeographic reconstruction of the circum-Mediterranean area is dependent on the palinspastic rearrangement of tec tonic units and on well-defined stratigraphic corre lation between different sedimentary basins. Detailed i nformation for such a rearrangement is, however, limited. For the tectonic requirements, information is available from around the Apennines, Alps, and Carpathians, but is not as accurate for the Eastern Mediterranean region . An updated and refined stratigraphic correlation for the Mediter ranean and Paratethys is presented, based on new palaeomagnetic, radiometric, and biostratigraphic data from different research groups. Current palaeogeographic questions that are discussed cen tre on the intracontinental mammal migrations between Europe, North America, and Africa in the
Oligocene and Miocene for specific regions, such as ( l ) the European-Asian connection, (2) the
B ering landbridge, ( 3 ) the " Gomphotherium" land bridge in the Near East, and (4) the Gibraltar land bridge . Finally, Norman MacLeod (Chapter 36) describes changes to the world's oceans during the lvliocene epoch and earlier. The oceanographic changes in the Oligocene through to the Miocene are among the best understood of any ancient time interval and illustrate the complex interplay of tec tonic, circulation, climatic, and watermass factors. Tectonically, the two most important palae oceanographic events of this period were the pro gressive isolation of the Antarctic landmass at the South Pole and the sinking of the Iceland-Faeroe ridge between the North Adantic and Arctic Oceans. The former led to the isolation of the Antarctic continent with its progressive refrigeration as development of a circum-Antarctic current sys
tem prevented warm waters from the southern mid latitudes from migrating south. The latter changed the nature of deep-water circulation throughout the Atlantic basin such that relatively warm saline waters originating in the Arctic Ocean surfaced south of the Antarctic Convergence. By the middle Miocene, these two factors combined to produce extensive glaciation of the Antarctic continent, a substantial lowering of sea level, and increase in overall oceanic salinities. These changes in turn affected mid- and low-latitude circulation as well as properties of oceanic water masses. In the biotic realm several noteworthy changes can be either directly or indirectly tied to these palaeoceanographic events. These include a dramatic increase in plankton productivity in the Southern Ocean, with the accompanying diversification of plankton-feeding clades (for example, mysticete whales); increases in mid- and low-latitude endemism among planktonic biotas; and increases in terrestrial biotas adapted to a mixed woodland-grassland and savannah-like biome as increases in the latitudinal temperature gradient dur ing the Miocene forced these habitats to replace the Oligocene and early Miocene tropical forests.
The Terminal Tethyan Event: _L\ Critical Review of the Conflicting Age Determinations for the Disconnection of the Mediterranean from the Indian Ocean C. GEOFFREY ADAMS,
DERYCK D .
:\lore than a hundred years ago, S uess ( 1 8 9 3 ) introduced Tethys a s a broad seaway separating Europe and North Mrica, and stretching eastwards across the area now occupied by Turkey, Syria, Iran, Pakistan, northern India, Tibet, Burma, and Thailand, thus uniting the Atlantic with the Indian and Pacific Oceans. Geologists and palaeon tologists have since made numerous attempts to date the event that caused the final disruption of Tethys and, with it, the creation of the proto Mediterranean Sea. It has long been known that the connection -vvith the Pacific was lost at about the end of Cretaceous times and that the relict easterly ann of Tethys finally disappeared during the Eocene when India collided with Asia. The earliest attempts at dating the terminal dis connection were based on the distribution of plants and terrestrial animals (mainly mammals) in north ern Mrica, Arabia, and southwest Asia, it being realised that these organisms could not readily cross any seaway of significant breadth. The result was a general consensus of opinion that a landbridge from Arabia to southwest Asia must have come into existence during Miocene times (Termier and Termier, 1960 ) . Dating the relative sedimentary sequences was, however, always a problem and it was not until planktonic foraminifera became widely and successfully used in zonation (from
BAnrss, AND JoHN E.
\NHITTl\I<ER
1957 omvards), and modern chronostratigraphic methods were introduced, that accurate age deter minations of the critical marine successions became possible. In recent years, three attempts have been made to date the disconnection more accurately than hitherto and although all are based on the distribu tion of marine invertebrates (wi.tl1 or without a consideration of lithofacies), they have produced conflicting results. Drooger ( 19 79, 1993 ), on the basis of mor phometric data obtained trom studies of generally accepted evolutionary lineages in three families of foraminifera (the Miogypsinidae, Cycloclypei idae, and Lepidocyclinidae ) , concluded that the terminal disconnection occurred at about the end of Oligocene times-that is around 2 3 . 8 million years ( Ma) ago, according to Steininger et al. ( 1 994). Steininger and Rogl ( 1 979) and Rogl and Steininger ( 1 983, 1984), utilising previously pub lished evidence on the distribution of marine and terrestrial sediments plus some scattered palaeonto logical data, concluded that although a disconnec tion may halT occurred in the early Miocene, the seaway ,,·as re-opened later and not finally closed until the middle L\liocene. Adams et al. ( 19 8 3 ) , basing their conclusions o n the distributions o f all foraminifera in the mid-Tertiary of genera of
Copyright © 1999 by Yale University. All rights reserYed. ISB?\i 0-300-07183-3
-
C. G. ADAMS ET
AL.
the Mediterranean, Middle East, and Indian Ocean
glomerates. The most cursory examination of the
areas, in the context of the best-known sedimentary
A. howchini show that howchini. Com pare, for example, Adams ( 1968: pl. 2, figs 3, 4, and 7) with Abid and Sayyab ( 1989: pl. l , fig. 2 ) , which, i n fact, is a perfectly good A . asmariensis. All records of A. howchini from the Middle East
sequences, decided that the late early Miocene ( B urdigalian) was the most probable time of dis connection. Clearly, all three conclusions cannot be correct. The purpose of this chapter is, therefore, to review
specimens they illustrate as
they in no way resemble the true
and explain the evidence on which the various con
should tl1erefore be treated with suspicion, unless
clusions depend and to assess the relative merits of
tl1ey can be verified.
the authors' claims in the light of data obtained since
1983.
First, however, it is necessary to con
sider briefly four important problems that have complicated and confused the literature for the past
50
years.
Assumptions that the Ranges of Certain Marker Taxa Are Wel l Established Most palaeontologists appear to believe tl1at the first and last appearances of certain marker species are well established, but this is not true. For exam
BIOSTRATIGRAPHIC PROBLEMS Failure to Identify Critical Species Accurately When this occurs, sediments are wrongly dated and correlations inaccurate. The foraminiferal genus
Aust1·otrillina provides
many good examples of
inaccurate identifications. Before
1968,
only two
species of Austrotrillina were known from the Mid
A. paucialveolata Grimsdalc and A. how chini ( Schlumberger). But Adams ( 1 968) then showed that the true A. howchini from the Indo dle East,
West Pacific region was quite different (structurally more advanced), and also younger, than the torms described under this name tram the Middle East. He theretore described a new species,
sis,
Orbulina mturalis Br6nni Borelis melo ct-trdica Reichel in the Mid
ple, tl1ey believe that
A. asmarie.n
whi ch accommodated most of the wrongly
mann and
dle East appeared simultaneously and that they, by definition, mark the base of tl1e middle Miocene
(N 9).
0. suturalis, it is not B. melo curdica, which is present in the upper part of N 8 and thus straddles the boundary (Adams, 1 9 8 3 , 1992; Adams et al. , 1 98 3 ) . Bore/is melo curdica can thus be regarded as a reliable mid While this is true for
true tor
dle Miocene marker only when it is accompanied by otl1er marker species. Other so-called markers are far less well known than
B. �nelo curdica.
Absolute Ranges of Larger Foraminiferal Species For some strange reason, micropalaeontologists tend to accept statements about the ranges of
named forms previously described from the Middle
important taxa when they are made by well-known
East. Despite this, many subsequent authors contin
specialists, even when they are unaccompanied by
ued to use
A. howchini in
the wrong sense and
to perpetuate the belief that this species is known
proof of any kind. Adams ( 1 992 ) has discussed this problem with reference to
Miogypsina
and has ques
in the Middle East. There are a tew records of
tioned the upward extension of its range in the
A. howchini from the western Mediterranean ( see Adams, 1968), but these require re-examination in
Indo-Pacific region from N8 to
the light of its absence from the Middle East.
the evolution of the
Abid and Sayyab
( 1 989) ,
in a valid attempt to
( 1963 ) ,
Nl4.
Drooger
in a well-known review article, stated that
1l;iiogypsina lineage
appeared to
be over in the Mediterranean area just before the
01·bulina,
determine the age and provenance of the derived
appearance of
blocks in the basal conglomerate of the Euphrates
dence has appeared to prove him wTong. Yet Prazak
Limestone, claimed to have found tour species of
( 1978)
Austrotrillina
Chattian/Aquitanian faunas of northern Jraq are
including
A. howchini
in the con-
and no subsequent evi
and other authors, while accepting that the
THE TI&'\11::-lAL TETHYAN
clearly Mediterranean in type, then go on to record Jfiogypsina (species usually unnamed) from strata of middle Miocene or younger age ( see Prazak, 1978: tables 2 and 5 ) explain this anomaly by citing the already questionable extended range for the genus in the Indo- Pacific, thus implying a north "·ard transgression of the Indian Ocean during the middle or later Miocene that allowed one Indo Pacific genus to migrate northwards, but not any of the others. Such a scenario is improbable to say the least. .
THE OLIGOCENE/ MIOCENE BOUNDARY This has always been a difficult boundary to define in shallow-water marine sediments because the are larger foraminiferal faunas in the tv,ro \·ery similar, and, for our purposes here, the recently published proposals of Steininger et al. ( 1 994) are accepted. These state (inter alia) that the boundary coincides approximately with the FAD (first appear ance datum) of the planktonic foraminifer Para globorotalia kugleri ( Bolli). This requires some modification to the same boundary as defined in the East Indies Letter Classification (for example, Adams, 1 970, 1 9 8 3 ) for shallow-shelf, carbonate environments, but the necessary changes can be accommodated. Basically, the boundary vvould be marked by the first appearances of },fiogvpsina gunteri Cole and/ or Miogypsinoides bantamensis Tan Sin Hole.
THE THREE DISCONNECTION HYPOTHESES Disconnection at the End of Oligocene Time (Drooger's Hypothesis) Drooger ( 1 979) regarded the disconnection as hav ing occurred at about the end of the Chattian. He used data from morphometric studies of the larger foraminiferal genera Miogypsina s.s. and Lepidocy clina ( Nephrolepidina ), together with th e extinction of Cycloclypeus in the Mediterranean region, as the
basis tor his conclusion. He repeated his claim in 1 99 3 , but \Yithout mentioning that there were two alternati,·e hypotheses. The three taxa on \\-hich Drooger relied all appeared in the Mediterranean region in Oligocene times and all three are known to haYe survived far longer in the Indo-Pacific region. "\11 show gradual evolutionary changes ( trends) in the jm·eniie parts of the shelL The morphometric approach to taxonomy involves the study of the evolutionary trends shm,,n in various shell features and their quantification based on counts or measurements of different para meters . Given sufficient specimens, a statistically significant mean can be calculated for each parame ter and the results plotted in the torm of graphs for assemblages from different stratigraphic horizons. Using this method, results have so far been obtained from lineages within genera belonging to several different families, the best coming from the main Miogypsina lineage ( several other lineages occur within this genus but for various reasons are not applicable to the solution of the disconnection problem; see Drooger, 1993 ) . Drooger has been able t o show that the Medi terranean lineage of Miogypsina differs from tl1e its equivalent in the Indian Ocean. Similarly, the lineages within Lepidocyclina ( Eulepidina) and L. ( Nephrolepidina) seem to differ in tl1e two regions after the end of Oligocene times. These differences do not, however, become significant until the B ur digalian. Cycloclypeus was represented in both regions during the Oligocene, but became extinct in the western Tethys during Chattian time; it lives on in the Indo-Pacific today. The reason for the extinction of this genus in the western Tethys is unknown. It could, as Adams ( 1967) suggested, and as Drooger ( I 993 ) now appears to belie\·e, be linked to the disconnection of the two areas. B ut it also disappeared from the western side of the Indian Ocean at about the same time and this can hardly be attributed to the closure of the Tethys seaway. Our knowledge of Cycloc�vpms i n the ;.. rediterranean region is probably 110\Y complete thanks to the intensive research undertaken in this
-
C. G. ADAMS ET li.L.
area during the past 6 0 years. The same cannot be said of the Indian Ocean where much work remains to be done. Comparison of Cycloclypeus lineages in the two areas is therefore impossible at present. Species based on morphometric studies have the advantage of being sharply defin ed. Thus, Mio gypsinoides bantamensis Tan is described by Drooger ( 19 9 3 ) as possessing from 1 0 to 1 3 chambers in the initial whorl whereas Miogypsin oides deharti van der Vlerk has 8 to more than 1 0 . The element of subjectivity in identification is therefore removed, but morphometric species are not Linnean species and no biological reality is implied by these names. VVhether Linnean names necessarily imply a greater degree of reality is an open question. Lepidocyclina is the other genus in which the evolution of the embryonic apparatus has been shown to be susceptible to morphometric investi gation, and work in the East Indies and in the Mediterranean regions appears to indicate that the l ineages were different in the two regions (see D rooger, 1 99 3 , for a summary). Morphometric studies of the subgenera Eu.lepidina and Nephrole pidina have not, however, yet yielded the fine mor phometric distinctions claimed for Miogypsina and correlation is therefore less precise. Drooger's hypothesis ( 1 979, 1993) for the dat ing of the terminal disconnection may be said to rely on two unproven assumptions : 1 . That the main Miogyp.>ina lineage developed differently in the Mediterranean and Indo Pacific regions after the end of Oligocene times. This is true, but the main changes were i n post-Aquitanian times and not, therefore, necessarily relevant to this discussion, although, be expected as Drooger says, a time-lag to occur. 2 . That Lepidocyclina ( Nephrolepidina) and L. (Eulepidina) also developed differently in the two areas after Oligocene times. Unfortunately, no comparative study of the lepidocyclines of the Middle East and the western part of the Indian subcontinent yet exists . Beretti and
A.mbroise's ( 1 9 8 0 ) work on tl1e Madagascar faunas fails to elucidate the situation.
Disconnection in Middle Miocene Times (Steininger and Rogl's Hypothesis) Steininger and Rog] approached the disconnection problem in a quite different way. Following a pre liminary paper (Steininger and Rogl, 1 979 ), they ( Rogl and Steininger, 1 9 8 3 , 1 984) subsequently constructed a series of impressive palaeogeographic maps for the Mediterranean area and Middle East on the basis of previously published lithological data, but apparently without checking whether the sediments concerned were accurately dated. They concluded that a marine connection between the Indian Ocean and the Mediterranean area existed throughout the Oligocene, that it was dis rupted in the Middle East at some time during the early Miocene, when evaporites were widely deposited, and briefly restored again in early mi ddle Miocene times. No lithostratigraphic evidence was presented for the middle Miocene reconnection of the two areas and indeed none exists, reliance being placed on the supposed occurrence of two or three Indo-Pacific species of planktonic foraminifera in Paratethys. This hypothesis, as Adams et al. ( 1 9 8 3 ) pointed out, requires a deep-water seaway o f mid Miocene age along the line of the present Zagros Range. While being theoretically possible, tl1is is highly unlikely, especially as simpler explanations are available. Another objection made to the Steininger and Rogl hypothesis is that the deep-water trough nec essary for the passage of planktonic foraminifera could not have existed without the presence of shallow-water tacies on either side. Larger fOrami nifera would certainly have "migrated" along such shallow-water shelves, but there is little evidence that tllis occurred. Rogl and Brandstatter ( 1 993 ) reported Amphistegina man·mzilla (Fichtel and Moll) from central Poland and Austria, but of all possible larger foraminifera, Amphistegina is the least impressive as a potential migrant from the Indian Ocean because the genus was present in the Mediterranean region at a much earlier date and
THE TERMINAL TETHY&'l EVENT
could therefore have evolved in situ. In fact, these authors admit that there is at least one questionable record of this species from Israel. Convincing evidence of reconnection can be provided only by the discovery of Indo-Pacific genera such as Alveolinella or Flosculinella or of species of Lepidocyclina or Cycloclypeus) the imme diate ancestors of which are known not to have been present i n the Mediterranean region previ ously. Such evidence has not yet been forthcoming.
Disconnection During the Early Miocene (Adams, Gentry, and Whybrow's Hypothesis) Adams et al. ( 1983) investigated and correlated the distribution of critical marine carbonates and terres trial sediments from the northern limits of the Indian Ocean to the Mediterranean and tabulated their associated faunas. They concluded that the continuity of marine sedimentation across the Mid dle East appears to have been interrupted during Aquitanian times. By the mid-Burdigalian at the lat est, a definite barrier to the dispersal of marine organisms existed between the Indian Ocean and the Mediterranean, and that this could have been the landbridge needed for the Orleanian dispersal of mammals. They also firmly refuted Steininger and Rogl's contention that a middle Miocene re connection occurred. In the past 12 years new information has been published that has a critical bearing on the argu ments and this is added belm:v. In 1989 Abawi firmly dated the Euphrates Formation of northwest Iraq showing that it falls within the planktonic zones N 7-N 8 (late Burdi galian-early Langhian) , but the depositional envi ronment indicated by the smaller benthic forami nifera was shallow-water, open marine. This is further evidence that there was no deep-water connection between either the Indian Ocean or the Mediterranean in Burdigalian times. The overly ing Jeribe Limestone is dated not by planktonic foraminifera but by larger foraminifera, the assump tion being that Bore/is melo curdica Reichel is a
!:Ell
marker tor the middle Miocene. This is not true, however, as this taxon occurs on both sides of the early/middle Miocene boundary (Adams, 1983, 1 992; Adams et al. , 198 3 ) . It is clear, therefore, that further evidence is needed betore a middle Miocene age can be ascribed to this limestone unit. In this context, it should also be remembered that the uppermost part of the Gachsaran Forma tion (Lower Fars of many authors) in Iran ( Lurestan and Khuzestan) also contains B. melo curdica (the furthest south this Mediterranean subspecies occurs) and is thus virtually coeval with the Jeribe Lime stone. McCall et al. ( 1994) described the early Mio cene (Aquitanian) coral limestones from the Maluan Mountains of southern Iran. Situated within an arm of the "proto- Arabian Gulf", the corals (and associ ated foraminifera examined and identified in part by Adams) have especial palaeobiogeographical and palaeogeographical interest because they were col lected from localities close to areas of Miocene uplift associated with the severance of the Tethyan seaway. Both corals and foraminitera show an almost entirely Indo-Pacific affinity, indicating significant biogeographical separation of the Mediterranean from the Indo-Pacific region even before the final closure occurred. In addition to this coral and foraminiferal evidence, a study of hemiasterid echinoids by N eraudeau ( 1994) concluded that the Indian Ocean became disconnected from the Mediterranean in early Miocene (Aquitanian) times. Robba ( 1 987) and Piccoli et al. ( 1991 ) pre sented results of detailed studies of molluscan fau nas in relation to the final occlusion of Tethys, and these should be noted here. Robba's findings are based on a faunal analysis at supraspecific level from Italy and Indonesia (Java) and Piccoli et al.'s mathematicaJ modelling is constructed trom identi fications from many scattered parts of Tethys and beyond (20 areas) , from the Paris Basin in the \vest to Fiji in the east. In his abstract, Robba states "the final Tethyan event was already concluded in the Burdigalian, between 18 and 19 Ma ago"; fur ther on (p. 409), he writes that the data proYided
!liPJ
C. G. AD1\.MS
ET AL.
by the benthic molluscs are in agreement with the larger foraminiferal evidence of Adams et al.
( 1 98 3 ),
indicating faunal divergence took place
Adams et al .
( 1983)
accepted that there was
never a good marine connection between d1e Mediterranean and Indian Ocean after the Oligo
during the Aquitanian and that the Tethyan seaway
cene, Aquitanian time being represented mainly
(quoting the latter) "ceased to be an effective dis
by evaporitic deposits. Nevertheless, they thought
persal route for most marine taxa, and that a land
that the Burdigalian faunas and sediments across
bridge between Arabia and Southwest A.sia existed
the region provided possible evidence for the con
by mid-Burdigalian times". In their text, Piccoli
tinuance of a marine link. As a consequence of
et al.
( 199 1 : 1 8 7 )
agree \vith Robba
( 1 987)
about
the final occlusion of Tethys occurring about
1 8.5
Ma ago, but their palaeogeographical sketch maps, prepared on the basis of recent literature, for the late Oligocene
(25
Ma), early Miocene
and the middle Miocene
(15
(20
Ma),
Ma) all show an
interrupted, if restricted seaway through the Mid
Austrotrillina howchini Floscu linella, and associated genera such as Cyclodypeus, Alveolinella, and Lepidocyclina ( Nephrolepidina), to d1e absence of the true
(Schlumberger) in d1e region, the failure of
penetrate further north than the southern part of
( Guri Limestone) and of Borelis melo curdica Reichel ( a Mediterranean sub
the Persian/Arabian Gulf
dle East connecting eastern and western Tethys.
species) to occur in the Indian Ocean , it is now
Occlusion is indicated on the map only for the
thought that there was no continuous Burdigalian
Messinian
(5.5
Ma). Their three land/sea distribu
seaway across the Middle East. On present evidence, effective disconnection
tion maps ( late Oligocene to middle "Miocene) have an " u nreal" appearance and are unconvincing, espe
of the Mediterranean from the Indian Ocean
cially the long, narrow seaways and the parallel,
occurred, if not precisely at the end of the Chattian (as Drooger,
elongate peninsulas. Finally, it must be said, in fairness to others, that Adams et al.
( 19 8 3 )
1 979 ,
postulated on other evidence) ,
soon after i n early Miocene ( Aquitanian) times.
based their conclusions
Such evidence seems to fit well with Why
( 1 984: 1 9 5 )
on the distribution of all published records of
brow's notion
larger foraminifera across the Middle East lmown
mal-bearing Hadruk:h Formation of Saudi Arabia
1982 ) .
on the age of the mam
up to that time. The wealmess of that approach
(v\7hybrow et al. ,
is that some records are not verifiable, because
connection between Arabia and Iran/Iraq might
Thus an early Miocene
material is unavailable, so certain identifications are
have been available for "inter-migrations" of i\fro
questionable.
Arabian and Southwest Asian continental faunas.
SUMMARY AND CONCLUSIONS
NOTE
Drooger's hypothesis
( 1 979, 1 99 3 )
has the merit of
internal consistency. It relies only on morphometric studies of a few taxa, disregards the rest and takes no account of the sedimentary record. Steininger and Rogl
1983, 1984)
( 1 979 ; Rogl and Steininger,
accept the sedimentary record at its
l.
C. Geoffrey Adams died on
6
February
1 99 5 .
His unfinished manuscript (to have been pre sented by C.G.A. in a complete form at the First International Conference on the Fossil Vertebrates of Arabia, held in the Emirate of Abu Dhabi, March
1 99 5 )
has been completed
face value and do not question the age determina
and updated by two of his colleagues as a mark
tions of previous authors. They postulate a mid
of respect for his considerable knowledge and
Miocene reconnection of the Mediterranean and
his unswerving pursuit of accuracy and preci
Indian Oceans, arguing that the marine s ediments
sion in Cenozoic stratigraphy. The opinions
that would provide the necessary evidence are now
expressed in this chapter, however, are those of
buried below the Zagros Range.
C. Geoffrey Adams.
THE TERMIKAL TETHYAK EVENT
Gumbel a Madagascar, mesure du
REFERENCES m e?-
z z I w
...J
.... z 8. UJ c. U - ::I Q -
:a:
10
f---
M
13
1-
1(/)
w
0
w
c rn
"53
u
c rn
N
Messi nian
NN 11
c-----1
NN
z 0::: w
Vith a description of a new
Calotomus
species from the middle 1Vliocene (Badenian) of Addicott, W.
1974.
Giant pectinids of d1e eastern
North Pacific margin: Significance i n Neogene zoo-
Annalen Natm·histo1risches Museztm VVien 92A: 55�71.
Austria.
- F. R6GL Berggren, W. A., Kent, D. V., Swisher, C. C., III, and Aubry, M.-P.
1995 .
A revised Cenozoic geochronol
Geochronology, Time Scales and Global Stratigraphic Correlations: A Uni fied Temporal Framework fm· an Historical Geology, pp. 129-212 ( ed. W. A. Berggren, D. V. Kent, and J. ogy and chronostratigraphy. In
Hardenbol) . SEPM (Society for Sedimentary Geol ogy) Special Publication no.
54.
...'-'J..""''urucll,,
E., Mayhew, D. F., Meulen, A. J. van
der, Schmidt-Kittler, N . , and Telles Anumes, M .
1992.
Report o f tl1e RCMNS working group o n fossil
mammals, Reisenburg
Cahuzac, B . , Alvinerie, J., Lauriat-Rage, A . , Montenat, C., and Pujol, C.
Tulsa.
1 990. Newsletters on Stratigra
phy 26: 65-1 18.
1992.
Palaeogeographic maps of the
northeastern Atlantic Neogene and relation with the Boccaletti, M . , Calamita, F. , Centamore, E . , Chioc chini, U . , Deiana, G., Micarelli, A. , Moratti, G., and Potetti, M.
1986.
Evoluzione dell'Appennino tosco
umbro-marchigiano durante i l Neogene.
Giornale di
Geologia 48 : 227-33.
Paleontologia i Evolucio, Institut de Paleontologia «Miguel Crusafont)) 24-25: 279-9 3 . Mediterranean Sea.
Charollais, J . , Hochuli , P. , Oertli, H . J., Perch
Nielsen, K , Toumarkine, M., Rogl, F., and Paris, J . -L.
Boccaletti, M., Ciaranfi, N., Cosentino, D., Deiana, Gelati, R., Lentini, F., Massari, F., Moratti, G., Pescatore, T., Ricci Lucchi, F., and Tortorici, L.
1990.
1980.
Les marnes a foraminileres et les schistes a
Meletta des Chaines subalpines septentrionales (Haute-Savoie, France ) .
Edogae Geologicae Helvetiae
73: 9-69.
Palinspastic restoration and paleogeographic
reconstruction of the peri-Tyrrhenian area during the
Checchia-Rispoli,
Palaeogeography, Palaeoclimatology, Palaeoe cology 77: 41-50.
dens" Cuvier dei dintorni di Bugio in provincia di
1'-lc:og:ene.
Boger,
and Dermitzakis, M.
1 987.
G . 1 914.
Sui "mastodon angusti
Giornale di Scienze Naturali ed Economiche di Palermo 30: 285-96.
Girgenti.
Neogene
palaeogeography in the central Aegean region.
De Giuli, C., Masini, F., and Valleri, G.
A nnates Instituti Geologici Publici Hungarici, A Mag yar Allami Fiildtani Intezet Evkiinyve 70: 2 1 7-20.
Paleogeographic evolution of the Adriatic area since
1 987.
Rivista Italiana di Paleon tologia, Milano 93: 1 09-26.
Oligocene to Pleistocene.
Boersma, A., Premoli Silva, I . , and Shackleton, N. J .
1987.
Atlantic Eocene planktonic foraminiferal
paleohydrographic indicators and stable isotope paleoceanography.
Paleoceanography 2: 287-3 3 1 .
Dercourt, ] ., Zonenshain, L. P., Ricou, L.-E., Kazmin,
V. G., Le Pichon, X., Knipper, A. L , Granjacquet,
C . , Sborshchik:ov, L M., Boulin, ]., Sorokhtin, 0., Geyssant, J., LepvTier, C., Biju-Duval, B., Sibuet,
Bonis, L . de, Brunet, M . , Heintz, E . , and Sen, S .
J.-C., Savostin,
1992. L a province
198 5 . Presentation de 9 cartes paleogeographiques au 1 /20.000.000 s'etendant de !'Atlantique au Pamir pour la periode du Lias a l'ActueL Bulletin de la Societe Geologique de France, Paris 1 : 637-52.
greco-irano-afghane e t I a
tion des faunes mammalienes a u Miocene superieur.
Paleontologia i Evolucio, Institut de Paleontologia '"Miguel Crusafimt» 24-25: 103-12. Bruijn,
H.
de, <md Zachariasse, W. J .
1 979.
The cor
Dizer, A.
199 1 .
Westphal, M., and Lauer, J.-P.
The biostratigraphy of the Langhian
Proceed 71-8 1 .
relation of marine and continental biozones of Kastel
and Serravallian in North Kahramanmaras.
lios Hill reconsidered.
Annates Geologiq$tes des Helteniques 1 : 2 1 9-26.
ings, Ahmet Acar Geology Symposium,
Bruijn, H. de, Daams, R, Daxner-Hock:, G.,
Dumittica, P.
Fahlbusch, V., Ginsburg, L., Mein, P., Morales, J.,
Paratethys. In
Kosovien, pp.
pp.
1978. Badenian radiolaria from Central M4--Badenian (lvloravien, Wielicien, 23 1-61 (ed. A. Papp, I. Cicha, J. Senes,
OLIGOCENE .WD MIOCENE PAlAEOGEOGRAPHY
-
and F. Steininger). Chronostratigraphie und Neostra
the Hazeva Formation (Israel) and the paleogeography
totypen no. 6. ·vEDA, Bratislava.
of J\liocene Africa.
Newsletters on Stratigraphy 20:
25.
Dumitrica, P., Gheta, N . , and Popescu, Gh. 1975 .
Goriir, N. 1989. Timing the opening of the Black
New data on the biostratigraphy and correlation of
Sea: Seclimentological evidence from the Rhodope
the J\1iddle Miocene in the Carpathian area.
Dari de Seama ale Sedintelor, Institutl de Geologie si Geofizica
Pontide fragment. In
61 ( 1973-1974): 65-84.
Advanced Study Institutes, ser. C, no. 259. Kluwer
Region,
Tectonic EJ'olution of the Tethyan
pp. 1 3 1-36 (ed. A. M. C. Sengor). �ATO
Academic, Dordrecht. Erentoz,
L.
1956. Stratigraphie des bassins neogenes
de Turquie, plus specialement d'Anatolie Meridionale
Hagn, H. 1978. Die iilteste Molasse im Chiemgau/
et comparaisons avec le Domaine Mecliterraneen dans
ostliches Oberbayern (Katzenloch-Schichten, Pri
Publications de l)Institut dJEtudes et de Recherches Minieres de Turquie no. 3 : l-54.
abon). Mitteilunge der Bayerischen Staatssammlung for Palaontologie zmd Historische Geologie 18:
son ensemble.
Erkal, T. 1987. Sedimentation in the strike-slip North
Annates Insti Utti Geologica Publici Hungarici, A Magyar Allami Foldtani Intezet EvkonyPe 70: 23 5-44.
Anatolian fault zone, Thrace, Turkey.
Fejfar, 0. 1990 ( 1989). The Neogene VP sites of
167-2 3 5 .
Neogene Palaeogeographic Atlas of Central and Eastern Europe. Hungarian Geological
Hamor, G . 1988.
Institute, Budapest. Haq, B.
U.,
Hardenbol,
J.,
and Vail, P. R. 1987.
Czechoslovakia: A contribution based in rodents. In
Chronology of fluctuating sea levels since the Triassic.
European Neogene Mammal Chronology, pp. 2 11-36 V. Fahlbusch, and P. Mein).
Science 2 3 5 :
(ed. E . H. Lindsay,
Plenum Press, New York.
1 1 5 6-67.
---. 1988. Mesozoic and Cenozoic chronostratig raphy and cycles of sea-level change. In
Sea-level
Fermeli, G . , and Ioaldm, Ch. 1993. Biostratigraphy
Changes: An Integrated Approach, pp.
and palaeoecological intetpretation of Miocene succes
Special Publication no. 42. Society of Economic Pale
sions in the molassic deposits of Tsotylion, Mesohel
ontologists and Mineralogists, Tulsa.
Pale ontologia i Evolztcio, Institut de Paleontologia ((Jlfiguel Crusafont·v 24-25 ( 1992) : 199-208.
71-108. SEPM
lenic Trench (Grevena area, northern Greece).
Kissel, C., Laj, C., Mazaud, A., Poisson,
Y., Simeakis, K., Fraissinet, C . , and Mercier,
Savascin,
J. L.
1989. Paleomagnetic study of the Neogene forma
Gagic, N. 1 9 8 3 . Representatives of the genus
Borelis
in the Badenian and lower Sarmation of Yugoslavia.
Anuarul Instituutlui de Geologie si Geofizica, Bucaresti
Tethyan Region,
Tectonic Evolution of the
pp. 1 37-57 (ed. A. M . C. Sengor) .
XATO Advanced Study Institutes , ser. C, no. 259. Klmver Academic, Dordrecht.
59: 1 69-81. Gelati, R. 197 5 . J\liocene marine sequence from the Lake Van area, eastern Turkey.
Paleontologia, Milano 8 1 :
tions of the Aegean Sea. In
Rivista Italiana di
427-52.
Konigswald, W. 1981 . Palaogeographische Beziehun gen der Wirbeltierfauna aus der altrertiaren Fossil lagerstatte Messel bei Darmstadt .
jahrbt1-ch Hessen
Geologisches
1 09: 85-1 02.
Goldsmith, N . F., Hirsch, F., Friedman, G . M . , Tchernov, E . , Derin, B . , Gerry, E . 1988. Rotem
mammals and Yeroham crassostreids: Stratigraphy of
Kotlarczyk,
J.,
and Lesniak, T. 1990. Lower part of
the Menilite Formation and related Futoma diatomite
!fl;J F . R6GL member in the Skole unit of the Polish Caq.1athians. Akademia Gorniczo-Huntincza) Instytut Geologii i Suro1vcow Mineralnych) pp. 1-74.
McGowran, B . 1 979. Some .Miocene configurations from an Australian standpoint. Annates Geologiques des Pays Helleniques 3: 767-79.
Krhovsky, J., Adamova, J . , llladikova, J., and Maslowska, H. 1 99 1 . Paleoenvironmental changes across the Eocene/Oligocene boundary in the Zdan ice and Pouzdrany units (vVestern Carpathians, Czechoslovakia) : The long-term trend and orbitally forced changes in calcareollS nannofossil assemblages. In Proceedings 4th INA Conference, Prague 1 991Knihovnicka Zemni Plyn Nafta.-Hodonin, pp. 1 05-87 ( ed. B. Hamrsmid and J. Young).
Mein, P. 1989. Die Kleinsaugerfauna des Unter miozan ( Eggenburgien) von Maigen, Niederoster reich. Annalen des Naturhistorischen Museums, Wien. 90A: 49-58.
Krutsch, W., and Lotsch , D. 1958. Uberischt i.iber die palaogeographische Entvvicldung des zentraleuropiiis chen AlttertHirs (ohne Tethys-Raum). Bericht de1· Geolrtqischm Gesellschaft in der Deutschen Demokratis chm Republik) Berlin 3: 99-1 10. Larsen, A. R. 1978. Phylogenetic and paleobiogeo graphical trends in the foraminiferal genus Amphiste gina. Revista EspaFiola Micropaleontologia 10: 2 1 7-43. Lipps, J. H. 1 964. 1\liocene planktonic toraminitera from Newport Bay, California. TuJane Studies i1t Geol ogy and Paleontology 2: 109-3 3 . Made, J. van der. 1992. African lower and middle 1\'liocene Suoidea (pigs & peccaries). Paper pn:seJated at VIII Jornadas de Paleontologia Barcelona, 8-10 de Octobre de 1992, Barcelona, sect. 2: Taxonomy of the Listriodontinae (Suidae, pigs). Summaries, pp. Barcelona. Martini, E. 1971 . Standard Tertiary and Quaternary calcareous nannoplankton zonation. In Proceedings IInd Planktonic Conference, Roma, 1 970, pt 2, pp. 738-85 ( ed. A. Farinacci) . Technoscienza, Rome. . 1990. The Rhinegraben system, a connection between northern and southern seas in the European Tertiary. Veriiffentlichtmgen aus den·t Ubersee-Aiuseum Bremen A 10: 83-98 .
---
Miller, K. G., M. D., Wright, J. D., and Clement, B . M. 199 1 . 1\'liocene isotope reference sec tion, Deep Sea Drilling Project Site 608: �'lli evalua tion of isotope and biostratigraphic resolution. Pale oceanography 6: 33-52. Miller, K G., Katz, M. and Berggren, W. A. 1992. Cenozoic deep-sea benthic foraminifera: A tale of three turnovers. In Studies in Benthic Foraminifera, BENTHOS (90) Sendai, pp. 67-75. Tokai University Press. Nevesskaya, L. A., Bagdasarjan, K. Nosovosky, M. F., and Paramonova, X P. 1975. Stratigraphic distribIn Report ution of Bivalvia in the Eastern on Activity of the R.C.lvl.N. S Working C:rro ups (1 971-1975), pp. 48-74 (ed. J. Senes) . Bratislava. Nevesskaya, L. A., Goncharova, I. lljina, L. B . , Paramonova, N . P., Popov, S . V., Vorouina, A . A., Chepalyga, L., and Babak, E. V. 1987. History of the Paratethys. Annates Instiutti Geologici Publici Hun garici) A ivfagyar Atla�rti Foldtani Intezet Evkih1yve 70: 337-42. A., lljina, L. B., Paramonova, N. P. , Popov, S. V. , Chepalyga, L., and Babalz, E . V. 1984. Istoriya
Nevesskaya, L. A. , Voronina, A. A. , Goncharova, I.
Paratetisa (History of the Paratethys) . Doklady, 27th International Geological Congress Moscow, Paleo okeanologiya, Koll. 03 3: 91-101 . G., Sartori, S., Franchino, A., Pedron, R. , Claudio, L., and Natale, A. R. 1 99 1 . Mathematical model of faunal spreading in benthic palaeobio ge()graptly ( applied to Cenozoic Tethyan molluscs) .
OLIGOCENE AND lvlrOCENE PAL'\EOGEOGRAPHY
Palaeogeo_qraphy) Palaeoclirnatology; Palaeoecology 86:
1 39-96. Pickford, M. 1990 (1989). Dynamics of Old World biogeographic realms during the Neogene: Implica tions for biostratigraphy. In European Neogene Mmmnal Chronology> pp. 41 3-42 E. H. Lindsay, V. l"ahlbusch , and P. Mein) . PlenUlll Press, New York. Popescu, Gh. 1976. Phylogenetic remarks on the genera Candorbulina; Velapertina and Orbulina.
Dari de Seama ale Sedintelor, Institutl de Geologie si Geofizica 62 ( 1974-75), 3, Paleont. : 61-167.
---. 1979. Kossmljan foraminifera in Romania.
Memoires> Institut de Geologie et Geophysique, Bucarest
29: 5-64 .
Rio, D., and Fornaciari, E. 1994. Remarks on middle to late lvliocene chronostratigraphy. Neogene Newslet
ter 1 : 26-34. Rogl, F. 1996. Stratigraphic correlation of the Para tethys Oligocene and Miocene. Mitteilungen der
Gesellschaft der Geologie und Be'!'lfbaustttdenten in Osterreich) Wein 41: 65-73. Rogl, F., and Brandstatter, F. 1993. The foraminifera genus Arnphistegina in the Korytnica Clays ( Holy Cross Mts, Central Poland) and its significance in the lvliocene of the Paratethys. Acta Geologica Polonica
43: 121-46. Rogl, F. , and D�'ffier-Hock, G. 1996. Late Miocene Paratethys correlations. In The Evolution of Western Eurasian Neogene lMammal Faunas> pp. 47-55 (ed. R. L. Bernor, V. Fahlbuseh, and H.-W. Mittmarm). Columbia t;niversity Press, New York. Rogl, F., and MUller, C. 1976. Das Mittelrniozan und die Baden-Sarmat Grenze in Walbersdorf (Burgen land). Annalen des Naturhistorischen .Museum:>) Wien
80: 221-32.
-
Rogl, and Steininger, F. F. 1983. Vom Zerfull der Tethys zu Mediterran und Paratethys: Die neogene Palaogeographie und Palinspastik des zirkum-mediter ranen RaUllle s. Annalen des Naturhistm�ischen lvfuse
ums) Wien 85A: 1 35-63. Rogl, F., Bernor, R. L., Derrnitzakis, M. D . , Muller, C., and Stancheva, M. 199 1 . On the Pontian correla tion in the Aegean ( Aegina Island). Newsletters on
Stratigraphy 24: 1 3 7-58. Rogl, Steininger, F . F . , and MUller, C. 1978. Middle Miocene salinity crisis and paleogeography of the Paratethys (Middle and Eastern Europe). Initial
Reports Deep Sea Drilling ProJect 42: 985-90. Russell, D. E., and Tobien, H. 1986. Mammalian e\lj dence concerning the Eocene-Oligocene transition i n Europe, North America and Asia. In Terminal Eocene EJJents, 299-307 ( ed. Ch. Pomerol and I . Premoli Silva). Elsevier, Amsterdam. Savage, R. J. G. 1990 (1989). The African dimension in European early Miocene mammal faunas . In E�t1�o pean Neoge1te L�1ammal Chronology) pp. 587-99 (ed. E. H. Lindsay, V. Fahlbusch, and P. Mein ). Plenum Press, New York. Schmidt-Kittler, N. ed. I987. International Sympo sium on },Jammalian Biostratigraphy and Paleoecology of the European Paleogene-Mainz) February 18-21 1 987. Miinchner Geowissenschaften Abhandlungm> ser. A, 10: 1-312.
Schrader, H.}. 1973. Cenozoic diatoms fi·om the northeast Pacific, 18. Initial Reports Deep Sea
Drilling Project 18: 673-797. Serova, M.Y. 1986. Karagansky section (USSR I(ara gansky Island, East Kamchatka). In Terminal Eocene pp. 147-5 I (ed. Ch. Pomerol and I . Premoli Silva). Elsevier, Amsterdam.
miiiJ
F. R6GL
Steininger, F. F., and Rogl, F. 1984. Paleogeography
results in the Atlantic Ocean: Continental marginB
and palinspastic reconstruction of the Neogene of the
and paleoenvironment. Maurice Ewing Series (AGU)
Mediterranean and Paratethys. In The Geological Evo
3 : 275-96.
lution of the Eastern Mediterranean, pp. 659-68 (ed. ]. E. Dixon and A. H. F. Robertson) . Special Publica
Tobien, H. 1987. The position of the "Grande
tion of the Geological Society no. 1 7 . Blackwell,
Coupure" in the Paleogene of the Upper Rhine Graben and the Mainz Basin. In International Sym
Oxford.
posittm on Mantmalian Biostratigraphy and Paleo Bernor, R L., Sen, S . , and Agusti, J. I996. Circum
ecology of the European Paleogene-Llfainz, February
marine-continental chronologie correlations of Euro
197-202.
Eurasian Neogene Mammal Faunas, pp. 7-46 ( ed.
Torre, D . , Ficcarelli, G., Rook, L., Kotsakis, T.,
Columbia University Press, New York.
nary observations on the paleobiogeography of the
Steininger, F. F., Berggren, W. A., Kent, D. V.,
Mediterranean Neogene (Miocene and Pliocene)
pean mammal tmits. In The Evohttion of �stern
R L. Bernor, V. Fahlbusch, and H.-W. Mittmann) . Steininger, F.
Bernor, R L., and Fahlbusch, V.
1990 ( 1989 ) . European �eogene marine/continental chronologie correlations. ln European Neogene Mam mal Chronology, pp. 15-46 (ed. E. H. Lindsay, V.
18-21 1987 (ed. N. Schmidt-Kittler). Munchner Gemvissenschaften Abhandlxmgen, ser. A, 10:
Masini, F., Mazza, P. , and Sirotti, A. 1995. Prelimi Central Mediterranean during the late Miocene. Paper
presented at the International Conference on the Biotic and Climatic Effects of the Messinian Event on the Circum Mediterranean, Benghazi. Technical pro grantme and abstracts, 63-64.
Fal1lbusch, and P. Mein). Plenum Press, �ew York. Veselov, A. A. 1979. To the accurate definition of the Steininger, F.
Rabeder, G . , and Rogl, F. 1985 .
stratigrafical correlation of the Oligocene-lower
Land mammal distribution in the Mediterranean Keo
Miocene border-marking hmizons of the Eastern and
gene: A consequence of geokinematic and climatic
events . In Geological Evolution of the Mediterranean
Central Paratethys . Annates Geologiques des Pays Hel
Wezel ) . Springer Verlag, New York.
Voronina, A. A., and Popov, S. V. 198 5 . Main fea
leniques, hors ser., fasc. 3 : 1243-52.
Basin, pp. 559-71 (ed. D . ]. Stanley and F.-C.
tures of the evolution of the Eastern Paratethys in the Tchernov,
Ginsburg, L. , Tassy, P., and Goldsmith,
Oligocene and lower Miocene. Annates Unive1·sitatis
N. F. 1987. Miocene mammals of the Negev (Israel).
Scientiarum B�tdapestensis de Rolando Eotvos Nomi
Journal of Vertebrate Paleontology 7 : 284-3 1 0 .
natae, sect. Geologica 25 ( 19 8 3 ) : 87-95.
Thenius, E . 1979. Afrikanische Elemente i n der
Whybrow, P.
miozaenen Saegetierfatma Europas (African elements
from Arabia for mantmal "migrations" between Asia
J.
1984. Geological and faunal evidence
in the Miocene mammalian fuuna of Europe) .
and Africa during the Miocene. Courier Forschungsin
A:nnales Geologiques des Pays Hel!iniques 3 : 1201-8.
stitut Senckenbe1lf 69: 1 8 9-98.
Thiede, J. 1979. History of the North Atlantic
Ziegler, P. A. 1982. Geological Atlas of �stern and
Ocean: Evolution of an asymmetric zonal paleo
Central Europe. Shell International Petroleum
environment in a latitudinal ocean basin. Deep drillin g
Maatschappij B.V. , The Hague.
Oligocene and Miocene Palaeoceanography-A Review NORMA.'< �MACLEOD
Much of our present understanding of Oligocene
In turn, patterns of oceanic circulation strongly
and Miocene climates and palaeoceanography stems
influence the Earth's climate by determining the
from the highly successful Cenozoic Palaeoceanog
manner in which heat is distributed within the cou
raphy Project ( CENOP) .
pled oceanic-atmospheric system. The single most
CENOP
was a multidiscipli
nary, multi-institutional effort charged 'Nith the
important factor thought to be responsible for late
generation of palaeoceanographic reconstructions
Oligocene-Miocene oceanographic and climatic
of the Miocene ocean utilising information gained
changes is the progressive isolation of Antarctica at
from the study of m icrofossils collected as a result
the southern rotational pole and the consequent
of the Deep Sea Drilling Project ( DSDP) . The
i nitiation of circum-Antarctic circulation. Shallow
Oligocene-Miocene was chosen as the focus of
water circulation around Antarctica developed in
CENOP investigations
two stages:
because it represented the
(I)
the opening of a shallow proto-Tas
interval during which modern ice sheets first
man Sea in the late Cretaceous about
became established in the polar regions. As such,
years ( Ma) ago and
(2)
80
million
the opening of a shallow
a detailed appreciation of overall Miocene palae
proto-Scotia Sea in the middle-late Eocene
oceanographic and climatic history was thought
Ma ) . Before
to be crucial to the formulation and testing of
water flowed to the north of Australia.
hypotheses seeking to explain the development
70
( 40
Ma southern high-latitude deep
A deep-water connection within the Tasman
and climatic importance of regional events-for
Sea dates from the early Oligocene
example, the closure of the Tethyan gateway and
a space developed between South Tasman Rise and
Paratethyan physiographic evolution.
( 30
Ma) when
East Antarctica. Meanwhile, the Drake Passage within the Scotia Sea opened to shallow-water flow in the late Oligocene
(22
Ma) when the Antarctic
TECTONIC SEHING
peninsula m oved past the tip of southern South
Oceanic circulation patterns are primarily controlled
middle Miocene
by the distribution of continental landmasses over
Antarctic deep-water circulation had undoubtedly
the Earth's surfuce, eustatic sea level, and various
developed.
America. Lawver et al.
climatic factors (for example, evaporation ) that might determin e the density of large water masses.
(20
( 1992)
estimate that by
Ma) a vigorous circum
At the same time as the Scotia and Tasman Seas were opening to circum-Antarctic circulation,
Copyright © 1999 by Yale University. All rights reserved. ISBN 0 300-071 83-3
mm
N. MACLEOD
the Tethyan Sea was closing to circum -equatorial
sea ice as well as enhancing the production of cold
circulation. The initial phase of this event was the
bottom waters. Subsequent refrigeration of the
elimination of circum-equatorial deep-water circula
entire Antarctic region would steepen the global
tion due to deep closure of the Gibraltar gateway in
latitudinal thermal gradient and intensity oceanic
the late Cretaceous. Before tllis closure circum
circulation patterns world wide.
equatorial shallow and deep-water circulation was unimpeded, resulting in the establishment of a
Progressive glaciation of Antarctica, together •vim intensification of the circum-Antarctic Cur
"Supertethyan" climatic belt. Progressive deep
rent, resulted in the formation of the Antarctic con
water closure of the Gibraltar gateway during
vergence during the early Miocene . At that time
the Maastrichtian resulted in degradation of the
tl1e Cretaceous-Paleogene pattern of low-latitude
S upertethys and initiation of the Cretaceous
interocean circulation was reorganised into the
Tertiary faunal turnover event (Jolmson and Kauff
modern geometry of surface-water mass circulation
man,
1996).
Closure of a deep-water connection
belts. This circulation pattern extend'> to the pre
between the Tethys and the Indian Oceans fol
sent day although me exact latitudinal position of
lowed in the latter part of the early Miocene ( 1 8
these belts has varied during the Neogene.
Ma) l e ading to development of the Para tethys and progressive isolation the Mediterranean Basin. Oceanographically important events were also occurring in the Nortl1 Atlantic during the earliest
In the far north, subsidence of me Iceland Faeroe Ridge was also drastically changing me nature of deep-water oceanic circulati on. By the late Oligocene the early stages of ridge subsidence
Neogene. The Iceland-Faeroe Ridge, which bad
allowed North Atlantic surface waters to flow into
previously isolated the Arctic Ocean from me
the Norwegian Sea. This water (which was derived
Atlantic, began to subside below sea level during
from me warm, salty, shallow-water outflmv of
the early Miocene. By me middle Miocene no bar
the Tethys at Gibraltar) became much denser as
rier existed to communication between the Atlantic,
it cooled, sank, and made its way back into the
Arctic, and North Sea waters at eimer shallow or
Atlantic as new bottom water. The increased den
intermediate depths . Finally, deep-water circulation
sity of tllis new Norwegian overflow resulted in
mrough the Indonesian gateway stopped during
enhanced circulation velocities and funda111entally
me nliddle Miocene as a result of tectonic activity
changed the character of deep-sea sedimentation
associated with subduction of the Pacific plate
patterns in tl1e North Atlantic.
i nteracting with Australia's normward drift.
During tl1e middle Miocene the Iceland-Faeroe Ridge subsided to the point where circulation
CHANGES IN OCEANIC CIRCUlATION PATTERNS
between the Arctic and Atlantic Oceans was more or less lmimpeded. This event changed the nature of Atlantic deep-water circulation via me creation of North Atlantic Deep Water ( NADW). Before
While cyclonic g}Tes probably existed in the south
this event water drawn north from the topics was
ern high latitudes during the Cretaceous and
cooled in me North Atlantic, mixed with Antarctic
Paleocene, mese would have been eliminated by
Bottom Water (AABW) and flowed back south as
me advent of circum-Antarctic circulation iu the
North Atlantic Intermediate Water (NAnV) to the
Oligocene or early Miocene. Before the opening of
equatorial regions. Once a connection benveen the
the Tasman Sea tl1e Ross Sea would have been fed
North Atlantic and the Arctic Ocean was estab
by relatively warm waters tl:om the East Australian
lished, warm Atlantic water flmving north to higher
Current. \Vim the Oligocene opening of the Tas
latitudes was cooled more intensively, becoming
man Sea, though, cool surface waters from the
NADvV on re-entry into the Atlantic Ocean Basin.
southern Indian Ocean would have moved over me
By me middle Miocene the volume of Ni\DW was
Ross Sea embayment triggering the tormation of
such that it flowed the entire length of tl1e Atlantic,
0LIGOCE>IE iL'iD MIOCENE PALAEOCEANOGRAPHY
-
where it surfaced as warm, saline water south of the
bottom water in a zone of upwelling south of the
Antarctic Convergence.
Antarctic Convergence. Heat contained in this water upwelling in a cold area would have been converted to latent heat on its evaporation, result
CLIMATIC CHANGES
ing in even h igher evaporation rates . This created a positive feedback mechanism that provided a new
Stable isotopic data suggest that the oceanographic
source of atmospheric moisture in the high south
isolation of Antarctica did not result in the imme
ern latitudes from the middle Miocene onwards.
diate formation of a glacial ice cap. The most
Schnitker ( 19 8 0 ) argued that upwelling NADW
convincing direct evid ence for only partial Antarc
provided the moisture necessary to construct the
tic glaciation during the Oligocene comes from
Antarctic ice sheet. Before th e introduction of
Antarctic palynological data. These indicate that
NADW, prevailing climatic conditions ( which
substantial areas of Antarctica were covered with
had been in place since the late O]jgocene) were
a low- diversity
unable to produce extensive glaciation in Antarctica
Nothofagus and Podoca1-pus flora
(Kemp, 1 9 7 5 ) during the late Oligocene and early
because there was no source for precipitation. Once
Miocene. Taxonomically similar floras are found in
KADW began to upwell around Antarctica, how
modern New Zealand, which is not extensively
ever, conditions favouring the development of
glaciate d . A variety of additional isotopic, faunal,
extensive ice sheets quickly established themselves
and lithological evidence supports this interpreta
and forced a progressive intensification of the entire
tion.
process.
Although there is an abundance of evidence
As glacial conditions became established in
against the development of extensive Oligocene
Antarctica during the middle Miocene the global
Antarctic glaciation, establishment of the East
climate cooled. Sea l evel also began to fal l as large
Antarctic ice sheet does seem to have taken place
amounts of water were removed tro m the oceanic
by the middle Miocene ( 1 4 Ma ago ) . Evidence tor
circulation system . These factors, in turn, intensified
this event includes a sharp increase in the
marine circulation patterns. Cooling was particu
8180 val
ues of planktic and benthic foraminifera, the com
larly pronounced in the waters off Antarctica,
mon occurrence of ice-rafted sediments in th e
resulting in a northward migration of the Antarctic
regions around Antarctica, and an increase in the
Convergence (see below).
frequency of deep-sea hiatuses. Presence of these
The late Miocene glacio-eustatic fall in sea level
hiatuses indicates a shoaling of the calcite compen
has been ,�ariously estimated at 40-50 metres.
sation depth (CCD) t h at may be connected to
Altho ugh a eustatic sea-level drop of tllis magni
decreased temperature of AAB\V. Moreover, Bre'>l'
mde would be expected to h aYe major sedimento
ster ( 1980) suggested that atmospheric circulation
logical, biotic, and climatic repercussions, fortuitous
intensified during this time as reflected by an
com bination of this oceanographic eYent with mid
increase in Antarctic biogenic sedimentation (see
dle European tectonism rna\' ha\'e contributed to
below ) .
desertification of the Mediterranean Ocean basin
The apparent delay o f 20 M a b etween the
( Ryan, 1 9 7 3 ; Hsii. et
1 977; Ryan and Cita,
advent of circum-Antarctic circulation and the for
1 9 7 8 ) . \side fi·om the profound local and regional
mation of extensive Antarctic ice sheets may seem
climatic implications surrounding the recognition of
puzzling at first, but can be accounted for by exam
extensive late Miocene deposition of marine evapor
ining the effect of oceanographic events taking
ites in t h e Mediterranean Basin, removal of these
place in the northern hemisphere. Submergence
salts from the global marine system would h ave
of the Iceland-Faeroe Ridge in the early-middle
lowered salinity of the world ocean by approxi
6%,
Miocene resulted in the creation of KADW and the
mately
ultimate appearance of this relatively >Varm saline
have raised the freezing temperature of sea water,
to 2 per mil. This, in turn would
. N. MACLEOD thereby increasing production of sea ice, raising the Earth's albedo and, through these factors, further lowering the Earth's mean surface temperature. Thus, glacio-eustatic lowering of sea level during
The vertical structure of water masses can also
be inferred via comparative stable isotopic data if
the comparisons are made between different plank tic and benthic foraminiferal species vvithin the
the middle Afiocene precipitated the onset of
same fauna. These data indicate that, as latitudinal
another positive feedback mechanism that helped
temperature gradients increased throughout the
move the Earth's climate towards a late Miocene
Miocene, surface waters also become more stratified
glacial maximum.
into surface, intermediate, and deep
near ther-
mocline) zones. Early Miocene surface waters were
WATER TEMPERATURES AND VERTICAL STRUCTURE General aspects of the physical structure of Ceno
uniformly warm vvith little temperature differentia tion between surface and deep regions . During the middle Miocene, however,
au
intermediate
zone developed. The size of this intermediate zone apparently fluctuated throughout the middle and
zoic oceans can be inferred via comparative stable
late Miocene, expanding during warm climatic
isotopic analysis of calcareous microfossil tests tram
intervals and contracting during cool climatic inter
different localities. The CENOP project subdivided the late Oligocene-early Miocene interval into
three time slices, faunas from which were analysed from multiple deep-sea cores, and
19
time series.
Time slices provide detailed information on spatial
vals. An east-wes t biogeographic structure is also
present in the planktic toraminiferal taunas of the
early-middle Nliocene that Keller
(1985)
has inter
preted as reflecting intensified equatorial circulation during cool climatic events. Disappearance of
variation among water masses while the time series
east-west fimnal ditierentiation within the Pacific
provide an overview of temporal changes and the
during the late Miocene is thought to stem from
character of global ocean throughout the sampled
circulation changes brought about by Antarctic
time interval. Although there is some ambiguity in
glaciation and tectonic closing of the Indonesian
the interpretation of these data ( for example, o180
gateway.
values may change as a function of temperature
variation and/or removal of 160 due to glaciation), they remain the standard palaeoceanographic refer
ence for the early Neogene. If changes in the stable isotopic ratio due to glaciation are ignored, Pacific surface waters appear
BIOGEOGRAPHY AND BIOTIC EVENTS Throughout the late Oligocene-Miocene interval the spatial organisation of planktic faunas under
to have warmed during the early middle Miocene
went a dramatic change. Oligocene and early
while retaining a differentiation between eastern
nas were well differentiated along an east-west gra
( 18-14.5
Ma
throughout the middle latitudes
Miocene tropical Pacific planktic foraminiferal fau
(cool) and western (warm ) regions. This warming
dient. During this interval western localities had a
occurs just before a cooling of deep waters in these
warm-water fauna dominated by shallow-dwelling
same ocean basins. Increasing latitudinal tempera
species ( for example,
ture gradients in surface waters are evident in the
faunas were dominated by cooler-water species ( for
Globigerinoides)
Glob01·otalia) .
while eastern
( 1 98 5 )
middle Miocene where western tropical sites show a
example,
warming and high southern latitude sites show a
that these longitudinal distinctions resulted tram
cooling that continues into the late Miocene. At
sluggish gyral circulation witl1in the Pacific Ocean
Kennett e t al.
infer
the same time , the east-west temperature differenti
basin coupled with either warmer western surface
ation l argely disappears. These patterns are thought
waters or a distinctly deeper thermocline along this
to reflect the onset of major continental accumula
basin's western margin.
tion of ice sheets, beginning in the middle Miocene (see above ).
Beginning in the middle Miocene this biogeo graphic structure began to be replaced by progres-
OLIGOCENE AND NfiOCENE PALAEOCEANOGRAPHY
sively stronger distinctions between equatorial and
mila
tl1e middle Oligocene in the south Atlantic, which
tropical planktic foraminiferal faunas. By the late
suggests that at least some localised areas of high
Miocene little trace of the previous east-west spatial
planktic productivity existed in this area. Neverthe
differentiation existed among these populations
less, the first major radiation of this group does not
while a distinct, warm-water, shallow-dwelling
take place until the early Miocene. This correlates
equatorial assemblage (for example, Globorotalia
well with initial phases of the northward expansion
menardii, G. limbata, and Globigerinoides) had
of siliceous oozes following the submergence of the
developed. These changes are thought to result
Iceland-Faeroe Ridge and increased glaciation of
from a strengthening of gyral circulation caused by
Antarctica. In this case, it would seem as though
steeper contrasts in latitudinal temperatures and
abiotic tectonic, oceanographic, and climatic events
intensification of the e quatorial countercurrent after
can be closely tied to major evolutionary diversifica
the closure of the Indonesian gateway.
tion in a convincing empirical manner.
The distribution of siliceous plankton also
Finally, it is worth noting that Miocene climatic
underwent major changes during the Miocene.
changes driven by a combination of tectonic and
During the late Oligocene siliceous oozes were only
palaeoceanographic events also had substantial
deposited close to the margins of Antarctica. But
implication for the evolutionary history of our own
throughout the Miocene these sediments came pro
lineage. The Miocene increase in latitudinal temper
gressively to dominate more northerly areas. In
ature gradients signalled the entry of the Earth's
modern oceans the zone of enhanced siliceous pro
present climatic regime. This global cooling trend
ductivity marks the approximate position of the
strongly affected the biogeography of most plant
Antarctic convergence. North of this zone carbon
groups, one of the most obvious examples of which
ate productivity predominates whereas south of this
was the replacement of the East African tropical
zone sediments are mostly siliceous in composition.
forests with mixed woodland and grassland habitats
1975 ).
This zone of siliceous productivity is closely tied to
(Andrews and Van Couvering,
the upwelling of nutrient-rich intermediate water
with this shift in the character of the East African
(
=
NADW) just south of the convergence. A' the
rate of this upwelling increases so does the level of
Correlated
plant biota is the first appearance of bipedal pri mates.
siliceous productivity. Moreover, as the convergence shifts to more northerly or southerly latitudes, the zone of siliceous sedimentation undergoes a parallel migration. This allows the approximate position of the convergence to be mapped through time. Rates of upwelling intensity
siliceous pro-
CONCLUSIONS The reconstruction and integration of Oligocene and Miocene palaeoceanographic, palaeoclimatic,
ductivity) reflect climatic cycles that alter the
and biotic events constitute one of the most thor
strength of westerly winds that move surface waters
oughly documented and compelling examples of
away from i\ntarctica. The space left by these
the power of global change research programmes to
surface waters is filled by the upwelling NADW.
gain insight into the processes that have shaped
Throughout the Miocene the rate of siliceous pro
(and will continue to shape) Earth history. �While
ductivity is also inversely correlated with the rate of
development of Oligocene-Miocene oceans, cli
equatorial planktic productivity, once again reflect
mates, and biotas could have been studied in isola
ing the intensified nature of oceanic circulation that
tion from one another, the union of these tradi
resulted from the increase in latitudinal thermal
tionally separate research programmes has provided
gradients.
a level of understanding that is demonstrably
An interesting evolutionary by-product of enhanced Miocene high-latitude productivity
greater than the sum of its parts. Early Neogene tectonic events in the northern
can be seen in the cetacean fauna of this interval.
and southern high latitudes altered the structure of
Baleen whales (suborder Mysticeti) first appear in
marine ocean basins in such a way as to force
. N. .MACLEOD the reorganisation of marine circulation patterns.
specific and predicable environmental changes that
These changes should be seen as the natural con
resulted ( at least in part) from specific contempora
sequences of tectonic trends that had been in
neous tectonic and oceanographic changes.
operation throughout the Paleogene, late Meso zoic, and beyond. Nevertheless the fact that these trends culminated in the simultaneous modification of Miocene ocean basins at either pole, coupled with the presumably fortuitous isolation of a large continental l andmass at the southern pole, pro vided sufficient S)lnergistic momentum to "flip" the
ACKNOWLEDGEMENTS This contribution is part of The Natural History Museum/University College London Global Change and the Biosphere Project.
Earth's oceanic and climate system between two mutually exclusive stable states in a remarkably short period of time. The fact that there was a
REFERENCES
long-term drift of the Earth's climate toward a
Andrews, P. J . , and Van Couvering, J.
progressively cooler state throughout the Paleo
Palaeoenvironments in the East Afiican wliocene. In
gene, including intervals when this process under
Approaches to Pt·imate Paleobiolog}'>
went substantial periods of intensification (for
S . Szalay). Karger, Basel.
pp.
H. 1975 . 62-103
( ed. F.
example, the Pal eocene/Eocene and terminal Eocene events ) , should in no way detract from the
Bre\\'Ster, N. A.
significance of the Miocene catharsis.
mentation in the Antarctic Ocean, based on two
Since the world ocean functions as the princi pal mediator of our planet's heat budget, funda mental changes in the ocean's circulation pattern inevitably precipitate correspondingly intense cli matic changes. Effects of these climatic changes in the Oligocene-Miocene were magnified via the operation of at least two major positive feedback mechanisms: upwelling and evaporation of NADW along the Antarctic coast; and isolation of the
1 980.
Cenozoic biogenic silica sedi
Deep Sea Drilling Project sites. Bulletin of the Geologi cal Society ofAmerica 9 1 : 337---47 . Hsu, K. J . , Montadere, L, Bernoulli, D., Cita, M . B.,
Erickson, A., Garrison, R. E., Kidd, R B . , Milieres,
F., Mueller, C., and Wright, R Mediterranean salinity crisis.
1 977.
History of the
Nature 267: 399---40 3.
Johnson, C . , and Kaufmann, E. G.
1996.
Maastricht
Tethys basin \Vith consequent changes in salinity
ian extinction patterns of Caribbean province rudis
values of the world ocean. Both of these operated in such a way as to intensifY further the latitudi
tids. In The Cretaceous-Tertiary Mass Extinction: Biotic and Environmental Events, pp. 2 3 1-73 (ed.
nally organised temperature gradients.
MacLeod and G. Keller). Norton, New York.
N.
Interestingly, overall biotic response to these climatic changes was not as profound as might
Keller, G.
have been expected. To some extent tllis may be
The M.iocene Ocean: Paleoceanog1·aphy and Biogeography ( ed. J. P. Kennett). Menwirs of the Geological Society of America) Washington 1 63 : 1 77-95.
linked to the fact that an extended faunal turnover had just occurred in the late Eocene-early Oligocene ( see MacLeod,
1 990
and references
1 98 5 .
Depth stratification of planktonic
foraminifers in the j\;liocene ocean. In
therein ) . Regardless, the late Oligocene-Miocene represents an interval of sustained evolutionary
Kemp, E . M .
diversification to which the concurrent expansion of
Deep Sea Drilling Project. In
1975 .
28 drill sites, Initial Reports of the 599-623 ( ed. D . E.
Palynology of
climatic heterogeneity could only have contributed .
Deep Sea Drilling Project, pp.
Indeed, several prominent evolutionary even ts-for
Hayes, L. A. Frakes,
example, diversification of filter- feeding whales and
C . Pei-Hsin, A. B. Ford, A. G. Kaneps, E. M . Kemp,
the evolution of bipedal primates-can be linked to
D. W. McCollum, D. J. W. Piper, R. E. Wall, and
L.
A. Barrem, D. A. Burns,
OLIGOCENE AND MIOCENE PAL"'-EOCEANOGRAPHY
P. N. Welds). United States Government Printing Office, Washington, D.C.
-
History: An Interdisciplinary Conference on Impacts)
Volcanism and Mass Mortality ( ed. V. L. Sharpton and
P. D. Ward) . Geological Society ofAmerica Special Kermett, J. P., Keller, G., and Srinivasan, M. S. 1 9 8 5 . Miocene planktonic foraminiferal biogeography and
Pnper no. 247: 595-606.
paleoceanographic development in the Indo-Pacific
Ryan, W. B. F. 1 973 . Geody11amic implications of the
region. In The lv'fiocme Ocean: Paleoceanography and
Messinian crisis of salinity. In Messinian Events in the
Biogeog1'aphy ( ed. J. P Kem1ett). Mm�oirs of the Geo
Mediterrnnenn) pp. 26-38 (ed. C. W. Drooger),
logical Society ofAmerica) Washington 1 63 : 1 97-236.
North Holland, Amsterdam.
La\vver, L. A, Gahagan, L. M . , and Coffin, M. F.
Ryan, W. B. F., and Cita, lv1. B. 1 978. The nature
1 992. The development of paleoseaways around
and distributions of Messinian erosional surfaces
i\ntarctica. In The Antarctic Paleoenvironment: ll Per
indicators of a several-kilometre-deep Mediterranean
spective on Global Change) part 1 , pp. 7-30 (ed. J.
in the !vliocene. Ma1·ine Geology 27: 1 9 3-230.
Kennett and D. A. Warnalce). American Geophysical Union, Washington, D.C.
Schnitker, D. 1980. Quaternary deep-sea benthic
MacLeod, N. 1990. Effects of late Eocene impacts on
Review of Earth and Planetary Scintce 8 : 343-70.
foraminifers and bottom water masses. Annttal planktic formanifera. In Global Catastrophes in Earth
INDEX
Numbers in
italics indicate
pages on which information is given only in
an
illustration or table; footnotes are
indicated by "n" following the page number.
Abu Dhabi, Emirate of: fossiliferous sites,
xxvi, 467-468 . See also Bay
nunah Formation; Shuwaihat For mation
and names of individ$tal
Agrim>emys, 1 5 3
AYn Boucherit, Algeria: chelonian, 1 44 Aln Qannas, Saudi Arabia: stone artifacts, 3 8 1 Albulidae,
places
462
Abu Dhabi Project, 6 , 1 6
AI-Busaidi, S., 4 3 0-442
Abu Dhabi-The Missing Link (fil m),
Alcelaphini , 309 a! Darmaki, His Excellency Saif Ali
19
Abudhabia, 188, 415, 4 1 6 ; A. bay nunensis, 9 , 20, 1 08 , 1 88-1 9 1 ,
Dhab'a, 1 5
mammilla; A. mammilla, 4 8 0 , A. radiata, 492
492;
Amran Formation, 454, 456
Amyda: A. cartilaginea, 1 38 , 140; "A. grega1·ia" , 1 3 8 Anancus, 224, 2 30; A . kenyensis, 2 30 , 42 3 ; A . siFalensis, 2 3 0 Anchitherium: migrations, 486, 490, 494
Alestinae, 436, 462 , 463
i\.ndrews,
1 9 5 , 196, 3 2 1 , 325, 4 1 3, 4 1 5 ,
Algae, 323, 467
A.ndrews , P. , 1 6 , 1 9 , 20, 3 38 -3 5 3
467;
Algeria: cercopithecoids, 198, 199;
Andrianos Quarry, Samos, Greece:
A. kabulense, 1 9 5 , 4 1 5 , 4 1 6 ; A. pakistanensis, 4 1 3-4 1 7 Abu Hureyra, Syria: stone artifacts, 382
chelonian, 144; hipparions, 2 36 , 2 40, 241, 242, 2 4 3 , 24 8 ; hip
Acacia, 9, 323, 467
popotamuses, 2 7 1 , 274-275;
Acanthopterygii, 465
rodents, 412, 415; suids, 255
Aceratherium, 465
AI Ghuddah, Saudi Arabia, 322;
A'n1 , 1 5 , 1 9
397
Gazella,
-fauna, 4 3 7; chelonian, 1 5 4; croco dilians, 1 64; hyracoids, 4 3 3 ; mam mals, 4 3 3 , 436
Felidae, 1 7, 1 09, 206-207, 326, 465, 467, 468
Feri1�estrix, 2 0 3 See Pisces Flosculinella, 4 8 1 ,
Fishes.
309;
Gebel Zelten.
See
Jebel Zelten
290- 3 1 6 482, 492
Flynn, L. J . , 4 1 2-419
Gentry, Anthea, 2 0
Geochelone,
1 5 5 , 1 5 6, 435, 439, 463,
309-3 1 2 , 327, 466, 467, 468; sp.
indet., 1 09 , 294-2 9 5 , 308, 3 6 3 , 366, 367, 468
Giraffoidea,
465, 466
Globigerina, 489 Globigerinoides, 504, cus, 49 1
505;
G. bisphcri
- INDEX carbonates, 359, 360, 3 6 1 ,
tampbibius, 272-273, 284,
magnetostratigraphic studies, 8 2 ;
protampbibius subsp. andmvsi,
Golunda, 4 1 3
palaeomagnetic studies, 7 6 , 79, 8 0 ,
2 7 5 ; H. sahabiensis, 1 09, 274,
Gomphotlm�ium, 465 ; G. angustidens,
82-8 5 , 84; sediments,
Globorotalia, 5 04; G. kugleri, 489; G. limbata, 5 0 5 ; G. menardii, 5 0 5
466, 490; G. cooperi,
G. stein -
beimensis, 226
stone
artifacts, 1 7, 336, 374, 376-378,
3 8 1-3 8 2 , 3 8 3-384; stratigraphic
-migrations, 486, 487, 490, 492,
chelonians, 1 3 6; fishes, 124, 1 2 8 ;
490--49 1 . See also Tethys seaway,
hipparions, 2 3 6, 2 3 8 , 249; hip
closure
popotamuses, 277, 278, 280;
Gongylopbis, 464
mustelid, 203; pecorans, 294, 296,
Goniasteridae, 435
30 1 , 303, 305, 306, 308; pro
Graeco-Iranian region, 290; pecorans,
boscidean, 2 1 3-2 1 7; suids,
290-29 1 , 294-295, 299, 304, 306, 308, 309, 3 1 2 "Grande Coupure " , 486, 488 -489,
2 5 7-2 5 8 , 2 6 1 , 266 Harmiyah, Abu Dhabi, xxl'i, 3 2 5 - -fauna, 4 67 -46 8 ; fishes, 1 2 8 ; hippopotamuses,
494 Grays, Essex, United Kingdom: cervids, 290, 307 fauna: cercopithecoids, 1 9 8 ;
pecorans, 304,
305, 308 Harrat Hadan, Saudi Arabia, 4 6 0 , 461 Hasnot, Pal-> L:ill � � � : � l ilA � � UhW I �!>lA .,..U .� ly.J I r- l:!'.¥ 1 i::Sj �� I �>'� ..:...l:i \+> .I.)"'!"',J-1 Wl.1 W.,.>L ,..;.. l._,_,.h . '.¥4 � I � � I .I,J7,J I� � .>� -..J! .1� ..:... L:. t�,., t�l �I V.. � L:. l� w;: . �}�' 1 �fiJ I � � 1� � Jl:,.:; '.¥,., ...::..:; Ls: Ukl..J I wl ..:... L,u,.,.h� � L:.i� .l.i,., '"':-',;-i.JI,., J.);.J I � �,.,.h�I ..:... L... I.>..JI �>'...).::>..I I � ��1 .>,.,1U,J �>,.,Z..:. -..J � ..:... .1i �I ..:;...\.:!.1� 1 . ��� � � �� ��I �l'_.�.rJI ..u� l r- 4-)U >.!.J.i:i,J � j'i:"•2 .�l ilA v.. li�l,., .�LJ..l �I t->-> L>fl � ....., .a;