Geomorphology and sedimentology of estuaries

DEVELOPMENTS IN SEDIMENTOLOGY 53

Geomorphology and Sedimentology of Estuaries

FURTHER TITLES IN THIS SERIES VOLUMES 1-11, 13-15, 17,21-25A, 27,28, 31,32 and 39 are out of print 12 R.G.C. BATHURST CARBONATE SEDIMENTS AND THEIR DIAGENESIS 16 H.H. RlEKE Illand G.V. CHlLlNGARlAN COMPACTION OF ARGILLACEOUS SEDIMENTS 18A G.V. CHlLlNGARlAN andK.H. WOLF, Editors COMPACTION OF COARSE-GRAINED SEDIMENTS, I 186 G.V. CHlLlNGARlAN and K.H. WOLF, Editors COMPACTION OF COARSE-GRAINED SEDIMENTS, II 19 W. SCHARZACHER SEDIMENTATION MODELS AND QUANTITATIVE STRATIGRAPHY 20 M.R. WALTER, Editor STROMATOLITES 25B G. LARSEN and G.V. CHILINGAR, Editors DIAGENESIS IN SEDIMENTS AND SEDIMENTARY ROCKS 26 T. SUDO and S. SHIMODA, Editors CLAYS AND CLAY MINERALS OF JAPAN 29 P.TURNER CONTINENTAL RED BEDS 30 J.R.L. ALLEN SEDIMENTARY STRUCTURES 33 G.N. BATURIN PHOSPHORITES ON THE SEA FLOOR 34 J.J. FRIPIAT, Editor ADVANCED TECHNIQUES FOR CLAY MINERAL ANALYSIS 35 H. VAN OLPHEN and F.VENIALE, Editors INTERNATIONAL CLAY CONFERENCE 1981 36 A. IIJIMA, J.R. HEIN and R. SIEVER, Editors SILICEOUS DEPOSITS IN THE PACIFIC REGION 37 A. SINGER and E. GALAN, Editors PALYGORSKITE-SEPIOLITE: OCCURRENCES, GENESIS AND USES 38 M.E. BROOKFIELD and T.S. AHLBRANDT, Editors EOLIAN SEDIMENTS AND PROCESSES 40 B. VELDE CLAY MINERALS-A PHYSICO-CHEMICALEXPLANATION OF THEIR OCCURENCE 41 G.V. CHILINGARIAN and K.H. WOLF, Editors DIAGENESIS, I 42 L.J. DOYLE and H.H. ROBERTS, Editors CARBONATE-CLASTICTRANSITIONS 43 G.V. CHlLlNGARlAN and K.H. WOLF, Editors DIAGENESIS, II 44 C.E. WEAVER CLAYS, MUDS, AND SHALES 45 G.S. ODIN, Editor GREEN MARINE CLAYS 46 C.H. MOORE CARBONATE DIAGENESIS AND POROSITY 47 K.H. WOLFand G.V. CHILINGARIAN. Editors DIAGENESIS, Ill 48 J. W. MORSE and F.F. MACKENZIE GEOCHEMISTRY OF SEDIMENTARY CARBONATES 49 K. BRODZIKOWSK1andA.J. VAN LOON GLACIGENIC SEDIMENTS 50 J.L. MELVIN EVAPORITES, PETROLEUM AND MINERAL RESOURCES 51 K.H. WOLF and G.V. CHILINGARIAN, Editors DIAGENESIS, IV 52 W. SCHWARZACHER CYCLOSTRATIGRAPHY AND THE MILANKOVITCH THEORY

DEVELOPMENTS IN SEDIMENTOLOGY 53

Geomorphology and Sedimentology of Estuaries Edited by G.M.E. PERILLO lnstituto Argentino de Oceanografia, 8000 Bahia Blanca, Argentina and Departamento de Geoiogia, Universidad Nacionai del Sur, 8000 Bahia Bianca, Argentina

E LSEVl E R Amsterdam - Lausanne - New York - Oxford - Shannon -Tokyo

ELSEVIER SCIENCE B.V. Sara Burgerhartstraat 25 P.O. Box 211,1000 AE Amsterdam, The Netherlands

First edition: 1995 Second edition: 1996

ISBN: 0-444-88170-0

0 1995 Elsevier Science B.V. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the publisher, Elsevier Science B.V., Copyright & Permissions Department, P.O. Box 521,1000 A M Amsterdam, The Netherlands. Special regulations for readers in the USA - This publication has been registered with the Copyright Clearance Center Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923. Information can be obtained from the CCC about conditions under which photocopies of parts of this publication may be made in the USA. All other copyright questions, including photocoping outside of the U.S.A., should be referred to the copyright owner, Elsevier Science B.V., unless otherwise specified.. No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. This book is printed on acid-free paper. Printed in The Netherlands

V

PREFACE

In the Solar System there is a strange planet that occupies the third position from the Sun. It is the only planet in the system that has liquid water covering about three-quarters of its surface. The planet is so strange that even its name is reversed, instead of being named Oceanus, it was named after a minor characteristic: Earth. Although dynamic processes over the oceans and continents of this planet are strong, there is nothing to compare with the energy of the interaction between the atmosphere, the sea and the continent at the area of contact between the latter two. The coastal zone, tremendously dynamic, is where forces are continuously changing in an abrupt fashion, depending on the local and also the distant climatic conditions. Storms that ram the deep ocean produce waves that a few days later impinge beaches located several thousand kilometres away. In general, these swells help to build up the beach by transporting sand ashore. Waves generated by storms at or near to the shore tend to be destructive to the beach, moving sand seaward. In any case, littoral transport is important in developing spits, barriers and other morphological features that tend to close embayments, modifying inlets and redistributing the sediment introduced by the rivers. Inland precipitation (either rain or snow) is the actual source of river water that follows the river valley until it finally debouches into the ocean. Normally, higher river discharges are associated with larger sediment transport (the converse is also true). This sediment is deposited at the river mouth forming a coastal plain that contains (or not) a delta, or on the adjoining continental shelf, or, in a few cases, is carried out directly to the abyssal plains. The larger the relative energy of the sea (as compared with the river sediment discharge) at the coast, the higher the chances that the sediment input by the river is redistributed along the adjacent shoreline. As the river encounters the sea, fresh and salt water mix within the lower river valley forming estuaries. Although the dilution of sea water is a distinctive characteristic of estuaries, there are many other factors equally as important. For instance, the tides are also fundamental to the development of estuaries since they provide, in general, most of the energy to establish the mixing process of both water masses, but also play a definitive role in establishing the morphology of the environment and the distribution of sediments. Other factors, such as waves or wind, render major parts in microtidal estuaries or at particular places within other estuaries. Practically all processes that take place in an estuary are related, at least to a minor extent, to the general or particular shape and sediment input, output, distribution and transport. For instance, tidal wave propagation is strongly dependent on the variations in depth and the ratio convergence/friction offered by the channel. Biota-

vi

PREFACE

sediment interactions are always present as environmental conditions and geomorphology are changed. Modifications in channel depth by dredging induce variations in salt intrusion and the resulting thermohaline circulation, often defining sectors of turbidity maxima, and sediment deposition may increase several times resulting in the need for more dredging. Furthermore, many biological and chemical pollutants are commonly associated with fine sediment particles transported in suspension. Particular geomorphologic settings may establish the hydrodynamic conditions to force deposition of the contaminated particles, thus affecting the benthic fauna of the area. The few and brief examples outlined in the previous paragraphs have been taken from real cases occurring in different estuaries throughout the world. They are not isolated cases, but facts that are commonly reported in the estuarine literature. All of them are actually dependent on the geomorphology and sedimentology of the environment. Nowadays there is a large number of books on the market dealing with different aspects of the biology, chemistry and physical characteristics of estuaries and processes occurring in estuaries. There also is an increasing amount of literature describing the general processes and modelling of sediment transport. However, to my knowledge, there is no book that specifically covers the basic geomorphology and sedimentology of these coastal water bodies. In textbooks and other books resulting from scientific meetings which deal with estuarine problems, the way geomorphology affects all other processes is discussed summarily and, on many occasions, is disregarded as a minor part. However, it is my view that the particular shape of the environment and the constitution of its boundaries actually play a decisive role in the outcome of any process occurring there. Commonly, this situation arises because all processes are quite complex and their interactio’ns with the boundaries are strongly nonlinear, becoming still more difficult to model. Therefore, the aim of the book is to provide a detailed view of the geomorphology and sedimentology of estuaries. The matter will be presented in such a way that it can be utilized not only by specialists of the subject, but also by other researchers requiring the background to put their own work into an adequate perspective. The new generation of researchers, now graduate students, will benefit from this book. It will help them to understand that an estuary is a complex entity that cannot be analyzed only at the level of a single science. Multi- and interdisciplinary approaches are a must. Furthermore, an adequate knowledge of the geomorphology of estuaries is also required for a relatively new and most needed science: coastal management. The book is based on a new definition and morphogenetic classification. The new definition of estuaries covers, for the first time, the basic characteristics required for all disciplines dealing with these coastal environments. Moreover, the morphogenetic classification actually resumes the most modern approaches provided by renown specialists in geomorphology (e.g., Rhodes Fairbridge), plus it also introduces a criterion that relates the degree of modifications produced by the sea. The balance between the terrestrial and marine forces are a definitive conditioning of the resulting morphology. Leading experts have provided in-depth descriptions of the geomorphology, sedimentology and interactive processes associated with each category in individual

PREFACE

vii

chapters. Their exposition is directed to present the state-of-the-art in a format adequate for the researcher, but also of use as a textbook for graduate students. It is also worthwhile mentioning the quality of the specialists that have accepted to write the different chapters. This international ensemble has, in conjoint, an expertise only paralleled a few times in other books of similar scope. Each author is active both in research and teaching (most of them are senior researchers and/or full professors at their respective institutions). I tried to be very careful in their selection to cover both research and teaching aspects assuring a didactic rather than purely scientific form of presenting the facts and examples. The first two chapters give an introduction to the study of the geomorphology and sedimentology of estuaries and present a review of the most common definitions and geomorphologicclassifications. Specificallyin Chapter 2, a new definition of estuaries is introduced with an open criterion. I see this definition as a step further to finding out a still more comprehensive definition that will arrive after we have obtained a thorough knowledge of estuaries. Chapters 3 to 9 are devoted to the description of the geomorphologic and sedimentologiccharacteristics of the elements that form the classification on which this book is based. Chapters 10 to 13 cover major features that are normally present in estuaries, although they are also common in open coasts. Finally, Chapter 14 provides a review of the most common sediment transport processes that occur in estuaries. From the moment I first had the idea about this book until the writing of these notes, several years have passed and many colleagues have encouraged me to continue, alongside, in particular, my wife Cintia and my children, Mauricio and Vanesa, who put up with the long hours of work necessary for the book. My special thanks go to the authors of each chapter who believed in the project and made special efforts to meet the deadlines. I would also like to express my sincere gratitude to the reviewers of the individual chapters, listed here in alphabetic order: Henry Bokuniewicz, Diana G. Cuadrado, James M. Coleman, Clifford Embleton, G. Evans, Rhodes Fairbridge, Eduardo A. Gbmez, s. Susana Ginsberg, John McManus, M. Cintia Piccolo, H. Postma, Donald J.P. Swift, J.J.H. Terwindt, Federico Was, Eric Wolanski and another five reviewers who wished to remain anonymous. All of them contributed profoundly, providing new insights and criteria that increased the value of each contribution. I would also like to thank Elsevier Science, especially Drs. Martin Tanke who accepted the idea right from the beginning and encouraged me all the time he was in charge of the production. Mr. Dominic Vaughan received the ‘hot potato’ halfway and handled it most efficiently. Mrs. Maria Ofelia Cirone was very efficient in editing the original manuscripts and arranging them in a unique editorial format. Gerard0 M.E. Perillo Bahia Blanca, September 1994

A tidal creek in the reclaimed salt marshes of the Petitcodiac River, Bay of Fundy. The dykes were originally constructed by French Acadians during the 17th century. Much of the original dykes have been eroded by relative sealevel rise and by tidal channel migration (foreground). Turn the photo upside-down for a view of a mud esker. (Photograph taken by R. Belanger, Bedford Institute of Oceanography.)

8

ix

LIST OF CONTRIBUTORS

CARL L. AMOS, Geological Survey of Canada, Atlantic Geoscience Centre, Bedford Institute of Oceanography,Dartmouth, Nova Scotia, B2Y 4A2 Canada ROWLAND J. ATKINS, Hay and Co. Consultants Inc., 1 W 7th Ave., Vancouver, British Columbia, V5Y 1L5 Canada PIETER G.E.F. AUGUSTINUS, Netherlands Centre of Coastal Research (NCK), Institute for Marine and Atmospheric Research Utrecht, Utrecht University, PO. Box 80 115,3508 TC Utrecht, The Netherlands HENRY BOKUNIEWICZ, Marine Sciences Research Center, State University of New York, Stony Brook, New York 11794-5000,USA PATRICE CASTAING, Departement de GCologie et OcCanographielURA 197, UniversitC de Bordeaux I, Avenue des FacultCs, 33405 Talence, Cedex-France ROBERT W. DALRYMPLE, Department of Geological Sciences, Queen’s University, Kingston, Ontario, K7L 3N6 Canada KEITH R. DYER, Institute of Marine Studies, University of Plymouth, Plymouth, Devon PLA 8AA, UK JONATHAN W. GIBSON, Department of Geography, Simon Fraser University, Burnaby, British Columbia, VSA 156 Canada ANDRE GUILCHER, DCpartement de GCographie, Universite de Bretagne Occidentale, B.P. 814,29285 Brest, France BRUCE S. HART, Department of Geosciences, Pennsylvania State University, University Park, Pennsylvania 16801, USA FEDERICO I. ISLA, CONICET-UNMDP, Centro de Geologia de Costas y del Cuaternario, C.C.722, 7600 Mar del Plata, Argentina. JOHN L. LUTERNAUER, Geological Survey of Canada, 100 W Pender St., Vancouver, British Columbia, V6B 1R8 Canada ANNE I. MOODY, AIM Ecological Consultants Ltd., 100 Mile House, British Columbia, VOK 2E0 Canada

X

LIST OF CONTRIBUTORS

GERARD0 M.E. PERILLO, Instituto Argentino de Oceanografia, Av. Alem 53, 8000 Bahia Blanca, Argentina, and Departamento de Geologia, Universidad Nacional del Sur, San Juan 670,8000 Bahia Blanca, Argentina MARIO PIN0 QUIVIRA, Instituto de Geociencias, Universidid Austral de Chile, Casilla 567, Valdivia, Chile ROBERT N. RHODES, COA Coastal Ocean Associates, Inc., 7 Coral Street, Dartmouth, Nova Scotia, B2Y 2W1 Canada JOHN SHAW, Geological Survey of Canada, Bedford Institute of Oceanography, Dartmouth, Nova Scotia, B2Y 4A2 Canada JAMES P.M. SWITSKI, Geological Survey of Canada Bedford Institute of Oceanography Dartmouth, Nova Scotia, B2Y 4A2 Canada JOHN T WELLS, Institute of Marine Sciences, University of North CarolinaChapel Hill, Morehead City, North Carolina 28557, USA HARRY EL. WILLIAMS, Department of Geography, University of North Texas, Box 5277, Denton, Texas 76203-0277, USA

xi

CONTENTS

Preface .......................................................................................... List of Contributors.. ............................................................................

Chapter 1.

v ix

GEOMORPHOLOGY AND SEDIMENTOLOGY OF ESTUARIES: AN INTRODUCTION

.............

.............

1

Evolution of estuaries in the geological time scale ......... Factors influencing the geomorphology and sediment distribution. ................................ 13 Summary.. ....................................................................................... 14 References. ..... ........ ........... .... 15 Chapter 2.

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONS OF ESTUARIES G.M.E. Perillo., ..................................................................

............. Introduction ............................................................... Previous definitions .............................................................................. A proposed new definition of estuaries ....... .. Previous geomorphological classifications of estuaries. ............................................ Physiographic classification ............................................. Classification by tidal range.. ........................................... Evolutionary classification. ....... Morphological classification.. .... ...................... A proposed new morphogenetic class ............. ................................................................................

17 17 18 26 27

36 37 40

s of estuaries in dictionaries and encyclopedias References .......................................................................................

46

SEDIMENTARY SYSTEMS OF COASTAL-PLAIN ESTUARIES .... ...................... H. Bokuniewicz ......

49

Chapter 3.

Introduction ..................................................................................... Coastal plains.. ... ..... ......................................... es ..................................... Sedimentological classification of estuaries. ....................................................... Summary ... .... ...................... References ..................................................................

49 50

58 64

xii

CONTENTS

.

Chapter 4

GEOMORPHOLOGY AND SEDIMENTOLOGY OF R I M P.Castaing and A.Guilcher .......................................................

Definition and included areas..................................................................... Regional description ............................................................................. Northwestern and northern coasts of the Iberian Peninsula (Spain) ............................ Brittany (France) ............................................................................. Provence (France) ............................................................................ Southwest England and possible other areas in the British Isles................................ Korea ........................................................................................ Southeast China and Shandong ............................................................... Argentina..................................................................................... Red Sea shanns and their worldwide extension ................................................ Messinian rias in the Mediterranean sea....................................................... General considerations........................................................................... Ria evolution ................................................................................. Sedimentary processes ........................................................................ Pluri-annual sedimentary budget .............................................................. References ....................................................................................... Chapter 5.

SEDIMENTOLOGY AND GEOMORPHOLOGY O F FJORDS J.P.M. Syvitski and J . Shaw ........................................................

Introduction ..................................................................................... Character ........................................................................................ Oceanographic characteristics .................................................................... World distribution................................................................................ Short-term depositional processes ................................................................ Ice-influenced fjords. ............................................................................. Ice-front melt ................................................................................. Glacifluvial processes ......................................................................... Iceberg calving and rafting .................................................................... Ice-front movement........................................................................... Land-based fjord valley deposition ............................................................ Sea-ice influence.............................................................................. River-influenced fjords ........................................................................... Fjord river discharge.......................................................................... Sediment transport ........................................................................... Fjord deltas................................................................................... Fjord river plumes ............................................................................ Hemipelagic sedimentation ................................................................... nrbidity currents ............................................................................. Wave- and tide-influenced fjords.................................................................. Tidal processes ............................................................................... Wave processes ............................................................................... Fjords dominated by slope failure ................................................................ Release mechanisms .......................................................................... Mass transport processes...................................................................... Deep-water renewals and anoxic fjords ........................................................... Deep-water renewal .......................................................................... Renewal and sedimentation ................................................................... Anoxia ....................................................................................... Long-term depositional trends.................................................................... Stages of fjord infilling........................................................................ Relative sea-level fluctuations.................................................................

69 69 70 70 75 82 83 85

89 89 90 92 94 95 98 101 107

113 113 113 118 119 122 124 124 124 127 128 130 130 131 131 132 133 136 137 140 143 144 146 147 148 150 152 152 154 154 155 155 156

...

CONTENTS

xlll

Climate and sedimentation .................................................................... Numerical models ............................................................................ Progress .......................................................................................... Summary......................................................................................... References .......................................................................................

162 163 164 167 168

Chapter 6.

TIDE-DOMINATED ESTUARIES AND TIDAL RIVERS J.T. Wells .........................................................................

179

Introduction ..................................................................................... T h e classification problem .................................................................... Physical processes in tide-dominated estuaries .................................................... Effects of tide on sediment dynamics .......................................................... Formation of a tidal turbidity maximum ....................................................... Morphologic and sedimentologic character ....................................................... Estuarine morphology ........................................................................ Estuarine sedimentology ...................................................................... Fluid-mud deposits ........................................................................... Estuarine infilling............................................................................. Holocene examples: tide-dominated estuaries ..................................................... Gironde Estuary .............................................................................. Severn Estuary ................................................................................ Ord Estuary .................................................................................. Cobequid BaySalmon River Estuary ......................................................... Holocene examples: tidal rivers ................................................................... Rio de la Plata ................................................................................ Amazon River ................................................................................ Summary......................................................................................... References .......................................................................................

179 179 181 183 184 185 185 188 189 191 192 192 193 194 195 197 197 198 200 202

DELTA FRONT ESTUARIES B.S. Hart .........................................................................

207

Chapter 7.

Introduction ..................................................................................... Delta morphology and growth .................................................................... Alluvial feeder systems........................................................................ Receiving basin characteristics ................................................................ Deltaic environments ............................................................................. ................................................ Channels ................................. River mouths ................................................................................. Interchannel areas ............................................................................ Summary ......................................................................................... References ....................................................................................... Chapter 8.

207 207 209 210 211 212 217 221 223 224

STRUCTURAL ESTUARIES M . Pino Quivira ...................................................................

227

.......................... Introduction ....................................................... General classifications of structural estuaries ..................................................... Morpho-tectonic classification................................................................. Neotectonic influence on the formation of estuaries............................................... Summary......................................................................................... References .......................................................................................

227 228 230 232 237 237

CONTENTS

XiV

.

Chapter 9

COASTAL LAGOONS El . lsla ...........................................................................

Introduction ..................................................................................... Origin of coastal lagoons ......................................................................... Geomorphology.................................................................................. Sedimentology ................................................................................... Conditioning factors for the development of coastal lagoons ...................................... Climate effects................................................................................ Tectonic effects ............................................................................... Biogenic effects............................................................................... Wind-wave effects............................................................................. Tidal and wave effects ........................................................................ Longshore-drifteffects........................................................................ Related environments............................................................................ Tidal inlets ................................................................................... Tidal deltas ................................................................................... Barriers....................................................................................... Tidal flats..................................................................................... Marshes ...................................................................................... Mangroves.................................................................................... Coastal lagoon evolution ......................................................................... Summary......................................................................................... References....................................................................................... Chapter I0.

SILICICLASTICTIDAL FLATS C.L. Amos ........................................................................

241 241 242 243 244 245 246 249 250 250 252 253 254 254 255 261 263 263 264 265 266 267

273

The classification of tidal flats .................................................................... 273 Siliciclastic tidal flat research ..................................................................... 275 The zonation of tidal flats and relative elevation .................................................. 279 Tidal flat sedimentation a comparison between the Wash and the Bay of Fundy ................ 282 Mud flat deposition and sediment supply ...................................................... 282 Mud flat erosion .............................................................................. 288 Sand flat stability and the transport of non-cohesive sediment ................................. 290 A model for sediment accretion/erosion on the tidal flats of the Wash and Minas basin ........ 293 The influences of waves on tidal flats ............................................................. 298 References....................................................................................... 301

-

Chapter I I .

SALT MARSHES J.L. Luternauer. R.J. Atkins. A.1. Moody. H.EL. Williams and J.W. Gibson ........ 307

Introduction ..................................................................................... Overview of coastal marsh morpho-sedimentology................................................ Estuarine marsh dynamics ........................................................................ Modelling estuarine marshes ..................................................................... Summary......................................................................................... References.......................................................................................

.

Chapter I2

GEOMORPHOLOGY AND SEDIMENTOLOGYOF MANGROVES F!G.E.E Augustinus ..............................................................

Introduction ..................................................................................... Global distribution of mangrove species .......................................................... Composition and zonation of mangroves..........................................................

307 309 318 326 328 329

333 333 333 336

xv

CONTENTS Mangrove species and their environmental constraints ............................................ The influence of mangroves on hydrodynamics.................................................... Sedimentation and sediment in estuarine mangrove forests ....................................... The influence of mangroves on soil stability....................................................... Mangroves and geomorphology ................................................................... Conclusion ....................................................................................... References ....................................................................................... Chprer 13.

ESTUARINE DUNES AND BARS R.W. Dalrymple and R.N. Rhodes .................................................

Introduction ..................................................................................... Dune classitication ............................................................................... Distribution of dunes............................................................................. Controlling variables .......................................................................... Distribution within estuaries .................................................................. Dune size ........................................................................................ Water depth/boundary-layer thickness ......................................................... Current speed and grain size .................................................................. Water temperature and sediment availability .................................................. Unsteady flow ................................................................................ Summary ..................................................................................... Dune shape ...................................................................................... Profile shape.................................................................................. Plan shape .................................................................................... Dune orientation ................................................................................. Variability of current direction ................................................................ Non-uniform migration ....................................................................... Discussion .................................................................................... Superimposed dunes ............................................................................. Morphological response to unsteady flow ......................................................... Dune migration rates............................................................................. Internal structure of dunes ....................................................................... Simple dunes ................................................................................. Compound dunes ............................................................................. Estuarine barforms ............................................................................... General characteristics and classification ...................................................... Repetitive barforms ........................................................................... Elongate tidal bars ............................................................................ Delta-like bodies.............................................................................. Internal structures ............................................................................ Summary and research needs ..................................................................... References .......................................................................................

Chuprer 14.

339 341 346 349 349 352 353

359 359 359 363 363 365 366 367 371 372 373 374 374 375 382 386 386 388 389 391 392 399 401 402

404 406 406 407 410 413 413 416 417

SEDIMENT TRANSPORT PROCESSES IN ESTUARIES K.R. Dyer.........................................................................

423

Introduction ..................................................................................... Tidal effects ...................................................................................... Qpes of estuary .................................................................................. Highly stratified estuary ....................................................................... Partially mixed estuaries ...................................................................... Well mixed estuaries .......................................................................... Modes of sediment transport .....................................................................

423 424 427 428 428 428 429

xvi

CONTENTS

Mud properties.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430 Flocculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430 Settling velocity.. . .. . . .. .. Erosion. . . . . . . . . . ........ ... ... ... .. ... ... .. .. .. .. .... ..... .. . . . .. .. .. . .. . . .. . .. . Transport of mud in t Turbidity maximum.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Processes forming the turbidity maximum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .. .. . . . . Residual circulation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lag effects.. ... Horizontal fluxes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Estuarine trapping . . . . . . . . . .. .. .. ... ... . . ... ... . .. .. .. . . .. .. ... .. . . .. .. ... . .. , .. .. . . . Summary. . . . . . . . . . . . . . . . . . . . .. .. .. . .. . . ... ... ..... . .. ... ... ... .. . .. . .. .. ... ... ... . .. . References.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . .

433 435 438 438 439 442 443 446 447

Geographic Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451 Subject Index .. . . . . . .. . ... . . . . . .. . . . . . .. . . . . .. . .. . . . .. . . . . . ... . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . 459

Geomorphology and Sedimentology of Estuaries. Developments in Sedimentology 53 edited by G.M.E. Perillo 0 1995 Elsevier Science B.V. All rights reserved.

1

Chapter 1

GEOMORPHOLOGY AND SEDIMENTOLOGY OF ESTUARIES: AN INTRODUCTION GERARD0 M.E. PERILLO

INTRODUCTION

Geomorphology is concerned with the study of earth-surface forms and with their evolution in time and space due to the physicochemical and biological factors acting on them. Most of the evolution is the product of a cyclic process based on erosiontransport-deposition of sediment particles. Added to this are the combinations that may occur from the meteorization of a hard rock until the particle is permanently buried and becomes part of a new sedimentary rock. In particular, the coastal environments are subjected to the most energetic conditions on the earth surface. Modifications of geoforms and the characteristics of sediment distribution may occur in very short time periods. Nevertheless spatial and time scales may range from few seconds and centimeters to centuries and thousands of kilometers (Table 1-1). Estuaries are one of the most important coastal features subject to strong processes that fully cover the space-temporal scale. Geomorphologic and sedimentologic changes are continuously occurring within and around estuaries that effect their specific characteristics. Normally estuaries occupy the areas of the coast least exposed to the marine action. In this way, wave activity is generally quite reduced, allowing the development of harbors, recreational facilities, or appropriate aquaculture initiatives. Nevertheless, within the estuaries the dynamical processes are rather strong and impose a remarkable stress over the biota, either permanent or temporary, the morphology and the civil works. Some authors have indicated that “estuaries have been uncommon features during most of earth’s history...” (Russell, 1967), simply because “estuarine deposits rarely can now be delimited unequivocally from other shallow water marine deposits in the geological record because of their limited areal extent, their ephemeral character and their lack of distinctive features” (Schubel and Hirschberg, 1978). Nevertheless, as Table 1-1 Measurement units on the space-temporal scale (after Perillo and Codignotto, 1989)

Space Time

Megascale

Macroscale

Mesoscale

Microscale

km century

km yearhonth

m days/h

cm min/s

2

G.M.E. PERILLO

Table 1-2 Schematic sequences of sedimentary lithofacies in a transgressive estuarine environment for (a) axial and vertical trend, and (b) lateral and vertical trend (After Nichols and Biggs, 1985). (a) Axial and vertical sequence in the estuarine environment River

Seaward

Sea

ESTUARINE FLUVIAL

ESTUARINE

ESTUARINE MARINE Coarse marine sands massive with abundant cross-bedding, tidal current ridges with low angle cross-bedding in fine sands with silt laminae

Silt and clay with sandy lenses and laminae, massive silt and clay deposits Massive silt and clay with abundant plant and roots, sandy lenses, and laminations, grading downward into sand, gravel and cobble (b) Lateral and vertical sequence in lower estuary Shore SHORELINE DEPOSITS

Mid-channel SUBTIDAL FLATS

ESTUARINE MARINE Coarse marine sands massive or with abundant cross-bedding (as above)

Laminated and massive muddy sands and sandy muds Sand, gravel, and shell with or without washover complex and muds with plant frangments and basal peat

long as a river was present in any paleocoast being affected by tidal action inducing changes in salinity distribution within its valley, an estuary existed. By the time Russell (1967) proposed his opinion, there were few unifying models of estuarine deposition and geologist had difficulties to identify them from other shallow marine environments. However, Nichols and Biggs (1985) have provided axial and lateral sequences of estuarine lithofacies in transgressive conditions (Table 1-2). Figure 1-1 is a schematic representation of the evolution process due to high river-load discharge. In the present time, estuaries are very common features in most world coasts. For instance, Emery (1967) estimated that 80-90% of the Atlantic and Gulf coasts and 10-20% of the Pacific coasts of United States are occupied by estuaries in the broad sense. The large variety of estuaries that exist depends on the local climatological, geographic, geological and hydrological characteristics. But also their

GEOMORPHOLOGY AND SEDIMENTOLOGYO F ESTUARIES: AN INTRODUCTION

3

A

C

Fig. 1-1.Schematic evolutionary sequence of an estuary associated with a large ratio of river-load input to sea-level rise. A) Flooding by the sea of the fluvial valley; B) progradation of the coastal plain; C) developing of barriers by littoral transport, and D) developing of a river delta.

present position and future evolution largely relies on the variations in sea level, sediment supply and structural activity. Therefore, the aim of the present chapter is to consider the basic geomorphologic and sedimentologic characteristics of estuaries in relation with its global distribution, factors that influence them and to provide some clues to identify estuaries in the geological record.

HISTORICAL BACKGROUND

Since river mouths have served as natural harbors from the beginning of civilizations, knowledge of the shallows and channels, tides and currents, and the extent of salt water penetration has been empirical for the first navigators, city founders and engineers. Nevertheless, the first morphological charts were introduced by W. Bourne

4

G.M.E. PERILLO

in 1578. He described the genesis and geomorphology of coasts, including the first indication of the presence of shoals at river and estuarine mouths. As geomorphology was initiating in the last decades of the 19th century, much work was done in coastal environments and, specially, in rivers. They were made following the Davisian model associated to time evolution stages (youthful-matureold) of landscape. However, estuaries were not regarded as a particular separated entity from the river. Actual interest in estuaries started at the beginning of the ~ O ’ S after , a series of papers by Pritchard (1952), Stommel (1953) and Stommel and Farmer (1953) that followed the basic paper by Kuelegan (1949). However, most of these papers only considered the geomorphology of the estuaries in analyzing the constrains that the borders introduce in their circulation. Pritchard (1952) introduced the first physiographic classification, modified by the same author in 1960 (see discussion by Perillo, this volume). His classification is still being considered as a good preliminary approach to the understanding of the general structure of these coastal bodies. Interest in the geomorphology, sedimentology, and sediment transport of estuaries has increased steadily since them. Classical papers like those produced by Postma (1961, 1967), Allen et al. (1980) and more recently Nichols and Biggs (1985) or books by Davis (1985) and Dyer (1986) stand out from a remarkable list. Even though the extensive literature and the numerous experiments carried out in many estuaries in the world, precise knowledge of the actual processes that shape estuaries, distribute its sediments and control the fate of pollutants and biological species is still elusive. Integrated approaches has to be devised to understand individual estuaries or even some particular feature within an estuary.

OCCURRENCE AND DISTRIBUTION OF ESTUARIES

As long as freshwater is discharged into the sea in a channeled form, there is potential for the development of an estuarine environment. Figure 1-2 shows the distribution of the most important estuaries in the world associated to the tidal range and climatic zones (many of the estuaries mentioned in the following chapters have been included in the map). Most estuaries developed in former river valleys are located on subtropical and temperate regions and associated with mesotidal conditions. Those related to previous glacial valleys have formed in polar and subpolar climates. Pure coastal plain estuaries appear in areas where sediment load provided by the rivers are relatively small when compared with the dynamic forces that redistribute the material. Deltas, on the contrary, are found in places where these conditions are reversed. Although delta tributaries may behave as estuaries themselves. On the other hand, fjords are concentrated in high latitudes and mostly on rocky shores, meanwhile the few existing fjards are observed on low-lying coasts of northern Sweden. Rias are detected in rocky or cliRy shores where alpine glaciation did not reach into the inundated valley or its modifications cannot be revealed from the river influence. Structural estuaries cannot be related to any climatic or tidal range criteria, but to areas presently active like the western boundary of the American continent.

GEOMORPHOLOGY AND SEDIMENTOLOGY OF ESTUARIES: AN INTRODUCTION

5

6

G.M.E. PERILLO

Table 1-3 Factors controlling the formation of estuaries. Climate

Polar and subpolar Temperate Tropical and subtropical

Type of coasts

Trailing edge Collision Marginal sea

Coastal lithology

Hard-rock Soft-rock (sedimentites)

Tidal range

Macrotidal Mesotidal Microtidal

Coastal stability

Submerging Emerging Stable

Neotectonism

Present Absent

River discharge and sediment load

High Low

Marine diffusive forces (waves, littoral currents, tidal currents, etc.)

High Low

Atmospheric influence (winds, temperature, humidity, etc.)

High Low

Finally, coastal lagoons present a complete different criteria. They are the product of marine action that totally cleared the original valley by providing its particular morphology. In general, coastal lagoons are associated with micro and mesotidal coasts where littoral processes are presently, and/or in the near past, dominant. According to Emery (1967) these features are characteristic of coastal plains where minor sea level increases may inundate large surfaces. In summary, there are several criteria that control the presence or absence of estuaries and, in the former case, their type. Some of the most important are presented in Table 1-3. The listing is not complete and it has not been ordered in any specific manner. Evidently adequate combinations of these factors will produce characteristic types of estuaries which in themselves have particular circulation patterns. Although most factors have been quantified, there is still no clear correlation between any combination of these parameters and the resulting estuary.

EVOLUTION OF ESTUARIES IN THE GEOLOGICAL TIME SCALE

Being coastal features, the position of estuaries depends on the location of the shoreline, which itself is conditioned by sea level oscillations, tectonism, isostasy, etc.

GEOMORPHOLOGY AND SEDIMENTOLOGY O F ESTUARIES: AN INTRODUCTION

7

A stable coast is the product of the balance between forces that tend to move it either landward or seaward. If the delicate balance becomes modified, the result is a transgression or a regression of the sea. Bowen (1978) suggests that sea level may change due to one or several of the following processes: long-term tectonism, glacial isostasy, hydro-isostasy, geoidal modifications and glaciation. General falling of the sea level during the Tertiary period can be related to worldwide tectonism and orogeny. Uplift implied deepening of ocean basins since ocean floor material must have been used to fill up the elevations. Although the tectonic effect on sea level is important in itself, a consequence of the formation of high mountain ranges is the major changes that occurred on the climatic pattern of the Earth. Notably is the formation of the Antarctic ice cap 5 Myr ago. As suggested by Tanner (1968), the mid-Cretaceous sea level was some 130 m higher than at present. The sea level reduction occurred in two steps: about 50 m were reduced in 70 Myr due to the tectonism during the late-Cretaceous-earlier Tertiary. The second step spreaded for another 25 Myr with a 75 m sea level drop that may have been produced also in another two processes. These were, first an isostatic rebound due to erosion of the mountain ranges, and second, and more important for our purposes, was the growth of the Antarctic and Greenland ice sheets. If the latter process did not occur, sea level should be about 68 m higher than it is now. This is coincident with Russell (1964) observation that melting of the Antarctic and Greenland ice caps would produce a rise of sea level between 60 and 75 m. There is general agreement that four major glaciation periods occurred during the Pleistocene (since 2.8 Myr BP). Fairbridge (1961) scheme (Fig. 1-3) considers that sea level was reduced from a maximum of about +80 m during the Aftoninan interglacial to -100 m (Kraft and Chrzastowski, 1985) during the Wisconsin, some 15-18,000 yr ago. Although some authors (i.e., Emery, 1967) place the lowermost sea level stand at -130 m. The passage from glacial to interglacial periods and back was marked by numerous oscillations. Employing oxygen isotopes analysis, Shackleton and Opdyke (1973) found out nine glacial and ten interglacial events within the last 700,000 yr, while Beard et al. (1982) proved the occurrence of eight interglacial and the same

L

mow0

I

2oooO0 Yr

1ooOOo

I

0

DP

Fig. 1-3. Mean sea level variations within the Pleistocene due to the different glacial and interglacial periods. Note the general sea level trend that clearly shows a marked long-term reduction. (Modified from Fairbridge, 1961.)

8

G.M.E. PERILLO

number of glacial events for the whole Pleistocene. Anyway, the largest glaciation and the one that concerns us the most is the previously mentioned Wisconsin (Wurm, as it is named in Europe). Glaciations occur when the water that normally flows to the sea is retained on the continent as ice. The lack of runoff and the associated strong evaporation on the sea, product of dry atmospheric conditions that tend to accompany glaciations, lower the sea level. Although ice sheets developed around the poles, this simple process affected the world ocean on each glacial period. This is specially true during the Wisconsin which apparently covered the largest surface than any previous glaciation. The increment in the atmospheric temperature produced the melting of the ice, originating thus a rise in sea level. Most authors agree that sea level raise was very rapid during the first 12-15,000 yr until about 3,000 yr BP (Fig. 1-4). Since then, the rate of change of sea level has diminished significantly reaching in the present rates on the order of, for instance, 2 mm/yr in the eastern coast of US (Hicks, 1980) and 1.6 mm/yr in the Argentine coast (Lanfredi et al., 1988). Further evidence presented by Fairbridge (1961) suggested that the rising process was also marked by strong oscillations. Some of them that occurred within the last 7,000 yr moved the sea level above the present stage. As an example, Gonzalez (1989) has displayed a series of four transgressive episodes that occurred between Y E A R S BEFORE

I$ 12 I

10 -

-

8 -

-

-

6 -

4

PRESENT

2

0

Fig. 1-4. Mean sea level curves from various authors for the last 12,000 year. Fairbridge (1961) curve shows several fluctuations above the present mean sea level which later was confirmed for the Southern Hemisphere (see Figs. 1-5 and 1-6).

GEOMORPHOLOGY AND SEDIMENTOLOGY OF ESTUARIES: AN INTRODUCTION

9

1

-1.4 mm/yr

9.4 mm/yr

1.6 mm/yr -5

-10

-15' 0

1

2

3

4

5

6

7

I

6

years (Thousands)

Fig. 1-5. Estimated mean sea level curve for the Bahia Blanca Estuary (Argentina) showing a very high (up to 7 m) sea level stand above the present condition (modified from Gbmez and Perillo, 1994).

5,990 and 3,560 yr BP in the Bahia Blanca estuary (Argentina). The maximum and oldest transgression left beach and tidal flats deposits at about 7 m above the present sea level. Aliotta and Perillo (1985, 1990) have described a series of wave-cut terraces between 13 and 16 m below datum level near the mouth of the same estuary which were formed during a lower still stand 8,000 yr BF! G6mez and Perillo (1992, 1994) have described similar terraces at depths of 15 m outcropping from beneath shoreface-connected linear shoals. Based on the information provided by Aliotta and Perillo (1985,1990), G6mez and Perillo (1992) and Gonzalez (1989), G6mez and Perillo (1994) developed a minimum sea level variation curve. The curve shows the different rates of sea level evolution during the last 8,000 yr for the Bahia Blanca Estuary (Fig. 1-5). It was made by using the minimum depth at which the macroterraces were found and assigned them an age of 8,000 yr, and the lowest level of occurrence of each transgressive stage mentioned by Gonzalez (1989) giving to each of them their probable geological age. The resulting composite curve shows a sharp increase, roughly 1 cm/yr in the first 2,000 yr; having about the same rate assumed by most authors for the period 15,000 to 6,000 yr BP (Schubel and Hirschberg, 1978). The Late Pleistocene-Early Holocene delta complex of the Desguadero-Colorado rivers (Perillo, 1989) was rapidly covered by the sea; becoming for over at least 4,000 yr a shallow inner shelf zone. The calculated rate of 1.4 mm/yr considers as if the sea level dropped continuously until 90 yr ago, giving a minimum rate, from which we used Lanfredi et al. (1988) estimate. Obviously this rate may be much larger if we consider that upward movement of the sea level must be occurring for at least 400-500 yr as has been recently proposed by Gonzalez and Weiler (1994), but there is no enough evidence to support this. The curve given here compares quite well with the general structure of the curves given by Isla (1989) (Fig. 1-6) for different sites on the Southern Hemisphere and specially along the Argentina coast where sea level above the present has been repeatedly recorded.

10

G.M.E. PERILLO

GEOMORPHOLOGY AND SEDIMENTOLOGY OF ESTUARIES: AN INTRODUCTION

11

As the sea level stood at its minimum position, most of the world continental shelves were converted in extensive plains. As estimated by Emery (1967), the average shelf break (-130 m) is at or near the predicted values for the lowest sea level. During almost all the Tertiary and Pleistocene, rivers were restricted to the present day hinterland. Then during the glaciation period, they found their way through the continental shelves driven by a lower base level. Both, the rivers and glaciers that occupied previous river valleys in high latitudes, cut down more definite and deeper valleys on the continental shelves. In many cases, they reached the shelf break where they originated submarine canyons (i.e., Hudson, Baltimore). To the present, there has not been found any evidence of a connection between the very few submarine canyons existing on the continental slope of the Argentine shelf and present day rivers. It is considered that during the last glaciation and even up today, the Patagonian climate was dry. Therefore, rivers had relatively low discharges that prevented them from reaching the shelf break that is over 200 km up to 850 km away. During the lowest sea level, estuaries occupied the border of the continental shelves. They were, in general, scarce and limited in their areal distribution. In effect, estuaries were mostly restricted to valleys bordered by abrupt walls. The most immediate effect of the thawing of continental ice was felt by river discharges which also raised substantially the sediment load input to the sea. Due to the high gradient valleys in the canyons, sediment was not deposited in them. Bypassed sediments formed abyssal cones and partially contributed to the building of the continental rise. Similar situations are observed today with the abyssal cones formed by, for instance, the Ganges (India) and Mississippi (USA) rivers. There is little evidence of estuarine deposits in the proper canyons. If there are, many deposits originated during this period may be easily confused with those formed by fluvial action. Why? As a general approximation, we can infer that the tides against the Wisconsin coasts were small as it occurs near present-day ocean islands having steep accesses. Also, based on the water equivalent ice volume estimated by Flint (1961), average salinity must have been about l%o higher than present. Therefore, the circulation on the mouth of the estuaries that occupied the “canyon” valleys must have been of the salt wedge type. However, tidal effect in the inner part must have been important. It is expected that because of the strong convergence and relatively low friction, these estuaries were of the hypersynchronous type resulting in a continuous increment in tidal height and tidal current headward. Consequently, we may estimate that mixing of water masses occurred only at the mouth and sedimentation within the “canyon” may appear as fully fluvial although affected by tidal influence. High river runoff resulted in a sea level rise. After surpassing the shelf break, the transgression front found the extensive, low gradient (on average 7’ slope) shelf plains. Therefore, the channeled river valleys were replaced by the development of quite ephemeral coastal lagoons (Fig. 1-7), tidal flats and salt marshes similar to those presently observed on the east coast of USA and northern Europe. Only those places where rivers have cut down a deep valley across the shelf may have retained the classical estuarine type.

12

G.M.E. PERILLO

I

15

10

Years BP x

6

0

Shoreface

Continental Shelf

Shelf Break

lo3

Fig. 1-7. Estimated relationship between continental shelf slope and type of estuary resulting from a sea level rise: A) general trend of sea level rise in the last 15,000 years, and B) scheme of a continental shelf. (Modified from Emery, 1967; Nichols and Biggs, 1985.)

Due to the low relief, minor elevations of the sea level should have produced large inundations on the continental shelves; therefore, the lagoon type deposits cannot be too thick. Emery (1967) suggests that many sand ridges found presently on the continental shelves as described by Swift et al. (1978) have trends, shapes and sizes analogous to the sand bars and barrier islands that close present day lagoons. Field and Duane (1976) also indicated that barrier islands occurred in many places of the continental shelves and that they migrated continuously in time but discontinuously in space toward the present coastline. The dynamical conditions acting on these estuaries were probably similar to those observed on the present microtidal estuaries, specially concerning wave and littoral sediment transport. However, general tidal range must have been higher than before the sea level passed over the shelf break, and average salinity values were reducing slowly due to major input of fresh water. Further sea level raise allowed the transgression to reach the inner shelves which gradients (about 17’) are larger than those of the middle and outer shelf. Here the presence of valleys, now formed by river, glacial and (in a lower number) neotectonic activity, lead to the appearance of some classical estuaries but mixed with lagoons (Fig. 1-7). Their areal distribution was dependent upon the local shoreface gradient. Allowing for the fluctuations mentioned earlier in this section, there is general agreement that sea level reached about the present position 3,000 yr BP. Today estuaries have then reached their present position. From then on estuaries have adapted to the particular conditions of each coast, river and climate in which they have developed. The search is now toward an equilibrium that most probably will never attain. Here is where we can introduce the idea of the ephemeral conditions of estuaries from the geological time scale standpoint. Considering the cyclicity of the Pleistocene glaciations, many authors agree that we are in an interglacial period. Schubel and Hirschberg (1978) even stress that interglacial periods occurred only during 8% of the time in the last million years; each lasting 10,000 f 2, 000 yr. Then it should be only a matter of time before the return of the glaciers. However, the present situation differs from that during the Sangamon or earlier interglacials because of the presence of the “industrial man.” Through the combustion of fossil fuels, man is changing the COz cycle and thus intensifying a greenhouse effect with an associated artificial

GEOMORPHOLOGY AND SEDIMENTOLOGY OF ESTUARIES: AN INTRODUCTION

13

rise of Earth’s temperature. Prediction as to the behaviour of the temperature for the next few thousands years based on what happened in the last one and half centuries seems uncertain. However, if the present trend is firm, ice caps will be slowly retreating and consequently coastal areas will be invaded by the sea. Hoffman (1984) predicts an increase in mean sea level of the order of 1 m within the next 60-150 yr. More recent estimates indicate that value will not be larger than 0.3-0.5 m (Carter et al., 1992). Nevertheless, any increase in sea level will move estuaries further inland. Transgression over trailing edge coasts that have extensive plains may result in developments of coastal lagoons and tidal flats rather than typical estuaries. Meanwhile, collision or subduction coasts will produce very short estuaries of the ria type. However, eustatic modifications are not the only way in which estuaries evolve. Once they are formed, estuaries become sediment traps (Nichols and Biggs, 1985). First, let us imagine that a coast is stable, that is, there is no coastal migration and no eustatic changes occur. Therefore, the interplay is between the sediment introduced and the estuarine circulation that should export it to the continental shelf. The circulation within the estuaries is restricted due to the reversing nature of the tidal currents. Only the residual fluxes, which are strongly dependent on the density structure and tidal asymmetry, drive the sediment within the estuary and the material is not always exported. As a consequence, residence time of the sediment particles may increase exponentially to infinity (ultimate deposition) from the values in the river. In a stable coast, this process results in the filling of the estuary and, later on, the river bypassing it and discharging directly into the shelf. If the coast is affected by subsidence, filling up of the estuary will then depend on the balance between sediment supply and rate of subsidence, either due to isostasy or eustasy. If supply is larger than subsidence, we have the same result as described in the previous paragraph (ie., formation of deltas). When subsidence is equal to or larger than supply, we have the “eternal” estuary since it will never be filled up as long as the general conditions do not change.

FACTORS INFLUENCING THE GEOMORPHOLOGY AND SEDIMENT DISTRIBUTION

A detailed description of the dynamic factors that influence the geomorphology and sediment distribution of estuaries is beyond the scope of the present chapter. There is a large bibliography that provides deep insight on these factors, for instance, the books by Dyer (1973, 1986) and Officer (1976). Specific influences related to particular types of estuaries and major environments commonly found in them are included in the respective chapters of the book. Dyer (this volume) describes the sediment transport process occurring in estuaries. Nevertheless, it is important to mention here the most significant factors that induce the formation of estuaries or act on their evolution. As prime responsible of the estuarine characteristics are the hydrodynamic factors, namely tides, river inflow, estuarine circulation, waves and atmospheric forcing. The resulting estuary is primarily a consequence of the combination of these factors

14

G.M.E. PERILLO

acting over all the estuary or in specific parts of it. Interactions between the different factors with the borders are complex; mostly non linear. Evidences of them are the geomorphologic changes that occur in the estuary associated with the sediment transport processes. The general sedimentology of a specific estuary is the consequence of many conditions. One of the most important is the sediment source, which may be from the river, the adjacent shelf, transported by littoral currents and introduced into the estuary by tidal action or littoral drift. Erosion of inner estuary rocks or pre-estuary sediments and biogenic material is also significant in relation with the particular geological setting of the estuary or the climatic situation of the region. Furthermore, within the estuary proper, sediment distribution is extremely variable reflecting the hydrodynamic conditions and the particular transport processes dominant on each portion of it. All these aspects are treated in detail on the corresponding chapters of the book.

SUMMARY

Normally estuaries occupy the areas of the coast least exposed to the marine action. In this way, wave activity is generally quite reduced, allowing the development of harbours, recreational facilities, or appropriate aquaculture initiatives. Nevertheless, within the estuaries the dynamical processes are rather strong and impose a remarkable stress over the biota, either permanent or temporary, the morphology and the civil works. Although the number of examples of estuaries observed in the geological record is small yet, there are increasing evidences that they were a common feature of the planet. It is only a matter of common sense to accept this concept, since river and sea have interacted from the Precambrian period to the present. Still, their cast is difficult to find due to the fact of their little regional span and the variety of facies that can be confused with other environments. The interplay of elements like climate and type of setting may define the basic structure of the estuary during its formation. However, once formed, further evolution depends on many factors that act at different scales in time and space. The most important are the physical parameters and the input of sediment. The former will act to modify the original shape to attain an equilibrium form, while the latter is either deposited within the basin or exported to the shelf. Whichever prevails, the estuary disappears or becomes a permanent feature in the coast as long as the sea level does not change dramatically.

ACKNOWLEDGEMENTS

Partial support for the present article has been provided for National Geographic Society Grant 4540/91 and CONICET PID 3886/92. Instituto Argentino de Oceanografia, Contribution No. 280.

GEOMORPHOLOGY AND SEDIMENTOLOGY O F ESTUARIES: AN INTRODUCTION

15

REFERENCES Aliotta, S. and Perillo, G.M.E., 1989. Terrazas submarinas en el estuario de Bahia Blanca. Actas J. Geol. Bonaerenses, 1: 217-230. Aliotta, S. and Perillo, G.M.E., 1990. Antigua linea de costa sumergida en el estuario de Bahia Blanca, provincia de Buenos Aires. Rev. Asoc. Geol. Arg. 45: 300-305. Allen, G.P., Salomon, J.C., Bassoulet, P., DuPenhoat, Y . and DeGrandpre, C., 1980. Effects of tides on mixing and suspended sediment transport in macrotidal estuaries. Sediment. Geol., 26: 69-90. Beard, J.H., Sangree, J.B. and Smith, L.A., 1982. Quaternary chronology, paleoclimate, depositional sequences, and eustatic cycles. AAPG Bull., 66: 158-169. Bowen, D.Q., 1978. Quaternary Geology: a Stratigraphic Framework for Multidisciplinary Work. Pergamon Press, New York, 221 pp. Carter, TR , Parry, M.L., Nishioka, S. and Harasawa, H., 1992. Preliminary guidelines for assessing impacts of climatic change. Intergovernamental Panel for Climatic Change Rep. CGER-1005/92, 28 PP. Davis, R.A. (Editor), 1985. Coastal Sedimentary Environments. Springer-Verlag, New York, 716 pp. Dyer, K.R., 1973. Estuaries: a Physical Introduction. Wiley and Sons, London, 140 pp. Dyer, K.R., 1986. Coastal and Estuarine Sediment Dynamics. J. Wiley and Sons, Chichester, 342 pp. Emery, K.O., 1967. Estuaries and lagoons in relation to continental shelves. In: G.H. Lauff (Editor), Estuaries. AAAS, Washington, DC. pp. 9-11. Fairbridge, R.W., 1961. Eustatic changes of sea level. Phys. Chem. Earth, 4: 99-185. Field, M.E. and Duane, D.B., 1976. Post-Pleistocene history of the United States continental shelf significance to origin of barrier islands. Geol. SOC.Am. Bull. 87: 691-702. Flint, R.F., 1971. Glacial and Quaternary Geology. J. Wiley and Sons, New York, 892 pp. Gbmez, E.A. and Perillo, G.M.E., 1992. Geomorphologic evolution and sea level changes of the Bahia Blanca Estuary, Argentina. Wolfville '92, Geol. Assoc. Can. (abstract). Gbmez, E.A. and Perillo, G.M.E., 1994. Sediment outcrops underneath shoreface-connected sand ridges, outer Bahia Blanca estuary, Argentina. Quat. South Am. Antartic. Penn., 9(3) (in press). Gonzalez, M.A., 1989. Holocene levels in the Bahia Blanca estuary, Argentine Republic. J. Coastal Res., 5: 65-77. Gonzalez, M.A. and Weiler, N.E., 1994. Argentinian Holocene transgressions: sideral ages. J. Coastal Res., 10: 621-627. Hicks, S.D., 1981. Long-period sea level trends for United States through 1978. Shore Beach, 49: 26-36. Hoffman, J.S., 1984. Projecting future sea level rise, methodology, estimates to the year 2100, and research needs. Office of Policy and Resource Management, EPA 230-09-007, Washington, DC, 121 pp. Isla, F.I., 1989. Holocene sea-level fluctuations in the Southern Hemisphere. Quat. Sci. Rev., 8: 359-368. Kraft, J.C. and Chrzastowski, M.J., 1985. Coastal stratigraphic sequences. In R.A. Davis (Editor), Coastal Sedimentary Environments. Springer-Verlag, New York, pp. 625-663. Kuelegan, G.H., 1949. Interfacial instability and mixing in stratified flows. J. Res. Natl. Bureau Stand., 43: 487-500. Lanfredi, N.W., D'Onofrio, E.O. and Mazio, C.A., 1988. Variations of the mean sea level in the soutwestern Atlantic Ocean. Cont. Shelf Res., 8: 1211-1220. Nichols, M.M. and Biggs, R.B., 1985. Estuaries. In: R.A. Davis (Editor), Coastal Sedimentary Environments. pp. 77-125. Officer, C.B., 1976. Physical Oceanography of Estuaries and Associated Coastal Waters. Wiley and Sons, New York, 465 pp. Olausson, E. and Cato, I. (Editors), Chemistry and Biogeochemistry of Estuaries. Wiley, New York, 518 pp. Perillo, G.M.E., 1989. Estuario de Bahia Blanca: definicidn y posible origen. Bol. Cent. Naval, 107: 333-344. Perillo, G.M.E. and Codignotto, J.O., 1989. Ambientes costeros. In: G.E. Bossi (Editor), l o Simposio de Ambientes y Modelos Sedimentarios, Bol. Sediment., 4: 137-159.

16

G.M.E. PERILLO

Postma, H., 1961. Transport and accumulation of suspended matter in the Dutch Wadden Sea. Neth. J. Sea Res., 1: 148-190. Postma, H., 1967. Sediment transport and sedimentation in the estuarine environment. In: G.H. Lauff (Editor), Estuaries. AAAS, Pub. 83, pp. 158-179. Pritchard, D.W., 1952. Estuarine hydrography. Adv. Geophys., 1: 243-280. Pritchard, D.W., 1960. Lectures on estuarine oceanography. B. Kinsman (Editor), J. Hopkins Univ., Baltimore, 154 pp. Russell, R.J., 1964. Techniques of eustacy studies. Z. Geomorph., 8: 25-42. Russell, R.J., 1967. Origins of estuaries. In: G.H. Lauff (Editor), Estuaries. AAAS Pub. 83, Washington, DC, pp. 93-99. Schackleton, N.J. and Opdyke, N.D., 1973. Oxygen isotope paleomagnetic stratigraphy of Equatorial Pacific core V-28-238, oxygen-isotope temperatures and ice volumes on a 105 year and 106 year scale. Quat. Res., 3: 39-55. Schubel, J.R. and Hirschberg, D.J., 1978. Estuarine graveyard and climate change. In: M. Wiley (Editor), Estuarine Processes, Vol. I, pp. 285-303. Stommel, H., 1953. Computation of pollution in a vertically mixed estuary. Sewage Ind. Wastes, 25: 1065-1071. Stommel, H. and Farmer, H.G., 1953. Control of salinity in an estuary by a transition. J. Mar. Res.,l2: 13-20. Swift, D.J.P., Parker, G., Lanfredi, N.W., Perillo, G.M.E. and Figge, K., 1978. Shoreface-connected sand ridges on american and european shelves: a comparison. Est. Coastal Mar. Sci., 7: 257-273. Tanner, W.F., 1968. Multiple influences on sea level changes in the Tertiary. Paleogeogr. Paleoclimatol. Paleoecol., 5: 165-171.

Geomorphology and Sedimentology of Estuaries. Developments in Sedimentology 53 edited by G.M.E. Perillo 0 1995 Elsevier Science B.V. All rights reserved.

17

Chapter 2

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONS OF ESTUARIES GERARD0 M.E. PERILLO All sciences started with Philosophy asking the questions, and they spread out on the minds of humanity. When all answers are achieved, everything will collapse again in Philosophy.

INTRODUCTION

In the last 40 years many definitions and classifications of estuaries have been put forward. Before attempting to develop a new definition, I analyzed more than 40 different ones provided by common dictionaries and encyclopedias as well as by specialist in the different disciplines associated to estuaries. A structured account for disciplines of the most important definitions is given in the Annex 2-1. From definitions found in dictionaries and encyclopedias it is sometimes difficult to obtain any valid interpretation of their actual meaning. This is specially true for dictionaries. However, in thematic encyclopedia the problem is not the lack of a clear definition but the contradiction among them, even though they may pertain to the same collection. The contradictory and interpretative problems are not language constrained since examples given in the Annex cover the three most common languages in the western hemisphere. The only difference is that in Spanish, the term ria is employed more often than estuario to represent the same thing, although this is only valid in Spain since in Latin American countries only the latter is used. Most dictionary definitions and some others restrict an estuary to the mouth of a river or a tongue of the sea reaching inland. While others may carry the estuary out to the continental shelf (Ketchum, 1951) or even include all the Northern Pacific Ocean (McHugh, 1967) as long as there is dilution of sea water or the presence of euryhaline species. Between these extremes, there is a wide range of alternatives that may be grouped within specific disciplines. However, estuaries are no longer the domain of any individual discipline. Within the last 15-20 years, it has been evident that interdisciplinary research is needed to obtain an adequate understanding of a single estuary, or even of a particular reach within an estuary. The lack of a definition that covers all the characteristicsof estuaries, nevertheless, has not prevented researchers from studying them. On the contrary, despite the multiplicity of definitions our knowledge of world estuaries has been increasing steadily. Notable progress can be measured by the number of papers published every year in scientific journals, and the growing number of books that are concerned with the subject. Most major publishers have a book collection related to estuaries. Then, if we have lived without a single, comprehensive definition that covers all

18

G.M.E. PERILLO

aspects of estuarine characteristics, why to bother in making one? The answer lies in urgently needed management and legislation (see, for instance, the definition used in the US Public Law 92-500, Annex) of estuaries and other coastal environments. From the viewpoint of coastal management, it is necessary to have a unambiguous, mutually exclusive definition that can provide a clear understanding of these coastal bodies, but also give an adequate framework to establish administrative priorities, pollution control, fishery regulations, recreation facilities among other things. In addition when multi and interdisciplinary research are planned, it is required that all components of the team should have the same understanding of the water body to be explored. Looking back to the relatively short history of estuarine research, I am convinced that no definition will ever satisfy all members of the estuarine community. Nevertheless, for over 25 years, Cameron and Pritchard’s (1963) definition (a modified version of the original Pritchard (1952) definition) has been used by many specialists. Although this has many interesting and useful features, as we will discuss in the following section, it has some shortcomings that impede a better generalization. The aim of the first part of the present article is to provide a new and more comprehensive definition that essentially covers all disciplines involved in estuarine research. The second part of the chapter will deal with a new morphogenetic classification. The latter is based on a structured relationship between the form and the origin of the different morphological constituents of estuaries. The interaction between the marine and terrestrial forces in shaping the present morphology is also considered. As an introduction to the new classification, a discussion of previous classifications is also presented.

PREVIOUS DEFINITIONS

From a general viewpoint, one can say that each estuary is unique since every estuary has its own intrinsic characteristics that make it different from all the others. Consequently, as it happens with other objects, to establish a definition and classification is a very hard task. However, we need a base from which to proceed. Etymologically, estuary derives from the latin word aestus which means “of tide”. That is to say that the term estuary has to be applied to any coastal feature in which the tide has special significance. Although estuaries may be regarded only by their physiographic parameters: that is, their geomorphology and hydrology, their biological and chemical components should also be considered. Any comprehensive definition must necessarily include these aspects. Definitions presently available to the estuarine researcher do not fulfil all these criteria. Each of the many disciplines that study estuaries has at least one definition, but normally one can find between three and ten different definitions. Some of them are strongly contradictory. The variety of definitions within one discipline may be due to several reasons, but the two most important may be: 1) different background of the researchers producing the definition, and 2) the location of the estuaries upon which their definition is based (Perillo, 1989b). An example

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONSOF ESTUARIES

19

can be drawn from the existing geological and physical definitions. For instance, coastal plain estuaries are better known than other estuaries, and most definitions and classifications implicitly consider them as the classical estuaries. Perhaps most geomorphologists have considered only those estuaries associated to a typical river mouth (Lyell, 1834; Lee, 1840: both in Schubel and Pritchard, 1972). This bias is reflected in most dictionary definitions (Annex 2-1) as well as in many of the early definitions of estuarine oceanography. Fairbridge (1980) calls attention to this point when he discussed the definition by Pritchard (1967): “This [the definition] excellently describes certain estuaries familiar to him, but it has totally lost the original, and critical, tidal and river qualifications. ... Pritchard’s model is thus completely unrealistic for a globally acceptable definition”. A general review of geomorphological and dynamical estuarine definitions was made by Schubel and Pritchard (1972). They analyzed more than ten classical definitions introduced by geologists, geomorphologists, geographers, physical oceanographers and biologists. Even though all of them address important characteristics of estuaries, the authors consider that all these definitions are “either too exclusive or too inclusive”. Schubel and Pritchard (1972) make a case in favour of the definition given by Pritchard (1967). The later is also the most common used in physical oceanography (e.g., Dyer, 1973; Officer, 1976); but also in several biological textbooks (e.g., Perkins, 1974; McConnaughey and Zottoli, 1983). Nevertheless, it is necessary to comment that the first definition by Pritchard (1952) was different from the later one, since it indicated that “An estuary is a semi-enclosed coastal body of water having a free connection with the open sea and containing a measurable quantity of sea water.”

Obviously this definition expands upon the first physical and chemical definition of estuaries that I was able to detect: that given by Ketchum (1951) as “An estuary is a body of water in which the river water mixes and measurably dilutes sea water.”

The first mention of the newer version definition was made in a review paper by Cameron and Pritchard (1963) (hereafter CP); although is common usage to attribute it to the second author. Their definition says: “An estuary is a semi-enclosed coastal body of water having a free connection with the open sea and within which sea-water is measurably diluted with fresh water derived from land drainage.”

This definition addresses four major characteristics of estuaries, from which others concepts have to be implied. 1) The estuary is a coastal feature corresponding to a morphologically controlled (semi-enclosed) water body but always open to the sea. This means that its lateral borders have to be clearly defined and have also a strong influence on the circulation within the feature. 2) There must be a continuous provision of salt water coming from the adjacent sea. The salt is introduced into the estuary either by advection or diffusion.

20

G.M.E. PERILLO

3) The dilution of sea water must be measurable. 4) Fresh water is generally provided by rivers and creeks discharging into the body of water. But non-channelized sources like groundwater cannot be forgotten, especially in sandy shores with large precipitation rates (e.g., Biscayne Bay; Bly Creek, Kjerfve and Wolaver, 1988). Day (1980) introduces an important variation over CP’s definition. Again the influence of the type of estuaries in which the author has worked becomes a substantial constraint in the elements contained in the definition: “An estuary is a partially enclosed coastal body of water which is either permanently or periodically open to the sea and within which there is a measurable variation of salinity due to the mixture of sea water with fresh water derived from land drainage.”

The above definitions do not take explicitly into account one of the most important features of estuaries, and from which derives its name: the tide. It is apparent from both definitions that the tide was averaged out and only the time-mean salinity structure and the gravitational circulation are considered. It is thus, that the mean salinity distribution is actually the basis for Pritchard’s physical classification (Pritchard, 1967). Nevertheless, the tide is the major mechanism providing energy input for mixing in practically all estuaries. Sometimes wind influence may overpower tidal mixing (e.g., Oden estuary, Bokuniewicz, pers. commun., 1993) although this is normally related with local climatic conditions that enhance the diversity of estuarine characteristics. An estuary is necessarily a coastal feature. According to Shepard (1973), the landward boundary of a coastal environment reaches as far as the marine influence into the continent. Therefore, the idea of tidal action even into the fluvial reach of the estuary, discarded by Cameron and Pritchard (1963) and Day (1980), cannot be eliminated from the definition. Tidal action is not only relevant for salt related processes, but also is associated, for instance, to the erosion, circulation and deposition of sediments contributed by the rivers. The rise and fall of the tide in the fluvial reach produce major changes in river discharge, degree of exposure of the fluvial margins, etc., thus modifying the characteristics of the transport of sediment and other related organic or polluting substances, as well as the conditions for the biota living on the freshwater tidal flats. In addition, many tidal sedimentary structures are commonly found in the fresh-water tidal zone (Dalrymple et al., 1992). In summary, we can suggest that the geomorphologic evolution and the biological conditions of the upper reach of the estuary is heavily dependent on tidal dynamics, even though salt may not reach so far landward. As an example, the estuary of the Rio de la Plata (Argentina-Uruguay; Fig. 2-1) has salinity intrusion up to the line Punta Piedras-Montevideo, and it may arrive further inland along the northern coast (e.g., Colonia) and rarely up to La Plata city on the southern coast (Boschi, 1988). However, many features (e.g., ebb and flood sinus, etc.) of the banks in the upper reaches are formed by tidal action. Although it may be small, all large saline water bodies (e.g., Mediterranean, Baltic, Aral, Caspio seas) have tides, either by direct astronomical effect, by cooscillating

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONS O F ESTUARIES

U

R

U

G

U

A

21

Y

Fig. 2-1. Rio d e la Plata estuary (Argentina-Uruguay), an example of a wide tidal river estuary. Salinity intrusions are found up to the line Punta Piedras-Montevideo. Some of the banks in the inner estuary show ebb and flood sinuses, products of tidal currents.

processes or through wind generated seiches that, to the effect, have similar properties than tides. Therefore, as long as the proposed estuary has any interaction with another saline water body having tidal movements, it can be considered an estuary (of course, if the other required elements also hold). Obviously, as it is discussed later, tidal effect has to be strong enough to provide significant modifications to the different components of the estuary. CP and Day definitions contemplate only those estuaries discharging directly into the adjacent sea. Estuaries flowing into other estuaries are not included into their idea; although, the most important contributions by Pritchard were made from studies of the Chesapeake Bay (Fig. 2-2). The later constitutes an excellent example of a complex and hierarchical estuary were tertiary estuaries (e.g., Elizabeth and

22

G.M.E. PERILLO

Fig. 2-2. Chesapeake Bay (USA), an example of hierarchical estuary. In the main estuary (actually the Susquekahama estuary) flow other estuaries such as James river, Potomac. The latter estuaries have other estuaries flowing into them.

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONSOF ESTUARIES

23

Lafayette Rivers) discharge into a secondary estuary (James River) which itself flows into the primary estuary (Chesapeake Bay). On the other hand, Day (1980) proposed the inclusion of intermittent estuaries within his definition. Although the idea is interesting, the circulation processes and type of biological occurrences (or survivals) differ whether the connection is open or closed. In these circumstances, this type of “blind estuaries” should be considered as estuaries only when they have an open connection, otherwise they become an albufera without any resemblance to an estuary. Furthermore, the fact that estuaries must be connected either directly to the open sea or any other saline water body rules out the idea proposed by Herdendorf (1990), and partly supported by Odum (1990) and Dyer (1990), which rivers discharging into freshwater lakes subject to tidal action or other tide-like water-level movement (e.g., seiches) are also estuaries. It is not enough that changes in the chemical characteristics of the lakes’ and rivers’ waters are significant to induce an estuarine circulation pattern, even though all other elements proper of an estuary are present. Even if either CP or Day’s definitions are regarded as the most adequate for describing estuaries in general, the word “measurably” should be changed to “significantly”. Measurable means that a researcher ought to have an instrument sensitive enough to detect the dilution; otherwise, if a certain degree of dilution (not specified in the definition) cannot be measured, he is not in an estuary. The word measurable puts a restriction in the definition based on the “most available present day technology”. We can further ask, what is the degree of precision required to detect any dilution? Fig. 2-3 is a crude example showing the possible differences between researchers in developing (Fig. 2-3A) and advanced (Fig. 2-3B) countries may consider what measurable actually means. Also, in very extreme conditions, we need to have continuous information on the salinity of sea water being introduced into the estuary during the measurement period. Average salinity values of the adjacent sea are not adequate for estimating the amount of dilution. Additionally, even if there is a certain dilution and it can be measured, it can be so small that it does not provide the necessary density gradient to drive any thermohaline circulation. Hence, it is essential that the dilution must be large enough, not only to be detected, but to produce a gravitational movement of water masses. Furthermore, the use of “significantly” introduces a statistical criterion within the definition. That is to say that one single measurement (as it can be literally interpreted from “measurable”) it is not enough to establish the particular condition of the water body. Day (1980) proposed the inclusion of hypersaline estuaries, which called “negative estuaries” in Pritchard (1952) scheme. Normally, hypersaline conditions occur when freshwater input does not exist or is very small. These estuaries are normally associated with very dry, continental climates that only provide land drainage in specific occasions along the year, after long drought periods or when evaporation is much larger than runoff. As long as freshwater is introduced into the coastal embayment, a dilution of the marine water is occurring. Consequently, hypersaline embayments (that fulfil the other requirements necessary to be an estuary) that

24

G.M.E. PERILLO

A

indicate that salinity gradient is I O‘5 %o

B Fig. 2-3. Interpretation of the word “measurable” depending on the available technology. A) In a developing country salinity measurements may be made with quite primitive instruments providing only a rough estimation of salinity. B) However, the degree of sophistication found in instruments in advanced countries may provide information much deeper than the actually required.

receive freshwater are not excluded from the estuarine definitions (including the one proposed in the next section). Extreme evaporation is a local climatic factor that is superimposed over the relationship between the amount of fresh and seawater that enters the estuary, and should not be taken into account as it occurs with the wind or air pressure. For instance, Piccolo et al. (1990) found salinities up to 39%0 at the mouth of the Sauce Chico estuary (the main freshwater input for the Bahia Blanca estuary) with typical average river discharge (3.8 m3/s). The hypersaline conditions are produced here by the tidal washing of a back-estuary salt flat (Piccolo and Perillo, 1990); a local attribute independent of basic estuarine processes.

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONS OF ESTUARIES

25

Fluval of

Marine or

r r t u a ry c

Wine

a-

Ocaanic dominance mainly caitwatw

Salr- freshwater mixing ( b r a c k i s h )

T i d a l intlurnce only (tidal bar63 common1 frrrhwatrr

2011.3

Fig. 2-4. Description of the parts of an estuary as proposed by Dionne (1963).

On the contrary, if there is not freshwater input, then the hypersaline body does not cover a basic premise required to be included into the category of estuaries. Moreover, we can accept the criteria for the existence of “intermittent estuaries”. Meaning that coastal water bodies that fulfil the conditions to be an estuary only part of the time should be judged as estuaries only in those periods. Another definition to be discussed here is that given by Dionne (1963, in Fairbridge, 1980) which says: “An estuary is an inlet of the sea reaching into a river valley as far as the upper limit of tidal rise, usually being divisible into three sectors: a) a marine or lower estuary, in free connection with the open sea; b) a middle estuary, subject to strong salt and freshwater mixing; and c) an upper or fluvial estuary, characterized by fresh water but subject to daily tidal action.” (Fig. 2-4)

In my understanding, Dionne’s statement is properly speaking a definition only in the first sentence, where it does not differ too much from almost all other geological and geomorphological definitions, plus many of those encountered in dictionaries. The division into three major sectors is, at best, a description of what is expected in an estuary. The main importance of this definition is that it is the one that best summarizes the different criteria given for most other geological definitions (Annex 2-1). More recently, Dalrymple et al. (1992) introduced a new, geologically-oriented definition developed as the base for constructing an estuarine facies model. “The seaward portion of a drowned valley system that receives sediment from both fluvial and marine sources, and contains facies influenced by tide, wave and fluvial processes. The estuary is considered to extend from the inner limit of tidal facies at its head to the outer limit of coastal facies at its mouth.” (Fig. 2-5)

Only water bodies that are formed in valleys effected by relative sea level rise can be accepted as estuaries if this definition is followed. Therefore, those developed by the action of littoral transport with no definitive valley or those existing where the local (relative) sea level is descending (as described by Pino, this volume) cannot be estuaries. Likewise, the Bahia Blanca estuary should be eliminated as an estuary because in the long and short term averages does not receive sediment from outside its mouths. On the contrary, in the last 3,000 years associated to a lowering of

26

G.M.E. PERILLO 32% SALINITY ES BOUNDARY BETWEEN URINE SAND BODY AND HAL MARINE SEDIMENTS

FACES BOUNDARY BETWEEN MARINE (TIDALLY-) INFLUENCED AND FLUVIAL SEDIMENTS

/ ; 4

SEDIMENT SOURCE

’\\

MARINE

MARINE &-ESTUARY

ESTUARY (Dalrymple et 01,1992) (Pritchard,1967) REVIR- - - - -&-

Fig. 2-5. Description of an estuary as proposed by Dalrymple et al. (1992). Note that it does not differ substantially from that of Dionne (1963) (Fig. 2-4).

the local sea level as described by G6mez and Perillo (1992b), the estuary is in a strong erosional stage and all its internal coasts (formed by tidal flats) are retreating. Sediment is continuously exported into the inner shelf and toward the coast of the Buenos Aires Province to the north of the estuary (Perillo, 1989a; Perillo and Cuadrado, 1990).

A PROPOSED NEW DEFINITION O F ESTUARIES

From the foregoing general analysis of the most used definition and others that subsume the arguments found in many other definitions, a new definition of estuaries is proposed here: “An estuary is a semi-enclosed coastal body of water that extends to the effective limit of tidal influence, within which sea water entering from one or more free connections with the open sea, or any other saline coastal body of water, is significantly diluted with fresh water derived from land drainage, and can sustain euryhaline biological species from either part or the whole of their life cycle.”

The definition has derived from previous ones proposed by Perillo (1989b) (see Annex 2-1) where only the geomorphological and physical elements were considered and by Perillo (1992). Besides including parts of some previously cited definitions, this definition considers other aspects not incorporated before. First of all is the existence of hierarchical estuaries like Chesapeake Bay in which there are primary to tertiary estuaries. Second, there is the explicit indication of more than one free connection. In this form, coastal lagoons or the so called bar-built estuaries, both having significant dilution, are clearly included in the definition. Contrary to most geological definitions, the present one does not incorporate the character or origin of the

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONS OF ESTUARIES

27

depression in which the estuary has formed. Normally those definitions explicitly say “a river valley,” thus excluding coastal features not originated only by fluvial action as fjords and some bar-built estuaries. The later are sometimes not originated by fluvial action but related to alongshore transport of sediments closing an existing bay. In addition, the coexistence of tidal action and intrusion of sea water is now formally established. In effect, the estuary extends inland up to the effective limit of tidal action, but it is within the segment that stretches from that inland point to the mouth in which seawater dilution can occur. This model permits the differentiation within the estuary of the three sectors proposed by Dionne (1963) and further described by Dalrymple et al. (1992), and also allows for estuaries that have only one or two of the sectors. For instance, the Amazon river then may be considered as an estuary (a tidal river estuary in the morphogenetic classification proposed later) that only has the upper or fluvial sector. The suggested definition has a quality that makes it different from all others previously proposed: it spans all basic disciplines dealing with estuaries. Both geomorphological and physical criteria have been common in many definitions, and the chemical criterion is met by the part related to the dilution of salt water (meaning that there is a change in the elementary composition from the standard seawater solution). The biological aspect is uncommon in estuarine definitions. Most biological definitions as described in Annex 2-1 clearly represent the estuary as “...primarily a hydrographical phenomenon” (Barnes, 1974). But in the new definition the biological criterion is specifically included when the estuary can be the habitat of species that are adapted to resist important changes in salinity as has been first proposed by Ringuelet (1962) (see Annex 2-1). The euryhaline (from greek eury = wide, broad) term is used here just to describe biological species that can withstand those modifications in salinity and have no relation with any specific salinity range.

PREVIOUS GEOMORPHOLOGICAL CLASSIFICATIONS OF ESTUARIES

Estuaries may be classified as any other object: after defining the object, it is necessary to characterize and order its outstanding parameters. The next step is to define the viewpoint of the classification, that is, which are the criteria and objectives of the classification. Since this book is devoted mainly to the geomorphology and sedimentology of estuaries, I will only consider the parameters related to these disciplines. Within the geological parameters the most important in this case are the genetic, geomorphologic and sedimentologic criteria. While the physical concepts may involve all those parameters that can be measured in an estuary (i.e., salinity, temperature, tides, wind, currents, etc.). Although all of them may be employed, usage of one or a combination of parameters requires that it/they must be common to all estuaries and also must have some kind of differentiation from one estuary to another. Sediments, for instance, are common enough to all of them; nevertheless, their variation within a single estuary may be so large and dynamical and geomorphologically dependent that

28

G.M.E. PERILLO

a classification based only on sediment distribution patterns seems impracticable. The same occurs with tidal current intensity or winds. In the present section a review of several of the most common classifications is presented. The objective of the descriptions that follow is two fold: to introduce the classificationperse, but further on is to introduce the readers with the basic terminology and the particular environment that will be tackled in the following chapters. As a result, the particular description given for each element of any classification is composed from what the author originally indicated plus general interpretations added from other authors and myself. Each subtitle will be accompanied by the name of the researcher(s) that developed the classification. Afterward, a new morphogenetic classification is introduced. Physiographic classification (Pritchard, 1960) The first known classification of estuaries from a geomorphologic point of view is due to Pritchard (1952) who divided the estuaries in three groups: drowned rivers, fjords and bar-built estuaries. Later, Pritchard (1960) completed the classification by including a fourth category that contemplated those formed by tectonic processes. Some features of the estuaries included in this classification will be discussed at length since they will be employed also in the following classifications.

Drowned river valleys This term has been wrongly employed in many occasions as synonymous of coastal-plain estuaries. They are basically what everybody thinks an estuary should be. They were formed by sea flooding of Pleistocene-Holocene river valleys during the Flandrian transgression. In Fig. 2-6 a schematic view of a classical drowned river estuary is exhibited. Normally they have a funnel shape with an exponential increase of the cross-section toward the mouth (Fig. 2-6a). The longitudinal profile shows a seaward gradient which is, in general, not interrupted by a sill (Fig. 2-6b) formed by either the original material of the valley or a barrier deposited previously to the drowning of the valley. On the average, these estuaries are about 10 m deep reaching some 20-30 m at the mouth. The valley has an acute V-shape when formed on

.'iJ

Ground W & ? w

... .. . ... ...,......_.. . . .. .'..'.,

-~

..... (C) '

..

Fig. 2-6. Schematic diagram of a drowned river valley estuary: a) plan view; b) longitudinal profile; c) cross-section profile.

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONSOF ESTUARIES

29

mountain and cli@ coasts, but the classical coastal plain estuary has a more open V-shape restricted only to the channel. Normally the valley presents “shoulders” or terraces either on one or on both sides. Modifications of this general description may be produced by the regional setting of the feature, the climate and the type of rock in which they were carved. Therefore, the width to depth ratio may vary over a large range being from the order of 10-100 in ria valleys (northern Spain coast) to 1000 (Chesapeake Bay) to 20000 (Rio de la Plata) in coastal plain estuaries. In general, drowned river valleys exhibit important sediment deposits and the exponential dependence of the cross-section seems (although it has not been proved) to be related to a long-term adjustment between sedimentation and erosion toward an equilibrium shape. Most estuaries of the world correspond to this category. The classical examples are Chesapeake and Delaware Bays, and the Thames and Gironde rivers. Fjords As drowned river valleys estuaries have developed in low and middle latitudes, fjords are associated to high latitudes which were covered by the Pleistocene ice-sheets (northern Europe and Canada) or coasts affected by alpine glaciation (southern coast Chile). Usually the glacial tongue invaded a previous river valley and by its effective and characteristic method of erosion carved a totally different new valley. As the glacier retreated, the sea advanced drowning these glacial valleys. The general physiographic characteristics of a fjord type estuary are presented in Fig. 2-7. Valley width is relatively uniform (Fig. 2-7a) and in cross-section it has an U-shape (Fig. 2-7c). However, a variety of drowned glacial valleys called fjards have developed in the low-relief rocky coast of northern Sweden, having cross-sections with less steep walls and presenting some lateral terraces which may be confused with strandflats. Another major difference between fjards and fjords, which is also due to the different coastal relief, is that the former has highly irregular inner shores and the tributaries are mostly lateral. One outstanding feature of most fjords is the presence of a shallow sill near or at their mouth, that closes the very deep valley (Fig. 2-7b). While the sill can be as shallow as 4 m, as in the Norwegian coast or as deep as 150 m (British Columbia

Fig. 2-7. Schematic diagram of a drowned glacial valley estuary: a) plan view; b) longitudinal profile; c) cross-section profile.

30

G.M.E. PERILLO

coast), the valley can be normally between 200 and 800 m deep, reaching maxima of 1200 m as in the Mercier Channel (Chile). In general, the sill corresponds to the most advanced frontal moraine formed within the valley. Minor sills can be found within the inlet produced by other frontal moraines either due to fluctuations during the main glacier retreat or by discharge of tributary glaciers. The latter will appear nearly parallel to the main valley sides though, and may be confused with relicts of the lateral moraine. Because of their regional setting, fjords are located in rocky shores and sediment supply is relatively scarce and seasonally variable. Coarse sediments are found normally at the head of the estuary, near the main river entrance. Meanwhile bottom material appears as a veneer of mud deposited in a reducing environment. The muds are the product of the settling of suspended sediments through water column as water circulation is very low or null. The level of recirculation of the water column below the sill level is dependent on the depth of the sill and the depth of the valley.

Bar-built estuaries These estuaries are also called Coastal Lagoons. Most bar-built estuaries are located on river valleys of very low relief coasts with small tidal ranges and river discharges. Although there are examples in meso- and macrotidal shores, littoral processes appear as dominant in the local environment. Consequently, dynamical dominance is produced by wind and littoral transport which can build up a barrier that encloses the lagoon (Fig. 2-8A). Although the most commonly described bar-built estuaries (Eastern and Gulf coast of USA) respond to the previous characteristics, there are many other examples worldwide in which the lagoon is located on previous (Mar Chiquita lagoon, Argentina; Dos Patos lagoon, Brazil) or present (Queule and Lenga estuaries, Chile) embayments restricted by the formation of a barrier. There are many differences between both types of barriers, being the most remarkable their length, width and number of inlets. South American lagoons occupy more restricted areas (although Dos Patos lagoon is the world largest) and are closed by a relatively short and wider barrier with only one inlet. Overwashing of the barrier seldom occurs even during the strongest storms. The lagoons proper are normally shallow (about 2 m deep) bordered on the land side by either the original coast (microtidal environments) or tidal flats but most commonly by salt marshes or mangroves in tropical climates. Highly sinuous tidal channels are developed on the muddy bottom sediments. Only the inlets, where tidal currents are stronger due to the jet-like behaviour, are deeper and sometimes limited in both extremes by tidal deltas. Tectonic estuaries The last category in Pritchard’s classification is, as defined by the same Pritchard (1960) “.4 m (Fig. 2-8).

Microtidal estuaries Dynamically, microtidal estuaries (Fig. 2-8A) are dominated by wind and wave action. If rivers are important, their influence can be decisive in the rapid evolution of the feature toward a deltaic environment. Tidal influence is felt mainly at inlets. This type of estuary may be associated to the bar-built estuaries of Pritchard or wavedominated of Dalrymple et al. (1992). Nevertheless, some major rivers discharge on microtidal coasts (e.g., Mississippi, Nile). Chesapeake Bay is also a microtidal estuary which only in broad terms can be fitted into Hayes’s classification. The principal forms of deposition are flood deltas, wave built features (spits, bars, beaches, etc.), storm deposits (overwash fans) and river deltas. Mesotidal estuaries These estuaries are probably the most common and widely studied estuaries in the world (Fig. 2-8B). Many estuaries on the southeastern and western coast of USA and some others elsewhere (e.g., Orinoco, Niger, several in Indonesia, Bahia Blanca, etc.) are located on mesotidal coasts. Tidal currents are dominant as a form-generating agent over other marine, fluvial or climatic agents. The major forms are tidal deltas (both flood and ebb), salt marshes and tidal flats. Macrotidal estuaries They are the least studied (Fig. 2-8C), although there is a strong tendency toward their analysis within the last two decades. Some examples are the Bay of Fundy (Canada), Tay (Scotland), Gironde (France), Rio Gallegos (Argentina). Hayes (1975) considered that these estuaries are broad-mouthed and funnel-shaped with linear sand bodies occupying the central portion and extensive tidal flats and salt marshes bordering the coast. Tidal currents are overall dominating and wave action may be important, as in all other cases, at the mouth.

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONS OF ESTUARIES

33

If Pritchard’s classification is considered as purely genetic, Hayes’ one is totally geomorphologic although is based on only one physical parameter: the tide. It is important to point out that Hayes intention was not to produce a classification but to correlate depositional sand bodies in estuaries with different coastal environments represented by distinct tidal ranges. Another drawback of this classification, is that there are other factors that control the morphology of an estuary that were not taken into account. The examples and the morphological elements contained in each of them were all taken from the eastern coast of North America, where there is a general continuity of morphological patterns that no necessarily repeats itself in other coasts of the world.

Evolutionary classification (Dalrymple et al., 1992) Closely related to the one developed by Hayes (1975), Dalrymple et al. (1992) estuarine classification is part of a more complex facies model that combines the relative importance of river outflow, waves and tides with time. The result is a triangular prism that represents the different coastal environments associated to the three essential parameters (Fig. 2-9A). A cut through the prism reveals a single time-independent triangle that correlates the percentages of each environment for a particular sea level condition. Deltas (river dominated environment) are located at the fluvial apex while strand plains and tidal flats are positioned along the wave-tide side. Differentiation between them and also in the two types of estuaries is based on terms of wave or tidal dominance (Fig. 2-9B).

Wave-dominated estuaries The energy and facies distribution for wave-dominated estuaries is presented in Fig. 2-10. Waves are strongly dominant at the mouth producing littoral transport and normally developing some kind of barrier that partially closes the mouth. Tidal influence may be observed in its capability to maintain open the inlet(s), becoming practically null toward the head, where only the river input is dominant (Fig. 2-10A). The resulting facies distribution (Fig. 2-10B) clearly corresponds to a bar-built or microtidal or coastal lagoon estuary from other classifications. At the mouth of the barrier-inlet system and adjacent areas, it is possible to find flood deltas and washover fans. In the central portions sedimentation of fine sediments is dominant in a shallow basin crossed by tidal channels where the major process is the resuspension of the bottom material by local waves produced by the passage of storms. At the head, the river forms a delta as it enters a basin with very low capability of reworking and redistributing its input.

Tide-dominated estuaries Tidal dominance does not require necessarily of strong tidal currents or large tidal ranges, although those conditions make the analysis more clear. Simply lack of any wave activity is enough even in microtidal coast to produce tidal-dominance. Tides and waves may have similar amount of energy at and near the mouth, but tides are much stronger than both waves and river discharge in the middle and upper

34

G.M.E. PERILLO

SPiTF = STRAND PLAIN/

WhvCa

TIDES RIVERS

B Prograding: Fluvial

Embayed Mixed Sediment

DOMINATED

DOMlNni tu

\ Prograding. Marine Sediment

WAVES

R w l a t i v w Power W a v w / T i d w

TIDES

Fig. 2-9. Classification of coastal environments associated to estuaries according to Dalrymple et al. (1992). A) General classification structure considering river input, wave and tidal processes and their variation in time (sea level changes); B) a cross-section through the prism presented in (A) showing the classification of estuaries in wave- and tide-dominated.

estuary. River influence becomes progressively larger within the river valley proper as friction drains tidal energy (Fig. 2-llA). As the energy is about the same along the estuary, sand sediments and facies are found also respectively distributed (Fig. 2-llB). Obviously the larger concentrations are found at the mouth, being reduced to the tidal channels as we move landward. Finer sands are found at the zone of minimum energy. Fine sediments are distributed on tidal flats and salt marshes. Dalrymple et al. (1992) classification is purely geological rather than geomorphologic. No consideration of fine sediments transported in suspension is given since their movement is independent of the zonation. Nevertheless, fine sediments deposited from this transport make more than 60% of the sediment facies in most estuaries and in some up to 90%. Separation between wave and tide dominance may be useful if one considers only the estuarine mouth or a system quite small. Waves and wave-related sedimentary structures are only important at the mouth even if there is not tide at all. Local waves within the central basin are occasional and seldom produce major sedimentary

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONS OF ESTUARIES

'

A

!T

MARINE-DOMINATED

'

E S T U A R Y MIXED-ENERGY

I

RIVER- DOMINATED

35

' 100

/ - - - - I -

- 50

W

U

O

L

0

I

I

B CENTRAL BASIN

Fig. 2-10. Wave-dominated estuaries: A) distribution of dynamical processes along the estuary; B) distribution of major morphological components (modified after Dalrymple et al., 1992).

A

ESTUARY 100

-50

0

Fig. 2-11. Tide-dominated estuaries: A) distribution of dynamical processes along the estuary; B) distribution of major morphological components (modified after Dalrymple et al., 1992).

36

G.M.E. PERILLO

structures other than some stratification formed by fine sand layers, originated by the winnowing and resuspension of the fine material, intercalated in mud sediments. Even in wave dominated mouths, tidal influence is important since tides are necessary to develop tidal deltas and the tidal channels within the central basin. A question to ask is: how can waves dominate river action at the head, if they do not reach that part of the basin but for local, low energy waves?. Furthermore, there is not entrance for river dominated estuaries (e.g., delta-front or tidal rivers, see proposed classification) because they are directly assumed as deltas out of the estuarine part of the classification, or estuaries where sedimentation processes may be relatively poor in comparison with the basin (e.g., rias and fjords). In summary, Dalrymple et al. (1992) classification is very useful to establish the spatial and temporal correlation among river, waves and tides and from then on to define the facies distribution within the estuary. However, it does not cover enough elements to be an effective geomorphologic classification.Furthermore, there is even no clear differentiation between this and Hayes’ classification: if the names are taken out, both are considering the same structured classification. The only difference is that Dalrymple et al. (1992) make a good case in pointing out that there is a continuous evolution between the two extreme cases while in the case of Hayes (1975) one ought to assume such continuity.

Morphological classification (Fairbridge, 1980) More recently Fairbridge (1980) provided the embryo of a new and more comprehensive physiographic classification of estuaries. It is based on both physiographic and hydrodynamic factors. The physiographic categories were organized according to their relative relief and degree to which the circulation is restricted at the mouth. The seven categories are presented in Fig. 2-12 and described by the author very summarily as follows: (la) High relief estuary with U-shaped valley profile = f o r d . (lb) Moderately high relief estuary = fiard, firth,sea loch. (2) Moderate relief estuary with V-shaped valley profile and winding valley = ria, aber; and those formed on karst coasts = calenque, cala. (3) Low relief estuary with branching valleys and funnel shaped plan view = open coastal plain estuaries; those flask-shaped and partly blocked by bars or barrier islands = barrier (semi-enclosed)coastal plain estuaries. (4) Low relief estuary, L-shaped in plan with lower course parallel to the coast = bar-built estuaries. (5) Low relief estuary, seasonally blocked by longshore drift and/or dunes, with/or without eolianite bars = blind estuaries. (6) Delta front estuary in ephemeral distributaries = deltaic estuaries; in interlobate embayments = interdeltaic estuaries. (7) Compound estuary, flask-shaped, ria backed by low plains = tectonic estuaries. In this classification, the geodynamical conditions are related to the long term relationship between the sea level changes, estuarine-fluvial dynamics, and neotectonics. Fairbridge (1980) considered that “disequilibrium” estuaries “...are mainly

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONS OF ESTUARIES

High reliefShallow rill,conrtriction

37

ged Stmndliner ~

( 2 ) Ria

f unnol .hap0

€ in dry reason8 ( 7 ) Toctonic Estuary

intordoltaIc

Fig. 2-12. Morphological classification of estuaries as introduced by Fairbridge (1980).

due to the early Holocene sea-level rise where it has been offset in some way by tectonics...” While “equilibrium estuaries” are constructional (e.g., delta channels). Following Jennings and Bird (1967), Fairbridge (1980) indicates that the dynamical environmental factors that produce regional variabilities are: 1) fluvial hydrology, 2) wave energy, 3) tidal range, 4) biological sedimentary factors, 5 ) sedimentology and mineralogy, and 6) geotectonics and neotectonics. Here, Fairbridge (1980) defines neotectonics as any youthful structural change in the height of the earth’s crust.

A PROPOSED NEW MORPHOGENETIC CLASSIFICATION

Although the classification by Hayes seems quite coherent, clustering of estuaries only by tidal range does not reveal more specific differences (e.g., setting, relief, etc.) between them. The method is partial because it does not consider some dynamical factors such as river discharge, littoral processes, etc. They have been contemplated by Dalrymple et al. (1992) but, in both cases, there is no correlation with the previous structure and relief in which the estuary has formed. On the other hand, Fairbridge’s classification is more thorough but less detailed than the others discussed. All previous classificationscan in general be considered as too inclusive since many

38

G.M.E. PERILLO

Former River Valleys

U

a

Former Glacial Valleys

3 I-

v)

a!

zn

River Dominated

Structural

lwl

SECONDARY ESTUARIES

Coastal Lagoons Fig. 2-13. Morphogenetic classification of estuaries introduced in the present paper.

different estuarine types can fit within one category. Then a new classification is introduced here, which opens much more the spectra by covering all possible categories of estuaries that are established by the definition given before. This classification is based on genetic and morphological considerations.The first division is the necessary genetic differentiation of estuaries as either primary and secondary estuaries (Fig. 2-13) following the criteria given by Shepard (1973) in his classification of shorelines. Primaiy estuaries: the basic form has been the result of terrestrial and/or tectonic processes and the sea has not changed significantly the original form. Specifically, these are those estuaries that have essentially preserved their original characteristics up to the present. Secondaiy estuaries: the observed form is the product of marine processes and their relative influence over river discharge acting since the sea level has reached nearly its present position. Further discussion on the different categories will be limited only to new aspects not addressed for categories of the same or similar names in previous classifications. Nevertheless, detailed descriptions of them are give in Chapters 3 to 9, this volume.

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONSOF ESTUARIES

39

Former fluvial valleys: formed by sea flooding of Pleistocene-Holocene river valleys during the last postglacial transgression. This category corresponds to the drowned river valleys of Pritchard. According to their coastal relief, they have been divided in two subcategories: Coastal plain estuaries: normally occupy low relief coasts produced mainly by sedimentary infilling of the river(s). Typical examples are Thames (UK), Gironde (France), Yangh-Tse (China). Rim: are former river valleys developed in high relief (mountainous or clif€y) coasts. Examples of these are the Pontevedra (Spain) and Deseado (Argentina) rias. Former glacial valleys: formed by sea flooding of Pleistocene glacial valleys during the last postglacial transgression. Also, based on the coastal relief to which they are associated, they are divided in two subcategories: Fjords: occupy glacially formed troughs located in high relief coasts. Examples are Oslo (Norway), Mercier (Chile). Fjurds: occupy glacially formed troughs in low relief coasts. Examples are those formed in the northern coast of Sweden. River-influenced: in high discharge rivers like the Amazon, Mississippi and de la Plata the valley is not presently drowned by the sea. However, the circulation in the lower portions of the river is highly affected by tidal dynamics, including reversing currents, resulting in characteristic morphological patterns. They have been divided in two subcategories: Tidal rivers: include those rivers that are affected by tidal action but salt intrusion may be limited to the mouth or it is totally absent within the valley. Normally these estuaries are associated to large discharge rivers that either by their coastal setting (e.g., de la Plata river) or the relatively strong coastal dynamical processes occurring at their mouth (e.g., Amazon) do not develop a delta. The degree of salt intrusion is seasonally and climatically dependent; however, tidal processes are very important in sediment transport dynamics and morphological evolution within the valley. Delta-front estuaries: this category includes the estuaries found in the portions of deltas affected by tidal dynamics and/or salt intrusion. The classic example is the outer Mississippi channels. Tidal rivers and delta-front estuaries’ subcategories have seldom been taken as part of the estuarine environment which may have occurred due to the influence of Pritchard’s definition. When they were included, the chosen category was coastal plain estuaries. In line with the viewpoint of the definition introduced in the present article, tidal influence is as important as salt intrusion in establishing the characteristics of an estuary. As suggested, high discharge rivers may have their valleys undrowned by the sea. Some drowning may have occurred during high sea level stands but that is not today situation. However, river discharge is affected by tidal action large distances upstream. In general, the interrelation between river and tide generates characteristic sedimentary processes such as the large shoals with marked ebb and flood sinus observed at the mouth of Rio de la Plata (Fig. 2-1).

40

G.M.E. PERILLO

Structural: their valleys were formed by neotectonic processes such as faulting, vulcanism, postglacial rebound, isostasy, etc. occurred since the Pleistocene. Pritchard and the other authors (e.g., Fairbridge, 1980) employing Tectonic or Structural terms have not included an important argument in their consideration of this type of estuaries: time. All the structural processes that give place to the formation of the valley must be active in the present time or being occurring from the Pleistocene. Otherwise, since almost all rivers are controlled by structural (e.g., faults) conditions their corresponding estuaries should all be tectonic. Examples are San Francisco Bay (USA) and Valdivia river (Chile). Coastal lagoons (after Kjerfve and Magill, 1989): inland water bodies usually oriented parallel to the coast separated from the sea by a barrier and connected to the ocean by one or more restricted inlets. In the present classification I included the subdivision suggested by Kjerfve and Magill (1989) based on the nature of the entrance: Choked: only one long and narrow entrance (Dos Patos, Brazil; Mar Chiquita, Argentina). Restricted: few inlets or a wide mouth (Pamlico Sound, USA, San Sebastian Bay, Argentina, Terminos, Mexico) Leaky: large number of entrances separated by small barrier islands (Belize Lagoon, Mississippi Sound). The coastal lagoons as proposed by Kjerfve and Magill (1989) and sustained here correspond to the bar-built and blind estuaries mentioned earlier. However, in the classification given by Kjerfve and Magill (1989) and in the present one blind estuaries are not considered. As indicated during the discussion of Day’s (1980) definition, water bodies whenever they are not connected to either the sea or any other saline coastal water body are not longer an estuary. It becomes an estuary as the inlet opens again. This is a common process occurring not only in South Africa but along the Atlantic coast of Uruguay where there is a series of Choked type lagoons that are closed during part of the year.

SUMMARY

There is a clear need for a definition that spans all disciplines related to the study of estuaries. Analysis of over 40 definitions show that none of those developed to the present fulfil this basic requirement. Neither they state the basic criteria necessary to establish the existence of an estuary, which are: coastal bodies, border control, tidal action, uni- or multiple connection with adjacent sea or a coastal saline water body, freshwater input that produces a statistically and circulation-wise significant dilution of the seawater, the existence of characteristics species that live in the estuary either through part or the whole of their life cycle. All these aspects are included within the definition proposed here as: “An estuary is a semi-enclosed coastal body of water that extends to the effective limit of tidal influence, within which sea water entering from one or more free connections with the open sea, or any other saline coastal body of water, is significantly diluted with fresh water

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONSOF ESTUARIES

41

derived from land drainage, and can sustain euryhaline biological species from either part or the whole of their life cycle.”

Following the criteria introduced by the definition, a morphogenetic classification of estuaries has also been presented (Fig. 2-13). The main concepts involved are: 1) preservation or not of the original features of the valley; 2) coastal relief; 3) process that originate the valley. In any case both are necessarily a step to reach a final definition and classification of estuaries. They can evolve (as they actually have done in the last five years) or being strongly modified by changing the phrasing or adding some further elements. However, the criteria in which they have been built cannot be overlooked in future definitions and classifications. Furthermore, I consider this article as practically the beginning of an international discussion that can actually carry us, devotees of Estuarine Oceanography, to that ultimate consideration. Having a unique definition and a unique classification will show us that we finally reach the point to say: “we really know what an estuary is”.

ACKNOWLEDGEMENTS

This article greatly benefited by a long, epistolary, discussion with Henry Bokuniewicz, even though some “significant” terms still require to be “measured”. Many other colleagues have made comments at different stages of the evolution of both the definition and classification, they all provided a different perspective that resulted in the ideas presented here. Partial support for the present article has been provided for National Geographic Society Grant 4540/91 and CONICET PID 3886/92. Instituto Argentino de Oceanografia, Contribution No. 281.

ANNEX 2-1. DEFINITIONS OF ESTUARIES IN DICTIONARIES AND ENCYCLOPEDIAS

Concise Oxford Dictionary Tidal mouth of a large river. Webster’sNew 20th Centuly Dictionary An arm of the sea; a frith or firth; a narrow passage, or the mouth of a river or lake, where the tide meets the current. Webster’sNew International Dictionary a) A passage, as the mouth of a river or lake where the tide meets the river current; more commonly, an arm of the sea at the lower end of a river, a firth. b) In physical geography: a drowned river mouth, caused by the sinking of land near the coast. Ediciones Garriga, 1958. Enciclopedia General del Max Barcelona Estuario: lugar donde entra y sale la marea a1 flujo y reflujo. Ria: canal o embocadura de rio o brazo de mar que se interna en la tierra donde suben las mareas y se mezclan las aguas dulces y saladas. Estuary: a place where the tide enters and leaves by flow and ebb.

42

G.M.E. PERILLO

Ria: a channel or river mouth or arm of the sea that penetrates inland where the tides rise and fresh and seawater mix.

Editorial Larousse, 1967. PequeAo Larousse de Ciencias y Tknicas. Buenos Aires Desembocadura de un rio por el cual penetra el agua del mar a1 subir la marea. Se distingue de la ria por el mayor caudal del rio correspondiente. Mouth of a river through which seawater penetrates as tide rises. It is distinguished from the ria by the larger discharge of the corresponding river. Grindley, J., 1969. Estuarine sedimentation. In: EI. Firth (Editor), The Encyclopedia of Marine Resources. Van Nostrandt, Reinhold, Co. New York The area in which sea water and freshwater have mutual influences. Encyclopedia Americana, 1970. New York Where a shoreline is sinhng or has been recently depressed, the rivers, unless large and heavily charged with sediments, have their valleys invaded by the encroaching sea, forming roughly funnel-shaped bays. Such bays are called estuaries... Real Academia Espa Aola, 1970. Diccionano de la Lengua Espaiiola. Madrid Estuario: estero de la orilla de una ria. Estero: (del lat. aesterium) terreno inmediato a la orilla de un rio por el cual se extienden las aguas de las mareas. Ria: Penetraci6n que forma el mar en la costa debido a la sumersion de la parte litoral de una cuenca fluvial de laderas mas o menos abruptas. Ensenada amplia en la que vierten a1 mar aguas profundas. Estuario: estero at the bank of a ria. Estero: (latin: aesterium) land at the bank of a river over which tidal waters extend. Ria: Penetration that forms the sea on the coast due to the drowning of the littoral part of a fluvial basin which sides are more or less abrupt. Wide mouth in which deep waters flow into the sea. Fairchild, J.E., 1972. In: Collier's Encyclopedia A geographical and geological term for an unusually broad river mouth. Stevenson, R.E., 1972. Estuarine hydrology. In: R.W Fairbridge (Editor), The Encyclopedia of Geochemistry and Environmental Sciences. Van Nostrandt, Reinhold, Co. New York An estuary is a wide mouth of a river, or arm of the sea, where the tide meets the river current, or flows and ebbs. La Grande Encyclopedia Larousse, 1973. Paris Bras de mer entrant dans les terres a l'embouchure d'un fleuve ou une riviere. Arm of the sea that penetrates inland at the mouth of a river. American Geological Institute, 1976. Dictionary of Geological Terms. Anchor Press, New York Drainage channel adjacent to the sea in which the tide ebbs and floods. Some estuaries are the lower course of rivers or smaller streams, others are no more than drainage ways that lead seawater into and out of coastal swamps.

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONS OF ESTUARIES

43

Berthois, L., 1978. Estuarine sedimentation. In: R.W Fairbridge and J. Bourgeois (Editors), The Encyclopedia of Sedimentology. Dowden, Hutchinson and Ross, Inc., Stroudsburg, PA An estuary is that part of the river subject to oceanic influence. Libraire Larousse, 1979. Larousse de la Langue FranGaise, Paris SinuositC du littoral, qui n’est couverte d’eau qu’in maree haute. Golfe form6 par l’embrochure d’un fleuve. Partie aval du lit d’une riviere oh se font sentir les martes. Littoral sinuosity, that it is covered by water only in high tide. Gulf formed by the mouth of a river. External part of a river bed where tides are felt. Hachette, 1980. Dictionnaire Hachette de la Langue FranGaise. Paris Embouchre d’un fleuve, formant un golfe profond et Ctroit. Mouth of a river, shaping a deep gulf and strait. Grand Dictionnaire Encycloptdique Larousse, 1983. Paris Embrochure fluviale, soumise B la marCe formant une indentation profonde dans la track littoral. Mouth of a river affected by the tide forming a deep indentation on the littoral. Encyclopedia Britannica, 1984. Chicago An estuary is a partly enclosed body of water that forms where river water is mixed with and diluted by sea water. Allabr, M., 1984, A Dictionary of the Environment. Translation in Spanish. Ediciones Piramide, Madrid Valle fluvial cubierto por agua a causa de 10s cambios en el nivel del mar con respecto a la tierra despuks que el rio ya ha excavado su canal. Fluvial valley covered by water due to changes in sea level in relation with land after the river has excavated its channel. Physical and geological definitions Lyell, C., 1834. Principles of geology, Vol.3. London Inlets of the land, which are entered both by rivers and the tides of the sea. Lee, C.S., 1840. Elements of geology Inlets of the sea into the land. The tides and fresh-water streams mingle and flow into them. They include not only the portion of the sea adjacent to the mouths of the rivers, but extend to the limit of tide-water on the streams. Ketchum, B.H., 1951. Thepushing of tidal estuaries. Sewage Ind. Wastes, 23: 198-209 An estuary is a body of water in which the river water mixes and measurably dilutes sea water. Pritchard, D.W , 1952. Salinity distribution and circulation in the Chesapeake Bay estuarine system. J. Mal: Res., 11: 106-123 An estuary is a semi-enclosed coastal body of water having a free connection with the open sea and containing a measurable quantity of sea water.

44

G.M.E. PERILLO

Emery, K.O. and Stevenson, R.E., 1957. Estuaries and lagoons. I. Physical and chemical characteristics. In: J.W Hedgpeth (Editor), Treatise of Marine Ecology and Paleocology Geol SOC.Am. Mem., 6T 673-693 Bodies of water bordered and partly cut off from the ocean by land masses that were originally shaped by non-marine agencies. Also: The wide mouth of a river or an arm of the sea where the sea water meets the river current or flows and ebbs. Dionne, J.C., 1963. Towards a more adequate definition of the St. Lawrence estuary. Z. Geomolph., 7: 36-44 An estuary is an inlet of the sea reaching into a river valley as far as the upper limit of tidal rise, usually being divisible into three sectors: a) a marine or lower estuary, in free connection with the open sea; b) a middle estuary, subject to strong salt and freshwater mixing; and c) an upper or fluvial estuary, characterized by fresh water but subject to daily tidal action. Cameron, WM.and Pritchard, D. W , 1963. Estuaries. In: M.N. Hill (Editor), The Sea, Vol.2. Wiley-Interscience,New York,pp. 306-324 An estuary is a semi-enclosed coastal body of water which has a free connection with the open sea and within which sea water is measurably diluted with fresh water derived from land drainage. Pritchard, D. W,1967. What is an estuary: physical viewpoint.In: G.H. Lauff (Editor), Estuaries. A A A SPub. 83, Washington,DC, pp. 3-5 An estuary is a semi-enclosed coastal body of water which has a free connection with the open sea and within which sea water is measurably diluted with fresh water derived from land drainage. Gorsline, D.S., 196% Contrasts in coastal bay sediments on the Gulf and Pacific coasts. In: G.H. Lauff (Editor), Estuaries. A A A SPub. 83, Washington,DC, pp. 219225 An estuary is an indentation in a coast in which tidal circulation meets land runoff and generally prevails over the land contributions. Morgan, J.P, 196% Ephemeral estuaries of the deltaic environment. In: G.H. Lauff (Editor), Estuaries. A A A SPub. 83, Washington,DC, pp. 115-120 An estuary is any coastal embayment periodically affected by brackish oceanic waters. Vissel; WA. (Editor), 1980. Geological nomenclature. R. Geol. Min. SOC. The Netherlands. M. Nijhofi The Hague, 540pp. A more or less funnel-shaped river mouth, affected by the tides. Kjerfve, B. and Magill, K.E., 1989. Geographic and hydrodynamic characteristics of shallow coastal lagoons. Mar: Geol., 88: 187-199 An inland river valley or section of a coastal plain, drowned as the sea invaded the lower course of a river during the Holocene sea-level rise, containing sea water measurably diluted by land drainage, affected by tides, and usually shallower than 20 m.

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONS OF ESTUARIES

45

Perillo, G.M.E., 1989. New geodynamic definition of estuaries. Rev. Geofis., 31: 281287 An estuary is a semi-enclosed coastal body of water that extends to the upper limit of tidal influence, where sea water entering from one or more free connections with the open sea, or any other saline coastal body of water, is significantly diluted with freshwater derived from land drainage. Dalymple, R.W, Zaitlin, B A . and Boyd, R., 1992. A conceptual model of estuarine sedimentation. J. Sedim. Petrol., 62: 1130-1146 The seaward portion of a drowned valley system which receives sediment from both fluvial and marine sources, and which contains facies influenced by tide, wave and fluvial processes. The estuary is considered to extend from the inner limit of tidal facies at its head to the outer limit of coastal facies at its mouth. Biological and ecological definitions Odum, El?, 1959. Fundamentals of ecology, 2nd ed. WE.Saunders Co., Philadelphia, Penn An estuary is a river mouth where tidal action brings about a mixing of salt and fresh water. Ringuelet, R.A., 1962. Ecologia acuatica continental. EUDEBA, Buenos Aires, 138pp. Un cuerpo de agua permanente o temporalmente abierto, con intercambio entre el curso fluvial y el mar, poiquilohalino y favorable para la vida de organismos eurihalinos y anfibioticos. A water body permanent or temporarily open, with interchange between the river and the sea, poiquilohaline and favourable for the life of euryhaline and anfibiotic organisms. Barnes, R.S.K, 1974. Estuarine biology E. Arnold Ltd., London, 77pp. An estuary is a region containing a volume of water of mixed origin derived partly from a discharging river system and partly from the adjacent sea; the region usually being partially enclosed by land mass. Perkzns, E.J., 1974. The biology of estuaries and coastal waters. Academic Press, London, 678pp. Uses Cameron and Pritchard definition. Day, J.H., 1980. What is an estuary? South Afi J. Sci., 76: 198 An estuary is a partially enclosed coastal body of water which is either permanently or periodically open to the sea and within which there is a measurable variation of salinity due to the mixture of sea water with fresh water derived from land drainage. McConnaughey, B.H. and Zottoli, R., 1983. Introduction to Marine Biology. C.V Mosby Co., St Louis, 638pp. Use Cameron and Pritchard definition.

46

G.M.E. PERILLO

Chemical definitions Portmann, J.E. and Wood,RC., 1985. The UK national estuarine classification system and its application. In: J.G. Wilsonand W Halcrow (Editors), Estuarine Management and Quality Assessment. Plenum Publ. Co., pp. 173-186 An estuary is the transition zone along the quality of water changes from that of freshwater, characteristic of inland river water, to that of saline water, characteristic of the open sea. Freshwater estuaries Bates, R.L. and Jackson, J.C., 1980. Glossary of Geology Am. Geol. Inst., Falls Church, E,2nd ed., 749pp. A freshwater estuary is the lower reach of a tributary to the lake that has a drowned river mouth, shows a zone of transition from stream water to lake water, and is influenced by changes in lake level as a result of seiches or wind tides. Offshore estuaries MeHugh, J.L., 1967. Estuarine nekton. In: G.H. Lauff (Editor), Estuaries. A A A SPub. 83, Washington,DC, pp. 581-620 Offshore estuaries are limited by the salinity front rather than the boundaries.

REFERENCES Barnes, R.S.K, 1974. Estuarine Biology. E. Arnold Ltd., London, 77 pp. Boschi, E.E., 1988. El ecosistema estuarial del rio de la Plata (Argentina y Uruguay). An. Inst. Cienc. Mar Limnol., 15: 159-182. Cameron, W.M. and Pritchard, D.W., 1963. Estuaries. In: M.N. Hill (Editor), The Sea. WileyInterscience, New York. 2: 306-324. Dalrymple, R.W., Zaitlin, B.A. and Boyd, R., 1992. A conceptual model of estuarine sedimentation. J. Sediment. Petrol., 62: 1130-1146. Davis, J.L., 1964. A morphogenetic approach to world shorelines. Z. Geomorph., 8: 127-142. Day, J.H., 1980. What is an estuary? South Afr. J. Sci., 76: 198. Dionne, J.C., 1963. Towards a more adequate definition of the St. Lawrence estuary. Z. Geomorph., 7: 36-44. Dyer, K.R., 1973. Estuaries: a Physical Introduction. Wiley and Sons, London, 140 pp. Dyer, K.R., 1990. The rich diversity of estuaries. Estuaries, 13: 504-505. Fairbridge, R.W., 1980. The estuary: its definition and geodynamic cycle. In: E. Olausson and I. Cat0 (Editors), Chemistry and Biogeochemistry of Estuaries, Wiley, New York, pp. 1-35. Gbmez, E.A. and Perillo, G.M.E., 1992a. Geomorphology of the Largo Bank, Bahia Blanca Estuary entrance. Mar. Geol., 105: 193-204. Gbmez, E.A. and Perillo, G.M.E., 1992b. Geomorphologic evolution and sea level changes of the Bahia Blanca Estuary, Argentina. Wolfville '92, Geol. Assoc. Can. (abstract). Hayes, M.O., 1975. Morphology of sand accumulation in estuaries: an introduction to the symposium. In: L.E. Cronin (Editor), Estuarine Research, Vol. 11. Academic Press, New York, pp. 3-22. Herdendorf, C.E., 1990. Great lakes estuaries. Estuaries, 13: 493-503.

DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONS O F ESTUARIES

47

Jennings, J.N. and Bird, E.C.F., 1967. Regional geomorphological characteristics of soe Australian estuaries. In: G.H. Lauff (Editor), Estuaries. AAAS Washington, DC. Pub. 83, pp. 121-128. Ketchum, B.H., 1951. The flushing of tidal estuaries. Sewage Ind. Wastes, 23: 198-209. Kjerfve, B. and Wolaver, T.G., 1988. Sampling optimization for studies of tidal transport in estuaries. Am. Fish. SOC.Symp., 3: 26-33. Kjerfve, B. and Magill, K.E., 1989. Geographic and hydrodynamic characteristics of shallow coastal lagoons. Mar. Geol., 88: 187-199. McConnaughey, B.H. and Zottoli, R., 1983. Introduction to Marine Biology. C.V. Mosby Co., St. Louis, 638 pp. McHugh, J.L., 1967. Estuarine nekton. In: G.H. Lauff (Editor), Estuaries. AAAS Washington, DC. Pub. 83, pp. 581-620. Medeiros, C. and Kjerfve, B., 1993. Hydrology of a tropical estuarine system: Itamaraci, Brazil. Est., Coastal Shelf Sci., 36: 495-515. Odum, W.E., 1990. The lacustrine estuary might be a useful concept. Estuaries, 13: 506-507. Officer, C.B., 1976. Physical Oceanography of Estuaries and Associated Coastal Waters. Wiley and Sons, New York, 465 pp. Perillo, G.M.E., 1989a. Estuario de Bahia Blanca: definici6n y posible origen. Bol. Centro Naval 107: 333-344. Perillo, G.M.E., 1989b. New geodynamic definition of estuaries. Rev. Geofisica, 31: 281-287. Perillo, G.M.E., 1992. A new definition of estuaries. Joint ECSA/ERF Estuar. Conf., Plymouth (abstract). Perillo, G.M.E. and Cuadrado, D.G., 1990. Nearsurface suspended sediments in Monte Hermoso beach (Argentina): I. Descriptive characteristics. J. Coastal Res., 6: 981-990. Piccolo, M.C. and Perillo, G.M.E., 1990. Physical characteristics of the Bahia Blanca estuary (Argentina). Est. Coastal Shelf Sci., 11: 303-317. Piccolo, M.C., Perillo, G.M.E. and Arango, J.M., 1990. Hidrografia del estuario del rio Sauce Chico (Bahia Blanca). Geoacta, 17: 13-23. Perkins, E.J., 1974. The Biology of Estuaries and Coastal Waters. Academic Press, London, 678 pp. Pino, M., Perillo, G.M.E. and Santamarina, P. 1994. Residual fluxes in a cross-section of the Valdivia River Estuary, Chile. Est. Coastal Shelf Sci., 39: 491-505. Pritchard, D.W., 1952. Salinity distribution and circulation in the Chesapeake Bay estuarine system. J. Mar. Res., 11: 106-123 Pritchard, D.W., 1960. Lectures on estuarine oceanography. B. Kinsman (Editor), J. Hopkins Univ., 154 pp. Pritchard, D.W.,1967. What is an estuary: physical viewpoint. In: G.H. Lauff (Editor), Estuaries. A A A S Washington, DC. Pub. 83, pp. 3-5. Ringuelet, R.A., 1962. Ecologia Acuitica Continental. EUDEBA, Buenos Aires, pp. 138. Schubel, J.R. and Pritchard, D.W., 1972. What is an estuary. In: J.R. Schubel (Editor), The Estuarine Environment: Estuaries and Estuarine Sedimentation. Am. Geol. Inst., Washington, DC, pp. 1-1 1. Shepard, F.P., 1973. Submarine Geology. Harper and Row, New York, 517 pp.

This Page Intentionally Left Blank

Geomorphology and Sedimentology of Estuaries. Developments in Sedimentology 53 edited by G.M.E.Perillo 0 1995 Elsevier Science B.V. All rights reserved.

49

Chapter 3

SEDIMENTARY SYSTEMS OF COASTAL-PLAIN ESTUARIES HENRY BOKUNIEWICZ

INTRODUCTION

While there have been many excellent studies of individual estuaries, comparative studies among estuaries are relatively rare. Emery and Uchupi made that point in 1972 and it is still largely true today. At a time when the definition of an estuary is being reconsidered, we may also wish to reconsider basic, dynamic characteristics of estuaries that invite comparison. The inherent assumption is the existence of a few, fundamental parameters that define the state of the estuary. If the parameters are not too numerous, they may be used to identify common behavior among different estuaries. To do this, we not only need the best definition of an estuary that we can devise, but also must explore parameterizations of the fundamental processes by which the behavior of estuaries can be classified. Our basic definition of an estuary was explained in an earlier chapter of this book (Perillo, this volume): “An estuary is a semi-enclosed coastal body of water that extends to the effective limit of tidal influence within which sea water entering from one or more free connections with the open ocean, or any other saline coastal body of water, is significantly diluted with fresh water derived from land drainage and can sustain euryhaline biological species for either part or the whole of their life cycle”. In this article, I will discuss the expression of fundamental, estuarine characteristics in a particular geologic setting - the coastal plain. Coastal-plain estuaries are those that occupy former river valleys along low relief coasts (Perillo, this volume). As a result of the Holocene sea level rise, such a geomorphic classification corresponds to the drowned valleys of rivers crossing the coastal plain (Curray, 1969; Dalrymple et al., 1992). It is conceivable, however, that during episodes of falling sea level, estuaries could be presumably re-established in what are today canyons and channels on the shelf. The scope of my topic excludes estuaries in the distributaries of deltas which were classified separately (Hart, this volume) and do not occupy former valleys. Coastal lagoons are also in a separate class, emphasizing the importance of river discharge in the behavior of coastal-plain estuaries. Both distinctions are sometimes ambiguous. Some coastal-plain estuaries, for example, have essentially filled their former valleys without yet creating either a submerged or a protruding delta. In another instance, an estuary may also reside in the channels of relict deltas. The mouths of others may be so modified by the growth of shore-paralleled spits and coastal barriers that the distinction between them and lagoons is more or less arbitrary. One conceptual difficultythat persists in the definition is the existence of drowned river valleys along the tideless marine coastal plain of Poland. The mixing of salt

50

H. BOKUNIEWICZ

water from the Baltic Sea into the mouths of rivers like the Oder and Vistula is accomplished primarily by meteorological forcing. These are considered estuaries by the local, scientific community and, although they are essentially tideless, I will include them in the class of coastal-plain estuaries. Perhaps the definition can be saved in the strict sense by considering the mechanism by which salt is introduced into the estuary as a meteorological tide. I will begin the article with a very brief survey of the major coastal-plain settings for estuaries. Often our ideotypes of coastal-plain estuaries are based on those that have been best studied, primarily those on the east coast of the United States or those in northwestern Europe. A more global viewpoint includes many estuarine systems that we know less about, but that must not be overlooked. The broadest perspective suggests to me that two estuarine characteristics deserve more attention - the pressure of littoral transport processes at estuary mouths, and the transition of the coastal-plain estuary to a delta estuary. I will also briefly review the basic parameters that are used to classifl estuaries. The hydrodynamic classifications are much better developed than sedimentological ones. I would like to suggest that an energy-based approach may help to bridge the gap.

COASTAL PLAINS

Coastal plains are the surfaces of unconsolidated sedimentary deposits at the margins of the continents. These units can either be fluvial ones formed of sediment delivered from the highlands or marine strata formed by deposition during transgressions. Coastal plains cover about 5.7 million km2 of the Earth’s surface (Colquhoun, 1968) and form a surface of low relief upon which the present drainage is superimposed. Most coastal plains are crossed by one or more major rivers. At the maximum of the last glacial period about 17,000 yr BP, sea level was about 135 m below present. At that time, rivers had the opportunity to incise valleys through the sedimentary deposits of their coastal plains in an attempt to reach a base level commensurate with lowered sea level. As sea level rapidly rose between 17,000 and 6,000 yr BP, these valleys were drowned and the well-defined estuarine characteristics appeared. Since that time, the rise in sea level has been more gradual and the ancestral estuaries have evolved under a set of processes that are less influenced by the rise and fall of sea level, but more sensitive to the hydrodynamics of the estuaries themselves and to the littoral processes impinging at their mouths. It is the expression of these processes that provide a basis for the parameterization of coastal-plain estuaries. The geographical habitat of coastal-plain estuary includes eight major coastal plains (Fig. 3-1; for a detailed location of most estuaries mentioned in the present chapter see Perillo, fig. 1-2). Except for the coastal plains in northern Russia, these areas are mostly characterized by subsidence. (1) The coastal plain along the Atlantic and Gulf coast of the United States covers an area of 940,000 km2 (Colquhoun, 1968). This coastal plain includes Cape Cod, Massachusetts and Long Island, New York, in the north, but the major coastal-plain

SEDIMENTARY SYSTEMS OF COASTAL-PLAIN ESTUARIES

51

Fig. 3-1. Index map. General locations of major coastal plains discussed by number in the text.

estuaries are located south of New Jersey. Very many coastal-plain estuaries are found along this coast and they have been the subject of many, excellent studies. Because of the great disparity in the volume of literature describing this region to other coastal plains in the world, there is a danger of these results having a disproportionate influence on our perspective of coastal-plain estuaries. Many basic principles concerning the behavior of coastal-plain estuaries have been distilled from studies in this region, nevertheless, some care must be exercised when applying these concepts to other settings. The tidal range generally increases from about 1 to 3 m southward along the Atlantic coast to Florida (Fig. 3-2). Waves are dominated by an east coast swell (Davies, 1980); wave energy decreases from north to south corresponding to a general decrease in wave heights from 1.6 m to 0.7 m (Nummendal, et al., 1977). The northern estuaries are in a relatively youthful stage of infilling and trapping both fluvial and marine sediments. Those in the south, however, are more mature and nearly filled (Meade, 1969). Delaware Bay is one of the largest estuaries on the east coast. Littoral sand is transported to the estuary mouth both from the north and south and a large shoal complex towards the northern shore restricts flow somewhat into and out of the estuary but strong tides have cut a deep channel on the southern side (Knebel et al., 1988). Further into the estuary a series of tidal channels separated by elongate shoals are found. In its upper reaches, the estuary is a partially-mixed one and fine-grained sediment deposits are found (Schubel and Meade, 1977; Oostdam and Jordan, 1972). Tidal salt marshes are extensive around the estuary’s shore (Kraft et al., 1979). Chesapeake Bay and its tributaries, such as the Potomac River, the Rappahannock River (e.g., Nichols, 1974) and the James River estuaries, is probably one of the most intensely studied, major estuarine systems in the world. The tide enters the bay over a complex series of channels and shoals. Zigzag shoals are formed from

52

H. BOKUNIEWICZ

Fig. 3-2. Approximate tidal conditions along the major coastal plains.

the interaction of strong tidal currents with the construction of a submerged spit by the littoral transport impinging on the bay mouth from the north (Ludwick, 1974). The channels merge into one main channel towards the head of the estuary and extensive deposits of fine-grained sediments are found in the deeper water (Nichols and Biggs, 1985). Fluid muds, which play such an important role in many estuaries in northern Europe, have been found in Chesapeake Bay (Nichols et al., 198l), the James River (Nichols, 1985), and the Rappahannock (Faas, 1981). Near the head of Chesapeake Bay at the Susquehanna River, fluvial sands become interbedded with silts downstream (Nichols and Biggs, 1985). The estuary essentially traps all the sediment delivered to it as do its tributaries’ estuaries (Biggs and Howell, 1984). Further to the south along the Atlantic seaboard, the coastal-plain -estuaries’ access to the sea is controlled by littoral transport and the dynamics of barrier beaches. The Chowan, Roanoke, Alligator, Pamlico, Tar and the Neuse River estuaries receive their salt water from the Pamlico-Albemarle Sound, a large lagoon complex along the wave-dominated coast. Further south, the estuaries of the Pee Dee, Waccamaw, North Sante, Sante rivers, Charleston Harbor, the Saluda River, St. Helena Sound and the Broad River, all in South Carolina, are tidally dominated and have direct access to the ocean. However, vigorous littoral sand transport exerts a strong influence at their mouths. The same is true for the Georgian estuaries. Some of the Georgian estuaries are tidally drained saltmarsh that fill former valleys and have developed behind the barrier island system; these include Wasaw, St. Catherine, Sapelo, Doboy and St. Simeons sounds (Frey and Howard, 1986). Others are within the mouths of rivers; these are the Savannah River, the Ogeechee River and Ossabaw Sound, the Altamaha River and Sound, the Satilla River and St. Andrews Sound and St. Mary’s River (Frey and Howard, 1986). Coastal processes and the dynamics of barrier islands and tidal inlets continue to dominate the estuaries along the Gulf coast. On the northern Florida coast, the St.

SEDIMENTARY SYSTEMS OF COASTAL-PLAIN ESTUARIES

53

Johns River discharges into the Atlantic and the Suwannee River into the Gulf of Mexico, but neither carries a significant alluvial load (Tanner, 1985). Further to the west along the Gulf coast, prevailing, southeasterly winds drive coastal sand to the west along a series of barrier spits and islands. The tidal range is relatively small (0.6 to 1 m; Fig. 3-2) but hurricane storm surges in excess of 4 m can occur. The sediment supply from rivers draining into bays behind this coast has not been sufficient to fill their ancestral valleys completely forming the estuaries of Perdida, Mobile, Biloxi and St. Louis bays (Nummendal and Otvos, 1985). Barrier islands also dominate the Texas coast but the sediment discharge of Texas rivers has been sufficient to fill the ancestral valleys of many of them. The estuary of the Rio Grande, for example, was filled by 4,500 yr BP (McGowen et al., 1976). The Brazos, Colorado, Guadalupe, Lavaca and Navidad rivers have likewise filled the deep valleys they occupied at the end of the Pleistocene. Some coastal-plain estuaries remain, however, generally behind the barrier island system. The lower reaches of the Sabine and Neches rivers and Sabine Lake are estuaries with mud deposits being accumulated between the bayhead deltas and the coastal marine sands (McGowen et al., 1976). Trinity and Galveston bays, Vavaca, San Antonio, Copano, Corpus Christi and Baftin bays are all estuaries whose access to the sea is completely controlled by the exchange of tidal inlets through the system of barrier islands. (2) The Caribbean coastal plain of Mexico covers 125,000 km2 in Tampico, Veracruz, Tabasco and the Yucatan. In addition, there is a relatively small coastal plain covering about 28,000 km2 along Costa de Mosquitos, Nicaragua and Honduras (Colquhoun, 1968). The shoreline is dominated by barrier islands and lagoons. Although some mangrove vegetation can be found on the U.S. Florida and Gulf coasts, mangroves are found all along the coastal fringe of the Caribbean plain. The sediment discharge of rivers draining this coastal plain tend to be large corresponding to 100 to 500 metric tons of sediment/km2/yr (Fig. 3-3; Milliman and

Fig. 3-3. Classes of general sediment yield from the major coastal plains (Milliman and Meade, 1983).

54

H. BOKUNIEWICZ

Meade, 1983), so that many of the rivers have filled the valleys that they had incised during the period of low sea level during the last glaciation. When ancestral valleys are partially unfilled, the longshore transport of sand has embayed the river mouths like those estuaries along the northern Gulf coast just described (Murray, (3) The coastal plain from the Orinoco in Venezuela to the Oyapock in French Guiana covers about 120,000 km2 (Colquhoun, 1968). The spring tides here achieve a range of 3 m and the coastline is characterized by mangrove swamps, chenier ridges, and mud flats that are composed not only of sediment delivered by the local rivers but also by fine-grained sediment driven northwestkard from the Amazon. The rivers crossing the narrow (25-35 km) coastal plain in Guyana are tidal (Schwartz, 1985) as presumably are the Marowigne, Suriname, Coppenane and Corantign rivers in Surinam. The estuaries in French Guiana include the Maroni, Approuque and Oyapock river estuaries which still maintain vestiges of their ancestral drainage system even in the face of the extensive deposition of fine-grained sediment along this coast (Turenne, 1985). (4) The coastal plain dominated by the Amazon delta in Brazil covers 245,000 km2 (Colquhoun, 1968). Much of the coast here is composed of lagoons, mangrove swamp, and salt marshes although some rivers, like the Paraquacu River, discharge through an estuary (Cruz et al., 1985). Sediment yields south of the Amazon system are generally low, less than 50 metric tons/km2/yr (Fig. 3-3). The coastal plain in Argentina encompasses 270,000 km2 (Colquhoun, 1968). It is dominated by the Rio de la Plata estuary which begins at the bayhead delta of the Parana River and exhibits a low generally marshy shoreline with extensive mud flats in Sanborombon Bay near its mouth (Schnack, 1985). The latitudes are too high here for mangroves. Marshland is again extensive in the vicinity of Bahia Blanca and further south between the Rio Colorado and the Rio Negro estuaries in the Anegada Bay area (Schnack, 1985). The very low supply of fluvial sediments coupled with energetic tides with a two-meter range and the influence of prevailing north and northwest winds place Bahia Blanca in an erosional mode (Perillo and Sequeira, 1989). Little sediment is supplied by the rivers and the circulation inhibits the import of marine sediment, so that the principal sedimentary activity is the redistribution of sediment internally from the erosion of tidal flats and channel banks. Along the Patagonian coast there are few rivers that reach the sea, but, due to the predominance of coastal cliffs, only the Chubut River forms a coastal plain (Perillo et al., 1989). Its mouth is controlled by a southward littoral drift of gravels. Further south only ria-type estuaries are found except for the Carmen Sylva and Grande Rivers located on the eastern coast of Tierra del Fuego. (5) The coastal plain of northern Europe covers about 156,000 km2 at the shore of the North and Baltic Seas in Belgium, the Netherlands, Germany and Poland (Colquhoun, 1968). Glacial, unconsolidated sediment predominates along this coast and coastal dunes, barrier islands and barrier spits have developed. Along the North Sea coast, the mean tidal range can reach 4 m (Fig. 3-2) and severe storm conditions are encountered. Deeply incised channels in Belgium were largely filled during the Holocene (Eqziabeher, 1992) leaving the Yser, Ede and Zwin river estuaries to cut through

SEDIMENTARY SYSTEMS OF COASTAL-PLAIN ESTUARIES

55

young coastal dunes to the sea. The tidal range tends to decrease along the coast of the Netherlands (Jelgersma, 1985) into Germany; the estuaries of the Scheldt, the Meuse and the Rhine, Wesser and Elbe have direct access to the North Sea while the mouth of the Ems estuary enters the shallow sea partially contained by barrier islands. The estuaries of the Netherlands have been the subject of many excellent studies and much of our understanding of estuarine sedimentation has been, and continues to be, based on these studies (e.g., Postma, 1967; van Leussen, 1991). The mechanisms by which high concentrations of suspended sediment are maintained in these estuaries by tidally induced transport (e.g., Dronkers, 1984) and the role of fluid muds in estuarine sedimentation are of particular importance. Fluid muds when the rate at which particles settle to the bottom exceeds the rate at which consolidation and dewatering can occur so that layers of mud with a very high water content and very low strength are formed. Such weak sediments are sensitive to changes in current velocities. They respond quickly to tidal currents and provide a large reservoir of sediment to the estuarine waters. As a result, estuaries containing fluid muds may show significant spring-neap cycles in sediment transport. During neap tides, rapid accumulation occurs with the formation of thick, nearly stationary fluid mud layers and thin depositional lamina under them, while during spring tides, the fluid mud is redispersed, suspended concentrations increase and may be accompanied by seaward escape of sediment (Nichols and Biggs, 1985). The tidal ranges along the Baltic coast are small, 0.2 m or less. Along the Polish coast, tideless estuaries from the Oder to the Vistula, including the Lupawa, Leba and Piasnica rivers, are maintained by meteorological forcing (Jasinska, 1990). Access to the Baltic Sea is restricted by littoral processes forming spits, bars and shallow bays. Jetties guard the entrances to the Polish estuaries to keep the mouths navigable in the face of the pressure of littoral sand transport. (6) The coastal plain of Mozambique covers 130,000 km2 (Colquhoun, 1968) from the shores of Zululand in the south to past the delta of the Zambesi River. Mangrove vegetation thrives along this coast. In Zululand, the plain is narrow (20 to 40 km) and rivers crossing it from the highlands carry a high sediment load (Orme, 1973). This part of the coast is microtidal with tidal ranges up to 1.8 m (Fig. 3-2) but incident wave energy is high and the sediment delivered to the shore is driven alongshore in spits and barriers. The longshore pressure has diverted the lower stretches of the river, such as that of the Tugela River, to run parallel to the shore a distance roughly proportional to their discharge (Orme, 1973). Fluvial discharge is strongly seasonal and river mouths may be closed during the dry season by longshore drift of sand. St. Lucia is reported to be the largest estuarine system in Africa and, because of the development of two sand spits and flood and ebb tidal deltas, dredging of the estuary mouth is needed to keep it open (Wright and Mason, 1991). Deeply incised Pleistocene valleys became estuaries during the mid-Holocene rise in sea level but, because of the large sediment supply, these estuaries were filled substantially to their present condition (Orme, 1973). The tidal influence greatly increases to the north where spring tides can exceed 6 m (Fig. 3-2). Both the Zambesi and the Save rivers have a sufficient sediment discharge to create deltaic coasts with extensive mud flats and shoals exposed at low tide and fringed by mangrove swamps (Tinley,

56

H. BOKUNIEWICZ

1985). Other river mouths, however, are more or less drowned. The Punque and the Buzi rivers also carry high sediment loads across the plain but discharge into the Bight of Sofala at Beira. The Limpopo River also delivers turbid water from the uplands through a ridge of high coastal dunes into the Bight of Limpopo (Tinley, 1985). Shorter rivers draining the sandy plain of southern Mozambique, by contrast, have relatively low sediment concentrations and the mouths of these estuaries are restricted by coastal processes along the adjacent sandy beaches. (7) The extensive coastal plains of southeastern Asia pose a particular problem for the classification being considered. The coastal zone here is dominated by deltas at all the major rivers the Indus, the Cauvery, the Krishna, the Godavii, the Ganga, the Irrawaddy, the Salween, the Mekong, the Huang Ho (Yellow), and the Changjiang (Yangtze). Not all the deltas are protruding ones. Although the high rate of sediment delivery from these rivers has filled the ancestral river valleys, some have retained a funnel shape which tends to focus tidal flows. The tidally dominated Fly River estuary in Papua, New Guinea, for example, delivers 85 million tons of sediment annually through a series of channels between extensive mangrove swamps (Harris et al., 1993). The Changjiang River is another example. Tides with a range exceeding 4.6 m (Fig. 3-2) in the mouth of the Changjiang maintain a geochemical estuary between 15 and 85 kilometers long depending on the river discharge. The small, tidal rivers occupying drowned valleys, such as can be found along the eastern coast of India and the south coast of New Guinea, are not well-represented in the literature. Hangzhou Bay on the East China Sea is a notable exception in this region. The Bay is funnel-shaped but the Qiantang River which discharges into its headwaters carries a sediment load insufficient to build a delta (Jin Changmao, personal communication). The tidal range can exceed 8.9 m which keeps the vertical salinity structure homogeneous. The major source of sediment infilling Hangzhou Bay is the finegrained sediment from the Changjiang River which discharges into the coastal waters immediately north of the Bay. In many other coastal-plain estuaries, it is the littoral transport of, primarily, coarse-grained sediment that dominates sedimentation but here the transport of fine-grained sediment into the estuary through its mouth that characterizes the sedimentary system. (8) The delta plain of the Lena River in Siberia is a relic of higher discharges that went to the Laptev Sea during the last glacial maximum (Zenkovich, 1985). Whether or not a drowned delta should be classified as a coastal-plain estuary is a moot point. Like rivers that flow over ancestral valleys that have already been filled with sediment, perhaps the Lena River estuary should be considered a deltaic one. Likewise the Omoloi, Yana, Indigirka and Alazea rivers have filled their bays although only small protruding deltas have been formed. The western Siberia coastal plain is a low plateau of Quaternary deposits drained by the Tazovc, Yenisey and Ob rivers. These open into relatively shallow muddy bays of which the Obskaya Guba is the largest. These major estuarine systems are poorly represented in the available literature, however, and deserve further attention.

SEDIMENTARY SYSTEMS OF COASTAL-PLAIN ESTUARIES

57

HYDROGRAPHIC CLASSIFICATION OF ESTUARIES

Hydrodynamic classifications of estuaries have been very successfully used for comparisons. They are based on salinity distributions, however, so they are not applicable to freshwater tidal reaches which are included in the definition (Perillo, this volume). Neither are they directly relevant to the geomorphology or sedimentology of estuaries upon which the distinction of coastal-plain estuaries is made. Nevertheless, they capture fundamental characteristics of estuarine behavior and, as I will discuss later, it may be possible to link them to parameterization of the relevant sedimentological processes through energy considerations. Estuaries have been widely classified according to their salinity distribution as wellmixed, partially mixed, stratified, or salt wedge estuaries (Cameron and Pritchard, 1963). The degree of mixing is usually controlled by the tide, so that a quantitative parameter which approximately discriminates among these classes is the mixing index which is the ratio of the amount of freshwater supplied in one tidal cycle to the tidal prism (Schubel, 1971; Schultz and Simmons, 1957). For well-mixed estuaries, this ratio is typically greater than one. At values of about 0.25 or 0.1 or slightly less, the estuary is usually partially mixed and stratified estuaries are usually found at values below 0.05. The mixing index is an expedient which very generally distinguishes these classes of estuaries. Since the mixing index is intended to represent a process by which some of the energy supplied by the tides appears as a change in the potential energy of the water column, an alternative parameterization may be defined by a stratification number which is based on the rate of tidal energy dissipation (Ippen and Harleman, 1961, as cited in Ippen, 1966). The stratification parameter is calculated as a ratio of the tidal energy dissipation rate (or tidal power) and the rate of gain of potential energy per unit mass of water due to the mixing of salt and fresh water as the water moves through the estuary. The latter value is dependent on both the vertical salinity gradients and the river discharge. As determined by Ippen and Harleman (1961), a well-mixed estuary would be characterized by a stratification parameter in excess of 200 while a stratified estuary would have a stratification parameter of less than 20. A two-parameter classification was proposed by Hansen and Rattray (1966) based on the steady-state, baroclinic circulation which results from the mixing of salt and fresh water and a consequent nodal point in the near-bottom velocity field. One parameter is a circulation parameter which is the ratio of the mean surface current velocity to the average freshwater velocity through any cross-section of the estuary. Lower values of this parameter tend to describe more well-mixed estuaries. The second parameter is a stratification parameter which is a direct measure of degree of stratification defined as the ratio of the difference between the surface and bottom salinity and the average salinity over the cross-section. Well-mixed estuaries, of course, have lower stratification parameters. This classification is a phenomenological one and can be applied to both tidal and tideless estuaries. In principle, tideless estuaries could be classified in a similar way, although the mixing of salt water and fresh water is accomplished by water level differences set

58

H. BOKUNIEWICZ

up by meteorological conditions. The tidal range in the Baltic is in the range of 0.02 to 0.10 meters along the Polish coast, for example, salinity gradients persist in the Oder River (Jasinska, 1990). The Lupawa, Leba, Piasnica and Vistula rivers have similar classifiable estuaries based on salinity distribution. Tidal, freshwater reaches of estuaries; however, are irrelevant to these classifications which use salt as a tracer of the mixing process. Even in the absence of salt, however, tidal mixing remains a relevant parameter for the transport of heat, suspended solids and other dissolved constituents. A more general parameterization may rely on evaluation of the turbulent energy.

SEDIMENTOLOGICAL CLASSIFICATION OF ESTUARIES

The definition proposed by Perillo (this volume) is basically a hydrodynamic one, but the expression of hydrodynamic processes in various geologic settings provides the foundation for sedimentological classifications. The characteristic parameters in this arena describe the sediment facies and the sedimentation over geologically relevant time.

Sedimentation Coastal-plain estuaries have inherited a geomorphology from their ancestral rivers and are to be found in various stages of an incomplete process of being filled with sediment. Their present sedimentological regime is in relative disequilibrium from the geologically long-term average regime. Fairbridge (1980) proposed the difference between erosion rates on two time scales as a measure of the degree of disequilibrium. One rate is the historical annual average erosion rate derived from hydrologic conditions and the other is the mean erosion rate for the area over appropriate geologic time. Comparison of recent marine or estuarine sedimentation rates with long-term averages may be a more relevant parameter indicative of the disequilibrium (Fairbridge, 1980). The sedimentation rate should equal the rate of denudation on the appropriate time scale if the estuary is an effective trap for sediment. Gordon (1979) used this approach to calculate the denudation rate in New England (USA) from deposition rates in a large estuary (Long Island Sound). The denudation rate derived from considering estuarine sedimentation was a reasonable one and nearly equal to the present, fluvial yield, indicating a long-term (8000-year) stability in both the average denudation rate and the estuarine sedimentation. Several special conditions in this estuary contribute to its persistent ability to trap sediments. As discussed by Gordon (1979), these were (1) a rate of sea level rise exceeding the upward growth of the sediment deposit, (2) the confinement of the estuarine salinity structure entirely in the estuary’s volume, and (3) a low rate of tidal and storm energy dissipation; sediment dispersion and loss should be expected if energy levels are too high. The acceptable energy level may be raised by biological processes which agglomerate and stabilize the deposits (Gordon, 1979). Alternatively, they could be lowered by the presence of fluid mud which may respond rapidly to increases in fluid power.

SEDIMENTARY SYSTEMS OF COASTAL-PLAIN ESTUARIES

59

Sediment facies Estuaries are characterized by marine sources of sediment as well as fluvial sources. Coastal plains are soft-rock coasts and substantial, littoral transport of sediment reaches the estuary mouth (Fairbridge, 1980). The combination of tidal currents and the estuarine circulation adds both littoral sand and marine suspended sediment to the fluvial discharge within the estuary. Since they are efficient sediment traps (e.g., Meade, 1972), this mix of marine and fluvial sediments along with the imprint of the action of waves and tides characterizes estuarine facies (Dalrymple et al., 1992) In general, three zones can be distinguished depending on the dominant source of energy for doing the work of sediment transport (Dalrymple et al., 1992; Fairbridge, 1980). The estuary mouth is always dominated by marine processes. Ocean waves drive sand toward the estuary mouth. If the wave energy is dominant, the circulation at the mouth of the estuary is restricted by the development of bars, spits, or barrier beaches (Roy, 1984, Dalrymple et al., 1992, Fairbridge, 1980). If the combination of river discharge and the tidal exchange is not sufficient to maintain inlets in the face of this lateral pressure, the estuary mouth may close producing a “blind estuary”, which can only maintain its estuarine status by being a temporary state. In estuaries dominated by strong tides, complex systems of tidal sand bars form which are elongated in the direction of the principal tidal currents usually perpendicular to the trend of the neighboring shoreline (Dalrymple et al., 1992). In either case, there is a landward transport of marine sands either by the overwashing of barriers by waves, the formation of flood tidal deltas, undirectional transport due to tidal asymmetry, or the superposition of the tides on the estuarine circulation (Officer, 1981). Landward-directed cross-bedding and other indicators of flood-tidal deltas and ovenvash deposits, the cross-bedded sands of bar sequences or parallel-laminates sublittoral sands might be expected. Lower energy conditions in the central reaches of an estuary allow fine-grained deposits to form. These may be submerged estuarine muds if the rate of deposition has been insufficient to infill the ancestral river valley. Often the deposits will be bioturbated and may contain abundant plant debris (e.g., Goldring et al., 1978). They may also occur as fluid muds if the rate of settling from the dilute suspension exceeds the rate at which the material reaching the sea floor can be consolidated either by gravitational self-compaction or by biologically mediated processes (e.g., the formation of fecal pellets). Low-relief coastal-plain coasts, however, favor the development of salt marshes or mangrove swamps which become more common as the estuary matures and fills its basin (Fairbridge, 1980; Roy, 1984). Interfingering of the sand and mud facies and rapid changes in facies both vertically and horizontally could be other indicators of the estuarine environments (Goldring et al., 1978). Near the head of the estuary, fluvial deposition predominates. Deltaic sequences may appear at the estuary head but some impression of the reversing tidal action can be preserved sedimentary structures. Saline or brackish fauna might also be found intermingled or interbedded with sediments showing their terrestrial sources with abundant plant debris (Goldring et al., 1978). A typical erosional surfaces, upward fining sequences, mud pebbles or many other indicators of flood events, interbedded

60

H. BOKUNIEWICZ

with tidally cross-bedded sediments or mud-draped surfaces could also be consistent with estuary-head conditions.

Sediment dynamics The schemes for classifying estuaries based on the hydrography have a proven utility. In principle, these fairly well-qualified parameters should also be relevant to the qualitative description of characteristic estuarine facies through models of sediment transport. Of course, the physics of sediment transport in estuaries has attracted its own well deserved attention (e.g., Officer, 1981; Dyer, 1986). There has been little attempt, however, to develop a classification scheme based on sediment dynamics. From a review of the venue of coastal-plain estuaries, there would seem to be at least three other important characteristics that need to be quantified to describe the state of the estuarine sedimentary system. These are the pressure of the littoral sand transport at the estuary mouth (Fairbridge, 1980), the presence or absence of fluid mud, and the trapping efficiency. Zapping eficiency The trapping efficiency of an estuary is one parameter of its sediment dynamics that has both been quantified and used for estuarine comparisons. An empirical relationship between the trapping efficiency and a “capacity-inflow” index originally developed to estimate the ability of man-made impoundments to trap sediments (Biggs and Howell, 1984). The capacity-inflow index is the ratio of the water volume capacity of a reservoir to the total water inflow (Bruun, 1953). This appears to be a useful expedient for US. coastal-plain estuaries (Fig. 3-4) even though it does not account for tidal variations, biologically mediated sedimentation processes (Biggs and Howell, 1984), or the other processes that cause estuaries to trap fine-grained sediment.

-

P

75

-

0.01

0.1

1

10

Capacitylln flow (yr ’)

Fig. 3-4. The capacity-inflow index for quantifying the trapping efficiency of estuary (Biggs and Howell, 1984). “The heavy line represents the best fit and the lighter lines represent the envelope that encloses the C j l ratio of 40 impoundments whose trapping efficiency was measured. Similar data, using MLW volume for C and potential runoff for I , along with measured trapping efficiency, are presented for Chesapeake Bay ( I ) , Rappahannock River (Z), Choptank River (3), James River ( 4 ) , and Mobile Bay (5)” (Biggs and Howell, 1984).

SEDIMENTARY SYSTEMS OF COASTAL-PLAIN ESTUARIES

61

The landward flux of energy and material is essential to the existence of an estuary. The penetration of tidal energy is a fundamental part of the definition and the characteristic estuarine sediment facies include marine sediments. For sediment particles, the landward flux at the estuary mouth is not a boundary condition; the net import is the difference between outward advection at the surface dispersion and the inward flux at the sea floor, so the import depends on the internal conditions. Parameters that describe this process would be basic ones. For suspended sediments, the concentrations inside the estuary are usually higher than those in the neighboring sea, so that there is a dispersion pressure to export material. This dispersion is augmented by a net outward flow of surface water and sediment and counteracted by the inflow of bottom water and sediment. Schubel and Carter (1984) used a two-layer box model representation including these processes to calculate the flux of suspended sediment across the estuary mouth and quantify a condition discriminating export from import. The flux of suspended sediment was:

[

QZ 1-

(Ac/F)(l+u*) - U* (As/S)(1+ u ) - u

(3-1)

where Q is the river discharge, S and F are, respectively, the average concentration of suspended sediment and the salinity in the lower layer, Ac and As are the differences in the concentration and salinity between the surface and bottom layers, u and u* is the fraction of the total seaward flux of salt and suspended sediment, respectively, that is balanced by dispersion. Whether sediment is imported or exported is determined by whether this expression is greater than or less than zero and the trapping efficiency is generally related to the hydrographic class of the estuary (Fig. 3-5). Conceptually, Eq. (3-1) can be moved closer to the physical parameters used in hydrographic classification though an energy-based approach. Energy is a convenient

t

I-

2-

I 2 TYPED TYPEB (SECTIONALLY (PARTIALLY HOMOGENEOUS) MIXED)

- 10,000

3

4

TYPE A (SALT WEDGE)

Fig. 3-5.The filtering efficiency of an estuary related to hydrographic classification (see text; Schubel and Carter, 1984). The “filter efficiency” is one minus the ratio of the suspended sediment flux (see text) to the fluvial suspended sediment input. The ratio of the surface water velocity ( u s )to the average is a measure of the strength of the gravitational circulation velocity of freshwater through a section (UF) (Hansen and Rattray, 1966).

62

H. BOKUNIEWICZ

and appropriate parameter for many reasons. It is a scaler, it is well defined, and it is conservative. In addition, classic sediment transport formulae rely on a fundamental relationship between the fluid power and the rate of sediment transport, and, for trapping to occur, the specific power dissipation in the estuary must be low enough to permit deposition (Gordon, 1981). In expression (3-1) As/S is related to the stratification number (Ippen and Harleman, 1966) or the ratio of the rate to tidal dissipation to the power used in mixing salt and fresh water which appears as a gain in the potential energy of the water column needed to maintain the salinity gradient. Ac/E can be similarly expressed in terms of the fluid energy. Although the vertical distribution of suspended sediment is usually calculated as a mass balance, it can also be described by an energy balance (Velikanov, 1955, as cited by Yalin, 1972) in which the concentration gradients can be calculated using the ratio of the specific vertically integrated power per unit weight to the specific fluid power (i.e., the fluid power per unit area).

Littoral sand transport The efficiency of coastal-plain estuaries as sediment traps leads to the characteristic fine-grained deposits in their central basins. The indicative sand facies at their mouths record the role of marine agents in the infilling of the ancestral valleys and the morphology of these deposits controls the access of the tides and salt water to the estuary’s interior. The character and strength of waves attacking the open coast (Fig. 3-6) define a pressure worlung to block the estuary mouth. Barrier beaches, spits and islands are found on all the major coastal plains (Fig. 3-6). In some coastal-plain estuaries, especially in the tropics, the seasonality of wave driven sand

Fig 3-6 Approxlmate distribution of barrier beaches (I) along the major coastal plains and general characteristics of the distribution of wave activity which control the littoral pressure to block the mouths of coastal-plain estuaries (Snead, 1980, Hayes, 1979, and Walker, 1975, as reported in Fairbridge, 1980).

SEDIMENTARY SYSTEMS OF COASTAL-PLAIN ESTUARIES

63

transport and the river discharge cause them to be isolated from the sea for part of the year; their existence as estuaries depends on the transient stability of their inlets. Inlet stability is an important topic in coastal engineering. As early as 1931, O’Brien published an empirical relation defining the tidal prism needed to maintain inlets against the pressure of littoral sand transport to fill them. The ratio of the tidal prism to the mean annual amount of littoral drift was proposed an indicator of inlet stability with values over 300 indicating a high degree of stability (Bruun and Gerritsen, 1960). In many estuaries, the fluvial discharge is also important in helping to maintain the estuaries’ access to the sea and a more fundamental parameter of inlet stability is the ratio of the gross supply of littoral sand to the inlet to the maximum rate of sand transport through the inlet, regardless of the driving agent (Battjes, 1967). Both of these quantities can be related to the energetics at the inlet mouth. The transport of sand has variously been quantified in terms of the fluid power (Bagnold, 1963) and the longshore drift of sand is usually forecast using the incident wave power (US. Army Corps of Engineers, 1977). It is conceivable that the formation of coastal sand dunes may also contribute to the pressures against which the estuary must contend to remain open. Dunes tend to accumulate vertically, but the volume of many coastal-dune systems testify to their importance on coastal sediment budgets (Goldsmith, 1989). It may not be unreasonable, therefore, to expect this process to have an impact in some estuaries. Information is not available, however, to assess the importance of this process.

Fluid muds The topic of fluid muds introduces a range of parameters relating to the deposition of fine-grained sediment. Fluid muds can be generated in two ways. They occur when the settling flux, that is, the product of the suspended sediment concentration and the settling velocity, is greater than the rate at which the particles are incorporated into the sea floor either by compaction (Parker, 1989) or biological processing (Gordon, 1981). They can also be generated when the resuspension rate of bed material is greater than the near-bed upward entrainment flux (Ross and Mehta, 1989). The issue of fluid muds, therefore, introduces the parameters describing the deposition rate, the resuspension rate, the settling flux (with a distinction between newly introduced particles and particles being recycled by resuspension), and biological processing. The physical fluxes are interrelated; the deposition rate should be the difference between the settling flux at the sediment water interface and the resuspension rate. Biological packaging of sediment particles can effect the fluxes by increasing the settling velocity, actively incorporating particles from suspension into the permanent deposits or altering the rates of resuspension. All of these values, however, are exceedingly difficult to measure and impossible to predict with certainty. They are also extremely variable so that any attempt to use them to describe the state of the estuarine sedimentary system must necessarily deal with averaged rates over comparable periods. This is not usually the case. Sediment traps, for example, may measure the settling flux over periods of days to months, while most resuspension rates are usually measured over a tidal cycle

64

H. BOKUNIEWICZ

at most, the biological processes are often dominated by seasonal variations, but are also punctuated by rapid, but short-lived, changes in production, and deposition rates are typically measured over time period of years or decades in estuaries. In the face of all the difficulties with the use of these parameters for classification and comparison, it is, perhaps, premature to be concerned with expressing them on the basis of their energetics. It is noteworthy, however, that Bagnold's theory of autosuspension equates the fluid power through an efficiency factor to the work needed to maintain the suspension against settling. This work is directly proportional to the settling flux.

SUMMARY

Even a brief survey of the settings of coastal-plain estuaries emphasizes the need to expand studies of many regions that are underrepresented in the literature. An understanding of coastal-plain estuaries may be biased by estuaries along the east coast of the US. and in the Netherlands. The information-base needs to be further expanded to a widened range of climates, hydrologic and geomorphic settings. This process can be expedited by continued development of a classification scheme for the sedimentary system of coastal-plain estuaries. The classification of estuaries based on hydrological parameters is well developed, has been widely applied, and has proven its usefulness. To bridge the gap between the oceanographic characterization of estuaries and facies models, the state of the estuarine sediment transport system must be defined, including (a) the stability of sand deposits at the estuary mouth, (b) the fluxes of fine-grained sediment at the estuary floor, and (c) the trapping efficiency of the estuary. The state variables that could be used to quantify the transport system could be: (1) the rate of tidal energy dissipation, (2) the rate of wave energy dissipation, (3) the power used in mixing, (4) the power needed to maintain the distribution of suspended sediment, (5) the settling flux (which would be calculated from the concentration of suspended sediment and the settling velocity or measured directly), (6) the deposition rate, (7) the resuspension rate (the power devoted to resuspension and transport), (8) the rate of longshore transport (or alternatively the incident wave power, (9) the rate of sand transport through the estuary mouth (which might involve a determination of the fluid power), (10) the rate of biological processing. For the purposes of comparisons, these parameters would have to be determined by some widely accepted method, especially since many of the estimates must involve the use of empirical constants and, because of the inherent variability in the processes, some averaging intervals must be chosen. This cannot be done over a wide range of coastal-plain estuaries at this time, but the search for such parameters would assist efforts to compare and contrast estuarine systems.

SEDIMENTARY SYSTEMS O F COASTAL-PLAIN ESTUARIES

65

REFERENCES Bagnold, R.A., 1963. Mechanics of marine sedimentation. In: M.H. Hill (Editor) The Sea. Interscience NY: VOI 3,507-528. Battjes, J.A., 1967. Quantitative research on littoral drift and tidal inlets, In: G.H. Lauff (Editor) Estuaries, AAAS, Washington, D.C., Publication No. 83, pp. 185-190. Biggs, R.B. and Howell, B.A., 1984. The estuary as a sediment trap: alternate approaches to estimating its filtering efficiency. In: V.S. Kennedy (Editor) The Estuary as a Filter, Academic Press, NY, pp. 107-129. Brunn, P. and Gerretsen, F., 1960. Stability of Coastal Inlets. North Holland Publishing Co., Amsterdam, 123 pp. Bruun, G.M., 1953. Trap efficiency of reservoirs. "Ians. Am. Geophys. Union, 34: 407-418. Cameron, W.M. and Pritchard, D.W., 1963. Estuaries. In: M.N. Hill (Editor) The Sea, Interscience NY, Vol. 2: 306-324. Cruz, O., Coutinho, P.N., Duarte, G.M., Gbmez, A. and Muehe, D, 1985. Brazil. In: E.C.F. Bird and M.L. Schwartz, (Editors), The World's Coastline. Van Nostrand Reinhold Co., NY, pp. 85-89. Colquhoun, D.J., 1968. Coastal plains. In: R.W. Fairbridge (Editor) The Encyclopedia of Geomorphology. Reinhold Book Corporation, NY, pp. 144-150. Curray, J.R., 1969. Estuaries, lagoons, tidal flats and deltas. In: D.J. Stanley (Editor), The New Concepts of Continental Margin Sedimentation: Application to the Geologic Record. Am. Geol. Inst., Washington, D.C., JC-111-1-JC-111-30. Dalrymple, R.W., Zaitlin, B.A. and Boyd, R., 1992. Estuarine facies models: conceptual basis and stratagraphic implications. J. Sediment. Petrol., 62: 1130-1146. Davies, J.L., 1980. Geographical Variation in Coastal Development, Longman. Dronkers, J., 1984. Import of fine marine sediment in tidal basins. Neth. Inst. Sea Res., Publ. Series 10: 83-104. Dyer, K.R., 1986. Coastal and Estuarine Sediment Dynamics. Wiley Interscience, NY., 342 pp. Emery, K.O. and Uchupi, E., 1972. Western Atlantic Ocean: Topography, rocks, structure, water, life and sediments. AAPG Mem., 17,532 pp. Eqziabeher, TG., 1992. The Holocene coastal plain evolution of Lo Area, Southwestern part of Belgium. Master's Essay, Free University of Brussels. 76 pp. Faas, R.W., 1981. Rheological characteristics of Rappahannock Estuary muds. U.S. Int. Assoc. Sedimentol., 5: 505-515. Fairbridge, R.W., 1980. The estuary: its definition and geodynamic cycle. In: E. Olausson and I. Cat0 (Editors), Chemistry and Biogeochemistry of Estuaries. John Wiley and Sons Ltd., London, pp. 1-35. Frey, R.W. and Howard, J.D., 1986. Mesotidal estuarine sequences: A perspective from the Georgia Bight. J. Sediment. Petrol., 56: 911-924. Goldring, D.W., Bosence, J. and Blake, T, 1978. Estuarine conditions in the Eocene of Southern England. Sedimentology, 25: 861-876. Goldsmith, V., 1989. Coastal sand dunes as geomorphological systems. Proc. R. SOC.Edinburgh., 96B: 3-15. Gordon, R.B., 1979. Denudation rate of central New England determined from estuarine sedimentation. Am. J. Sci., 278: 632-642. Gordon, R.B., 1981. Estuarine power and trapping efficiency. In: River Inlets to Ocean Systems. United Nations Publication, pp. 86-91. Hansen, D.V. and Rattray, M., 1966. New dimension in estuary classification. Limnol. Oceanogr., 11: 319-326. Harris, P.T., Baker, E.K., Cole, A.R. and Short S.A., 1993. A preliminary study of sedimentation in the tidally dominated Fly River Delta, Gulf of Papua. Cont. Shelf Res., 13: 441-472. Hayes, M.O., 1979. Barrier island morphplogy as a function of tidal and wave regimen. In: S.P. Leatherman (Editor), Barrier Islands from the Gulf of St. Lawrence to the Gulf of Mexico. Academic Press, NY, pp. 1-27.

66

H. BOKUNIEWICZ

Ippen, A.T., 1966. Salinity intrusion in estuaries. In: A.T. Ippen (Editor), Estuary and Coastline Hydrodynamics. McGraw Hill, NY, pp. 598-629. Ippen, A.T. and Harleman, D.R.F., 1961. One-dimensional analysis of salinity intrusion in estuaries. Committee on Tidal Hydraulics U S . Army Corps of Engineers, Waterways Experiment Station, Vicksburg, MS, Tech. Bull. No. 5. Ippen, A.T. and Harleman, D.R.F., 1966. Tidal dynamics of estuaries. In: A.T. Ippen (Editor), Estuary and Coastline Hydrodynamics. McGraw Hill, NY, pp. 493-545. Jasinska, E., 1990. Salt intrusion in tideless estuaries. Proc. 22nd Conf. Coastal Eng., Delft, pp. 28652879. Jelgersma, S.,1985. Netherlands. In: E.C.F. Bird and M.L. Schwartz (Editors), The World’s Coastline. Van Nostrand Reinhold Co. NY, pp. 343-352. Knebel, H.J., Fletcher, C.H. and Kraft, J.C., 1988. Late Wisconsinan-Holocene paleogeography of Delaware Bay; a large coastal plain estuary. Mar. Geol., 83: 115-133. Kraft, J.C., Allen, E.A., Belknap, D.F., John C.J. and Maurmeyer, E.M., 1979. Processes and morphologic evolution of an estuarine and coastal barrier system. In: S.P. Leatherman (Editor), Barrier Islands from the Gulf of St. Lawrence to the Gulf of Mexico, Academic Press, NY, 149-183. Ludwick, J.C., 1974. Tidal currents and zig-zag sand shoals in a wide estuary mouth. Geol. SOC.Am. Bull., 85: 717-726. McGowan, J.H., Brown, L.F., Evans, T.J., Fisher, W.L.and Groat, C.G. (Editors), 1976. Environmental Geologic Atlas of the Texas Coastal Zone. Bureau of Economic Geology, University of Texas at Austin, 7 volumes. Meade, R.H., 1969. Landward transport of bottom sediments in estuaries of the Atlantic coastal plain. J. Sediment. Petrol., 39: 222-234. Meade, R.H., 1972. Transport and deposition of sediments in estuaries. In: B. W. Nelson (Editor), Environmental framework of coastal plain estuaries. Geol. SOC.Am. Mem., 133: 91-120. Milliman, J.D. and Meade, R.H., 1983. World-wide delivery of river sediment to the oceans. J. Geology 91: 1-21. Murray, G.E., 1961. Geology of the Atlantic and Gulf Coastal Province of North America. Harper and Brothers, NY. 692 pp. Nichols, M.N., 1974. Development of the turbidity maximum in the Rappahannock estuary, Summary. Mem. Inst. Geol. Bassin d’Aquitaine, 7: 19-25. Nichols, M.N., 1985. Fluid mud accumulation process in an estuary. Geo-Mar. Lett., 4: 171-176. Nichols, M.N. and Biggs, R.B., 1985. Estuaries. In: R.A. Davis Jr. (Editor), Coastal Sedimentary Environments. Springer-Verlag, NY, pp. 77-186. Nichols, M.N., Harris, R. and Thompson, G.S., 1981. Significance of Suspended Trace Metals and Fluid Mud in Chesapeake Bay. EPA Report No. R806002-01-1, Annapolis, MD, pp. 1-129. Nummendal, D., Oertel, G.F., Hubbard, D.K and Hine, A.C., 1977. Tidal inlet variability - Cape Hatteras to Cape Canaveral. Coastal Sediments ’77. ASCE Charleston, S.C., pp. 543-562. Nummendal, D. and Otvos, E.G., 1985. Mississippi and Alabama. In: E.C.F. Bird and M.L. Schwartz (Editor), The World’s Coastlines. Van Nostrand Reinhold Co., NY, pp. 155-162. O’Brien, M.P., 1931. Estuary tidal prisms related to entrance areas. Civil Eng., 1: 738-739. Officer, C.B., 1981. Physical dynamics of estuarine suspended sediments. Mar. Geol., 40: 1-14. Oostdam, B.L. and Jordan, R.R., 1972. Suspended sediment transport in Delaware Bay. In: B.W. Nelson (Editor), Environmental Framework of Coastal Plains Estuaries. Geol. SOC.Am. Mem., 133: 143-150. Orme, A.R., 1973. Barrier and lagoon systems along the Zululand coast, South Africa. In: D.R. Coates (Editor), Coastal Geomorphology. State University of New York, Binghamton, pp. 161-180. Parker, W.R., 1989. Definition and determination of the bed in high concentration fine sediment regimes. J. Coastal Res., 5: 175-184. Perillo, G.M.E. and Sequeira, M.E., 1989. Geomorphologic and sediment transport characteristics of the middle reach of the Bahia Blanca Estuary (Argentina). J. Geophys. Res., 94: 14,351-14,362. Perillo, G.M.E., Piccolo, M.C., Scapini, M.C.and Orfila, J., 1989. Hydrography and circulation of the Chubut River estuary (Argentina). Estuaries, 3: 186-194.

SEDIMENTARY SYSTEMS O F COASTAL-PLAIN ESTUARIES

67

Postma, H., 1967. Sediment transport and sedimentation in the marine environment. In: G.H. Lauff (Editor), Estuaries. AAAS Washington, D.C., Publ. 83, pp. 158-186. Ross, M.A. and Mehta, A.J., 1989. On the mechanics of lutoclines and fluid muds. J. Coastal Res., 5: 51-61. Roy, P.S., 1984. New South Wales Estuaries: their origin and evolution. In: B.G. Thom (Editor), Coastal Geomorphology in Australia. Academic Press, Orlando, pp. 99-121. Schnack, E.J., 1985. Argentina. In: E.C.F. Bird and M.L. Schwartz (Editors), The World’s Coastline. Van Nostrand Reinhold Co., NY, pp. 69-78. Schubel, J.R., 1971. Estuarine circulation and sedimentation. In: J.R. Schubel (Editor), The Estuarine Environment: Estuaries and Estuarine Sedimentations. Am. Geol. Inst. Short Course Lecture Notes, Washington, D.C. Schubel, J.R. and Carter, H.H., 1984. The estuary as a filter for fine-grained suspended sediment. In: V.S. Kennedy (Editor) The Estuary as a Filter. Academic Press, NY, pp. 81-105. Schubel, J.R. and. Meade, R.H., 1977. Man’s impact on estuarine sedimentation. In: Estuarine Pollution Control and Assessment, Proc. Conf. Vol. 1. U S . Government Printing Office, Washington, D.C., pp. 193-209. Schultz, E.A. and Simmons, H.B., 1957. Fresh water-salt water density currents, a major cause of siltation in estuaries. Committee on Tidal Hydraulics U.S. Army Corps of Engineers, WES, Vicksburg, MS, Tech. Bull. 2, 28 pp. Schwartz, M.L., 1985. Guyana. In: E.J. F. Bird and M.L. Schwartz (Editors), The World’s Coastline. Van Nostrand Reinhold Co., NY, 103-104. Snead, R.E., 1980. World Atlas of Geomorphic Features. Robert E. Krieger Publishing Co., Inc., Huntington, NY and Van Nostrand Reinhold Co, NY., 301 pp. Tanner, W., 1985. Florida. In: E.C.F. Bird and M.L. Schwartz (Editors), The World’s Coastline. Van Nostrand Reinhold Co., NY, pp. 163-167. Tinley, K.L., 1985. Mozambique. In: E.C.F. Bird and M.L. Schwartz (Editors), The World’s Coastline. Van Nostrand Reinhold Co., NY, pp. 669-677. Turenne, J.F., 1985. French Guiana. In: E.C.F. Bird and M.L. Schwartz (Editors), The World’s Coastline. Van Nostrand Reinhold Co., NY, pp. 93-97 U.S. Army Corps of Engineers, 1977. Shore Protection Manual. CERC Rept. 008-022-00113-1,262 pp. van Leussen, W., 1991. Fine sediment transport under tidal action. Geo-Mar. Lett., 11: 119-126. van Leussen, W. and van Velzen, E, 1989. High concentrations suspensions. Their origin and importance in Dutch estuaries and coastal waters. J. Coastal Res., 5: 1-22. Walker, H.J., 1975. Coastal morphology. Soil Sci., 119: 3-19. Wright, C.I. and T.R. Mason, 1991. Sedimentary environment and facies of St. Lucia Estuary Mouth, Zululand, South Africa. J. Afr. Earth Sci., 11: 411-426. Yalin, M.S., 1972. Mechanics of Sediment Transport. Pergamon Press, NY, pp. 290. Zenkovich, V.P., 1985. Arctic USSR: In: E.C.F. Bird and M.L. Schwartz (Editors), The World’s Coastline. Van Nostrand Reimhold Co., NY, pp. 863-869.

This Page Intentionally Left Blank

Geomorphologv and Sedimentologv of Estuaries. Developments in Sedimentologv 53 edited by G.M.E. Perillo 0 1995 Elsevier Science B.V. All rights reserved.

69

Chapter 4

GEOMORF’HOLOGY AND SEDIMENTOLOGYOF RIAS PATRICE CASTAING and ANDRE GUILCHER

DEFINITION AND INCLUDED AREAS

The word ria, which is of popular use in Galicia, Asturias and the Basque country, north of the Iberian Peninsula, has been introduced in the general geomorphological literature by von Richthofen (1886). Followingvon Richthofen, rias are mountainoussided estuaries that are not glaciated, thus not fjords, but are subaerially eroded, former river valleys that have been drowned by Holocene rise of sea level. Davis (1915) proposed the term ria for “...any broad or estuarine river mouth, and not necessarily an embayment produced by the partial submergence of an open valley in a mountainous coast, in the sense that von Richthofen originally proposed”. Perillo (1989, this volume) distinguished between Coastal Plain Estuaries and Rias. Both are “Former Fluvial Valleys formed by sea flooding of Pleistocene-Holocene river valleys during the last post-glacial transgression”. According with their coastal relief, Perillo subdivided them in two categories: “Coastal Plain Estuaries normally occupy low relieve coasts produced mainly by sedimentary infilling of the river(s); Rias are former river valleys developed in high relief coasts”. These classifications are based only on geomorphological criteria and not from hydrodynamics and sedimentology. In the writers’ opinion, the word ria must be restricted, as a general rule, outside the Iberian Peninsula, to Brittany in France, Devon and Cornwall in the British Isles, Korea, parts of the Chinese and the Argentina coasts (Fig. 4-1). So, estuaries in coastal plains or low areas such as the Gironde in France and Chesapeake Bay in the United States need to be excluded. However, the writers sug-

Fig. 4-1. Distribution of rias in the world. Scattered sharms outside the Red Sea are not shown.

70

P. CASTAING AND A. GUILCHER

gest to include the drowned valleys bearing coral reefs and called sharms on the Red Sea shores, which have a strong resemblance with the rias. The valleys which were deeply cut around the Mediterranean and especially in southern France during the huge Messinian lowering of that sea, and subsequently filled up, will be also included.

REGIONAL DESCRIPTION

Northwestern and northern coasts of the Iberian Peninsula (Spain) The longest stretch of coasts where rias exist in Europe is found in the Atlantic region of the Iberian Peninsula, from Vigo near the Portuguese border, and in the Cantabric from Cab0 Ortegal to the Basque country on the French border (Fig. 4-2). The numerous drowned valleys which occur there are cut into high hills, plateaus or mountains, often several hundreds of metres high in the vicinity of the sea, but without any Pleistocene glacial influence in the morphology of the lower courses of the valleys, although glaciers have existed and were very efficient inside the country in the Cordillera Cantabrica (Picos de Europa, 2648 m): so that the drowned lower courses of the valleys are always quite different from fjords. The rocky material ranges from Palaeozoic to Tertiary, and includes in the Asturias, north Spain, large limestone outcrops which have been intensively karstified. The rias which have been most accurately investigated and described are those occurring in Galicia, which include in the southwest the "rias bajas" (Vigo, Pontevedra, Arosa, Muros y Noya, the best known ones); more to the north, the "rias centrales" (La Coruiia, Betanzos, Ares, El Ferrol); and, in northeast, three others (Cedeira, Ortigueira, El Barquero). From research by Scheu (1913), Birot and Sole Sabaris (1954), Nonn (1964,1966), Pannekoek (1966, 1969), Rey (1993) and others working with them or separately, in Miocene times broad valleys, related to pre-existing fault lines, existed in Galicia at the sites of the present river valleys, and were flanked by mountains. Dry land

.

.

I "b . Y",O

.

.

.

.

.

.

.

l . p . l n .

.

" . - r ' Y ' "

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

. .

.

.

.

.

.

.

.

.

$'

.

r.

.

.

.

.

.

.

.

.

.

.

1.

.

.

.

.

.

.

.R

.

.

.

.

.

.

.

.

.

.

.

.

.p

.

.

*ynaLcL

,A'

1

/.s' y p.s

.

.

.

.

.

.

.

.

.

.

R

.

.

.

.

.\

.

.

.

. .

.

.

.

. '

.

.

.

.

i..*m

.

.

. .

.

.

. .

.

.

.

.

.

.. '

.

.

.

.

.

. '

.

.

Fig. 4-2. General distribution of rias in the northwestern and northern parts of the Iberian Peninsula. In Asturias and Pais Vasco, many small rias are not shown.

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS

71

is assumed to have then extended farther westwards than now. Later on, a new erosional wave penetrated into the lower courses of the valleys, and these deepened valleys are thought to have reached their present-time width during the Pliocene. Still later on, Pleistocene erosion brought them down to their present depth, lower than present-time sea level, during glaciations. Thus, as a result of these sea-level shifts, the valleys were alternately cut and partly filled. Periglacial processes have been recognized in many places down to sea level. Along the outer coast between the rias (Nonn, 1966), remnants of Pleistocene beaches, assumed to be Eemian (or perhaps Holsteinian?) point to interglacial sea levels standing at a few metres above the present one. In the ria of Muros y Noya, seismic profiles by Herranz and Acosta (1984) (also: Acosta and Herranz, 1984) have shown, below the Holocene muds and sands, several erosional surfaces interpreted as events related to successive Pleistocene cold periods. The Tardi- and Postglacial rise of sea level, probably accompanied by minor oscillations which can continue in present time (Balay, 1956), resulted in a sedimentation of various types in the drowned lower courses of the rivers; these sedimentary types, their origin, and the related landforms are an essential part of the Iberian ria geomorphology. Figure 4-3 gives, according to Nonn (1966), the general distribution of sediments in the Rias Bajas, with consideration of sedimentological investigations by Margalef (1958), Parga Pondal and Perez Matos (1954), Sainz Amor (1962), which have been checked and precised by more recent work in a number of rias by Vilas (1983), Vilas and Nombela. (1985), Nombela et al. (1987), Junoy and Vieitez (1989), Rey (1993). If the outer reaches of the rias are left out, the distribution of the grain size shows quite generally an increasing gradient of the silt-clay fraction toward the inner part of the rias (the clay found on the continental shelf being left out). In the outer reaches, and, more generally, in parts of the rias exposed to the surf at high tide, sand predominates with beaches, spits and sometimes dunes. E.g., at Ria de Ortigueira, sand banks can become extensive, with a partly outer, marine origin, the shelly fraction being high sometimes. In the outer part of Ria de Ribadeo, pebbles appear, more or less worn according to their exposure to the surf. In the inner reaches, on the contrary, the main feature is the wide distribution of mud, with a typical distinction between low marshes (corresponding to the so-called slikkes of the Flanders) with numerous meandering tidal creeks; and high marshes (the so-called schorres of the Flanders) bearing an herbaceous vegetation and covered by the sea only at high spring tides; e.g. at La Coruiia (Nonn, 1966, fig. p. 362) and Betanzos (Nonn, 1966, figs. pp. 366-367). The mud in rias derives mostly from inland material and not from the sea, since it includes kaolinite which comes from Miocene alterations of inland rocks and clay minerals also found locally (Nonn, 1966). In Ria de Arosa, according to Dutch research (Arps and Kluyver, 1969), the heavy mineral composition reflects well the composition of the bedrock, especially in the deeper inlets of the ria. On the more exposed parts, the content is only a little more varied and shows a weak longshore transport. The detailed study of Ria de Vigo by Vilas (1983), Nombela et al. (1987), Rey (1993) has allowed to map a typical example of this distribution of sediments (Fig. 4-4). At the boundary between Galicia and Asturias, the Ria de Foz and the Ria de

72

P. CASTAING AND A. GUILCHER

Fig. 4-3. General distribution of sediments in the Galician rias (according to Nonn, 1966).

Eo (Ribadeo) give other examples of such a distribution (Nonn, 1966;Asensio Amor, 1960, and other papers; here Fig. 4-5). More eastwards, many other rias exist in the provinces of Asturias, Cantabria and in the Basque Country (Pais Vasco). In E Vilas opinion (pers. commun.), those are not rias, but typical estuaries if we attend to hydrodynamics and sedimentology. For example the oceanic influence is very limited compared with those from Galicia (N.W. of spain). However, based on geomorphological criteria they are typical rias. As far east as Castro Urdiales on the Vasco-Cantabric border, a large part of the

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS

73

medium sand

Fig. 4-4.Distribution of sediments in Ria de Vigo, Galicia (redrawn from Nombela et al., 1987).

rias is cut into wide planation surfaces called rasas, on which an extensive literature exists (principally: Barrois, 1882; Hernandez-Pacheco, 1950; Hernandez-Pacheco and Asensio Amor, 1960; Llopis Llado, 1956; Guilcher, 1974; Mary, 1967,1979; Mary and Medus, 1971). These rasas, defined by Hernandez-Pacheco as erosion surfaces cut inland into rugged mountains and ending seawards into high, steep cliffs, include in many places more than one step; they are cut into various rocks including Palaeozoic, Mesozoic and Cenozoic limestones and marls, with deep karstification in limestones, especially in the vicinity of the sea where dolines, invaded by the postglacial transgression, coexist with rias and make their pattern complicated. From the study of geomorphology and alterites found on rocks, Mary (1979, p. 180) distinguishes three planation levels, at 260-168 m, 155-100 m and 100-60 m, which he dates from three transgressions: Aquitanian, Lower Pliocene and Lower Pleistocene. Such is the general pattern into which the rias of this area were progressively cut, with Upper Pleistocene beaches found on the sea side at La Franca and at Castro Urdiales as in Galicia (Guilcher, 1955a, 1972) and pointing to the last sea level shifts before the postglacial transgression. These rias have thus had a complex history. More to the east, in the Basque country (Pais Vasco), in the surroundings of Bilbao, Ondarroa, San Sebastian, Pasajes, rasas disappear, but a number of rias occur again, cut into high hills or mountains several hundreds of metres high, at Zumaya, Lequeitio, Ondarroa, Bermeo (Ria Mundaca), etc. (Fig. 4-6). A general feature common to these northern rias, cut or not into rasas, is that their lower courses are currently filled at their mouths by a considerable amount of sand, which currently forms spits or bars, often bearing dunes, e.g. at the AvilCs ria near Gi-

74

P. CASTAING AND A. GUILCHER

Fig. 4-5. Sediments in Ria de Foz, northeastern Galicia (according to Nonn, 1966, and Junoy et al., 1987).

jon, at Agiiera and Ason west of Bilbao, at Cuchia and Rio de Pas west of Santander. An example of what can happen in rias on the Cantabric coast with sand spit growths is the situation and evolution in the Santander harbour (Losada et al., 1991). This harbour is located at the mouth of a (former) ria which has been largely sedimented by wide mud flats. These mud flats grew behind a spit, El Puntal, which deflects the channel to the west; its evolution can be followed since 1730. In such a situation, the mouth of the river hardly deserves the name of ria, since it tends to become merely a lagoon, with a lateral channel artificially preserved for the access to Santander harbour. The sand which invades these rias is considered to derive, in its siliceous fraction, from the erosion of Cretaceous rocks outcropping behind the coastal limestones (or closer to the coast where limestones are absent): it was transported by rivers to the continental shelf during the Pleistocene regressions, and carried again shorewards during the Holocene transgression. It continues now to be pushed into the rias by the powerful surf of Mar Cantabrico (Bay of Biscay). It has thus initially a continental

GEOMORPHOLOGY AND SEDIMENTOLOGY O F RIAS

75

Fig.4-6.Ria Mundaca, Basque country, located on Fig.4-2,cut into high hills (photo by A. Guilcher, 1972).

origin, but a large calcareous fraction has been added, with marine shells which currently form 25 to 35% of the sand. In the inner reaches of the rias, sand is, as usually, replaced by mud, probably of continental origin, with the usual superposition of vegetated high marshes and bare low marshes, e.g. in the Ria de GuernicaMundaca studied by Cuevas, 1990. The zonation at San Vicente de La Barquera, Cantabria, is particularly handsome (Gonzalez Lastra and Gonzalez Lastra, 1984).

Brittany (France) Brittany, which forms the western part of the Armorican Massif, bears on its northern, western and southern coasts small estuaries cut into plateaus which deserve everywhere the name of rias, except for the Loire estuary in the southeast where the general topography is lower. The Breton word aber is still or has been in use for many of them, especially in the northwest: Aber Wrac’h, Aber Benoit, Aber Ildut. Brittany, and the Armorican Massif as a whole, were folded during the Palaeozoic, and are thought to have been bevelled in the Trias, in the Eocene and probably in between. The altitudes are considerably smaller than in Galicia, Asturias and the Basque country, being everywhere lower than 400 metres. The coastal plateaus lie around 80-100 m in the north and west and 30-50 m in the south. Is has been found that some valleys at least were deeply cut into the plateaus well before the Pleistocene: this has been shown for the Aber Ildut, northwestern Brittany, which became a ria as soon as the Lower Oligocene (Hallegouet et al., 1976; Guilcher and Hallegouet, 1987); the Elorn valley, which ends into the Brest Roadstead, was already cut at least at 15 m altitude in the Upper Pliocene near Landerneau city

76

P. CASTAING AND A. GUILCHER

(Hallegouet, 1982); and, in the south, the Vilaine river seems to have cut its lower course at 6 m into the plateaus at Langon, as soon as the Oligocene (Guilcher, 1948, pp. 481-482). In other places, the beginning of the cut creating the present features is not precisely known. What is sure is that, as in Galicia, Asturias and the Basque country, a succession of cuts and fills occurred during the Pleistocene, resulting from the shifting sea levels which accompanied the glaciations. No glaciers existed in Brittany, due to the low altitude, but periglacial processes had an even larger influence than in the Iberian Peninsula and must be considered in the ria geomorphology and sedimentation. The present-time tidal range at largest spring tides increases from 5.50-6 m in the south to 7-8 m in the west, 9-12 m in the north and 15.40 m in the northeast in the Mount St Michael Bay. In north Brittany, quite typical rias exist, being represented in the east by the mouth of the Rance River, which has been dammed near its outer end for electric production, and has thus become an artificial feature. Westwards are successively found the Fremur, the Arguenon known for its tidal bore, the Trieux, the Jaudy, the Leguer, the Douron. The Trieux ria widens in its middle course, with smoothed slopes on both sides, because it crosses a strip of soft shales between harder rocks upstream and downstream. The recent sediments in that set of rias have not yet been so far investigated in detail. In south Brittany, the plateaus into which rias are cut are lower, because of the general asymmetry of the peninsula; nevertheless, typical rias are found, being, from west to east, the Odet, the Scorff, the Blavet, the Etel River, and the Loc’h. Between the coastline and the inner reaches, where the drowned valleys are narrow with steep sides, widenings occur in the middle courses of the Odet, Scorff, Blavet and Etel rivers; and a still larger widening forms the Morbihan, a Breton word which means little sea. These features have a tectonic origin, being related to a set of Cenozoic uplifts and depressions running in the general strike of the south coast of Brittany and continuing offshore on the inner continental shelf (Guilcher, 1948, pp. 163-214 and 382). Details on sedimentation in these southern Breton estuaries will be given after Gouleau (1975) in the section on sedimentary processes. In western Brittany, recent investigations have been more numerous, and it is possible, on a sedimentological point of view, to define there four types of rias (Guilcher et al., 1982). These types are: essentially pelitic rias; rias including a large sandy fraction coming from the sea; widely open rias where outer influences are still larger; microperiglacial rias located in two small areas of southwestern Brittany.

Pelitic nus Pelites are defined as sediments in which the mean diametre of particles is smaller than 50 pm. This first type is usually considered as the current type, in Brittany and elsewhere in the world. The sedimentological environments are distributed into slopes, tidal flats and channels. Details on general processes of deposition in this type of rias, which were investigated by Guilcher and Berthois (1957), are given in the second part of this contribution (sedimentary processes, tidal flat budget). In western Brittany (Fig. 4-7), the Morlaix ria, the Penze and the Aber Wrac’h in Leon country, the Elorn, Daoulas, I’Hbpital, Le Faou and Aulne rias ending into the

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS

77

Fig. 4-7. Distribution of the four ria types in western Brittany (according to Guilcher et al., 1982, with minor changes). Rias without figures still await classification.

Brest Roadstead, and probably the Odet river in Cornouaille (southwest) belong to this type. Details are given here for Penze, Aber Wrac’h and Aulne rias. The Penze ria (Auffret, 1968), some 10 km long, includes quite typical tidal flats in which only 4 to 20% of the sample sediments exceed 80 pm in diametre. The median is sometimes 50 pm, but in other samples 50% are less than 7 pm. Erosion cutting into lateral slopes provides particles of all sizes existing in the intertidal flats and creeks. In the innermost end, silty clays (less than 35 pm) rest upon periglacial sediments, a feature also found in the Daoulas and Faou rias, Brest Roadstead. Lower down the sandy fraction increases, with mean grain sizes ranging from 125 to 400 pm. At the mouth of the estuary, sandy muds occur (50 pm); generally, following the general rule, the sand fraction is larger in the main creek than on intertidal flats. The Aber Wrac’h, the name being more properly Aber Ac’h (Andrade, 1981; Glemarec and Hussenot, 1981; Guilcher et al., 1982) (Figs. 4-8 and 4-9), ends into the English Channel with depths exceeding 10 metres at lowest spring tides, including there pure sands without pelites, and medium sizes ranging from 125 to 350 pm. At the south of Terc’h Island, pelites begin to appear, forming less than 25%; further upstream, they reach 35 to 70% on flats in lateral bays (Les Anges, Keridaouen). As far inside as Moulin d’Enfer, rather coarse muds (25 to 35% of pelites) occur in the channel and on lateral flats. Between Moulin d’Enfer and Keradraon, the grain size

78

P. CASTAING AND A. GUILCHER

Fig. 4-8. Aber Wrac’h (Ac’h) and Aber Benoit, northwestern Brittany (redrawn from Guilcher et al., 1982).

becomes much finer on tidal flats (pelites up to 85%) than in the main creek (pelites: 3 to 20%; medium sizes 135 to 1200 pm). The contrast between lateral flats and main creek decreases again near Prat Paul, and reappears in the innermost part between Pont Krac’h and Diouris. All along the estuary, cliffs cut into periglacial deposits feed beaches in sand and gravel, and banks in finer particles. The Aulne ria (Fig. 4-10), the longest one (25 km) in western Brittany, is among the most accurately known on a sedimentological point of view (Francis-Boeuf, 1947; Bassoulet, 1979; Andrade, 1981). Completely outside the influence of the oceanic swell since it ends into the innermost part of the Brest Roadstead, this ria had, before the Postglacial transgression, deeply cut fine meanders into Palaeozoic rocks. Tidal currents are strong (up to 2 m s-’). Between Landevennec and Lanvian, i.e. in the two outer thirds of the ria, over 34 samples the sediments display a typical classic difference between the central tidal channel or geul(57% of average in weight above 50 pm, with a rather large amount of broken shells) and the lateral soft mud flats or low marshes (13% only). On the high vegetated marshes (schorres) which extend mostly in the inner reaches near Logonna-Quimerc’h, the particles above 50 p m decrease to 6%; in that inner area, the grain size in the central channel is as fine as on the surrounding high marshes, an exceptional fact probably

GEOMORPHOLOGY AND SEDIMENTOLOGY O F RIAS

79

Fig. 4-9. Lower course of Aber Wrac’h (Ac’h) (photo by A. Guilcher, 1993).

Fig. 4-10. Aulne and Faou Rias, western Brittany (redrawn from Guilcher et al., 1982). 1 = rocky flats (Faou Ria); 2 = low cliffs with reworked periglacial sediments at their basement; 3 = spit in Faou Ria; 4 = sand and mud; 5 = idem with coarse shell fragments; 6 = vegetated high marsh.

related to the absence of broken shells, because of a too poor salinity. Samples collected at high tide level at the foot of the periglacial cliffed banks show the usual large range in grain size, since they are, with the shells, the main source of the ria sedimentation.

80

P. CASTAING AND A. GUILCHER

Sandy rias This type includes in western Brittany, Aber Benoit, Aber Ildut, Quillimadec and Penfoul in the north, and Goayen, Pouldohan, Laita and Belon in the south. Aber Benoit, Goayen and Laita are selected here for description. Aber Benoit (Cotton de Bennetot et al., 1965; Fig. 4-8), 10 kms long, includes two very different parts. The outer, longer part, is quite sandy, with 0 to 2% of pelites, the sand being well sorted and the lime content ranging from 10 to 30%. The marine origin of this sediment is evident, the material being closely related to the calcareous (18 to 50%), coarse (170 to 1100 pm), well sorted sands found in the outer reaches of the ria. However, in the innermost part from Treglonou to Tariec, and in the lateral tributary of Locmajan, the sedimentation is completely different, deriving from the washing of periglacial cliffs, with pelites ranging from 35 to 75% and lime falling to 0-5%. The contrast with the nearby Aber Wrac'h (Fig. 4-8) is striking. Similarly, the Goayen (Cotton de Bennetot, 1967; Fig. 4-11), 6 km long, is sandy in its two-third outer parts, with grain size median between 200 and 400 p m or more, and well-sorted particles with 55 to 80% of lime content, increasing towards the sea. In the innermost part, however, between Kermalero and Pont Croix, the sediment characteristics are completely inverse: pelites ranging from 20 to 50%, grain size median below 100 pm, except in the channel, lime content falling below lo%, mica particles ranging from 10 to more than 50%, except in the channel again. We have to do with dynamics in which the southwesterly swell rules the outer part of the estuary while sediments are fed by erosion of lateral periglacial slopes in the inner part. The Laita or Quimperle river, 16 km long (Berthoql964; Oliviero, 1978) is a ria in which the main channel is sandy (median between 180 and 900 pm), the low marshes being also made of fine sand, pelites appearing only (50 to 70%) on high marshes. Calcium carbonate ranges from 5 to 25% in the third outer part, where marine pebbles are also present. In the inner reaches, calcium carbonate disappears,

I

am

OF WHT CROB

Fig. 4-11. Goayen Ria, western Brittany (redrawn from Guilcher et al., 1982).

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS

81

and small lateral tributaries are muddy and fine. It is again a ria into which outer marine sediments are pushed by the swell.

Bay-like n’as (Andrade, 1981; Guilcher et al., 1982) In this third type, including Lauberlac’h and Le Conquet rias, the outer marine influence is still larger, pushing marine sediments into the rias, and resulting in the formation of mid-bay spits on the sides of the rias. At Lauberlac’h in Brest Roadstead (Fig. 4-12), where the fetch is 10 km southwestwards, the direction of the dominant winds, resulting in the production of a great amount of coarse detrital material, the ria is divided into two parts by a mid-bay spit almost unique in the world (an imperfect another one exists in southeast Ireland), which penetrates deeply into the opposite coast leaving only a residual channel where ebb and flow currents are quite strong. This spit, and smaller ones on both sides of the outer ria, are fed by coarse pebbles deriving as usually from erosion of periglacial sediments in slopes. The transportation of these pebbles is made for a large part through the buoyancy of marine kelps. Outside the main spit, the sediments of the ria are very poorly sorted, with a large amount of pelites but also small stones, shells and Lithothamnion particles which thrive in Brest Roadstead. Inside the spit, where wave action is nil, sediments are considerably finer as expected, with medians of 25 to 30 pm, lime percentage 10 to 15% deriving from local fauna which is well fed in sea water (the river ending into the ria is quite small). Le Conquet ria (Guilcher et al., 1982; Fig. 4-13), including two tributaries, is also invaded by marine sands, directly from the outer sea through its outlet, and indirectly from sand dunes on its northern side. Symbol 8 on Fig. 4-13 shows in the outer part the large calcium carbonate content in sand, decreasing from west to east and

Fig. 4-12. Lauberlac’h Ria, Brest Roadstead (from Guilcher et al., 1982). 1 = rocky flats; 2 = cliffs with reworked periglacial sediments at their basement; 3 = spits made of poorly rounded pebbles; 4 = coarse heterogenous sediments; 5 = low flats, fine sediments; 6 = mixed sediments including Lithothamnion particles; 7 = vegetated high marsh.

P. CASTAING AND A. GUILCHER

82

BLANCS

SABLONS

BAY

SAND

DUNES BUNCS S

Fig. 4-13. Le Conquet Ria, western Brittany (redrawn from Guilcher et al., 1982). I = rocky flats; 2 = big blocks; 3 = pebbles; 4 = former pebble spit (now destroyed by man); 5 = sand; 6 = low mud flat; 7 = vegetated high marsh; 8 = CaC03 content (%).

becoming negligible in the inner end, where pelites are predominant in salt marshes. As in Lauberlac’h ria, two sand spits had been built by waves, owing to the strong outer surf; one has been replaced since a long time by St Christophe jetty which shelters the fishing harbour; another one existed at Croae, north side of the rias until the Second World War, when it was exploited by the German Army to build the Atlantic Wall.

Dwarflike, micropenglacial rias This curious type, which has been described by Guilcher (1948, pp. 322 and 426; 1982) and Schulke (1968, pp. 56-66), and has been called Kastentalna or Zwergria in German by Schulke, is represented in southwestern Brittany by Porz Lamat, Brigneau, Merrien and Doelan rias and by several other ones on the southwestern coasts of the Isle of Groix and Belle-Ile, southern Brittany (Fig. 4-14). The Dahouet ria, north coast of Brittany in Bay of Saint Brieuc, belongs to the same type. All are very short, always less than 2 kilometres long, cut into metamorphic schists along cliffs 20 to 50 metres high, and have flat bottoms very poorly sedimented, and steep sides resembling those of auges of glaciated fjords although no glacier occurred there. They had been filled by very thick periglacial deposits which have been largely or completely washed now by the surf in the outer and middle courses of the valleys, since all are quite exposed to the open sea. There is certainly a relation between the very steep sides and the nature and steep dips of the country rock. Provence (France) In Provence near Marseilles, southern France, six very short (1 to 2 kilometers) narrow, deep valleys called calanques, cut into hard Mesozoic limestones, can be

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS

83

Fig. 4-14. Microperiglacial ria, outer coast of Groix Island, south Brittany. Low tide. Sediments have been completely washed by surf at high tide. (Photo by A. Guilcher, 1981.)

considered as karstic rias. Their cut occurred mainly during Pleistocene low sea levels (Nicod, 1951; Schulke, 1968, pp. 75-79).

Southwest England and possible other areas in the British Isles Cornwall and Devon, southwest England, in which the geological evolution had been rather similar to what happened in Brittany, are girdled by a set of drowned valleys (general view in Steers, 1964, pp. 205-260), the main ones being, from north to south and west to east, the Taw at Barnstaple, the Camel at Padstow and Trebetherick, the Fa1 and the Carrick Roads at Falmouth, the Tamar and tributaries which form the Plymouth Roadstead, the Dart, the Teign, and the Exe. As Dewey (1948, p. 64) wrote “the drowned valleys of Cornwall and Devon are of the ria type and do not resemble fjords”. No Pleistocene glaciation occurred in southwest England. However, glaciers issued from more northern countries (Wales, Scotland) and flowing southwards through what is now the eastern part of St George’s Channel, reached the mouths of the Taw and the Camel and left there morainic deposits (Arkell, 1943; Clarke, 1969; Kidson, 1971; Coque-Delhuille, 1987, pp. 634-655); but the shape of these estuaries has nothing to do with glacial action. All these rias were deeply cut during Pleistocene low sea levels: the depth of the bedrock has been measured since a long time (Codrington, 1898) in a number of places, being generally located at some 30 metres below present sea-level near the mouths of the rias. At the same time, periglacial “heads” or rubble drifts were deposited on slopes, the word head itself having been introduced for Cornwall and Devon in the literature by De

84

P. CASTAING AND A. GUILCHER

la Beche (1839), father of the British Geological Survey. As in Galicia, northwest Spain and Brittany, the rias had been drowned as such by interglacial transgressions, numerous evidences for this deriving from Pleistocene beaches lying on their sides in their outer reaches. The precise age of these beaches remains a matter of discussion as in Brittany, being either Holsteinian or Eemian (called Hoxnian and Ipswichian in the British Isles), an Eemian-Ipswichian age being perhaps more probable in most sites (Stephens, 1966; Kidson, 1971; Coque-Delhuille, 1987, pp. 827-832; etc.). Contrary to Brittany, a detailed typology of the recent sedimentology of the rias of southwest England does not seem possible so far. As a whole, the current pattern of sandy sediments near the mouths, and mud flats-high marshes in the inner reaches, is valid. For precise measurements of sediment growth in Great Britain, particular reference should be made to the Severn estuary (Allen and Rae, 1988; Allen, 1990a, b) which is not a ria. Some local situations and distributions of sedimentary material will be summarized here. In Devon, the double ria of the Taw and the Torridge ending into Barnstaple bay (Steers, 1964, pp. 216-217) is fronted by “the greatest development of sand dunes in Devon and Cornwall”, a feature which reminds situations found in Asturias. Behind these dunes, the usual mud flats and marshes occur in both rivers. Farther south, in Cornwall, the Camel estuary (Steers, 1964, p. 226) bears a large mass of sand “due largely to the waste material in times past from the tin workings”, being thus here of human, not natural, origin. The Plymouth ria includes a complex of tributaries beside the main Tamar ria, with the usual inner mud flats, but the development of Plymouth harbour has widely reworked the outer part, which forms “The Sound” or Roadstead behind a breakwater. In the Dart, “borings have shown that the drowned valley is trough-like in form and had not reached base level, thus indicating a short and quick elevation of the land relative to the sea” (Steers, 1964, p. 245). At the mouths of the Teign and the Exe, east Devon, which are located at the boundary of the New Red Sandstone outcrops, complicated and changing patterns of sand bars and spits have been built, and investigated in detail (Steers, 1964, pp. 249-254, and previous authors summarized and discussed). This reminds to some extent what occurs in Asturias and in Barnstaple bay, but it must be said that the Teign and the Exe lie at the border of the old massif and could be considered as intermediate between rias and “ordinary” estuaries. Outside Devon and Cornwall in the British Isles, what kind of estuaries occurs in the near-by countries of Wales and south and southwest Ireland is a matter of discussion. These two countries are well known to have been glaciated during the Pleistocene, but their river mouths cannot be considered as fjords since they have not been deeply cut by ice. In Wales, River Loughor at Llanelly, River Tywi south of Carmarthen, Milford Haven at Pembroke, Afon Dyfi and Afon Mawddach on the east coast of Cardigan Bay (Steers, 1964) resemble rias in spite of a Pleistocene glacial cover. In Ireland, the very fine estuary of Cork, which cuts at right angle across the alternately hard and soft Palaeozoic rocks, with successive straits and widenings, could perhaps be considered as a ria since ice action has not been very efficient. More to the west, Kenmare River, which follows, on the contrary, the strike of the rocks, resembles a ria after its general shape; but the drumlins which occur in its middle

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS

85

course, and have been reworked into “drumlin and spit structures” (Guilcher, 1965) introduce a complication in classification. More northwards, Dingle Bay (Guilcher and King, 1961) is too widely open on the ocean to be considered as a whole as an estuary.

Korea As said in previous studies (e.g., Lautensach, 1945; Guilcher, 1976a), Korea is a country where typical rias are found (Fig. 4-15). These rias are not located on the east coast facing the Sea of Japan, where the tidal range, usually although not universally requested for real ria features, is insignificant (0.2 to 0.25 m), but on the south and west coasts where the tidal range in spring tides increases from 2.4 m in the southeast to 3.9 m in the south, 4 m in the southwest, 7.1 m in the west, and 9.3 m at Incheon in the northwest near Seoul. Although Korea as a whole is asymmetric, as emphasized by Lautensach, with the highest mountains in the east and lower relief in the west, the altitudes in the south and west are large enough to allow deep valleys, with mountains or hills 600 to 400 m high along the southeast coasts, 800 to 300 m in the south, and 200 to 150 m in the west (all these summits lying close to the shores). Rocks are varied, ranging from Precambrian gneiss, schists and quartzites in the west and southwest to Cretaceous sedimentary and volcanic rocks in the south, Jurassic and Cretaceous granites in different areas. These rocks have been deeply weathered, this evolution resulting in red soils 2 to 7 m deep, with much kaolinite pointing to Tertiary hot and humid climates. During the cold Pleistocene periods, periglacial actions gave way (Guilcher, 1976a) to flows of frost-shattered blocks, embedded or not in clay, quite similar to the slope deposits which are called head in Cornwall, Devon and Brittany. As in these countries, the coasts have not been glaciated in Korea. Interglacial (Eemian?) beaches have been found in the southeast, southwest and west (Guilcher, 1976b), testifying the same Pleistocene sea-level shifts with cuts and fillings as in Galicia, Asturias, Brittany and southwest England. The dissection of the Korean mountains by rivers under these successive climates, and the succession of Pleistocene sea-level shifts, common to all these countries, have resulted on the southern and western coasts in a ria morphology of a type different from the type found in Galicia, Asturias, the Basque country, and Brittany. The number of islands in front of the continent is exceptionally large, especially in the southwest, the south and the southeast, but also to some extent in the west (Figs. 4-15 and 4-16). As a matter of fact, islands in front of a ria coast exist also in northwestern and northern Brittany, but these Breton islands result from a differential alteration of granite before a dissection which removed the weathered parts and left residual hills; while in Korea the insular type of ria coast exists in all types of rocks, granitic or not. So that everywhere the drowned mouths of the rivers are surrounded by conical hills. The pattern of drowned valleys can be followed on marine charts as deep as 30 to 50 m off the western coast. These valleys are often 50 to 70 km long, being particularly conspicuous in the Mogpo area, southwestern coast. In the southeast, they are shorter, probably because their outer courses have been filled and bevelled by recent sedimentation, contrary to what happens in the southwest.

? 0

Fig. 4-15. General map of ria pattern on southern and southwestern coasts of Korea (redrawn from A. Guilcher, 1976).

2 e

GEOMORPHOLOGY AND SEDIMENTOLOGY O F RIAS

87

An important feature of the Korean ria coasts is the high energy relief, resulting from the altitude of big hills or small promontories close to the sea. See for example on Fig. 4-15, the altitudes in Namhae, Dolsan and Geumo Islands, on the south coast, and in Geoge Island on the southeast coast. This feature is common with Galicia, Asturias and the Basque country, but different from Brittany and British Cornwall. It coexists with a great width of the drowned valleys, which are comparatively wider than on other ria coasts (Fig. 4-16). Since the Holocene transgression, marine erosion has been generally insignificant, even in the southwestern, southern and southeastern archipelagoes, where, even in places where fetches are long, marine cliffs are uncommon. The wave action has only washed the weathered formations and discovered the solid rock which appears in steep slopes. Marine notches at the foot of these slopes are rare and insignificant. Nevertheless, pebble ridges exist in a number of coves and small bays, reworking periglacial slope deposits. Two areas are exceptional in this respect. One is the southern and southeastern set of islands off the continental coasts, i.e. the outer area of the ria pattern, where longer fetches allowed cliffs several tenths of metres high. The second case occurs where the formations resulting from the Tertiary weathering are deeper than the average, allowing the development of cliffs even in places where the fetch is short, as long as the solid rock is not reached by marine erosion (Guilcher, 1976a). Recent sedimentation along the inner parts of the rias has been studied in the Inchon area, west coast near Seoul (Wells et al., 1990), where the tidal range is particularly large as said previously. Mud flats are extensive, several km wide

Fig. 4-16. Garorim Ria, west coast of Korea. Low tide. Conical islands. Commercial salt pond in foreground, encroaching upon mud flats. (Photo by A. Guilcher, 1975.)

88

P. CASTAING AND A. GUILCHER

Fig. 4-17. Part of Seoul Ria, west coast of Korea, cut into high hills. Very narrow vegetated high marsh in foreground, low marsh below. (Photo by A. Guilcher, 1975.)

sometimes at low spring tide; but, there and along the other Korean rias, the usual high marshes bearing vegetation are rare or absent (Fig. 4-17). This is a result (Kwon, 1974) of extensive reclamations for rice cultivation, to face the growth of population. Large reclamations were also made to produce salt, especially in the Garorim Bay, west coast (Fig. 4-16). Maps surveyed at the beginning of the XXth century show that natural, vegetated high marshes or schorres were still very extensive at that time on the west coast. The absence of Holocene high marshes along the rias is thus an artificial, human feature. According to Kwon (1974) who has studied the western estuaries, the sediments come from upstream in the main courses of the rias, but, later on, the finer particles are distributed into smaller estuaries which are not well fed from the continent. Wells et al. (1990) and Adams et al. (1990) have described the channel geometry and intertidal sedimentation, which seem to be similar to what happens elsewhere in rias. Most channels appear to be ebb-dominated with respect to sediment transport, as said by Kwon (1974).

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS

89

Monsoonal storms (tropical cyclones) increase the ebb currents and remobilize the material. During one of these storms, a 5-10-cm thick layer of soft mud was eroded from a tidal flat surface. Such an erosion, and the daily dispersal by ebb currents, require a substantial return of sediments to maintain the tidal flat elevation; the process of such a return needs further investigation. Finally, the Korean rias, although displaying particular aspects related to climate or other local circumstances, fall really into the general ria type, as generally accepted by authors.

Southeast China and Shandong The coast of southeast China between Hang Zhu Wan (Hangshow Bay) and Zhan Jiang (Chanluang) is probably the longest stretch of ria coast in the world, and, in north China, the coast of the Shandong peninsula shows the same features and is thus to be added. Good old descriptions of these ria coasts have been given by the great German geographer von Richthofen (1877-1912; 1898) in handsome books difficult to find now, and recent work, although in progress, still needs development. Anyhow, it is quite sure that the southeastern Chinese coast is to be classified as a ria coast. It includes a lot of long-winding embayments at river mouths, cut into hills or low mountains, as for example the Fu Zhou (Foochow) ria at the mouth of River Min, the Xia Men (Amoy) ria at the mouth of River Kiulung, the Shan Tan (Swatow) ria at the mouth of River Mei. The well-known site of Hong Kong and Macao should also be considered as a typical ria environment. Similarly in Shandong peninsula, many ria sites can be quoted, and the wide bay of Qing Dao (Kiao Chow) has been compared, because of its narrow entrance and inner widening, to the Gulf of Morbihan in Brittany. The environment of high hills implicates that uplifts occurred, resulting in valleys cuts before the recent drowning. This sequence is checked by a recent Chinese paper (Li Congxian et al., 1991) which distinguishes, in the Chinese coastal environments, subsidence and uplift belts, the former including the lower courses of the Yellow and Yangtse rivers, and the latter, the south Chinese coastal area and the Shandong. It must be pointed out, however, that the uplift did not continue on a large rate as late as present time, unless it would have prevented the Holocene sea to drown the lower courses of the valleys as it really did.

Argentina In south Argentina, on the southern tip of Patagonia, there are four rias: Deseado, San Julian, Santa Cruz and Gallegos (Fig. 4-18). Unfortunately, very little is known about them (Piccolo and Perillo, in press). These rias formed by flooding of valleys occupying Tertiary sedimentary formations. The Deseado ria is oriented WSW-ENE and is 40 km long. Its mouth is very long, and the width of the last 18 km varies between 400 m and 2500 m. This is primarily due to the rias irregular shape which is linked to the presence of islands, tidal sand banks, and small bays limited by capes. Near the rias mouth, maximum depth varies between 30 m and 37 m, but it quickly decreases upstream to fall between 5 m and 20 m. Although the tidal range is around

90

P. CASTAING AND A. GUILCHER I

1

/

1

I

I

ATLANTfC OCEAN

R Chico

5OoS -

550s

Fig. 4-18. General distribution of rias in the southern part of Argentina (redrawn from Piccolo and Perillo, in press).

2.9 m during neap tides and around 4.2 m during spring tides, there are no tidal flats, except in some areas where currents are weak. Flood and ebb currents vary between 2.5 and 3 m s-l and are turbulent enough to induced a high turbidity of the water which is loaded with greyish-whitish clays of volcanic origin. The Santa Cruz and Gallegos (see Pino, Chapter 8 this volume) rias have similar features. Large tidal flats formed in response to the high tidal range (9.5 m during spring tides and 5.4 m during neap tides). They are associated with outcrops and pebbly beaches which are aligned along the internal and external rias margins. Maximum depths are over 20 m in the rias mouth, although ebb deltas, consisting mainly of pebbles and silts, are present in both rias. It should be noted that usually, these sedimentary formation are not found in macrotidal environments. The Santa Cruz ria ebb delta has two ebb tidal channels. The southern is the most active, and it actually cut off the delta front. No secondary flood channels are present in the two rias. This is an indication that the tide moves upstream as a whole, perhaps creating a tidal bore at the same time. The ebb deltas formed because the mouth of both rias is narrow (approximately 2 km) compared to other valleys situated upstream (width reaching 5 to 6 km). As a result, strong ebb currents are formed in the two rias.

Red Sea s h a m s and their worldwide extension Sharms (Arabic plural: Shurum) are drowned valleys bearing coral reefs, found on the Red Sea coast of Hedjaz, Saudi Arabia, which have counterparts on the

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS

91

Fig. 4-19. Sharm Abhur, cut into the Tehama, eastern coastal plain of the Red Sed (redrawn from Monnier and Guilcher, 1993).

Sudanese coast of the same sea where they are called marsas, and also in some other corallian areas (distribution in Guilcher, 1988, pp. 57-59). They are to be considered as a particular kind of rias. A rather important literature exists on those found in the Red Sea (Schmidt, 1923; Rathjens and von Wissmann, 1933; Guilcher, 1955b, 1985; Sestini, 1965; Mergner, 1967; Dalongeville and Sanlaville, 1981; Monnier and Guilcher, 1993). Sharm Abhur, the best known one (Fig. 4-19), can be selected as typical. Sharm Abhur, located at some 20 km at the north of the centre of the city of Jeddah, is a meandering, narrow gulf, 10 km long, 250 to 1410 m wide, sharply cut into the coastal plain of Tihamat at the foot of the Precambrian granitic mountains of Hedjaz. The Tihamat is made of fluviatile alluvions in its inner part, and, in its outer part, of Pleistocene coral reefs, their exact age being a matter of discussion. Sharm Abhur is assumed to have been cut by fluviatile erosion during the Pleistocene glacial regressions, probably before the last, Weichselian, one. Its longitudinal profile, which displays a great acceleration in its lower part, points to a large lowering of the Red Sea at the time of the cut. Since the Red Sea has no tides, especially in that area, no tidal currents occur, contrary to what is found generally in rias. However, other currents exist in it, as result of the sea winds which push the superficial water into the sharm, a counter current existing at depth: so that the environment is favourable to marine life, and, due to the high temperature throughout the year, fine coral

92

P. CASTAING AND A. GUILCHER

reefs grow on the sides of the sharm (or grew before the recent urbanization: see Monnier and Guilcher, 1993, in that respect). The upper sides of the sharm bear two superposed visors cut by bioerosion in fossil corals, which occur also in the nearby fossil coral reefs of the outer coast. In the channel, fine sediments occur of silt and clay size, poorly sorted since the current is quite slow, with coral debris provided by fish activity. So far this sedimentology has not yet been investigated in detail. Summarizing these features, Sharm Abhur is a fossil valley cut into the surrounding topography, as the other rias in the world, but without any tidal range; the growth of corals on its sides, depending on the climate, should not prevent to classify it among the rias. Sharm Abhur is not an isolated feature. Many others exist on the coast of Hedjaz (e.g. in Guilcher, 1988, fig. 37, p. 57). Yanbo Sharm has become a modern harbour because of its depth. Similar drowned valleys, very short, are also found in the Gulf of Aqaba or Eilat, on the coast of the Sinai Peninsula (Guilcher, 1979). Those called marsas on the Sudanese coast of the Red Sea (Dalongeville and Sanlaville, 1981) are also fossil valleys, but their outline is particular, consisting of an outer narrow course cut across a fossil Pleistocene fringing reef, and an inner widening extending into the depression behind the fossil reef. The Suakin and Port Sudan harbours have used that pattern which provides excellent sheltered sites for ships. Outside the Red Sea, a sharmlike morphology can be recognized in several corallian areas. It has been mentioned (Guilcher, 1988, pp. 57-59) in Kenya, Vanuatu and Hispaniola. On the coast of Kenya, the sites of Mombasa Harbour, Shimo and IOlifi are equivalents of the Sudanese marsas, with an inlet across the Pleistocene reef which widens landwards behind. In the same way, at Erromango Island in Vanuatu, formerly New Hebrides, sharms exist, especially at Ipota drowned valley in Cook Bay on the east coast, which has been cut into an emerged coral reef lying at 3-4 m above present sea level. A rather similar but smaller feature is found on the north coast of Haiti, Hispaniola Island, Caribbean, near Cap Haitien at Ducroix beach. Sharms have thus counterparts widely distributed in the tropical seas.

Messinian rias in the Meditewanean sea. During the Messinian (Miocene), the Mediterranean sea level was lowered down to 1500-2000 metres as a result of the closure of the Strait of Gibraltar, a fact shown by the thick salt layer discovered and investigated by the Deep Sea Drilling Project (Drooger, 1973; Cita and Ryan, 1978; Hsu et al., 1978a, b). This lowering determined a huge cut of the surrounding rivers, which were subsequently filled up as rias, in the northwestern Mediterranean, by Plaisancian (Pliocene) marine blue marls, when the connection with the Atlantic Ocean was established again. The most impressive Messinian ria is the Rhone ria, studied in detail by Clauzon (1975, 1982, etc.; here Figs. 4-20 and 4-21), which extend northwards as far as Lyons city over 300 km, and continues southwards below present-time sea level down to the salt deposits at some 1800 metres depth, with lateral tributaries, the main one being the lower course of the Durance river. Other similar but shorter cuts occur at the west of the Rhone in Languedoc and Roussillon (Tech and Tet rivers), and in the east

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS

93

Fig. 4-20. Messinian Rhone Ria and other Messinian rias on the French Mediterranean coast (mainly after Clauzon, 1982).

------ -----------

0

m 500

Messinian

,ooo 1500

2000

profile from brings

salt

-

-0--

UTY

-

Alluvial fan

Tertiary and older

,. I50

100

in Provence and NiGois at Saint Tropez, Frejus, Cannes (Siagne river) and principally near Nice with the Var ria which became a delta after the recent filling (Fig. 4-22). Shorter ones appear on the Ligurian coast, Italy. The western Corsican submarine canyons belong to the same family, the filling of Plaisancian blue mark appearing on dry land at their heads, e.g., near Ajaccio. The sedimentology of the marl filling will be characterized later in this paper.

94

P. CASTAING AND A. GUILCHER

J"1

N

. . 4 N€S

-44.

Catchment- areas

-*w

Messinian erosional

c-

.

+

+

-le

Fig. 4-22. Messinian rias in the Nice area, southeastern France.

As expected, similar cuts occurred during the Messinian crisis in other part of the Mediterranean, but the use of the word ria for them appears inappropriate or at least questionable. It has been suggested (Finckh, 1978) that the southern Alpine lakes (Como, Garda) in northern Italy could have been originated in the same way. The Nile valley was deeply entrenched as far as 1200 km inland up to the Aswan cataract, and beneath the Nile delta Messinian valleys appear down to 2500 m below present sea level (Ryan, 1978). But the shape of ria does not appear as in the Rhone valley, being completely obscured by recent sedimentation. In Messinian times, the Red Sea was probably connected to the Mediterranean and not to the Indian Ocean, after the Miocene evaporites and halites found off the coasts of Sudan, Egypt and Saudi Arabia (Hsii et al., 1978a, b), so that it should have been lowered in the same way. However, the above-described sharms are not at all related to that event, since they are cut into Pleistocene reefs and are thus much younger. Rias of the Messinian type do not seem to have been reported so far from the Red Sea, although they are to be expected to exist below more recent filling.

GENERAL CONSIDERATIONS

General lessons concerning ria evolution and sedimentological processes can be drawn from the regional study of rias. However, we will see that sedimentary pro-

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS

95

cesses observed in rias are similar to those identified in most estuaries. Therefore, we will focus herein on a few examples which will be used to illustrate the sedimentary filling mode of rias during the Late Quaternary on one hand, and modern sedimentary processes and how to model them on the other hand.

Ria evolution As in estuaries, the result of ria evolution over geological time is generally an infilling. Modalities of this infilling have been described, especially by Clauzon and Rubino (1990) in the Mediterranean basin and by Prior and Bornhold (1990) in Pacific estuaries. Description of processes will be based on the case of the Pliocene infilling of the Var Ria in France (Clauzon, 1978) and the Holocene infilling of Ria de Muros y Noya, Galicia, Spain (Herranz and Acosta, 1984).

Pliocene infilling The French Mediterranean coast in Provence and Nigois, and the adjoining Ligurian coast in Italy are mostly rocky and show a number of fossil rias, as said in the regional section. Figure 4-22 indicates the location of these rias on the Riviera, Nice area, where the largest one is the Var ria. It will be used here as a reference to illustrate one of the possible modalities of the past and future infilling of rias around the world. The Pliocene infilling in this ria, as in the nearby Mediterranean rias, is structured in so-called Gilbert deltas (Clauzon and Rubino, 1990). From base to top, silty bottomsets (marine facies), gravelly foresets and topsets forming alluvial cones (continental facies) can be found. The marine/continental transition separating the submerged clinoform levels from the subhorizontal emerged levels constitute a (frequently) ligneous cartographicable level. In present time, as a result of recent regional tectonic movements, this structuration is more or less distinct in different rias. An outline of the Pliocene infilling events modalities of these rias is shown, using the ria of the Var as an example (Fig. 4-23). This ria, as all the Mediterranean rias, has an erosional origin (subaerial canyons) depending on the Messinian salinity crisis (Ryan and Cita, 1978). Eustatism controlled the cut of canyons and the filling of rias (Clauzon et al., 1987). The cut (Fig. 4-23a) resulted as said previously from the closure of the strait of Gibraltar. The further sea-level rise, caused the submergence of the desiccated basin and the sedimentary filling of the rias between -5 Myr and -3,8 Myr (Fig. 4-23b, c). The average sedimentation rates, measured in the east of the basin, all facies considered, amounted to 60-75 cm for 1000 years, depending on the duration of the infilling chosen: 1.2 or 1.5 Myr (Clauzon et al., 1987). Today, the modern Var and Messinian valleys are distinct (Fig. 4-23d). The migration is considered to result from the vertical accretion of continental deposits during sea-level rise (Fig. 4-23c). A large part of the Pliocene sediments has been eroded. The original substratum was deformed and rifted by tangential tectonic movements which caused the surelevation of the filled fossil ria and determined an intensification of subaerial erosion.

96

P. CASTAING AND A. GUILCHER

C

-

a U l E R T DELTA TOP

LYSOZUC SUBSTRATUM

OPmNENTAL F M E S

=YYIIIEICONTINENTAL

MESSINIAN EROSIONAL SURFACE

TRWTKN

Fig. 4-23. Evolution of the Var valley, near Nice, since the Messinian (after Clauzon, 1975; 1982)

Holocene infilling The infilling events in the Ria de Muros y Noya are quite well known from seismic investigations by Herranz and Acosta (1984), Somoza and Rey (1991), Rey (1993). Their conclusions are summarized here. Today, the hydraulics of that ria are controlled by a mesotidal circulation of estuarine type in the inner reaches and by an asymmetric circulation of oceanic water. This oceanic water travels landwards along the south coast and seawards

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS

97

Fig. 4-24. Pleistocene and Holocene morphodynamic features of the Ria of Muros (Spain) (after Somoza and Rey, 1991).

Fig. 4-25. Schematic model representing the Holocene seismic sequence of the Ria of Muros and the continental shelf of Galicia. The transgressive system of prograding clinoforms ( S I to S7)is correlated with landward prograding deposits at the ria mouth, interpreted as a flooding sequence. (After Somoza and Rey, 1991.)

along the north coast. This circulation pattern is clearly reflected by the sand wave asymmetry (Fig. 4-24). High resolution records (uniboom system) of the ria deposits show three main seismic units (Fig. 4-25): (1) an acoustic basement, which represents the basal unit: Somoza and Rey infer that this unit is formed by Hercynian material;

98

P. CASTAING AND A. GUILCHER

(2) a Pleistocene unit, which forms the basement of the Quaternary sequence: it is characterized by seismic facies of complex and chaotic filling types; Somoza and Rey interpret it as a channelled fluvial sequence, the location of which coinciding with the present longitudinal axis of the ria; (3) a Holocene section overlying the Pleistocene unit and covered by a thin layer of Recent muds; two main types of seismic facies are differentiated in this unit, with different spatial distributions in the ria: the first type showing parallel to subparallel internal reflection patterns of high continuity and giving rise to prograding clinoforms with a well-differentiated morphology in the ria. The detailed analysis of the Holocene complex shows an architecture composed of a least 7 prograding clinoforms separated by discontinuities: (1) parallel progradational sets presenting S1 and S2 bodies which overlie chaotic and filling seismic facies units. The thickness of these bodies is about 4 m (Sl) and 8 m (S2). (2) sigmoid sets occur mainly in S3 and S4 bodies. The thickness reaches 15 m in S4 body. (3) oblique sets form the internal reflection patterns of S5 and S6 bodies. They represent the major prograding clinoform of the system, with a thickness of 25 m. This Holocene depositional sequence is interpreted by Somoza and Rey as the result of the general rise of sea level which changed the hydrodynamic conditions in the rias by a progressive invasion of oceanic water. The filling of the fluvial channels marks the start of sea-level rise after the lowest stage of regression in about 18,000 yr BP. The example of Ria de Muros Y Noya thus provides a model of a transgressive system connected with the Holocene sea-level rise. According to these authors, the three types of prograding clinoforms (parallel, sigmoid and oblique) which have been determined can be related to delta variability (Postma, 1990) depending on the depth (Fig. 4-26). The parallel prograding clinoforms are associated with shoal-water delta profiles where bed-load transport was predominant. The sigmoid patterns occurred with higher depth rates and can be related to “Gilbert type” fan delta profiles in more important homopycnal conditions. Oblique prograding clinoforms are related to a delta-fed submarine ramp system. These authors conclude that the variation of the clinoform types observed in that ria is directly related to sea-level rise, which controlled the ria filling. The progressive flooding of the ria changed the type of prograding clinoforms and controlled the basin depth, salinity rate and wave energy. The sedimentation inside the ria occurred during stillstands or inflexions in sea-level rise.

Sedimentaryprocesses General features In large estuaries, the fluvial-marine balance occurs more or less upstream according to fluvial discharge. Variations in tidal range have a minor influence on sedimentation processes. On the contrary, in most rias the phenomena related to oceanic tide predominate, since the fluvial discharge is always very small (Berthois

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS DELTA TYPES

99

P R O G R A D I N G C L I N O F O R M TYPES

PARALLEL F.

Fa

T

FLJODING

IUDC~CE

Fig. 4-26. Evolution of prograding clinoform patterns in the transgressive systems track of the Ria of Muros and its relation with delta type variability (Postrna, 1990) depending on the depth ratio (after Sornoza and Rey, 1991).

and Auffret, 1966). In almost all rias, three different sections can be distinguished from the point of view of dynamic processes as well as the sedimentological nature of the bottom: - a marine section which often extends to the longest part of the ria, - an estuarine section, - a fluviatile section.

The marine section In the coastal zone, substantial transfers of sandy material occur, caused by longshore currents resulting from the swell action, or by tidal currents when they are strong, as in the English Channel. This transport, performed by overthrusting due to the size of sediments, stops in areas where topography and depth slacken the current. This is the case, for example, for bays and rias which constitute extremely efficient sand traps. The penetration of marine sands is shown by different morphological, granulometric and organodetritic evidences. In the mouth of some rias such as the Goayen in France and Ria de Muros y Noya in Spain, asymmetrical sand waves are found with the smaller slope facing the ocean, indicating a transport landwards. In most rias, the grain size of the sand becomes finer landwards, showing that the source lies outwards. Likewise, in the outer parts, limestone algae and conchiferous debris are abundant, and their number decreases very quickly upstream. This is shown by the rates of CaC03 along the rias. Also, the salinity of the water resembles that in the sea, showing that the output of fluvial water is insignificant.

100

P. CASTAING AND A. GUILCHER

The estuarine section In this section, which occurs upstream, the sediments are made of a mixture of marine and fluviatile material, the influence of the former becoming increasingly poorer. It is a section where a substantial reduction of calcareous organogenic debris is usually observed. In the majority of rias, especially those lying in macrotidal environments, this section has a hydrosedimentary behavior identical to that of estuaries. The deposition and transport mechanisms are the same. In rias as in estuaries, the tidal asymmetry causes the trapping of maximum turbidity (Nichols and Biggs, 1985). This tidal asymmetry determines a flood velocity predominance. Coleman and Wright (1978) observed this phenomenon in the Ord River, an Australian macrotidal estuary, and Bassoulet (1979) found the same thing in the Aulne Ria, Brittany (Figs. 4-10,4-27).

*d Dam

f

Sedimentation

d k

Erosion

-

Ebb Flood

Fig. 4-27. Sedimentary processes in the Aulne Ria (France) during a semi-diurnal tide (after Bassoullet, 1979).

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS

101

The sides in this section are often occupied by tidal mud flats. Gouleau (1975) has described particularly well the physico-chemical mechanisms: they result in deposition of fine sediments on the banks of bays, rivers and rias. He especially shows that emersion after high tide leads to sediment fixation, since the water content of the top layer decreases through flow percolation and evaporation, thus increasing the density. The mud tidal flats are quickly thickened by the “transversal pulsation” process (Berthois, 1954). The turbid waters are pushed back toward the banks during ebb tide by the tidal current which reaches its maximum speed in the channel or main creek. As shown previously (Fig. 4-27), sedimentation is most active at high slack tide. At ebb tide, a part of the sediments which were deposited are put in suspension again and return to the channel. But at each tide, the net result is positive, since a thin film of sediment deposited at high tide subsists, so that the flats thicken progressively.

The fluviatile section Here the dynamics of flow are clearly controlled by the river current, and fluviatile sedimentation predominates. This area is located higher than the area of fluvialmarine balance. It is often characterized by a substantial silting-up deriving from the river input which is trapped there. Pluri-annual sedimentary budget In order to determine accurately the long-term sedimentation and erosion phenomena which occur in rias, various types of measurements and estimations have been made, especially in Brittany where they are based only on tidal flats and valley slopes (Guilcher and Berthois, 1957) or on the whole ria (CYavanc and Bassoulet, 1991).

Tidal flat budget From 1951 to 1955, Guilcher and Berthois (1957) carried out a five year survey of the tidal flat evolution in four selected Breton rias: Le Conquet, Le Faou, Keroulle and Aber Benoit (Fig. 4-7). They have shown from grain size and thermal differential analyses that mud settling in these tidal marshes derives from periglacial Pleistocene deposits covering the slopes, which are washed by waves at high spring tides. Concerning sediment deposition, the study consisted of measurements of upward growth on vegetated high marshes by means of sand patches spread on mud (a procedure previously used by others on Danish, Welsh and English marshes), and of successive photographs at fixed points in each ria. They did not comprise bare, unvegetated low marshes that cannot be studied by this method. It was found that the rate of deposition depends primarily on the altitude (level) of the marsh, and subsidiarily on the distance between the surveyed points and the main tidal creeks acting as feeders. Successive photographs of microcliffs (Fig. 4-28) show a disintegration of small blocks of hardened mud fallen down upon bare low marshes, their mud being again put into suspension and redeposited on vegetated high marshes, so that a real cyclic evolution of the mud may be observed. Simultaneous processes of deposition and erosion were also found in the Loire estuary by Gouleau (1975) and in Dutch marshes by van Straaten (1954). Moreover,

102

P. CASTAING AND A. GUILCHER

Fig. 4-28. Aber Benoit Ria, western Brittany (located on Fig. 4-7), upper course. Microcliff cut into high, vegetated, marsh, with fallen small mud blocks reworked by tidal current at high tide. (Photo by A. Guilcher, 1954.)

the latter shows how the higher part of a low marsh can in turn be undermined by a microcliff as the high marsh above it, leading to the formation of two superimposed microcliffs. This pattern was also observed by Guilcher and Berthois (1957) in Le Conquet ria. Therefore, tidal flats or parts of tidal flats are not all in the same stage of the cycle. There are young tidal flats with numerous creeks, as in a part of Le Faou ria; mature tidal flats with few creeks as in another part of the same ria; senile, decaying tidal flats as in Le Conquet ria. The senile stage is thus marked by a splitting of the high vegetated marsh into mounds of increasingly smaller size, although the older structures are still visible along the main creeks. For the cycle notion to be completely valid, destructions must be completely compensated by constructions. The five year survey of the four Breton estuaries did not enable Guilcher and Berthois (1957) to prove it, even on data from measurements in Le Conquet ria, the most evolved one. In fact, deposition still goes on upon old high marshes, but as their surface is continuously reducing, this is not sufficient and there must also be an upward growth of low marshes which will become vegetated high marshes later on. Even today, the “high low marshes” are not widespread, and they are themselves actively undermined in some places. A total compensation would apparently imply the building of more high low marshes than today. The problem may be raised whether there is not some loss of fine material in Le Conquet ria. In such case, the cycle would not be complete and a part of the fine sediments would be discharged seawards. The study of another ria, the Morlaix ria, northwestern Brittany (located on Fig. 4-7) brings some data in this respect.

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS

103

Overall ria budget Morlaix ria (CYavanc and Bassoulet, 1991) is 5.5 km long. It belongs to the classic type with sand in the channel and silty sand, sandy silt and silt in the intertidal zones. Swell penetrates in its outer part; tidal range is quite large, reaching 9.3 m at largest spring tides. Two small rivers flow into it. Their mean discharge, approximately 3 m3 s-l, is insignificant in the oscillating volume of water involved in the semi-diurnal tide. From 1988 to 1990, CYavanc and Bassoulet studied first the upstream-downstream movements of the fine sediments by monitoring graduated markers. They concluded that in a period of low river discharge, the fine sediments migrate upstream, creating an instability of the silty slope; in a period of high river discharge, the fine sediments which were previously stored upstream are resuspended by the current action and erosion of the lower silty marshes, and transported downstream. The detailed survey of the size of accretional and erosional areas in the ria allows to distinguish three different zones: - An overall balanced zone located in the upstream part of the estuary (Fig. 4-29a), characterized by high turbidities ( > 1 g/l) at low tide, substantial shifts in salinities and an asymmetry of the tidal wave. Sedimentation rate is low; accretion on the bottom does not exceed 0.15 m during the period considered. - A median zone (Fig. 4-29a), in evolution, with the same hydrodynamic characteristics as the former, but where the asymmetry between ebb and flow is smaller; the

Fig. 4-29. Long term (1929-1986) sedimentary processes in the Morlaix ria (Brittany, France): a. upstream and median areas; b. downstream area (after CYavanc and Bassoullet, 1991).

104

P.CASTAING AND A. GUILCHER

B Erosion Sedimentation

Fig. 4-30. Long term (1929-1986) sedimentary budget (m3) in the Morlaix Ria (Brittany, France) (after CYavanc and Bassoullet, 1991).

silty slopes are replaced by substantial flats on which the growth reached 0.40 m in 57 years (1929-1986). - A downstream zone (Fig. 4-29b), characterized by a substantial enrichment in the channel, of approximately 1 to 3 m on the average, and reaching 5 m at the upstream limit of the zero on marine charts. On both sides of the channel, a generalized erosion is observed on flats, reaching 1 m on the edge of the channel. The sedimentary budget for 1929-1986 is shown on Fig. 4-30. The volume of accretion in the upstream and median zones and in the channel of the downstream zone reached 2.14 x lo6 m3.Zone by zone, this volume increased from upstream to downstream to 70,000 m3, 0.87 x lo6 m3 and 1.2 x lo6 m3 respectively (1 m3 means 0.5 ton of dry fine sediment because of the high water content). The volume of eroded sediments on the flats of the downstream zone reached 1.8 x lo6 m3. A comparison with the deposited volume shows that the overall budget means a slight enrichment or perhaps a sedimentary balance if the relative inaccuracy of comparisons on maps is taken into account. However, although the sedimentary volume included in the ria has remained more or less unchanged since 1929, its distribution is quite different, since a filling of the bottom and the channel of the ria is observed, except near the mouth, together with an erosion of the flats in the

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS

105

downstream zone. The problem is to determine whether it is an irreversible process which, as in many estuaries, is represented by a filling moving from upstream to downstream, related to the general sea level rise.

Ria modelization A numerical modelization of hydrosedimentary processes occurring in estuaries has been made and well developed for several years. However, few authors have been interested in ria modelization. A few hydrological models, involving essentially the Galician rias, Spain, have been proposed to determine the water circulation diagram (Pascual, 1987). Recently, Prego and Fraga (1992) suggested a stationary model for the calculation of the water in the Vigo Ria. According to these authors, the concepts used for the study of estuaries (Dyer, 1973) must be adapted to individual rias. The model is built on the basis of the flow of freshwater and salinity as a tracer. In the Vigo Ria, the circulation belongs to the type of a partly stratified estuary. This ria is divided into five boxes, and a system of twenty equations is proposed, the solutions of which giving the residual outflows and inflows and the rise and mixing fluxes which occur in the ria. The proposed model enables also to introduce the wind influence, and the results match closely the in situ measurements. A numerical model of sedimentary movements in a ria has recently been developed by Le Hir et al. (1990). This model was developed to simulate the transport and distribution of the fine particulate sediments in the Morlaix Ria, northwestern Brittany. The basic principle of the model is classic. It consists of a local calculation of the sedimentary suspended mass resulting from the transport by currents (advection), turbulent mixture mechanisms (dispersion), drop of particles and exchanges with the bottom by erosion or deposition. The equation is numerically solved by a technique of finite differences in a network of meshes representing the interested area, divided into as many juxtaposed boxes. The model includes two main original aspects: - The possibility of transporting simultaneously several dissolved or particulate variables with possible interactions. - The capacity of monitoring the particulate variables in the superficial sediment, the rheological characteristics of which determining the erodibility of the soil. The model obtained in this way is quite adapted to the modelization of sedimentary processes on a monthly scale. The immediate results of this hydrosedimentary model are the space-time distributions of the suspended matter concentrations, Figure 4-31 illustrates the variation of these concentrations in the Morlaix Ria at spring tide with low river discharge. In the upstream section, an extreme variability of the concentration is observed, with a minimum of 10 mg/l at high slack tide and a maximum of a few g/l at low slack tide. The maximal concentration over a substantial part of the ria results from a resuspension of the fine sediments by the ebb current in a very small volume of water (the width of the channel at low tide in the upper part of the ria is around 10 m). Downstream, in the widest part of the ria, the turbidities are much smaller. Figure 4-31 shows the supply in suspension from upstream at the end of the ebb and the resuspension by the flow.

106

P. CASTAING AND A. GUILCHER

t

FLOOD

I

EBB-SLACK

EBB

I FLOOD-SLACK

I ore0 B3

0

’

Concentration Fig. 4-31. Numerical simulation of suspended solid matter concentration during spring-tide and low river flow in the Morlaix Ria (Brittany, France) (after Le Hir et al., 1990).

Tidal mud flats and marshes are common in both ordinary estuaries and rias. This is why the numerical model developed by Allen (1990b) to simulate the salt marsh growth and stratification of the Severn estuary, Great Britain, is applicable to tidal marshes and mudflats of the rias. Allen works on the principle that, “...theoretically, flat-marsh growth is determined by the rates of minerogenic and organogenic sedimentation, the rate of change and tendency of relative sea-level and the rate of ‘long-range’...’’ sediment compaction. A numerical simulation model “...is described and implemented for the Severn estuary on the basis of empirical knowledge of its tidal and fine-sediment regimes and the present-day order of magnitude of the deposition rate of fine sediment in its upper intertidal zone”. The model is relative to a tidal frame because it is the position of the sedimentary surface relative to tidal limits which controls: the rate of deposition of sediment from the tidal waters (mineral supply); and the level of plant productivity (organic supply). In agreement with Allen and referring to Fig. 4-32, “...the elevation E (m) relative to tidal datum (zero on a local tide gauge, approximately the level of the lowest astronomical low water) of the surface of a mudflat-marsh at a place changes annually according to the equation:

+

A E = ASmin(E) ASorg(E)- AM(t) - AP(t)

(4-1)

in which A E is the time-rate of change of elevation (Myr-I); Asmi, the time-rate of build-up by mineral sediment (Myr-l) autocompacted as a consequence of seasonal

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RlAS

107

level of extreme astronornicol tide

, base of mudflot-marsh, E=E0

tidal d at u m . E

=o

AM AM

Fig. 4-32. Definition diagram for vertical salt-marsh growth in a tidal frame of reference: Severn estuary, Great Britain (after Allen, 1990a, b).

drying; AS,,, the time-rate of build-up by the addition of plant-derived sediment (Myr-’) treated as autocompacted; AM the time-rate of change of relative sea level (Myr-’); t is explicit time; and A P the time-rate at which the surface is lowered (Myr-’) through long-range compaction”. Allen takes “...the implicit time-increment to be a year because it is the most convenient period over which to define the long-term average tidal regime”. Among the main results, Allen’s model predicts that the elevation-time curve describing mud flat-marsh growth rises very steeply during the earliest stages of build-up, but thereafter flattens off very rapidly. A marsh that is built during a period of rising relative sea-level (now for instance) reaches, after a certain maturation time, an elevation which is constant relative to the moving tidal frame but lower than the level of the highest tide. A stage of dynamic equilibrium is reached. In conclusion, Allen’s model predictions receive satisfactory empirical supports of various kinds outside the Severn estuary; the predicted form of the growth curve is supported by the pattern of marsh growth observed on the east coast of England. This model appears to be perfectly adapted to the prediction of the evolution of tidal flats occurring in rias around the world.

REFERENCES Acosta, J. and Herranz, P., 1984. Contribucih al conocimiento del Cuaternario marino en la ria de Muros y Noya. Thalassas, 2: 13-21. Adams Jr, C.E., Wells, J.T. and Park, Y.A., 1990. Internal hydraulics of a sediment stratified channel flow. Mar. Geol., 95: 131-145. Allen, J.R.L., 1990a. The Severn estuary in Southwest Britain: its retreat under marine transgression, and fine-sediment regime. Sedim. Geol., 66: 13-28. Allen, J.R.L., 1990b. Salt-marsh growth and stratification: a numerical model with special reference to the Severn Estuary, Southwest Britain. Mar. Geol., 95: 77-96. Allen, J.R.L. and Rae J.E., 1988. Vertical marsh accretion since the Roman period in the Severn estuary. Mar. Geol., 83: 225-235.

108

P. CASTAING AND A. GUILCHER

Andrade, B., 1981. Etude morpho-stdimentologique d’estuaires de la rade de Brest et de la cBte du Lton. Thesis, Brest, 144 pp. Auffret, G., 1968. Contribution i ]’Etude Stdimentologique de la Ria de la Penzt (Finistke). Thesis, Paris, pp. 127. Arkell, W.J., 1943. The Pleistocene rocks at Trebetherick Point, North Cornwall. Proc. Geol. Assoc., London, 54: 141-170. Arps, C.E.S. and Kluyver H.M., 1969. Sedimentology of the Northwestern shores of the Ria de Arosa, NW Spain. Leidse Geol. Med., 37: 135-145. Asensio Amor, I., 1960. Datos granulomttricos de las arenas de la Ria del Eo. Est. Geol., Madrid, 16: 93-97 and 187-189 (also: Est. Geogr., 1959,75: 251-262). Balay, M.A., 1956. Determination of mean sea level of Argentine Sea. Influences of the sea not caused by the tides. Int. Hydrogr. Rev. Monaco, 33: 31-65. Barrois, C., 1882. Recherches sur les terrains anciens des Asturies et de la Galice. Mem. SOC.Geol. Nord France, Lille 2, 630 pp. Bassoullet, P., 1979. Etude de la Dynamique des Stdiments en Suspension dans I’Estuaire de I’Aulne. Thesis, Brest, 137 pp. Berthois, L., 1954. Sur les dtplacements transversaux des eaux trks turbides dans I’estuaire de la Loire en ptriode d’ttiage. C.R. Acad. Sci., Paris, 239: 820-822. Berthois, L. and Auffret G., 1966. Dynamique de la stdimentation dans les rias et les estuaires des petits cours d’eau tributaires de la Manche. Cah. Octan., 18: 761-774. Berthou, PY., 1964. Etude Stdimentologique de la Laita et du littoral voisin de l’embouchure. Thesis, Paris, 177 pp. Birot, P. and Sole Sabaris, L., 1954. Recherches Gtomorphologiques dans le Nord-Ouest de la Ptninsule IbCrique. Mtm. Doc. CNRS, Paris, 4 7-61. Cita, M.B. and Ryan, W.B.F. (Editors), 1978. Messinian erosional surfaces in the Mediterranean. Mar. Geol., 27: 193-363. Clarke, B.B., 1969. The problem of the nature, origin and stratigraphical position of the Trebetherick boulder gravel. Proc. Ussher SOC.,2: 87-91. Clauzon, G., 1973. The eustatic hypothesis and the pre-Pliocene cutting of the RhBne valley. Init. Repts DSDP, 13,2: 1251-1256. Clauzon, G., 1975. Preuves et implications de la rtgression endortique messinienne au niveau des plaines abyssales: l’exemple du midi mtditerranten franqais. Bull. Ass. Gtogr. Franqais, 429: 317333. Clauzon, G., 1978. The Messinian Var canyon (Provence, Southern France). Paleogeographic implications. Mar. Geol., 27: 231-246. Clauzon, G., 1982. Le canyon messinien du RhBne: une preuve dtcisive du “dessiccated deep basin model”. Bull. SOC.Gtol. France, (7) XXIV, 3: 597-610. Clauzon, G., Aguilar, J.P. and Michaux, J., 1987. Le bassin pliocbne du Roussillon (Pyrtntes orientales, France): exemple d’tvolution gtodynamique d’une ria mtditerrantenne consecutive B la crise de salinitt messinienne. C.R. Acad. Sci., Paris, t. 304, Sir. 11, 11: 585-590. Clauzon, G. and Rubino, J.L., 1990. Eustatic control of Pliocene Mediterranean basin morphology and basin filling by Gilbert type fan deltas. IXth RCMNS Congr., Barcelona, pp. 99-100. Codrington, T, 1898. On some submerged valleys in South Wales, Devon and Cornwall. Quart. J. Geol. SOC.,54: 251-278. Coleman, J.M. and Wright, L.C., 1978. Sedimentation in an arid macrotidal alluvial river system: Ord river, Western Australia. J. Geol., 86: 621-642. Coque-Delhuille, B., 1987. Le massif du sud-ouest anglais et sa bordure stdimentaire, etude gtomorphologique. Thesis, Paris, 1039 pp., English abstract pp. 969-989. Cormault, P., 1971. Dttermination expdrimentale du dtbit solide d’trosion de stdiments fins cohtsifs. C.R. 14e Congrks de l’A.I.R.H., Paris, vol. 4, p. D2: 1-8. Cotton de Bennetot, M., 1967. Etude morphologique et stdimentologique de l’estuaire du Goayen. Thesis, Brest, 214 pp. Cotton de Bennetot, M., Guilcher, A. and Saint-Requier, A,, 1965. Morphologie et stdimentologie de 1’Aber Benoit. Cah. Octan., 17: 377-387.

GEOMORPHOLOGY AND SEDIMENTOLOGY OF RIAS

109

Cuevas, A.P., 1990. Utilizacidn de 10s Foraminiferos Bentdnicos y Ostrhcodos para un Mejor Conocimiento del Medio Ambiente en 10s Estuarios Viscainos: Aplicacih a las Rias de Guernica y Bilbao. Thesis, Euskal Herriko Unibertsitatea (Universidad del Pabs Vasco), Bilbao, 345 pp. Dalongeville, R. and Sanlaville, P., 1981. Les marsas du littoral soudanais de la Mer Rouge. Bull. SOC. Languedoc. Gtogr., Montpellier, 15: 39-48. Davis, W.N., 1915. The principles of geographical description. Ann. Assoc. Am. Geogr., 5: 61-105. De la Beche, H.T., 1839. Report on the geology of Cornwall, Devon and West Somerset. Mem. Geol. SUN.

Dewey, H., 1948. South-West England. British Regional Geology, H.M. Stationery Office, London, 72 PP. Drooger, C.W. (Editor), 1973. Messinian Events in the Mediterranean. North Holland Publ. Co., Amsterdam, 270 pp. Dyer, K.R., 1973. Estuaries: a Physical Introduction. Wiley, London, 140 pp. Finckh, P.G., 1978. Are Southern Alpine lakes former Messinian canyons? Mar. Geol., 27: 289-302. Francis-Boeuf, C., 1947. Recherches sur le milieu fluvio-marin et les dtpBts d’estuaire. Thesis, Paris, Ann. Inst. Octan., 196 pp. Glemarec, M. and Hussenot, E., 1981. Dtfinition d’une succession tcologique en milieu meuble anormalement enrichi en matitres organiques i la suite de la catastrophe de 1’Amoco-Cadiz. In: Amoco-Cadiz, Actes du Colloque International, Brest, 19-22 Nov. 1979, CNEXO Paris, pp. 499525. Gonzalez Lastra, J. and Gonzalez Lastra, J.R., 1984. Zonacion ambiental de la ria de San Vicente de La Barquera, Cantabria. Thalassas, 2: 43-48. Gouleau, D., 1975. Les premiers stades de la sCdimentation sur les vasitres littorales atlantiques. R d e de I’tmersion. Thesis, Nantes, 2 t., 241 pp. Guilcher, A., 1948. Le relief de la Bretagne mtridionale. Thesis, Paris, La Roche sur Yon, 682 pp. Guilcher, A., 1955a. La plage ancienne de La Franca, Asturies. C.R. Acad. Sci., Paris, 241: 1603-1605. Guilcher, A., 1955b. Gtomorphologie de I’extrtmitC septentrionale du Banc Farsan, Mer Rouge. Ann. Inst. Octanogr., Paris, 33: 55-100. Guilcher, A,, 1965. Drumlin and spit structures in the Kenmare River, Southwest Ireland. Irish Geogr., 2: 7-19. Guilcher, A,, 1972. La plage ancienne de Castro Urdiales, province de Santander, Espagne, et son int6rCt morphologique. Norois, Poitiers, 19: 365-367. Guilcher, A., 1974. Les rasas: un probltme de morphologie littorale gtntrale. Ann. Gtogr., 83: 1-33. Guilcher, A,, 1976a. Les c6tes rias de Corte et leur Cvolution morphologique. Ann. Gtogr., 85: 641-671. Guilcher, A,, 1976b. Prtsence de plages eemiennes/normanniennes dans I’Ouest de la Rtpublique de Corte et constquences gtomorphologiques. C.R. Acad. Sci., Paris, 282, Str. D, pp. 149-151. Guilcher, A., 1979. Les rivages coralliens de 1’Est et du Sud de la presqu’ile du Sinai. Ann. Gtogr., 88: 393-418. Guilcher, A,, 1982. Nouvelles observations sur les rias naines en forme de caisse (Kastentalrias) de I’ile de Groix (Morbihan). 107e Congr. Nat. SOC.Sav., Brest, Sect. de Gtogr.: 51-59. Guilcher, A,, 1985. Red Sea coasts. In: E.C.F. Bird and M.L. Schwartz (Editors), The World’s Coastline. Van Nostrand Reinhold Co., NY, pp. 713-717. Guilcher, A,, 1988. Coral Reef Geomorphology. Wiley, Chichester, 228 pp. Guilcher, A,, Andrade, B. and Dantec, M.H., 1982. Diversitt morpho-stdimentologique des estuaires du Finisttre. Norois, Poitiers, 114, Vol. 29, pp. 205-228. Guilcher, A and Berthois, L., 1957. Cinq anntes d’observations stdimentologiques dans quatre estuaires ttmoins de I’Ouest de la Bretagne. Rev. GComorph. Dyn., 8: 66-86. Guilcher, A. and Hallegouet, B., 1987. Histoire d’une vallte des environs de Brest. Le Gallo Commem. Vol., Brest, pp. 135-144. Guilcher, A. and King, C.A.M., 1961. Spits, tombolos and tidal marshes in Connemara and West Kerry, Ireland. Proc. R. Irish Acad., 61B, 17: 283-338. Hallegouet, B., 1982. Les formations de remblaiement de la vallte de 1’Elorn i Landerneau, Finisttre. Bull. Ass. Fr. Et. Quat., 19: 167-178.

110

P. CASTAING AND A. GUILCHER

Hallegouet, B., Ollivier-Pierre, M.F. and Esteoule-Choux, J., 1976. Dtcouverte d’un dtpBt Oligoctne inftrieur dans la haute vallte de I’Aber Ildut au Nord-Ouest de Brest. C.R. Acad. Sci., Paris, 283 D, pp. 1711-1714. Hernandez-Pacheco, E., 1950. Las rasas de la costa cantabrica en su segment0 asturiano. C.R. Congr. Int. Gtogr., Lisbonne, 2: 29-86. Hernandez-Pacheco, E. and Asensio Amor, I., 1959/1960. Materiales sedimentarios sobre la rasa cantabrica. Bol. Real SOC.Esp. Hist. Nat., 75-100 and 73-83. Herranz, P. and Acosta, J., 1984. Estudio geofisico de la ria de Muros y Noya. Bol. Ins. Esp. Oceanogr., 1: 48-78. Hsu, K.J., Montadert, L., Bernouilli, D., Cita, M.B., Erikson, A,, Garrison, R.E., Kidd, R.B., Melieres, F., Muller, C. and Wright, R., 1978. History of the Mediterranean salinity crisis. Init. Rep. DSDP, Washington, XLII, 1: 1058-1078. Hsu, K.J., Stoffers, P. and Ross, D.A., 1978. Messinian evaporites from the Mediterranean and Red Sea. Mar. Geol., 26: 71-72. Junoy, J. and Vieitez, J.M., 1989. Cartografia de 10s sedimentos superficiales de la Ria de Foz, Lugo. Thalassas, 7: 9-19. Kidson, C., 1971. The Quaternary history of the coasts of Southwest England. Essays in Honour of A. Davies, Exeter, pp. 1-22. Kwon, H.J. 1974. A geomorphic study of the tidal flats of the West coast, Korea. Geography, 10: 1-12 (in Korean, English abstract). Lautensach, H., 1945. Korea. Eine Landeskunde auf Grund eigener Reisen und der Literatur. Leipzig, 542 p. Le Hir, P., Guillaud, J.F., Bassoullet, Ph. and CYavanc, J., 1990. Application d’un modble stdimentaire au devenir des contaminants particulaires. Actes de Colloques “La mer et les rejets urbains”, Bendor, 13-15 Juin 1990, publ. IFREMER, Paris, 11: 205-220. Li Congxian, Chen Gang, Yao Ming and Wang Ping, 1991. The influence of suspended load on the sedimentation in the coastal zones and continental shelves of China. Mar. Geol., 96: 341-352. Llopis Llado, N., 1956. Los depositos de la costa cantabrica entre 10s cabos Busto y Vidio, Asturias. Speleon, 6: 333-347. Losada, M.A., Medina, R., Vidal, C. and Roldan, A,, 1991. Historical evolution and morphological analysis of “El Puntal” spit, Santander, Spain. J. Coastal Res., 7: 711-722. L‘Yavanc, J. and Bassoullet, Ph., 1991. Nouvelle approche dans I’ttude de la dynamique stdimentaire des estuaires macrotidaux a faible dtbit fluvial. Octanol. Acta, Proc. Int. Colloq. on the Environment of Epicontinental Seas, Lille, 20-22 March 1990, Vol. 11: 129-136. Margalef, P., 1958. La sedimentacion organica y la vida en 10s fondos fangosos de la Ria de Vigo. Invest. Pesqueras, Barcelona, 11: 67-100. Mary, G., 1967. Les niveaux marins fossiles de la rtgion de Otur (Luarca, Asturies). Bull. SOC.Linn. Normandie, 10: 38-52. Mary, G., 1979. Evolution de la bordure cbtitre Asturienne (Espagne) du Ntogtne I’actuel. Thesis, Caen, 288 pp. . Mary, G. and Medus J., 1971. Prtsence de Sparnacien B la base d’une rasa au Monte Granda B I’Ouest d’Aviles, Asturies. C.R. Somm. SOC.Geol. France, 125. Mergner, H., 1967. Ueber den Hydroidenbewuchs einiger Korallenriffe des Roten Meeres. Z. Morph. Oekol., Tiere, 60: 35-104. Monnier, 0. and Guilcher, A., 1993. Le Sharm Abhur, ria rtcifale du Hedjaz, Mer Rouge. Ann. Gtogr., 102: 1-16. Nichols, M.M. and Biggs, R.B., 1985. Estuaries, In: R.A. Davis (Editor), Coastal Sedimentary Environments. Springer-Verlag, NY, pp. 77-186. Nicod, J., 1951. Le p r o b l h e de la classification des calanques parmi les formes de cbtes de submersion. R. Gtmorph. Dynam., 2: 120-127. Nombela, M.A., Vilas, F.V., Rodriguez, M.D. and Ares, J.C., 1987. Estudio sedimentologico del litoral gallego. I11 - Resultados previos sobre 10s sedimentos de 10s fondos de la Ria de Vigo. Thalassas, 5: 7-19. Nonn, H., 1964. Los sedimentos antiguos de la Ria de Arosa. Algunas conclusiones geomorphologicas. Trab. Lab. Geol. de Lage, 16: 143-155.

GEOMORPHOLOGY AND SEDIMENTOLOGY O F RIAS

111

Nonn, H., 1966. Les rtgions c6tiCres de Galice (Espagne), ttude gtomorphologique. Thesis, Paris, Strasbourg, 591 pp. Oliviero, H., 1978. Dynamisme stdimentaire de I’estuaire de la Laita. Thesis, Nantes, 122 pp. Pannekoek, A.J., 1966. The ria problem. Tijd. Kon. Nederl. Aardr. Gen., 83: 289-297. Pannekoek, A.J., 1969. Additional geomorphological data on the ria area of Western Galicia, Spain. Leidse Geol. Med., 37: 185-194. Parga Pondal, I. and Perez Matos, J. 1954. Los arenales costeras de Galicia. I - La Ria de Lage. Ann. Inst. Esp. Edafol. Fisiol. Vegetal, Madrid, 13, 6: 483-513. Pascual, J.R., 1987. Un modelo de circulacidn inducida por el viento en la ria de Arosa. Buletin Instituto Espafiol de Oceanografia, 4, no. 1: 107-120. Perillo, G.M.E., 1989. New geodynamic definition of estuaries. Rev. Geofis., 31: 281-287. Piccolo, M.C. and Perillo, G.M.E., in press. Geomorfologia e hidrografia de 10s estuarios de la Republica Argentina. In: INIDEP (Editor), El Mar Argentino y sus Recursos Pesqueros. Postma, G., 1990. Depositional architecture and facies of river and fan deltas: a synthesis. Spec. Publ. Int. Assoc. Sediment., 10: 13-27. Prego, R. and Fraga, F., 1992. A simple model to calculate the residual flows in a Spanish ria. Hydrographic consequences in the ria of Vigo. Estuarine, Coastal Shelf Sci., 34: 603-615. Prior, D.B. and Bornhold, B.D., 1990. The underwater development of Holocene fan deltas. Spec. Publ. Int. Assoc. Sedimentol., 10: 75-90. Rathjens, C. and von Wissmann, H., 1933. Morphologische Probleme im Graben des Roten Meeres. Peterm. Mitt., 79: 113-117 and 183-187. Rey, J., 1993. Relacidn morpho-sedimentaria entre la plataforma continental de Galicia y las rias bajas y su evolucidn durante el Cuaternario. Instituto Espafiol de Oceanografia, publicationes especiales Madrid, no. 17, 233 p. Ryan, W.B.F., 1978. Messinian badlands on the Southeastern margin of the Mediterranean Sea. Mar. Geol., 27: 349-363. Ryan, W.B.F. and Cita, M.B., 1978. The nature and distribution of Messinian erosional surfaces, indicators of a several-kilometers-deep Mediterranean in the Miocene. Mar. Geol., 27: 231-246. Sainz Amor E., 1962. Estudio granulometrico y mineralogico de 10s arenales de la Ria de Vigo. Res. Cientif. SOC.Espan. Historica Natural, Madrid, pp. 77-92 and 172-194. Scheu, E., 1913. Die Rias von Galicien. Ihr Werden und Vergehen. Z. Ges. Erdk. Berlin, pp. 84-114 and 193-210. Schmidt, W., 1923. Die Scherms an Rotmeerkiiste von El-Hedschas. Peterm. Mitt., 69: 118-121. Schiilke, M., 1968. Morphologische Untersuchungen an bretonischen, vergleichsweise auch an Korsischen Meeresbuchten. Univ. des Saarlandes, Arb. Geogr. Inst., Bd XI, 192 pp. Sestini, J., 1965. Cenozoic stratigraphy and depositional history, Red Sea coast, Sudan. AAPG Bull., 49: 1453-1472. Somoza, L. and Rey, J., 1991. Holocene fan deltas in a “ria” morphology. Prograding clinoform types and sea-level control. Cuad. Geol. Iberica, Madrid, 15: 37-48. Steers J.A., 1964. The Coastline of England and Wales. Cambridge Univ. Press, 2nd ed., 750 pp. Stephens, N., 1966. Some Pleistocene deposits in North Devon. Biuletyn Periglac., 15: 103-114. van Straaten, L.M.J.V., 1954. Composition and texture of recent marine sediments in the Netherlands. Leidse Geol. Med., 19: 1-110. Vilas, F.V., 1983. Medios sedimentarios de transicion en la Ria de Vigo: secuencias progradantes. Thalassas, 1: 49-55. Vilas, F.V. and Nombela, MA., 1985. Las zonas estuarinas de la costa de Galicia y sus medios asociados, NW de la Peninsula Iberica. Thalassas, 3: 7-15. von Richthofen, F., 1877-1912. China, Ergebnisse eigener Reisen und darauf gegriindeter Studien. Berlin, 5 Vols., 2: Atlas. von Richthofen, E, 1886. Fuhrer fur Forschungsreisende. Berlin, Oppenheim (rias: pp. 308-31 0). von Richthofen, F., 1898. Shantung und seine Eingangspforte Kiautschou. Berlin. Wells, J.T., Adams Jr., C.E., Park, Y.A. and Frankenberg, E.W., 1990. Morphology, sedimentology and tidal channel processes on a high tide-range mudflat, West coast of South Korea. Mar. Geol., 95: 111-130.

This Page Intentionally Left Blank

Geomorphology and Sedimentolugy of Estuaries. Developments in Sedimentulugv 53 edited by G.M.E. Perillo 0 1995 Elsevier Science B.V All rights reserved.

113

Chapter 5

SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS JAMES P.M. SYVITSKI and JOHN SHAW

INTRODUCTION

Fjords are unique estuaries which represent a considerable portion of the Earth’s coastal zone. They are both an interface and a buffer between glaciated continents and the oceans, and have a wide range of environmental conditions, both in dynamics and geography. Fjords have unusual environmental problems, for example their (usually) slow flushing time, a feature common to many silled environments. Source inputs are easily identified and their resulting gradients are well-developed. This review aims to provide an overview of the sedimentology and geomorphology of fjords, updated from the more comprehensive earlier reviews of Nihoul (1978), Freeland et al. (1980), Farmer and Freeland (1983) and Syvitski et al. (1987).

CHARACTER

A fjord is a deep, high-latitude estuary which has been (or is presently being) excavated or modified by land-based ice. In Nordic usage, “fjord” is a generic name for a wide variety of marine inlets. Other designators used on marine charts include: loch or lough, lake (e.g. Lake Melville), river (e.g. Saguenay River), sound, inlet, arm, bay, reach and passageway. Fjords and fjord valleys may be considered synonymous features, the only difference being that fjords are submarine. Fjord-lakes are a subset of fjords discriminated by the fact that they contain only fresh water. Fairbridge (1968) advocated the Swedish name “fjard” for shallower, temperatezone fjord-estuaries. Embleton and King (1970) defined fjards as “coastal inlets associated with the glaciation of a lowland coast”. They lack the steep walls of fjord troughs and can be distinguished from rias in having rock basins. The description of Norway’s fjard coast (southern Oslofjord and the Skaggerak) by Bird and Schwartz (1985) differs slightly: “...where an undulating land surface with fissure valleys slopes gently into the sea, making an uneven coastline with numerous islands and islets with headlands and coves”. Fjords are products of the advance and retreat of glacial ice and relative sealevel fluctuations during the Quaternary. They are therefore immature, non-steady state systems, evolving and changing over relatively short time scales. Being partially ice-scoured, the typical fjord configuration (Fig. 5-1) is long, narrow, deep and steep sided, frequently branched and sinuous, but remarkably straight where ice once followed fault lines (Dowdeswell and Andrews, 1985). The fjord valleys are U-shaped, with walls often polished and striated, having formed from the plucking

114 A

B

J.P.M. SYVITSKI AND J. SHAW SINGLE FJORD BASIN

MULTIPLE FJORD BASINS

c

FJORD LANDSCAPE

.

~

.~

,,' BEDROCK ISLAND '.--.-ATROUGH OR BASIN ON CONTINENTALSHELF

Fig. 5-1. Simple features and dimensions of (A) a single-basin fjord cross-section; (B) a multiple basin fjord cross-section; and (C) map view of a fjord hinterland and coast.

action of glaciers on weakened bedrock surfaces and/or through subglacial fluvial erosion by meltwater carrying rock material under high hydrostatic pressure. Hanging valleys often occur as tributaries to the main fjord system. As a class, fjords are the deepest of all estuaries, and typically, but not inevitably, contain one or more submarine sills (Fig. 5-1). The internal basins defined by these sills determine many of the distinctive physical and biogeochemical characteristics of fjords. Sills at the mouth or within the main arm of a fjord may be comprised of exposed bedrock, morainal or other glacimarine deposits, and may appear as a series of islands or shoals, sometimes as a well defined ridge or a more lengthy threshold (Fig. 5-2). They may occur as a result of glacial over-deepening of the fjord basin relative to the adjacent shelf. Some fjords are just beginning to form, e.g. Columbia Glacier in Prince William Sound, Alaska, through the retreat of glaciers that largely fill their submarine basins. Fjords encompass a number of distinctive oceanographic environments: the nearsurface "estuarine zone", basically common to all estuaries, is underlain by marine water which, in silled fjords, may be physically restrained in enclosed basins. The circulation above and below the sill height is often poorly coupled, and, in deep

SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS

115

A

B AEOLIAN INPUT UPRAGLACIALINPUT

Fig. 5-2. Primary sediment inputs to (A) a nonglacial fjord (after Syvitski et al., 1987); and (B) a glacial fjord.

fjords, processes and reactions within the basins may be spatially and temporally separated from those occurring in the upper-zone estuarine environment. The resultant pronounced vertical hydrographic gradients in these deep fjords influence both biota and sediments. Fjords may sometimes contain fully oxygenated water masses at the surface to totally isolated anoxic regions at depth. Sediments derived from the continental shelf and transported into fjord basins are less abundant (Syvitski and MacDonald, 1982; Slatt and Gardiner, 1976) in comparison to other types of estuaries. The limiting factor for fjords is the effective barrier of the outer fjord sill. Additionally, the compensation current is not along the seafloor as in other shallow estuaries, but much closer to the sea surface. Hence it does not erode and transport sediment up-fjord. Biological material such as plankton may be transported into fjords by the compensation current and resulting plankton blooms may initiate a substantial flux of organic matter to the sediments. Greenland fjords, for instance, act as a sink for organic matter that largely originates from shelf waters (Petersen, 1978). Fjords have also acted as efficient sediment traps in recent geological times, retaining perhaps one quarter of the fluvial sediment delivered to the world ocean over the last 100,000 years (Syvitski et al., 1987). They exhibit a very wide range of sedimentation rates, from the highest recorded natural marine values, to rates approaching those characteristic of deep-sea basins. Fjords experiencing high rates of sediment accumulation are associated with ice-influenced hinterland erosion, and often exceptional high rates of uplift. Sediment inputs to temperate zone fjords include those from river and wind transported terrestrial sources, anthropogenic sources,

116

J.P.M. SWITSKI AND J. SHAW

Table 5-1 Parameters affecting fjord sedimentation (after Syvitski et al., 1987) A. Glacial

relative sea level history wet versus cold based glaciers floating versus tidewater versus hinterland glaciers style and rates of glacier advances and retreats basal shear stress

B. Fluvial

transport rates of bedload, suspended and dissolved loads runoff characteristics (e.g., jokolhlaup events) paraglacial history stratification and turbidity

C. Climatic

glacier movement including iceberg production sea ice conditions thermal stratification wind events (waves, upwelling, aeolian transport) terrestrial and marine biomass production

D. Geographic

fetch length fjord dimensions (e.g., basin and sill depths, width, volume) relative sea level history tidal characteristics Coriolis effect flushing dynamics

E. Geotechnical

frequency and size of slope failures mass transport process seiches and tsunami waves.

continental shelf sources and internal fjord sources (Fig. 5-2A). Ice-dominated fjords have additional sediment input sources (Fig. 5-2B). Fjord deposits have a good potential for providing a comparatively highresolution sedimentary record that reflects both local terrestrial and marine processes (Table 5-1). Stratigraphic interpretation of proxy climatic and paleoecologic signals, contained in well-dated and unbioturbated marine cores, can provide insight into the impact of past climatic and environmental conditions (Andrews and Syvitski, 1994). The combination of low salinity estuarine waters and high sedimentation rates common to fjord deltas results in an impoverished macrofauna such that physical structures tend to remain intact. For convenience in this chapter, we provide details on five sedimentological endmember fjords. As a word of caution, however, individual fjords often have more complex attributes. Additionally, fjords are not steady-state systems and may evolve from characteristics closer to one end-member group to those of other end-member groups later on. Our first end-member fjord is dominated by glacier ice and icemelt processes, in particular the discharge of submarine sediment-charged plumes, iceberg calving and ice rafting. Sediment input mechanisms (Fig. 5-2B) include: (1) supraglacial material (slumping off medial and lateral moraine till, supraglacial streams); (2) englacial materials (crevasse fills, englacial streams, and other englacial

SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS

117

sediment); (3) basal material (lodgement till, waterlain till, push and surge deposits); (4) iceberg rafted sediment; (5) sand and loess, blown off ice surfaces and along kame

terraces; and (6) lateral (kame) deltas. Our second end-member fjord is river influenced, often with its main fluvial input at the head, with only minor contributions from side-entry drainage basins. Fjord rivers transport erosional products from weathering, reworked glacigenic and raised marine deposits, and freshly produced glacial flour (Elverhoi et al., 1980). Additionally, temperate fjord rivers with dense vegetation cover in catchment areas supply terrestrial organic matter such as leaves, twigs and humic substances (Glasby, 1978). The grain size of the fluvial sediment will vary according to the parent material, the extent of erosion, the inclination of the river, the energy of the river water and the filtering effect of lakes. Hence, the sediment source material may range from clays to boulders. The coarser fluvial sediment is deposited within valleys as sand and gravel plains (sandur) and at fjord margins, forming deltas and outwash fans. An exponential decrease in sedimentation flux and particle size away from river sources is often observed with biogeochemical interactions controlling the vertical flux of suspended particles. Fjord sediment is composed predominantly (>95%) of inorganic particles, derived mainly from these fluvial sediment sources. Annual suspended load carried by fjord rivers can range from lo7 tonnes for large British Columbia rivers to lo4 tonnes for smaller Baffin Island rivers (Milliman and Syvitski, 1992). Our third end-member group comprises wave- and tide-influenced fjords, in which Holocene sediment deposits are largely sourced from the reworking of Pleistocene deposits, with sediment flux controlled by current or wave exposure and water depth. These fjords may receive t 2 5 % of their sediment fill from rivers. Here the main sediment supply is derived from waves or tidal currents reworking coastal deposits of older marine or glacigenic sediment. Cliff retreat rates can exceed 1 m ax1 near fjord mouths, decreasing to 25 cm acl in exposed inner-fjord areas (Piper et al., 1983). Fjords influenced by slope failure and mass sediment transport processes constitute our fourth end-member group. They may contain sediment displaying diverse sediment yield strengths, and in combination with variable basin morphology may provide for the development of a spectrum of elastic, plastic, and viscous subaqueous failures, triggered by a range of external factors. Fifty percent of the sediment fill within Hardangerfjord, Norway, for instance, is a result of slumps and turbidity currents (Holtedahl, 1965). Slope failures can occur near the fjord-head delta, the sidewall slopes, side-entry deltas, off sills and at junctions with tributary (hanging) valleys. Slide volumes may range from very small ( < lo 3 m3) to very large (> lo9 m3). The frequency of slope failures is controlled by the local rate of sediment accumulation and the frequency and force of the triggering mechanisms, and may range from annual events to rare catastrophic events. Anoxia-influenced fjords are the fifth end-member group. In certain near-stagnant fjords, a secondary source of sediment is from the precipitation of inorganic substances such as oxides, hydroxides, carbonates and sulfides. These substances are controlled by changes in redox conditions and pH. Iron and manganese may form oxidised precipitates above the redox boundary (Jacobs et al., 1985), while the same

118

J.P.M. SWITSKI AND J. SHAW

elements may form sulfides and carbonates, respectively, in the anoxic environment (Suess, 1979). Other end-member fjords exist, but are much rarer. For instance, in the polar desert regions of the Canadian Arctic Archipelago the contribution of wind-blown sand and silt to fjords is considered more important (Gilbert, 1983; Syvitski and Hein, 1991). Also during the last century, man-made products have played a significant role as a source of sediment in estuaries and fjords. Discharge of organic substances from sewage plants, pulp and paper mills, and of various solid wastes from the chemical and mining industries, has led to increased rates of deposition in some fjords (Pearson and Rosenberg, 1976; Nyholm et al., 1983; Skei et al., 1972). Population and industrial activities are traditionally concentrated along fjords, a result of near-perfect port conditions (deep water near the shoreline and limited fetch). In some extreme cases (Jossingfjord, Norway) the fjord bottom is entirely covered by industrial waste, up to sill depth (60 m from an initial basin depth of 96 m: Syvitski et al., 1987).

OCEANOGRAPHIC CHARACTERISTICS

Two-layer flow with entrainment of marine water into the surface plume has become synonymous with fjord-circulation: an outward flowing surface layer and an inward moving compensating current, replacing salt entrained into the surface zone. The force responsible for maintaining the flow of brackish water towards the sea originates from the pressure field associated with the seaward sloping free surface (Gade, 1976). Estuarine circulation is further complicated by: the Coriolis effect that forces flow to the right in the northern hemisphere; the centrifugal force, important along sinuous fjords; flow accelerations developed over major bathymetric elements and inlet constrictions;pressure gradients developed from meteorological conditions (changing wind structure or fresh water discharge); surface mixing from strong winds; energetics of breaking internal waves; and isohaline instabilities developed during the process of salt rejection during sea-ice formation. Many temperate fjords alternate between two-layer “fjord-style’’ circulation operative during the spring (snow-melt discharge), summer (ice-melt discharge) and fall (rain-storm discharge), and vertically homogeneous estuarine conditions of the winter (residual ground water discharge). In the polar regions where runoff is limited to a few months, fjords lack estuarine circulation for a large portion of each year. Deep fjords, in addition to the simple two-layer circulation, may have deeper circulation cells (Carstens, 1970) with alternating current directions. This complicates the dispersal of sediment (Syvitski and MacDonald, 1982). Multilayered currents may involve the entire water body in the fjord, in that they are frictionally controlled and sometimes frictionally driven (Gade, 1976). Multilayered circulation can form from current interactions with the sill, from other buoyant inputs from outside the fjord, and from wind stress within and outside the fjord. Wind-forced coastal circulation, with its geostrophic longshore currents, has a strong effect on circulation within the fjord. These geostrophic currents control the free surface and pycnocline displacement at the fjord mouth, thereby strongly affecting fjord circulation (Klinck et al.,

SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS

119

1981; Svendsen, 1977). Where the inflow from outside the fjord is volumetrically greater than the seaward-flowing surface layer, a strongly developed subpycnoclinal outflow results (Ebbesmeyer et al., 1975). Sills play an important role with respect to water structure, circulation, sediment transport and biological life in fjords. Shallow sills hinder a free exchange of water with the open ocean (Fig. 5-2). In extreme cases, this may lead to stagnancy in the deeper parts of the water and depletion of oxygen (Richards, 1965). This may have serious consequences for all macroscopic life, and causes a complete change in the chemistry of the water and sediment. Replenishment of a fjord’s deep water with more oxygenated shelf water is governed by density differences, meteorological conditions, internal waves and fjord geomorphology. Deep water renewal may take place when the density of the water at sill depth exceeds the density of the basin water inside the sill. This may occur frequently, annually or seldom. The sill depth and the density of the outside water are the critical factors. Some fjords have several sills and basins, and as the deep water overspills the first sill it gradually gets mixed with less saline water. Consequently, the innermost basins may not experience a deep water renewal. Deep water renewals do not generally exchange the entire volume of fjord water; replacements of 20 to 80% of the basin water are more common (Molvaer, 1980). Fairly intense vertical mixing of basin waters may take place (Gade, 1968), tides being a key source of energy, generating internal waves at the sill which subsequently convert to turbulence. The temperature structure in the shallow parts of fjord waters is not very different from that of other estuaries. A thermocline often corresponds with a halocline creating a strong pycnocline in the near surface water. The position of the thermocline may vary seasonally, with changes in the air temperature and the temperature of the river runoff. In deep fjords, basin water is distinctly different from that in other types of estuaries, remaining a more constant temperature year-round. The temperatures of the deeper basin waters depend on the temperatures of coastal shelf waters of similar depth.

WORLD DISTRIBUTION

Fjords are predominantly features of mountainous coastal regions which presently support, or have supported in the recent past, ice fields and valley glaciers. They have a world-wide distribution at mid to high latitudes: a belt north of 43”N and a belt south of 42”s. The principal fjord provinces occur along the coasts of North and South America; the Kerguelen Islands, South Georgia, the Russian and Canadian Arctic archipelagos, Svalbard and other high-latitude islands; the southwest coast of New Zealand’s South Island; Antarctica; Iceland and Greenland; and northern Europe, including the British Isles above 56“N. Some fjords are more “typical” than others, showing characteristic features which fit the definition of fjords. Other high-latitude estuaries are less fjord-like, exhibiting only a few of the characteristic features, but their overall natural setting allows their classification as fjords. Table 5-2 provides the salient and generalised (there are notable exceptions) characteristics

J.P.M. SYVITSKI AND J. SHAW Table 5-2 Generalized characteristics of the world's major fjord coastlines (after Syvitski et al., 1987) Fjord district

Number of fjords

Fjord stage

Tidal range

River discharge

Climate

Greenland

350

1,2

low

medium to high

subarctic to arctic maritime

0 to 2

medium to high

Alaska

200

1-4

high

low to high

subarctic maritime

3 to 7

medium to high

British Columbia

150

3, 4

high

medium to high

temperate maritime

6 to 9

medium to high

Canadian maritime

200

4,5

low to medium

low to high

subarctic to temperate maritime

-1.5 to 3

low

Canadian arctic

350

1-4

low to high

low to medium

arctic desert to arctic maritime

-1.5 to 0

low to medium

Norwegian mainland

200

3, 4

low

low to medium

subarctic to temperate maritime

Svalbard

35

2,3

low

low

arctic desert

-1 to 2

medium

New Zealand

30

4,5

medium

low to medium

temperate maritime

10 to 12

low to medium

200

2-4

low

low to high

temperate to subarctic maritime

6 to 9

medium to high

50

4, 5

low to high

low

temperate maritime

5 to 13

low

Chile Scotland

Basin water temperature

6 to 8

Sedimentation rate

low

Stage 1: glacier-filled; 2: retreating tidewater glaciers; 3: hinterland glaciers; 4: completely deglaciated; 5: fjords completely infilled. Low: t 2 m mean range; medium: 2-4 m; high: 2 4 m mean range. Low: 200 m3 s-l. Average water temperatures ("C)at or near the 200-m depth of fjord basins. Low: 4 cm a-l with average rates of 2 cm a-l (Haselton, 1965; Goldthwait et al., 1966; Matthews, 1981). The rate of ice terminus retreat or advance will impact on the accumulation of sediment at the ice front, whether from melting, from discharge, or from the calving of bergs. For a quasi-stable ice front position, sediment deposition will decrease rapidly with distance from the ice front. For an unstable ice position, sediment accumulation will be largely controlled by the rate of ice terminus retreat or advance (Powell, 1991).

130

J.P.M. SYVITSKI AND J. SHAW

Land-based fjord valley deposition Fjord valley glaciers carry basal debris derived from the subglacial bed, and, at higher levels (including their surface), debris derived from flanking mountain walls. The residual deposits comprise subglacial till, englacial eskers, supraglacial moraines and kames, and proglacial outwash in front of dump moraines. Sedimentary architecture depends on whether a glacier is advancing or retreating (Boulton and Eyles, 1979), and these glacigenic deposits often form the base of a fjord’s fill. When a valley glacier is stationary or advancing, deposition occurs along the ice margins and terminus, and at the sole of the glacier. Dump moraines accumulate as scree from the steep glacier front in association with mud flows and waterwashed sediment. If the supraglacial till cover is thin, the material is slumped off during retreat as a relatively thin and sporadic veneer over the progressively exposed subglacial surface (lodgement till, bed rock, or outwash). The thickness of the veneer is proportional to the rate of ice terminus retreat and ice velocity. If the supraglacial till cover is thick enough to slow the melting rate of the underlying ice, hummocky stagnation topography results. Melting of buried ice results in a pitted kame plain or outwash surfaces. The rapid buildup and decay of stream discharges has a strong influence on the character of glacifluvial sediments. The derived sediment closely resembles the parent till material, as all particles are transported and deposited en masse. Glaciolacustrine deposits are not uncommon in fjord valleys. The lakes are usually found in bedrock depressions formed during the glacial advance and exposed during retreat. Latero-frontal dump and push moraines, where extensive, can also form dams for valley lakes. Lake depths can vary from a few tens to several hundred meters, and often form in contact with the glaciers (0strem, 1975: Gustavson, 1975). Proglacial lakes remain turbid during the melt season, and sedimentation processes cover the lake floor with varved deposits: coarse-grained layers related to summer discharge maximum and finer-grained layers related to the lower discharge periods (Church and Gilbert, 1975; 0strem, 1975; Pickrill and Irwin, 1983). Varved proglacial lake deposits are apt to contain ice-rafted particles of all grain sizes which have been spread sporadically onto the lake floor.

Sea-ice influence The development of a winter ice cover leads to the establishment of a homogeneous surface layer due to the process of salt rejection from the freezing ice mass (Gade et al., 1974). As salt rejection continues, vertical mixing reaches increasing depths, eventually leading to gravity flows to the middle and lower layers (Lewis and Perkin, 1982). The onset of spring causes a cessation in ice growth and vertical circulation drastically decreases until ice break up (Lewis and Perkin, 1982). Duration and thickness of the ice cover depend on a variety of oceanographic and meteorologic conditions, but both generally increase with latitude. The higher latitude fjords may even be under permanent ice cover and are noted for their weak currents (Lake and Walker, 1976).

SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS

131

An important sedimentological consequence of sea ice is its ability to raft sediment. Fjord sediment can accumulate on or within sea ice by: (a) wind action, (b) stream discharge, (c) rock fall, (d) seafloor erosion, (e) wave and current wash-over, and (f) bottom freezing. Silts and sands transported from a sandur surface by aeolian action can be deposited on ice during winter storms (Gilbert, 1980, 1983). Nival melt can occur prior to the melt of sea ice and even before shoreline leads have had an opportunity to develop. Stream waters loaded with sediment may flood across the ice at high tides; high river discharges can deposit fluvial sediment a considerable distance over the still frozen fjord surface (Knight, 1971). Rockfalls, slides and dirty avalanches, released from the fjord walls by hydrofracturing during intervals of frequent freeze-thaw cycles (spring), supply colluvium to the ice surface along the entire length of a fjord (Gilbert, 1983). Drift-ice may become embedded with sediment at its base when dragged over intertidal flats with the rise and fall of the tides. Contemporaneously, waves and currents can wash considerable sediment onto the top of ice floes trapped on the intertidal flats, especially during break-up (Gilbert, 1983, 1990). Freezing of sediment to the base of ice in meso- and macrotidal environments has been recognized for some time (Gilbert, 1983). Large boulders are more likely to be pushed instead of rafted (McCann et al., 1981). It is expected that much of the ice rafted sediment is deposited reasonably close to the point where it came to rest on the ice surface: melting sea ice within a fjord shows little mobility during break-up (Gilbert, 1983, 1990). Ice-rafted boulders are ubiquitous within hemipelagic sequences in polar cores, although their distribution is unpredictable.

RIVER-INFLUENCED FJORDS

Many of the processes and products in fjord systems are closely related to the movement of water and sediment down fjord valleys. Often the rate of sediment accumulation is directly related to river dynamics. Fjord circulation and the transport of sediment are commonly dependent on the hydrological cycle. Herein we review the hydrological cycles common to fjords, sediment transport by fjord rivers, the general characteristics of fjord deltas, and the consequences of river plume generation and sedimentation.

Fjord river discharge The balance of water in a drainage basin is the simple balance of inputs and outputs with a slight modification for changes in storage, such as those caused by ice jams, log jams, sudden drainage (jokolhlaups), or the mass balance of a hinterland ice sheet. The spectrum of fjord alpine river hydrographs includes the following: (1) Arctic, nonglacial, nival regime: a large spring discharge from snow melt followed by lower summer flows punctuated by periodic rain-storm floods that are induced orographically. Lag between rainfall and river-mouth discharge maxima is of the order of minutes; this is significantly shorter than lags of hours or days characteristic of larger and lower latitude basins. Arctic rivers that have a glacier

132

J.P.M. SYVITSKI AND J. SHAW

melt component also receive peak discharges in the late summer related to air temperature. (2) Maritime, nonglacial, pluvial regime: the precipitation discharge is moderated by lakes, and thus the response time between peak rainfall and peak discharge is of the order of one or two days. Hydrograph peaks are directly related to precipitation events, thus the lowest discharge occurs during the dry summer months. (3) Continental nival regime: a large drainage basin with stable winter-time snow storage generates a large spring freshet often followed by shorter duration discharge events during a wet autumn. River flow is commonly year-round as a result of large ground-water and lake storage capacity. The hydrograph is considerably smoothed by the river’s slow response time. (4) Alpine, pluvionival, proglacial regime, with discharge peaks in early summer from snow melt, followed by glacier melt in mid to late summer. Proglacial regimes exhibit a discharge that continues to rise until late summer, as progressively higher zones on the glacier melt and become effective contributing portions of the watershed. A common hydrological phenomenon is the devastating flash flood, particularly in the autumn when an early frost is followed by heavy snowfall, rapid thaw, and warm rain. The resulting rapid runoff of surface water is unable to permeate the still-frozen ground. A rare flash flood might discharge 30 times more than the mean annual flood discharge. Discharge from glacier melt depends on the ablation characteristics of the individual ice field and is thus highly variable between drainage basins. The Decade River, flowing into Inugsuin Fjord, Baffin Island, drains a basin that is 68% glacier covered, yet precipitation appears to control the discharge hydrograph (0strem et al., 1967). At the other extreme, the Jostedal River, draining into Gaupnefjord, Norway, has only 27% of its watershed covered by glaciers. Here the runoff responds more directly to the glacier melt with a distinct diurnal periodicity (Relling and Nordseth, 1979). Proglacial rivers are also prone to sudden releases of water Cjokulhlaups) from ponds or lakes held back temporarily behind ice or snow dams. When the dam is breached, the peak discharge is great, up to 50,000 m3 s-l in the 1934 Grimsvotn jokulhlaup, Iceland (Nye, 1976). The amount of energy released during such an event is enormous ( 1019J over a few days or weeks (Tomasson, 1991).

Sediment transport Bed-load transport is controlled by stream discharge, hydraulic slope, bottom roughness, bed compaction, and grain properties. Bed-load transport can range from less than 5% of the total sediment load for lowland fjord-valley rivers to 55% for proglaciaI mountain streams (0strem et al., 1970; Church, 1972; Ziegler, 1973; Adams, 1980; Syvitski and Farrow, 1983; Bogen, 1983). The highest percentage of bed-load transport has been found in arctic proglacial fjord-sandur (Church, 1972). Bed-load deposition is rapid once the velocity of a stream falls below a corresponding threshold value for deposition of a particular grain diameter. Since many discharge events in fjord-rivers are short lived, bed load particles move stepwise down-valley in “trains” that would be remobilized only when a new discharge event of equal or

SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS ICE-MELTSEDIMENT SOURCE

133

SNOW-MELT SEDIMENT SOURCE

Fig. 5-6. Time dependence of fluvial rating curves that compare a river’s suspended sediment concentration and discharge (Syvitski et al., 1987)

greater magnitude occurs. Often this occurs during the annual flood event, but many years may pass before remobilization of a train, particularly if the train is lower down in the valley where threshold river velocities are seldom reached. Bed-material load is also dependent on stream discharge: as discharge increases, so does the quantity and coarseness of the suspended load material. Wash load is highly dependent on source area and supply conditions. Therefore suspended sediment discharge cannot be theoretically predicted from water discharge. Generally, the suspended sediment concentration (or its discharge load) increases exponentially with increasing stream discharge. The rate of increase is highest for glacial streams, lower for lowland streams draining silt and clay deposits (a function of the erodibility of the sediment), and lowest for high mountain streams because of restricted access to fine-grained material (Nordseth, 1976). Proglacial streams may transport 60 to 70% of their annual sediment yield during one day (Nordseth, 1976; 0strem et al., 1967). The rate of change may also change with the season as a result of new sources or changes in the sediment supply. For instance, nival rivers having a marked spring freshet have the greatest sediment yield in the spring (Fig. 5-6B), with the erosion of the recently weathered winter fines. The pattern is reversed for proglacial streams (Fig. 5-6A) with suspended concentrations increasing proportionally as the contribution of glacial meltwater increases in the late summer (Syvitski et al., 1987).

Fjord deltas The subaerial deposits of fjord deltas are controlled by: (1) the strength and periodicity of the fluvial discharge; (2) the river thalweg slope (gravity potential energy); (3) climate (periglacial vs. temperate conditions); (4) relative sea-level history; (5) sediment supply; (6) wave energy and direction; (7) tidal energy; and more rarely (8) tectonic activity. Fjord deltas have unique morphologies which reflect variable responses to these factors and basin accommodation space. Two broad categories of fjord deltas have been recognized (Syvitski et al., 1987): (1) wet, temperate deltas having features common to their open ocean counterparts; and (2) high-latitude deltas (sandur) strongly influenced by their lack of stabilizing vegetation, by glaciers, and by unique periglacial landforms. Sandur are not exclusive to high-latitude fjords, but they share many of the same features of arctic fjord deltas. Common features include strong winds, incomplete

134

J.P.M. SYVITSKI AND J. SHAW

vegetation cover, intermittent discharge pattern, and high competence resulting in large bed-load transport during short-lived events. Sandur are alluvial outwash plains undergoing rapid aggradation; they are crossed by braided streams that continually shift their pattern and course as local erosion and deposition occur (Church, 1972). Higher latitude fjord deltas have periglacial landforms developed through the response to intense frost, permafrost, nivation, strong winds, incomplete vegetation cover, and intermittent discharge pattern. Landforms may include frost-heaved boulder surfaces, ice-wedge and sandfilled polygons, and pingos. Low precipitation, freeze-drying of exposed sediment, sparse vegetation cover, and strong winds combine to make aeolian transport of sediment an important modifier on sandur deltas (Gilbert, 1983; McKenna-Neuman and Gilbert, 1986). The main season of aeolian action for the eastern Canadian Arctic is winter, when the sandur surface is dry and erosion is unrestricted. Fluvial transport of bed load dominates the development of sandur (Church, 1972), and flood events dominate the discharge pattern owing to the very high proportion of surface runoff. Between 25 and 75% of the total sediment transport may occur during the 4 or 5 peak flow days (Church, 1972). During a flood event, local aggradation causes channel division and braiding. Sandur surfaces consist of amalgamated flood deposits of river bars and channel fill, sandur levees, and sheet deposits. Grain size decreases and sorting increases toward the sea, yet there is a lack of pattern in the fines. The distal end of these periglacial deltas is mostly a continuation of the valley floor into the sea, especially for fjords having a low tidal range (Fig. 5-7). Deposition at the sandur delta front, although localized to the area around the river mouth, often extends relatively uniformly across the fjord width as a result of frequent channel switching. Temperate-fjord deltas, being both warm and wet, support a dense vegetation cover in their upriver valleys, usually a mixture of conifers and deciduous trees. The vegetation is partly successful in stabilizing river banks, and flood-derived driftwood may work to stabilize the delta surface. As a result, river channels are both deeper and narrower than those on arctic sandur. Vegetation and a wet climate limit aeolian transport. Temperate-fjord river channels widen and shoal toward the sea, resulting in a rapid decrease in bed load transport toward the river mouth. High discharge events result in levee development, crevasse-splay formation, and flood-plain deposition (Fig. 5-8). The delta plain can be divided into supratidal and intertidal components (Kostaschuk and McCann, 1983). Supratidal deposits develop over a forested plain during periods of high discharge. The intertidal length is a simple function of tidal range and river thalweg slope. Bell (1975) divided the fjord temperate delta intertidal zone into: (1) an upper tidal flat zone that marks the transition of marsh to forest, where sediment is deposited during flood-tide stage and horizontal (silty) laminations are preserved; (2) an intermediate zone, where sedge and grass trap fine silts and clays during periods of high tide and low river runoff - local bioturbators are present; and (3) a lower zone of mouth bar and sand flats that are reworked by tidal and wave forces - bioturbation is noticeably absent as a result of rapid sedimentation.

135

SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS

0

c,

km

0

DISTANCE (mi

50

km

0

DISTANCE(m)

50

jlw

Fig. 5-7.Itirbilung Fiord delta with bathymetry (in metres) superimposed on a NAPL air photo showing location of examples of hydrographic (echosounder) lines. Note the channels cut into the seafloor as seen on the sounder lines.

At high tide, distributary bars may form farther up the channel, where the sea water intrudes as a salt wedge along the river bed. The liftoff point at the head of the salt wedge is a place of rapid bed load deposition where a broad radial distributary bar may form. Over the bar there is a seaward transition from higher energy to lower energy bed forms with a concomitant decrease in grain size. This reflects the deceleration of the river over the distributary mouth bar (Kostaschuk and McCann, 1983). The low tide outlet has one or more distributary mouth bars that extend across the channel mouth: the bars slope gently landward and steeply seaward. The bars form on the leading edge of the delta and become subaerially exposed only during extremely low tides. The proximal part of the bar is composed of imbricate gravel grading distally into straight crested ripples of medium sand. Distributary mouth bars are ephemeral features (Syvitski and Farrow, 1983),and their positions may change from year to year.

136

J.P.M. SWITSKI AND J. SHAW

SETS

CUMBlffi RIPPLE CRDSS LAMINATDN

BURROWS

BUBBLE CAVITIES

*ccRETK)N

LAMINATED SAND HERRlffiBMlE CROSS LAM ERDSiVE UffiwlD MEEW RIPRES CUMBlffi RIPPLECROSS LAMINATDN

STACKED WAVE RIPPLES

@

--

CHANNLBAR W

E BED

m u n , m

-

0 LEME BUBSLE CAVITIES LlffiwlD MEMIPREB

0

STACKOD WAVE RIPPLES

Fig. 5-8. Intertidal zone and prodelta bathymetry of the Homathko delta (Bute Inlet, British Columbia) (left) and Klinaklini delta (Knight Inlet, British Columbia) (right). Note that the submarine channels of the Klinaklini prodelta line up closely with river distributaries and the channels of the Homathko prodelta stem from arcuate scarps (after Syvitski and Farrow, 1983). Also shown are the variable lithostratigraphy of five l-m long deltaic box cores.

Fjord river plumes Discharge of freshwater initially creates a hydraulic head near the river mouth and the effluent effectively flows downhill towards the sea. The gradient is calculated from the level or geopotential surface and the actual surface, and is typically of the order of 1 mm km-' (Farmer and Freeland, 1983). As the surface water flows seaward, it entrains marine water into its outflow (Fig. 5-3A). Surface layer turbulence arises initially from river flow instabilities and later by interlayer friction-induced turbulence, breaking of internal waves along the boundary between the two layers, and wind-induced surface turbulence. Entrainment of saline water is the process of one-way transport of fluid from a less turbulent to a more turbulent region. The effects of entrainment and acceleration balance to maintain a relatively uniform thickness of the surface layer along the fjord (McAlister et al., 1959). As saline water

SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS

137

is entrained into the outward flowing surface layer, new sea water must enter the fjord at depth. The return or compensating current is driven by a reverse internal pressure gradient arising from the sloping density field (Gade, 1976). It is generally assumed that the internal (baroclinic) pressure balances that of the sloping free surface (barotropic). Most fjord-valley rivers have relatively steep bed slopes. Thus these rivers tend to flow turbulently into the fjord (McClimans, 1978a). As a result, near the river mouth, the surface layer of the fjord is well-mixed, often surrounded by a brackish layer. The river plume spreads laterally to a width determined by down-fjord narrows. During its lateral spread, the surface plume passes through a zone of deceleration (Kostaschuk and McCann, 1983), a function of both spreading and mixing between the discharged river water and the surrounding brackish layer (McClimans, 1979). In the outer fjord, river plume circulation may also be influenced by the effects of the Coriolis force (which increases with latitude), centrifugal acceleration (particular to sinuous fjords), topographically-induced vorticity shedding, wind and tides. The surface plume may migrate from shore to shore and vary greatly in character. The surface waters become distinctly stratified, with salinity increasing seaward and downward. Wind or tidal interactions on an irregular shoreline can also induce vortices that incorporate freshwater into the brackish layer (Yoshida, 1980). Tidal currents may reverse the direction of the surface layer in a complex pattern (Huggett and Wigen, 1983), especially during periods of low discharge. Where opposing river plumes occur, shear between them can result in a three-dimensional current structure (McClimans, 1978a). Up-inlet winds can also impede or reverse the surface outflow, and even result in opposing cores of brackish water (Buckley and Pond, 1976). The direction of the surface layer, in the outer portions of some fjords, is best related to wind direction except in cases of high runoff (Farmer and Osborne, 1976; Buckley and Pond, 1976). Prolonged down-inlet winds can also remove the surface layer in a fjord (Hay, 1983), or in the case of up-inlet winds, pile the surface layer up onto the fjord-head delta (Farmer and Osborne, 1976).

Hemipelagic sedimentation The sediment load carried by a river separates into two components seaward of the river mouth bar. The bed-material load settles quickly onto the delta foreset beds, while the wash load is carried seaward within the river plume. The wash load is composed mostly of sand to clay-size mineral grains, and is often referred to as glacial or rock flour. These suspended particles undergo enhanced settling while mixing with the ambient saline water. The settling enhancement is initially due to flocculation, which begins within the brackish waters of a fjord plume. Once particles have joined together, the settling velocity of flocs is greater than that of their individual components. Flocculated particles may settle through the water column of a fjord in a matter of days, even though the water depth may be hundreds of metres. Particles smaller than 10 pm attain settling velocities of around 100 m day-' (Syvitski et al., 1985). This settling rate is some 10 to 1000 times larger than if the particles settled solo and as predicted by Stoke's Settling Theory (cf. Syvitski, 1991).

138

J.P.M. SYVITSKI AND J. SHAW

Fig. 5-9. Log-log plots of the size-frequency-distribution of: ( S P M ) suspended particulate matter (after Syvitski et al., 1985). Note the phytoplankton mode (20 pm) appearing in more seaward samples; ( S T ) material collected by sediment traps anchored above the seafloor (after Relling and Nordseth, 1979; Syvitski and Murray, 1981); ( B T ) seafloor samples collected for a river-influenced fjord (after Syvitski and MacDonald, 1982; Schafer et al., 1989; Syvitski and Hein, 1991; Hoskins and Burrell, 1972; Gilbert, 1983; Holtedahl, 1975).

For sand-sized particles greater than 100 p m in diameter, Reynolds Drag Law holds (cf. Syvitski, 1991). Particle settling near a fjord river mouth is also affected by the fluvial and tidal stage (Hoskin and Burrell, 1972; Hoskin et al., 1976, 1978; Phillips et al., 1991). The clay and very fine silt fractions are well stratified and confined mostly to the surface layer (Fig. 5-9). However, the medium and coarse silt fractions are able to breach the stratification, and thus are more influenced by the tidal stage and discharge dynamics (Syvitski et al., 1985). Away from the river mouth, the vertical flux of particles is controlled more by biogeochemical interactions such as planktonic pelletization of fine particles, flocculation (which occurs within rather than below the surface plume in contrast to the proximal zone), and agglomerative processes including the role of bacteria. In marine water, the flocs may continue to increase in size eventually developing into particles coated with mucous and suspended debris (Syvitski et al., 1985). At depth the filaments may form from bacterial growth on decaying planktonic fecal pellets. The down-fjord sedimentation rate decreases exponentially with distance from the river mouth (Hoskin et al., 1978; Relling and Nordseth, 1979; Smith and Walton, 1980; Syvitski and Murray, 1981; Bogen, 1983; Fig. 5-1OA). The sedimentation rates reflect the exponential decrease in SPM concentrations with distance from the source. In a silled fjord environment the settling path of a floccule has a near-vertical residual descent path once the particle has escaped the surface layer (Syvitski and MacDonald, 1982). Thus, changes in SPM concentrations within the surface layer will be reflected in the rates of sedimentation. There is also a close relationship between seasonal fluctuations in suspended sediment levels within the surface layer, seafloor sedimentation rate and mean grain size (Syvitski and Murray, 1981; Syvitski and Lewis, 1992; Fig. 5-10B).

SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS

A

139

GAUPNEFJORD 1oM

U

z

II lca

co

0 JULY 2-19.1976

0 JULY 2-19.1976

0 JULY 9-AVG 8,1976

0 JULY 9-AUG 8.1976

10 0.5

1.0

1.5

2.0

2.5

B

0.5

1.0

1.5

2.0

2.5

DISTANCE (km)

DISTANCE (km)

HOWESOUND 6E m

70

-

1977

Fig. 5-10. (A) The exponential decrease in sedimentation intensity (in g mP2 day-' determined from sediment traps anchored above the seafloor) and mean grain size (in pm) with distance out from a river mouth in Gaupnefjord for early and mid-summer (after Relling and Nordseth, 1979). (B) Seasonal variations in SPM concentration, sedimentation rate and mean grain size of sedimented material as observed in Howe Sound, B.C. (after Syvitski and Murray, 1981).

The exponential decrease in sedimentation flux away from a river source is associated with a concomitant decrease in the size of particles that settle out (Figs. 5-9 and 5-10). The size frequency distribution effectively changes from one of a coarse size mode with a fine-grained tail nearest the river mouth, to one of a fine size mode with a coarse-grained tail farthest from the source. In other words, fallout is dominated by single component sand nearest the outlet with an increasing component of silt floccules further out (Fig. 5-9). Seafloor samples also show this exponential decrease in grain size out from the river mouth (Fig. 5-9). New sediment sources, however, especially from sediment gravity flows, can completely alter the size character of the seafloor sediment as laid down from turbid river plumes (Schafer et al., 1989). Syvitski et al. (1988) developed numerical algorithms for predicting the seafloor particle size sedimented out of a fjord's river plume. The spatial distribution of different sized particles is determined using: (1) a velocity distribution developed to simulate a buoyancy-dominated, free, two-dimensional jet flowing into highlystratified marine basins; and (2) a particle-scavenging model that takes into account the biogeochemical effects on settling of particles, such as flocculation. The three dynamic zones of the river plume include: (1) a zone of flow establishment, nearest the river mouth, where the centre of the plume continues to behave as a plug flow;

140

J.P.M. SYVITSKI AND J. SHAW

(2) a zone of established flow where the axis velocity decreases as the plume spreads, and (3) a zone of constrained flow, where plume spreading is affected by the basin walls. Variations in the velocity of the surface layer will affect the ability of the surface layer to carry particles with higher settling velocities (i.e., sand). The concentration of SPM (suspended particulate matter) thus decreases both with depth and distance seaward.

Turbidity currents The bed-material load of a fjord river is deposited rapidly, below the low-low water line, and along foresets that slope 5" to 30" to depths between 10 m and 50 m. These foreset beds prograde seaward onto prodelta bottomset beds at dips between 0.1" and 5". Seasonal or semi-continuous failures of these typically sandy, possibly gravelly, foresets occur as numerous small-scale (lo3 to lo6 m3) displacements continually adjusting to maintain maximum slope stability (Prior et al., 1981a, b, 1987; Gilbert, 1982; Kostaschuk and McCann, 1983; Syvitski et al., 1988; Syvitski and Hein, 1991). These displacements form chutes along the delta lip, developed from small retrogressive slides or local liquefaction fronts generated through a combination of wave-induced cyclic loading and oversteepening after a recent period of rapid progradation (Carlson et al., 1992). The failed sediment masses, being rather coarse-grained, often completely liquefy and develop into turbidity currents. In many cases, turbidity currents flow within channels caused by erosion at the base of their flow and/or channels formed during the initial slide process (Fig. 5-11). High density sandy currents are relatively thin and fast, whereas low density muddy currents are relatively thick and slow (Bowen et al., 1984). Thus, turbidity flows that carry coarse sediment may be confined within the channel walls and will not overtop the channel levees. If a low-density turbidity flow spills over its channel, part of the flow will be stripped away from the main body and will undergo rapid flow spreading and sediment deposition. The channelized flow will, in turn, undergo a reduction in both velocity and sediment concentration. The velocity will also be reduced with decreasing slope. As a result, the channel crosssection will decrease downslope with the decrease in the turbidity current discharge that results from overspill and deposition (Fig. 5-7). During the erosive history of a turbidity current, channel walls may be undercut, initiating a new series of retrogressively developing slides. If these secondary slides add further volumes of liquid sand to the flow, the flow may be rejuvenated (Fig. 5-12). If the undercutting results in the addition of plastic mud and larger mud blocks, the turbulent flow characteristic may regress to that of a debris flow or a more viscous gravity flow. When a sandy turbidity current leaves the confines of the channel walls, such as when it reaches the floor of a fjord basin, the flow slows and spreads and the sand is deposited (Bjerrum, 1971). The time required for this deposition increases with decreasing permeability and therefore decreasing grain size (Terzaghi, 1956). In Queen Inlet, Alaska, surging slump-generated turbidity currents occur intermittently on the delta foreslope mainly when fluvial bed load reaches the delta brink

SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS

141

MAKTAK FIORD

E

-I 40l-

W a

0

a F50-

9 60 -

(b)

ITIRBILUNG FIORD

McBETH FIORD

-E

-

290 -

I 300-

l-

a 310-

MEGACHANNEL

a W 320Ia 3 330-

Fig. 5-11. Delta-front fjord channels. (a) V-shaped channels incising the Maktak prodelta (Baffin Island): note apparent levees. (b) U-shaped (flat floored) channels incising Itirbilung prodelta (Baffin Island). (c) Megachannel that runs 10 km along the length of McBeth Fiord, Baffin Island. Note the smaller leveed channel on the right. All three records are from high frequency sounder records run perpendicular to the fjord axis (for details see Syvitski and Farrow, 1989).

142

J.P.M. SYVITSKI AND J. SHAW

DAYS 1985

0

50

1M

150

203

250

OO

1M

203

3M

4w

TIME (minutes)

Fig. 5-12. Current meter records collected from Itirbilung Fiord over a 39 day period in 1985 (for details see Syvitski and Hein, 1991). (A) Current meter speed data put through a 0.5 hour filter and averaged over 1 hour. The 51 m water depth meter was moored 2 m above the seabed. Note that 9 gravity flow events were registered and that events 1 and 2 moved the mooring array into deeper water (as shown by the pressure sensor). (B)-(E) Details of current speed of the turbidity current events identified in (A). The complexity and duration increases from (B) to (E). Grain size of the suspended load as collected by sediment traps was largely poorly sorted fine sand.

during lower low spring tides (Phillips et al., 1991). Surges (up to 29 cm s-' ) last for a few minutes and carry more than 6 g 1-1 of suspended sediment (Phillips and Smith, 1992). Further offshore (2.7 km from the river mouth) a submarine channel was affected by a quasi-continuous turbidity current with average flows of 15 cm s-' and concentrations around 2 to 3 g 1-'. The near continuous nature of the turbidity currents is possibly caused by the attenuation and overlapping of numerous and variously sized surges generated by foreslope failure (Phillips and Smith, 1992). Delta-front failures that lead to the development of turbidity currents have many common seafloor characteristics (Syvitski and Farrow, 1989): (1) channels up to 100 m wide and 10 m deep cover the prodelta slope (Fig. 5-11a, b); (2) the channels

SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS

143

originate from one or more arcuate reentrants and chutes that have steep headslopes cut into the delta lip; (3) the channels, although not sinuous, may converge with or truncate one another; (4) channel widths and depths decrease downslope until the channel form disappears; (5) the channels, if active, contain rippled well-sorted sand; (6) the interchannel areas consist primarily of poorly-sorted and weakly compacted very fine sandy muds; (7) the channels are commonly lined with levees when the slope falls below 2"; (8) a percentage of the channels at any time are inactive, although there is a tendency for buried channels to be reactivated. Some fjord basins are fed sediment through one or two megachannels that have attained depths from 5 to >25 m and widths of 100 to 1,000 m (Fig. 5-11c; Gilbert, 1983; Syvitski and Farrow, 1983, 1989). These megachannels share some general characteristics: (1) the channel is commonly found on slopes less than 2"; (2) the channel decreases in depth and width with decreasing slope; (3) the channels are somewhat sinuous and may meander from fjord wall to fjord wall; (4) before a channel disappears into the flat of the basin floor, it develops levees; (5) upslope, where levees are not found, the channels have near vertical walls; and (6) if a megachannel is still active it contains sandy sediment in contrast to the surrounding hemipelagic basin muds. In Queen Inlet, Alaska, Hoskin and Burrell (1972) noted that its two megachannels had sediment modes of 205 p m and 44 pm, respectively, compared to the hemipelagic seafloor muds of l l p m . In Hardangerfjord, Norway, abundant graded beds, interpreted as deposits laid down by turbidity currents, gradually become finer with distance of transport (Holtedahl, 196.5, 1975). The graded beds are underlain by coarse, very poorly sorted material generated from side-wall slumps: the slumps contain littoral fauna and clay lumps. The turbidites are restricted to megachannels and cores taken outside the central channel did not contain turbidite layers. Fifty percent of the sediment column within the basins of Hardangerfjord has resulted from slumps and turbidity currents, with an average accumulation rate of s mm a-l. High concentration turbidity currents may be implied from graded layers, basal load casts and flute marks, flame structures, and ripple sequences (Syvitski et al., 1987). Evidence from seismostratigraphy indicates that basal erosive units occur within channel-fill sequences. Low concentration turbidity currents occur as thin (tl cm) layers of clean sand or silt. Turbidites seldom occur as single rare layers, and are more frequently found as thick units of amalgamated deposits (Hein and Syvitski, 1992). Ponded sequences of turbidites are common to the flat basin floors (Fig. 5-3C).

WAVE- AND TIDE-INFLUENCED FJORDS

Deep waters in fjords often have sluggish currents, and in some situations may be advectively isolated, yet mixing processes remain an integral part of the shallower water regions. In fjords where a sill is deep or absent, tidal currents may winnow or erode bottom sediments. For the shallow end-member fjords, especially when exposed to open ocean swells, wave reworking of the shoreline margins may result in

144

J.P.M. SYVITSKI AND J. SHAW

major contributions of sediment to the deeper basin areas. This section examines the influence of tides and waves on the sediment architecture of fiords. Tidal processes In the deep basins, tidal currents simply oscillate back and forth with the tidal wave form. They produce little residual flow and their effect decreases below sill depth (Pickard, 1961). Slack water occurs at times of high and low phases of the tide and maximum velocities occur around midtide. In the shallow reaches of some sills, tidal currents may become turbulent tidal streams, i.e. well-mixed, with a well-defined boundary layer flow. Tidal streams tend to follow local bathymetry and may generate eddies, whirlpools and upwelling domes (Thomson, 1981). Where stratification is well-developed, currents are strongest close to the seafloor, associated with the flood tide. Long and shallow entrances to fjords are friction dominated (McClimans, 1978b), and tides may be significantly dampened. The tidal stream is thereby driven by the hydraulic head caused by the variation in tidal heights between the coast and the fjord basin (Glenne and Simensen, 1963). Over shallow fjord sills, a variety of tide-related oceanographic features may develop (Long, 1980; Huppert, 1980): flow separation, lee waves, hydraulic jumps, jets, bores, and internal waves. In tide-influenced fjords, a turbidity maximum may develop with the entrapment of particles between the outflowing surface layer and inflowing compensation current (d’hglejan and Smith, 1973). In fjord-like estuaries, tidal resuspension tends to be depth-controlled: as the current energy decreases with depth, the current shear will pass below the threshold of movement of sediment grains. This critical depth may change with the stage of the tide and through the spring-neap tidal cycle. When the fjord channel is constricted, or where tributaries join the main channel, tidal flow and the critical erosion depth will increase. Below the zone of erosion there will exist an associated zone where seafloor sediments do not undergo erosion, yet suspended particles may not be deposited. In Cook Inlet, Alaska, the near-bottom turbidity maxima occur over thresholds, near the shallowing fjord head, and along the fjord walls (Feely and Massoth, 1982). If the crest of the sill lies above the critical erosion depth, sediment deposited on the sill during periods of slack tide will be eventually resuspended and transported into or out of the fjord basin. Where the currents are especially strong, the sill might be mantled with a gravel lag, or even consist of exposed bedrock (e.g., Gilbert, 1978; Syvitski and MacDonald, 1982). The tidal jet generated over the outer sill in Borgenfjorden, Norway, is reflected in the coarser sediment as compared to finer-grained basin sediments (Fig. 5-13A, Stromgren, 1974). Borgenfjorden shows the close relationship between grain size and bathymetry: decreasing grain size reflects decreasing current velocity with the increasing width and depth of the inlet (Fig. 5-13A). Turbulence generated by hydraulic jumps at the sill may create zones of erosion where the basin sediments abut with the sill. Bornhold (1983) provides such an example with conformable winnowing (unit A sediments on Fig. 5-14A) and erosion of some ten metres of basin sediment (unit B on Fig. 5-14A). Where tidal streams are

SEDIMENTOLOGY AND GEOMORPHOLOGY O F FJORDS

145

INNER U T F L4NDW4RDS SLOW SETTLlffi OF SUSPENDED MUD

DEPOSITION

EROSION

DEPOSITION

EROSION

P -

3. EXTREME STORM

SILT 8 SOME SAND

-

S ‘ ;\??ES

EROSION

Fig. 5-13. (A) Bathymetry and mean grain size (pm) for Borgenfjorden, Norway (after Stromgren, 1974). (B) Process models for low-sediment, wave-dominated fjords showing: (1) fair weather, (2) normal windy, and (3) exceptional storm conditions (from Piper et al., 1983).

proximal to a sediment source, zones of erosion and selective deposition may grade with zones of deposition: scour channels and stratigraphic wedging of units may result (Piper et al., 1983). Along the approaches to Makkovik Bay, Labrador, selective tidal stream erosion and winnowing of Holocene mud result in the formation of many of these features (Fig. 5-14B, Barrie and Piper, 1982). With the availability of coarser sediment (i.e., sand and gravel), powerful tidal currents may form an assortment of bedform groupings. Where the basin is deep, bedforms may be found along the fjord walls (e.g., St. Lawrence Estuary: Syvitski et al., 1983b) or on the basin floor if the fjord has no sill (e.g., Cook Inlet: Bouma et al., 1977, 1978). The flotation of sand is another tide-related but wave-limited transport process operative over intertidal flats. Sand will be picked up and floated on the sea surface with each rising tide, dependent on (Syvitski and van Everdingen, 1981): (1) proper

146

J.P.M. SYVITSKI AND J. SHAW

A DOUGLAS CHANNEL

.

HYDRAULIC JUMP EROSION

.

.

km

.

B MAKKOVIK BAY

.

. . .. .. . , . . . '

,

o

,

:.

.

0.5

.

.

km

C

YE '':Ah

.,ON-LAPPING

MAKKOVIK BAY BASIN FILL

km

Fig. 5-14. Schematics of seismo-stratigraphic sections that show contemporaneous scouring and/or winnowing of basin sediment. (A) Hydraulic jump erosion near Maitland Island sill, Douglas Channel, B.C. (after Bornhold, 1983). (B) Tidal current scour in the approaches to Makkovik Bay, Labrador (after Barrie and Piper, 1982). (C) On-lapping basin fill units as a result of wave erosion within Makkovik Bay, Labrador (after Barrie and Piper, 1982).

atmospheric conditions (no fog or precipitation); (2) rising water with intact surface tension (no surface turbulence); and (3) appropriate floatable sediment for the incoming water velocity. Sand, in patches as large of 100 x 100 m ,can float seaward as the tide begins to fall or under the influence of gentle land breezes. The annual tonnage of sand moved seaward will depend on the intertidal area that meets the above conditions but is typically of the order of lo5 to lo7 tonnes for macrotidal sandy fjords (Syvitski et al., 1988). The transport distance, however, is usually short (5 km a-' ), then because crustal response to glacier retreat is slow (

0.6

2

0.5

E9

0.4

Q

!W

M.H.W.N.T.

M.S.L.

I-

z

w 0.3

L

4 W 5

M.L.W.N.T.

0.2 0.1

M.L.W.S.T.

0.0

MARSH MUD FLATS MIXED FLATS

SAND FLATS

El CHANNEL SANDS

Fig. 10-1. A summary of the classification of tidal exposure and associated intertidal flat zonation. (A) Amos (1974; the Wash, U.K.): ( I ) salt marsh (silty clay); (2) higher mud flat (sandy silt); (3) inner sand flat (silty sand); ( 4 ) Arenicola sand flat (fine sand); ( 5 ) lower sand flat (fine sand); ( 6 ) channel sand (medium sand). ( B )Carling (1981, Burry Inlet, S. Wales); ( I ) salt marsh; ( 2 ) higher sand flat; ( 3 ) lower sand flat; ( 4 ) subtidal channel. ( C ) Zhuang and Chappell (1991, SE. Australia); ( I ) salt marsh; (2) mangrove mud flat; (3) upper sand flat; ( 4 ) seagrass muddy sand flat. (0) Knight and Dalrymple (1975; Cobequid Bay, Canada): ( I ) salt marsh; (2) sand/gravel beach or mud flat; ( 3 ) mud flat (sandy silt); ( 4 ) braided bar (sand); ( 5 ) sand bar (sand); ( 6 ) basal gravel. ( E ) Amos and Joice (1977; Minas Basin, Canada): ( I ) high water storm beach (sand); (2) salt marsh (silty clay); (3) higher mud flat (sandy silt); ( 4 ) inner sand flat (silty sand); (5)lower sand flat (fine sand); ( 6 ) channel sand and gravel (medium sand to gravel). ( F ) Martini (1991; Hudson Bay, Canada); ( I ) upper salt marsh; (2) lower marsh; ( 3 ) higher tidal flat; ( 4 ) upper sand flat; (5) lower sand flat. (C) Reineck (1972, German Bay, Germany): ( I ) salt marsh (clay); (2) mud flats (clayey silt); ( 3 ) mixed Hats (sand/silt); (4) sand flats; ( 5 ) channel deposits (mud and sand to gravel). ( H ) Evans (1965; the Wash, U.K.): ( I ) salt marsh (silty clay); (2) higher mud flats (sands and silty clay); (3) inner sand flats (very fine sand/silt); ( 4 ) Arenicola sand flats (very fine sand); ( 5 ) lower mud flat (sandy silt); ( 6 ) lower sand flat (fine sand). ( I ) Larsonneur (1975; Mont Saint-Michel Bay, France): ( I ) salt marsh (silt/clay); (2) higher mud flat (clayey silt); (3) muddy sand flat (sandy silt); ( 4 ) sand flat (fine sand); ( 5 ) biogenic sand (muddy sand); ( 6 ) biogenic gravelly sand. ( J ) Thompson (1968; Gulf of California, Mexico): ( I ) chaotic muds (clays); ( 2 ) brown laminated silt; (3)brown mottled mud (sandy silt); ( 4 ) gray burrowed clay; ( 5 )gray laminated silty clay. ( K ) Wang and Eisma (1988; Wenzhou region, China): ( I ) higher mud flat (silty/clay); (2) middle mud flat (fine sandy/silt); (3) lower mud flat (silt). ( L ) Belperio et al. (1988, Southern Australia): ( I ) samphire salt marsh; (2) beach ridges (sand); (3) samphire algal mud flat; ( 4 ) mangrove; ( 5 ) sand Hat; ( 6 )Zosteru flat and Posidoniu seagrass banks

SILICICLASTIC TIDAL FLATS

275

McCann (1980) suggested four criteria for the classification of tidal flats: (1) sediment composition (carbonate or non-carbonate); (2) hydrographic position (intertidal or subtidal); (3) tidal range (macro, meso, micro); and (4) physiographic setting (estuary, delta, exposed coastline and continental shelf). He showed that tidal flats predominate in mesotidal and macrotidal (Hayes, 1975) settings of abundant sediment supply and low wave energy. Dionne (1988) followed closely the views of McCann (1980) and suggested that tidal flats be classified on the basis of (1) tidal range; (2) geomorphological setting; (3) sediment type; and (4) geographic location. Ren et al. (1985) took a geomorphological approach to classify the tidal flats of China which are found in three distinct coastal settings: (1) embayment type; (2) estuarine type; and (3) open coast type. They also noted a distinction between: (a) prograding; and (b) receding types. China’s tidal flats have been further subdivided by Wang et al. (1990) into: (1) silt flats; and (2) clay flats. This latter sub-division may be of wide application as the so-called mud flats of Minas Basin fall nicely into the silt flat sub-division (Daborn et al., 1993). A summary of the gross geographical and geological factors leading to the development of tidal flats is given by Boyd et al. (1992) and Dalrymple et al. (1992). They propose that tidal flats prevail in regions sheltered from waves where the fluvial input is small; that is, they are the manifestation of progradation of sediments derived from a marine sediment source. According to Boyd et al. (1992), the morphological character and distribution of tidal flats depends on whether the coastline is transgressive or prograding; tidal flats on transgressive coasts are largely found in four geomorphic settings: (1) the low energy equivalent of a coastal strand plain on linear coasts; (2) the lateral portions of tidal-dominated estuaries; (3) the inner portion of wave-dominated estuaries; and (4) the inner portion of lagoons. Tidal flats on prograding coasts are more widely developed, but are largely found fringing the open coastline. Even the above elegant scheme is limited in application as it does not account for tidal flats on deltas such as those described by Kellerhals and Murray (1969) on the Fraser Delta and Wells and Kemp (1984) on the Mississippi Delta.

SILICICLASTIC TIDAL FLAT RESEARCH

Early scientific descriptions of tidal flats were based largely on observations made in the embayments and estuaries bordering the North Sea. An excellent review of this literature is provided by Klein (1976). In this review we are acquainted with the attributes of tidal flats though surprisingly a rigorous definition of a tidal put is not found. Although the term tidal put may have been self-evident within the context of research in mid-latitude European cases, the proliferation of recent tidal flat research to other climatic and geographic regions tends to blur our earlier notions. These early notions came from Hantzschel (1939) who equated tidal flats with wattenschlick (tidal slime or mud) and associated sandy deposits that are found between high and low water levels of the German Bight. He showed remarkable insight in recognizing that the source of the sediments to the flats was largely the

276

C.L. AMOS

offshore, and that these sediments were intensively reworked by sloughs (creeks) that crossed the intertidal region. Van Straaten in the early 1950's extended our knowledge of tidal flat morphology and composition, and postulated on mechanisms for the formation of flats in the Wadden Sea. He also considered that gullies (creeks) were a major factor in reworking of tidal flat deposits, arguing that their lateral migration would in time largely rework the original facies of the flats leaving behind a series of basal lag deposits, and inclined heterolithic foresets (longitudinal oblique bedding of Reineck, 1972) diagnostic of point bar formation; only the inner flats would be spared this process. Evans (1965, 1975) broadened our understanding of tidal flats through a detailed study undertaken in the Wash. He expanded on the observations of van Straaten, and proposed a stratigraphic sequence of upwardfining sediments resulting from the lateral progradation and superimposition of adjacent sub-environments and preservation in a manner not unlike that of deltaic sedimentation. His marsh, upper mud flat and sand flat comprise the top-sets where vertical accretion dominates, and the lower mud flat and lower sand flat constitute the foresets where lateral progradation of the flat takes place. In his view, creeks were restricted to narrow belts on the tidal flats and consequently were of less importance in reworking the flats than was postulated earlier. The creek deposits would thus form narrow prisms of reworked sediments that would be oriented shore-normal, and which would be couched within the regional progradational sequence described below. These prisms would have a surface expression not unlike a meandering fluvial system (Reineck, 1975) with well-developed lev6es along which the landward-situated sub-environments would extend. The progradational sequence is evident as a series of shore-parallel sub-environments more or less in equilibrium with hydrodynamic conditions and exposure. Kestner (1975) contested the view of steady progradation and suggested that tidal flats were inhibited in growth by the fixed position of the low water tidal channel. He speculated that progradation would take place only in the presence of a sedimentation umbra cast onto the flats through reclamation of salt marshes or channel entrainment. Kestner (1975) offered a further view of the role of creeks in tidal flat sedimentation. He proposed that the creeks enhanced vertical aggradation rather than lateral reworking; the levtes of creeks being the pathways along which the salt marshes and mud flats of the Wash prograde seaward beneath the entrainment umbra. This mechanism was put forward to explain the origin of the seaward edge of the inner sub-environments which, though shore-parallel at a distance, are cuspate in detail. The cusps follow the creek levtes seaward (Fig. 10-2). Kestner (1975) argued that the existence of cusps are diagnostic of a stable tidal flat in equilibrium with tidal inundation. Amos (1974) disputed this conclusion and proposed the opposite; that the cusps are diagnostic of active progradation: the larger the cusps, the greater the progradation rate. It follows that when no cusps are found the tidal flat would be either stable or in recession. Amos also proposed that creeks were responsible for a step-wise evolution of the tidal flats. Progradation would be rapid in those upper tidal flats fed by creeks, while the inter-areas would be relatively starved. In time, the inter-areas would capture the ebbing tidal flow, being relatively lower than the creeks, and the process of cusp development would begin again within the inter-areas.

277

SILICICLASTIC TIDAL FLATS . . . . . .

..

. . . . . . . . . . . . . .

:1: : : :

I

.

: : :I:

...

i

...

0SAND FLATS

.

::::

.

............. . . . . . . .1

-

: : : : . . . : . : : : : : : : : : : : : .I: . . . . . . . . . ./

I

NORTH SEA

MUDFLATS

E I l SALT MARSH

? , rnelres ,

4?0

Fig. 10-2. The cuspate pattern of the salt marshes and inner mud flats of the Wash, taken from Kestner (1975). Note that the cusps follow the creek IevCes seawards. A smaller cusp appears in the process of development in the inter-area between the two major creek systems. The cusps have developed largely because of reclamation in 1868 and 1953/54.

The dominance of mud flats in turbid environments results in an abundance of creeks. Wang et al. (1990), working on flats adjacent to Bohai Bay and Huanghai (Yellow) Seas, showed that creeks occupy 10% of the flats by area and are the pathways for the transport of what little sand crosses these flats. Yet the shoreparallel zonation of sub-environments (a pattern that typifies sand-rich tidal flats) is still evident (Wang, 1983; Ren et al., 1983). Wang et al. (1983) suggested that the lower and middle flats prograde in a seawards direction in a manner similar to that of the tidal flats of the Wash where fewer creeks are found (Evans, 1965). The implication of this mode of development and the shore-parallel zonation of sub-environments favours sedimentation processes related to tidal inundation rather than one of creek reworking. Tidal flats are found in three broad climatic regions (Dionne, 1988): (1) lowlatitude tidal flats in arid and wet tropical or subtropical regions; ( 2 ) mid-latitude tidal flats of temperate regions; and (3) high-latitude tidal flats influenced by ice.

278

C.L. AMOS

A review of the first group of tidal flats may be found in (but not restricted to) the collective works of Thompson (1968) in the Gulf of California; Neumann et al. (1970) in the Caribbean Sea; Belperio et al. (1988) and Zhuang and Chappell(l991) in south Australia; Semeniuk (1981) in northern Australia; and Wells and Coleman (1981a,b) off the Orinoco and Amazon rivers. Papers on the second group of tidal flats include those of Evans (1965), Evans and Collins (1975,1987) on the Wash; van Straaten and Kuenen (1957), Postma (1961), and Fitzgerald and Penland (1987) on the Wadden Sea; Klein (1963,1985), Middleton et al. (1976), Amos and Long (1980), Dalrymple et al. (1990,1991), on the Bay of Fundy; Larsonneur (1975), Caline et al. (1982) on Baie Mont Saint-Michel; Carling (1981,1982) on the Burry Inlet, S. Wales; and Berner et al. (1986), Reineck et al. (1986), Dieckmann et al. (1987) in the Jade estuary and eastern Frisian Islands. Recently, a considerable amount of information on the tidal flats around the Bohai and Yellow Seas has emerged. This includes the work of Ren et al. (1985), Wang and Eisma (1988,1990), and Zhang (1992) in China; and that of Frey et al. (1989), Adams et al. (1990), and Wells et al. (1990) in South Korea. The third group of tidal flats has largely been studied in the Americas by Champagne (1982), Anderson (1983), Grinham and Martini (1984), Dionne (1988), Smith et al. (1990), Martini (1991), and Isla et al. (1991). Recent research on tidal flats has altered in focus from studies of morphology and internal structure to measurements of tidal flat dynamics. We are becoming aware that a bewildering variety of factors influence tidal flat sedimentation and stability (Nowell et al., 1981; Jumars and Nowell, 1984). Early papers account for the origin and evolution of tidal flats on the basis of the properties of the tidal inundation. It is becoming more apparent that events that take place during tidal flat exposure may be as important as those during inundation (Ginsburg et al., 1977). Anderson (1979, 1983) recognised the effects of desiccation, rain pit dislodgement, solar heating, plant and animal activity, and ice effects on the development of a mid-latitude tidal flat in the American northeast. Paterson (1989), Paterson and Underwood (1990) and Paterson et al. (1990) made similar observations on the tidal flats of the Severn and Tamar estuaries, U.K. The significance of exposure is also supported by the observations of Amos et al. (1988) and Daborn et al. (1993) in the Minas Basin, Canada. Twenty-fold increases in bed strength were measured over a summertime period when low water coincided with solar noon. Also, solar heating (by 2°C) occurred to a depth of 0.4 m below the sediment surface during a single exposure event (Piccolo et al., 1993), with consequent blooms of microphytobenthos and mucopolysaccharide production. Daborn et al. (1993) have linked increases in mud flat stability to significant increases in microphytobenthos production, the consequent population explosions of the amphipod Corophium volututor (104/m2), and the subsequent frenzied feeding habits of the semipalmated sandpiper (Culidn's pusillu L.). Similarly, the feeding habits of the snow goose (Chen cuerulescens) appear to have an intense effect on the erosion of salt marshes in the Gulf of St. Lawrence, where deposition or ice effects normally dominate (Serodes and Troude, 1984). Faas et al. (1992) show graphic evidence of the effect of biostabilization in two photographs of quadrats of the mud flats of Minas Basin: one taken before application of poison to the quadrat region; and the other taken after poisoning.

SILICICLASTIC TIDAL FLATS

279

The once adhesive mud flat was transformed in hours through poisoning into a non-cohesive rippled silt flat. The loss in strength was due entirely to the removal of a biofilm of mucopolysaccharides; a diatom exudate (Grant et al., 1986a). Such effects are not restricted solely to the mud flats; Grant (1981), Gerdes et al. (1985), Grant et al. (1982, 1986b), Montague (1984), Grant (1988), Meadows and Tait (1989), and Emerson and Grant (1991) have found similar effects of bio-stabilization on tidal sand flats. The complexity of factors controlling tidal flat stability necessitates the use of innovative technologies and methodologies. The effects of microphytobenthos are largely restricted to the upper 2000 microns of sediment, so sediment indexes based on bulk properties are of limited use to explain them. This is perhaps most evident in the mismatch between measurements of the vane shear strength of marine sediments (o”), which is usually reported to be of order lo3 Pa (Christian, 1991), and the critical shear strength for erosion (re)which is usually of order 1-5 Pa (Amos et al., 1992). Given that re is equated with the shear strength of the sediment (Mehta and Partheniades, 1982), we must acknowledge a discrepancy of three orders of magnitude in measurement. The existence of fluid muds, gels and “fluff” layers are proving to be widespread in nature (Parker, 1987). The pseudo-plastic, non-newtonian, viscous behaviour of these sediment states is complex (Partheniades, 1984; Mehta, 1989,1991). It is strongly influenced by consolidation history and density (Hydraulics Research Station, 1980), physico-chemical activities within the sediment (Pamukcu and Tuncan, 1991), geochemical processes and redox state (Baeyens et al., 1991), as well as the rate of stress application (a rheological response, Faas, 1991; Julien and Lan, 1991). Opinions diverge on the influence of turbidity on the transmittal of fluid stresses to the bed and on the structure of the viscous sublayer, which is often millimetres thick. Consequently, a considerable amount of innovative work is in progress to determine the development of such bed states and the structure and density of slowly-consolidating seabeds at the micro-scale. New in situ devices such as INSIST (Christian, 1991), the Cohesive Sediment Meter (Paterson et al., 1990), the Sea Carousel (Amos et al., 1992), and benthic chambers (Buchholtz-Ten Brink et al., 1989) are providing information on bed stability and the complex links between biosphere, geosphere, hydrosphere and atmosphere. The recent upsurge in the development of multi-disciplinary field programs to monitor synoptically tidal flat processes and attributes (Gordon et al., 1986; Daborn et al., 1993; LISP-UK, 1992) offer exciting possibilities for future discovery. It is only through such discoveries that advances in our understanding of tidal flat evolution will occur.

THE ZONATION OF TIDAL FLATS AND RELATIVE ELEVATION

Virtually all tidal flats exhibit common variations in grain size, benthic floral and faunal diversity and abundance, surface morphology and slope that may be mapped into coherent sub-environments. In most cases, these sub-environments are

280

C.L. AMOS

oriented shore-parallel and occupy distinct positions with respect to exposure and tidal inundations (Evans, 1965; Klein, 1985; Dieckmann et al., 1987). The number of such sub-environments together with the specific attributes vary considerably. Figure 10-1 shows a variety of tidal flat sub-environments and their relative elevations above extreme low water ( h / R , where h is the height above extreme low water, and R is the extreme tidal range). Two major groups of tidal flats are apparent: (1) sandy tidal puts, where the mean inorganic suspended sediment concentration (SSC) of the inundating waters is generally less than 1 g/l (Fig. 10-1, references A-I, and L); and (2) muddy tidalputs, where the SSC is generally greater than 1 g/L (Fig. 10-1, references J and K). Group 1 salt marshes and mud flats dominate above MHWNT (the higher flats). Differences in the highest relative elevation of the mud flat are large ( h / R = 0.8-1.0), whereas the lower limit of the mud flat is relatively constant ( h / R = 0.75). The highest limit of the mud flat is predicated on the degree and type of its colonization as well as by its wave exposure (Kestner, 1975; Groenendijk, 1986). In some cases the mud flat is replaced by a wave-formed beach above MHWST (Amos and Joice, 1977; Knight and Dalrymple, 1975; Belperio et al., 1988); in other cases there is no marsh (Thompson, 1968; Wang and Eisma, 1988; Daborn et al., 1991). In the absence of a marsh, the maximum relative elevation is h / R = 0.91 (Kestner, 1975). The transition from a colonized marsh to exposed mud flat in a prograding situation is gradational as is the transition to a sand flat. The latter gradient results in the mixed flats. The mixed flats dominate between MHWNT and MSL (0.5 < h / R < 0.75). Though it is not evident in all the zonations shown in Fig. 10-1, it is nevertheless present in the form of a gradual transition from cohesive to non-cohesive surface sediments across the flats. The vertical extent of the mixed flats varies considerably ( S h / R M 0.02 in Evans, 1975, to Sh/R FZ 0.25 in Larsonneur, 1975). The large extent of the mixed flats reported by Larsonneur is due at least in part to lateral variations in sediment supply and wave activity; factors that also affect the zonation of Minas Basin tidal flats (Amos and Joice, 1977) as well as those of San Sebastian Bay, Patagonia (Isla et al., 1991). The sand flats are prevalent between MSL and MLWNT (0.25 < h / R < 0.5). The relatively small vertical extent of this zone is often masked by the wide areal expanse that is the result of its low slopes (1: 100 to 1:500). The sand flats are largely composed of fine and very fine sand. This explains the absence of large-scale bedforms, which are formed in medium sand or coarser (Middleton and Southard, 1984), and the dominance of small-scale wave-formed and current-formed ripples (Amosand Collins, 1978; Dingler and Clifton, 1984). The lack of relief of these sand flats is undoubtedly due to the fineness of the sand, which is readily mobilized as sheet flow (Tables 10-1 and 10-2). The lower mud flat of the Wash (Evans, 1965) stands out as a notable exception to the above trends. Found between MSL and MLWNT, it intermittently occupies a position within sandy sub-environments. The typical concentration of suspended particulate matter over the flats of the Wash is between 100 and 1000 mg/l (Evans and Collins, 1975; Collins et al., 1981). This range overlaps the concentration range detected over the flats of Minas Basin, Bay of Fundy (Amos and Long, 1980) where no lower mud flat exists. Biological colonization of the

SILICICLASTIC TIDAL FLATS

281

sand flats is especially prevalent in low latitudes. The presence and relative elevation of mangroves, algal mats, bacterial mats, halophyte grasses, sea-grasses and green algae is highly variable, though generally restricted to h / R > 0.3 (Ginsburg et al., 1977). Elsewhere, the edible mussel Mytilus edulis is responsible for the generation of vast quantities of pseudo-faeces that overprint the normal trends in tidal flat zonation. Being composed largely of fine-grained material, it is these pseudo-faeces that have formed the lower mud flat of Evans (1965) in the Wash, in a region where medium sand would otherwise dominate. The channel sands prevail below MLWNT ( h / R < 0.25). They are largely composed of medium sand or coarser material. The transition from the sand flat to the low water tidal channel is associated with an increase in slope and an increase in grain size. The coarser material in this region is less easily fluidized and may, therefore support the higher slopes (Komar and Li, 1986; Li and Komar, 1986) diagnostic of the low water tidal channels and the associated banks and bars. Furthermore, the coarser size of material together with higher flows results in the characteristic large-scale bedforms (sand waves and megaripples), bars and flood and ebb tidal channels described by Dalrymple (1977), Knight (1977), Lambiase (1977), Klein (1985) and Boothroyd (1985). Group 2 tidal flats are characterised by the dominance of mud flats and mixed flats and the lack of a sand flat. Even so, a seaward coarsening of surface sediment is apparent in the form of a clayey salt marsh that grades to a mud flat (mixed silt and clay) and ultimately to a silt flat near low water (Fig. 10-1, references G-I). There is a much lower diversity in biological colonization of this group than is evident in group 1. The upper (clayey) mud flat predominates above MHWNT ( h / R > 0.75). This is the turbid equivalent of the salt marsh and mud flats of group 1. Group 2 mixed flats are present between MHWNT and MLWNT (0.25 < h / R < 0.75). They cover a much broader range in elevations than do group 1 equivalents, and they occupy the position of the group 1 sand flat. Group 2 lower mud flats are found below MLWNT ( h / R < 0.25), where they occupy the position of the channel sands of group 1. The lack of sand on the group 2 tidal flats may be a function of supply rather than process. For example, the mud flats of China (which comprise approximately 50% of its coastline), the western coastline of South Korea, and off the Orinoco, Amazon, and La Plata rivers (all highly turbid environments) show marked differences from those which fringe the North Sea. High amounts of suspended silt and clay from the Huanghe, Changjiang and Zhujiang rivers (Wang, 1983) result in the development of extensive clay-rich mud flats bordering the Bohai and Yellow Seas and the virtual obliteration of the sand flat sub-environment. The work of Thompson (1968) on tidal flats in the Gulf of California gives insight into the factors controlling group 2 tidal flats. He found that his tidal flats were undergoing conversion from mud flats to sand flats due to a reduction in the supply of fines to the flats brought about by the construction of hydro-electric dams on the Colorado River. Wells and Coleman (1981b) also found active mud deposition under the turbid plume of the Orinoco River in a region of “moderate” waves. Is, therefore, the extent of mud flats mainly a function of supply and concentration? Also, are the hydrodynamic effects during tidal

282

C.L. AMOS

flat inundation and the effects of tidal exposure of second order importance only? To examine these issues we must look at the processes of tidal flat sedimentation. In the next section we examine two well-documented tidal flats: those of the Wash, U.K., and the Bay of Fundy, Canada.

TIDAL FLAT SEDIMENTATION - A COMPARISON BETWEEN THE WASH AND THE BAY OF FUNDY

Mud fiat deposition and sediment supply The dynamics of tidal flat aggradation and progradation by tidal inundation requires a knowledge of both cohesive and non-cohesive sediment behaviour within the water column as well as on the bed. This involves complex processes of erosion, transport, deposition and consolidation (Dyer, 1986). Many studies of the transport and deposition of tidal flat deposits exist, but less work is available on bed consolidation and the processes of subsequent erosion. The characterisation of tidal flat deposition began with the long-term, detailed observations of Inglis and Kestner (1958) who, on the basis of these observations, postulated that tidal flats grow only because of influences of marsh reclamation. Deposition rates on these flats are indeed generally low (10-20 mm/a) and time-variable (Amos, 1974). Also, this rate will vary with relative height across the flats (Kestner, 1975; Dieckmann et al., 1987). Dalrymple et al. (1991) in a paper on mud flat deposition in the Bay of Fundy, indicate that the history of mud flat deposition may be divided into two phases: a short-lived period of rapid aggradation, followed by a longer period of quasi-equilibrium in which accretion is slow and the deposits are more intensively bioturbated. For the Wash, Kestner (1975) proposed a similar evolution of mud flats following the exponential forms: 6h St

_ -- 0.836 - 0.136h

(10-1)

and h = 6.16 - 0.479e-0.136xt

(10-2)

where h is the elevation and x is distance across the flat. The relationship is purely empirical as it is independent of SSC, tidal current speed or wave exposure. Also the influence of creeks is unknown. Kestner (1975) measured accretion rates that were in excess of 60 mm/a adjacent to creeks, which suggests a possibly strong contribution from this source. Also he found that the accretion rate was accelerated by marsh colonization of Spartina alternifiora, although the maximum elevation for accretion remained the same as that of exposed mud flats (0.71 m below MHWST; Fig. 10-3). Accretion measurements made by Amos (1974) along three transects of the tidal flats of the Wash, and illustrated in Collins et al. (1981), show a shore-parallel arrangement in deposition rates. The highest rates (20-100 mm/a) are on the upper mud flats and sand flats, intermediate rates (10-20 mm/a) are on the marsh, and the lowest values (including erosion) are in the tidal channel. This pattern of accretion

283

SILICICLASTIC TIDAL FLATS

7.50

1

z

MHWST = + 3.780m ODN MLWST = - 3.139m ODN

I-

v)

3

2

>

0.711m

6.50

6.00 ARRIVAL OF SALT-MARSH PLANTS

5.50 I

I

I

5

10

I

I

15 20 YEARS

I

I

25

30

!I954 WINGLAND RECLAMATION BANK COMPLETED

Fig. 10-3. The accretion of the mud flats in the Wash that has resulted since the construction of the reclamation dyke shown in Fig. 10-2. The pattern is asymptotic to a maximum elevation of 0.71 m below MHWST (ODN = Ordnance Datum, Newlyn). Notice that marsh plants may accelerate the process of accretion, but the asymptote is the same as for bare mud flats.

suggests that the colonization by halophytic plants takes place with a reduction in the rate of sedimentation on a mud flat; a trend opposite to that of Kestner (1975). Furthermore, the pattern of accretion across the Wash tidal flat is not consistent with the long-term progradation of an equilibrium profile (where accretion rate is in direct proportion to the slope). It does, however, lend support to the original hypothesis of Inglis and Kestner (1958) that marsh reclamation dominates the longterm progradation of the tidal flats. Yet this hypothesis must be flawed, as it disallows the existence of tidal flats where no engineering schemes exist. So how do sediments move headwards onto the flats and what factors control deposition? The mechanics of tidally-driven sediment motion onto and across a tidal flat was postulated to be the product of “settling and scour lag” originally defined by Postma (1954, in Postma, 1961) and van Straaten and Kuenen (1957). These authors attempted to explain the enrichment of fine sediments in the deposits of the Dutch Wadden Sea relative to the source (the North Sea). Postma (1961, 1967) used similar arguments to explain the gradient in SSC in the Wadden Sea where no apparent residual flows were found to justify it. He attributed a net landward drift in suspended solids to a change in sediment behaviour from high to low tide. This, he reasoned, was due to a longer high water still-stand (and therefore greater deposition) at high tide than at low tide, and the development of yield resistance of the newly-deposited sediment... “Towards high tide, when the flood current velocity has decreased sufficiently far, nearly all material sinks to the bottom. The sediment is not again brought in suspension by the returning ebb current before the latter has reached a velocity considerably higher than that

284

C.L. AMOS

of the flood current at the moment of deposition. In this manner the material is resuspended in a water mass the relative position of which is farther inward than that of the water mass which carried the material during the flood. At low tide a considerable part of the material remains suspended and is thus not subject to a process similar to that at high tide, which would otherwise approximately compensate the latter. Consequently, over a whole tidal cycle, this material undergoes a net inward displacement.”

In short, it is the imbalance of the benthic (vertical) flux integrated over a tidal cycle that results in the shoreward residual motion of exotic material. Groen (1967) pointed out the short-comings of the advective approached described by Postma (1961) and warned that: “In reality, only the statistics of the behaviour of the suspended particles is described by the current.”

He used a diffusive approach to show that the shallow-water asymmetry of the flood and ebb current durations (while assuming the flood and ebb current speeds to be of equal magnitude, which is rarely the case) control vertical exchanges of sediment within the benthic boundary layer. These in turn produce vertical concentration gradients in the benthic boundary layer which influence the magnitude (not the direction) of the suspended sediment residual motion. A headward transport of suspended solids results, which may be up to 38% greater than the seaward motion. His explanation for this effect is: “the ebb current maximum is preceded by a much longer period of low current velocities than is the flood current maximum, so that during the former period there is much more time for the particles to settle down. And the ebb peak of the suspended load is the lower one because it has to be reached from a much lower preceding minimum.”

The residual flux, according to Groen, is sensitive to the settling lag. It increases as the particle settling rate increases and as the mean water depth decreases. Perhaps the greatest insight into the process of residual sediment motion onto tidal flats comes almost as an after-thought wherein Groen warns us that: “as soon as (even by this very process) gradients [longitudinal] of concentration of suspended sediment have been built up, the process of ordinary tidal and turbulent mixing will cause a down-gradient exchange of matter which eventually will counter-balance the action of the former process.”

Simply stated, the headward flux due to tidal asymmetry should be balanced by seaward diffusion due to a seaward-decreasing SSC-gradient. This, then raises several issues. Firstly, if such a balance exists then an equivalent equilibrium gradient in SSC should also exist. Secondly, if this equilibrium condition exists, then what is the mechanism of sediment import? Thirdly, if the equilibrium gradient in SSC is upset (for example by wave resuspension over the flats) can a largely-importing system export material? And if so, does this imply a (long) time-varying residual flux of material to and from the flats? The answer to the first question comes from synoptic measurements of SSC taken along the length of Cumberland Basin, Bay of Fundy by Keizer et al. (1976) over a

SILICICLASTIC TIDAL FLATS

285

SUSPENDED SEDIMENT CONCENTRATION ALONG 1978

I CHIGNECTO BAY 1100 A.S.T. :5 JUNE A HEAD OF SHEPODY BAY I

.-

\
60% (mean 32%; Fig. 13-14). There is considerable overlap, due largely to the variation between individuals in the 3D field, but the mean values are significantly different. We suggest that a height variability of 20% may prove to be a useful dividing line between 2D and 3D forms, but more data are required to test the generality of this value. In order to make this distinction useful in situations where the collection of troughline-elevation data is difficult or impossible, Rhodes (1992) also determined the crestline sinuosity index (straight-line span width divided by the sinuous crestline length, the value decreasing from one as the sinuosity increases) for each dune in the A,

+

R.W. DALRYMPLE AND R.N. RHODES

384

90

c

r

2DDUNES

0.74

I

8ol

*

n 56

3DDUNES

h

4 I-

40-

P 3

30-

I

* Ic

20--10

-

\

i * 00.85

I

I

I

I

I

0.87

0.89

0.91

0.93

0.95

I

1 0.97

* I

0.99

CRESTLINE SINUOSITY Fig. 13-14. Plot of crestline sinuosity (1.0 is perfectly straight) against the coefficient of variation of bedform height for individual dunes in two areas in the Minas Basin, Bay of Fundy. Height variability primarily reflects irregularities in the trough, the crestline elevation being nearly constant even in strongly 3D dunes (Figs. 13.2B-D and 13.13). The 3D dunes had a morphology intermediate between those shown in Fig. 13-2A, B. The diagonal line is the best-fit regression line. The dashed lines represent suggested divisions between 2D and 3D dunes. From Rhodes (1992).

Fig. 13-15. Small, simple dunes with prominent scour pits and lee-face sinuosity (i.e., 3D dunes) which nevertheless have relatively straight crestlines and excellent lateral continuity. Metre stick for scale.

ESTUARINE DUNES AND BARS

385

two fields, obtaining mean values of 0.97 for the 2D dunes (minimum value 0.93), and 0.93 for the 3D dunes (range 0.85-0.99; Fig. 13-14). Again there was considerable overlap, but a statistically-significantcorrelation exists between crestline sinuosity and the variability of the troughline elevation (Fig. 13-14). On the basis of the regression line, the distinguishing crestal sinuosity between 2D and 3D dunes is 0.96. Again the universality of this value requires further testing. It should be noted that dune lee faces are typically more sinuous than the crestline (Figs. 13-8; 13-15). Indeed, Dalrymple (1984) reported that the Zee-face sinuosity for a field of distinctly 2D, large dunes was 0.96 (on the 2D-3D boundary according to the data in Fig. 13-14). Thus, a lower value of the sinuosity index will be needed to distinguish between 2D and 3D dunes if the lee face is used instead of the crestline (e.g., sidescan sonograms generally show the lee face-trough contact more clearly than the crestline). The lateral continuity of individual dunes is another attribute which has received scant attention. It is generally believed that the crestlines of 2D dunes are continuous for considerable distances, whereas those of 3D dunes are less continuous; however, no quantitative data exist which would support or reject this widely-held belief. It is certainly true that large 2D dunes have high lateral continuity (Fig. 13-12B); for example, Dalrymple (1984) reports a mean value of 10 (maximum 18) for the horizontal-form index (flow-transverse span divided by the wavelength), and the dunes figured by Langhorne (1973, fig. 2, areas E l and E2) have horizontal-form indices of approximately 20-25, some crestlines being traceable for up to 1700 m. However, many tidal dunes with well-developed scour pits (i.e., 3D bedforms) may also exhibit good continuity (Figs. 13-2C, D, G; 13-15). It may be that the presence of reversing flow causes these bedforms to be more continuous than those in unidirectional (fluvial) flow, but comparative data are lacking. Allen (1968, fig. 4.48) has recognized several types of lateral bedform terminations: open: the crestline does not connect to a neighbour; zigzag: the two adjacent crestlines converge on each other, meeting at an angle generally >90"; and buttress: one crestline remains straight and the other curves abruptly, meeting the straight crestline at approximately 90". All of these are observed in both 2D and 3D dunes of all sizes, but the significance and relative frequency of each type is unknown. In areas with uniform flow conditions, branching appears to occur randomly; current-parallel, in-line rows of branches like those seen in straight-crested current ripples (Allen, 1968, pp. 194-197) have not been reported from dune fields. In areas with spatiallyvariable flow conditions (especially changing depth or current speed), the required changes in dune wavelength appear to occur by means of branching or merger of crestlines in narrow zones. Such zones mark the boundaries between some of the bedform fields delineated by Langhorne (1973, fig. 3) in the outer Thames estuary, and are evident in the sidescan-sonar mosaics produced by Aliotta and Perillo (1987, figs. 6, 7 and 9) for the Bahia Blanca estuary. A final planform pattern worth noting is the presence in rare(?) cases of what might be termed interference dunes. A spectacular example is figured by Hine (1975, fig. 17) from the flood-tidal delta of the Chatham Harbor estuary, Massachusetts. Here, two sets of equally-sized, medium to large, 2D dunes intersect at angles between 90" and 120". This pattern is ascribed to changes in current direction during

386

R.W. DALRYMPLE AND R.N. RHODES

the dominant flood tide, one set being transverse to currents early in the flood, while the other is transverse to currents during the middle of the flood. Other examples of dune interference have been figured from the flood-tidal delta in the Parker River estuary (Boothroyd, 1985, fig. 7-18a) and from a bar crest in Cobequid Bay (Knight, 1980, fig. 10.7c), but neither author comments on the pattern or its origin. This subject is considered further below.

DUNE ORIENTATION

It is widely assumed that dune crestlines are perpendicular to the strongest (dominant) current or residual sediment-transport direction. However, many authors have noted situations where large to very large dunes are oblique to both of these directions (Figs. 13-2D-G; 13-13; 13-16; Terwindt, 1970, 1971; Langhorne, 1973; Boothroyd and Hubbard, 1975; Dalrymple, 1984; Aliotta and Perillo, 1987; Fenster et al., 1990). As bedform orientation is commonly used to estimate the direction of net sediment transport, the factors controlling bedform orientation require examination.

Variabilityof current direction Based on experimental data, Rubin and Hunter (1987) and Rubin and Ikeda (1990) conclude that bedforms are oriented so as to maximize the gross (not net),

Fig. 13-16. Large, 2D compound dunes on an elongate tidal bar in Cobequid Bay, Bay of Fundy. The dominant flood currents flow from left to right, parallel to the channel in the foreground and perpendicular to the crestlines of the small dunes at the near edge of the bar; thus, the crestlines of the large dunes are oblique to the net transport direction. Note the in-line kinks in the large crestlines at the left side of the photo. The field of view is approximately 1.2 km wide.

ESTUARINE DUNES AND BARS

387

bedform-normal transport. For the simple case of sediment transport in only two directions (e.g., ebb and flood currents with little directional dispersion), this gross, bedform-normal transport is given by

T = D Jsinal

+ SI sin(y - a)l

(13-4)

where the terms are as defined in Fig. 13-17A. Note that the absolute values of the ebb and flood transport are added (i.e., the actual transport direction is ignored) to obtain the gross transport, whereas these two components would be subtracted to calculate the net transport direction. The bedform orientation (as measured by either a or 4; Fig. 13-17A) which maximizes T is dependant on the divergence angle between the dominant and subordinate transport ( y ) and the relative amount of sediment transported in each direction (the transport ratio). Solutions of eq. (13-4), which are supported by experimental data, are shown by the curved lines in Fig. 13-17B. In situations where the ebb and flood transport directions are variable a more complex procedure is needed to determine the preferred bedform orientation. A computer program to do this is supplied by Rubin and Ikeda (1990). Figure 13-17B shows that transverse bedforms (4 > 75") occur if the divergence angle ( y ) is either less than 90" or nearly 180", and at intermediate divergence angles only if the transport ratio is high (greater than about 8). For divergence angles between 90" and 180", longitudinal bedforms (4 < 15") occur if the transport ratio is approximately 1 (sub-equal ebb and flood transport), whereas oblique bedforms (15" < 4 < 75") exist for transport ratios between about 1.5 and 8. Transverse and longitudinal bedforms co-exist (i.e., interference dunes occur) at divergence angles in the vicinity of 90" for low transport ratios (Fig. 13-17B; Rubin and Ikeda, 1990). In most estuarine situations, confinement of the tidal currents by channel banks produces rectilinear flow ( y FZ 180"); thus, by Fig. 13-17B most tidal dunes should be nearly transverse to both the ebb and flood currents, and to the direction of residual transport. However, channel bends, shoreline irregularities and sand-bar crests may produce local situations where the divergence angle is less than 180"; in these cases either longitudinal or oblique bedforms will form, the exact orientation depending on the transport ratio. For example, Rhodes (1992) has documented a situation in the Minas Basin, Bay of Fundy, where the divergence angle is 130" and the transport ratio is approximately of 2 at spring tides. The resulting bedform crestlines are oriented at an angle of 40-45" to the resultant transport direction, exactly as predicted ( x in Fig. 13-17B). Divergence angles of nearly 90" are unlikely to occur in most estuarine settings, but are possible within one half of the tidal cycle due to changes in the influence of the surrounding topography as water depth rises or falls. Such a situation has been described by Hine (1975) from the flood-tidal delta of the Chatham Harbor estuary (Massachusetts). Here, the current direction during the early and middle part of the flood diverge by up to 70", producing two sets of dunes which intersect at approximately 90". Although Hine (1975) does not report the transport ratio for these two flows, it is likely that conditions fall within the field of interference dunes (Fig. 13-17B).

R.W. DALRYMPLE AND R.N. RHODES

388

Tt

A

DOMNAM TRANSPORT ID)

f

RESULTANT TRANSPORT

MVERGENCE ANGLE

IY) SUBORDINATE

e

5

I-4

K

Qz d

I - 2

1

0

20

40

60

80

100

120

140

160

180

DIVERGENCE ANGLE (7')

Fig. 13-17. (A) Definition of terms used in eq. (13-4) (see text) relating bedform-crest orientation to the sediment discharge in two directions. The lengths of D and S are proportional to the sediment discharge during the dominant and subordinate tides respectively; the ratio of D to S is the transport ratio, and the angle between them is the divergence angle ( y ) . Bedform obliquity (4) is the angle between the bedform crestline and the resultant transport direction. (B) Plot showing the bedform obliquity 4 (curved lines in diagram) as a function of the divergence angle and transport ratio: longitudinal bedforms (4 < 15'); oblique bedforms (15" 5 4 5 75"); transverse bedforms (4 > 75"). The co-existence of transverse and longitudinal forms creates interference dunes, but the limits of this field are poorly known. " x " indicates the conditions discussed in the text. After Rubin and Hunter (1987) and Rubin and Ikeda (1990).

Non- unifom migration Bedforrn orientation is also influenced by differences in the forward migration speed of adjacent parts of the dune crest. The cornmonly-used formula relating sediment discharge to bedforrn migration speed and size (van den Berg, 1987) can be

ESTUARINE DUNES AND BARS

389

reorganized and generalized for different degrees of obliquity to give (13-5) where UB = dune migration speed; 9s = (net) sediment discharge (dry weight); 4 = bedform obliquity (Fig. 13-17A); ps = sediment density; E = porosity; B = a factor to account for bedform shape and the bypassing of sediment in suspension ( ~ 0 . 6 ) ;and HD = dune height. Thus, along-crest variations in net sediment discharge and/or bedform height will lead to different migration speeds and oblique orientations relative to the effective current and net transport direction (Fig. 13-18). On the basis of detailed measurements Rhodes (1992) has suggested that this process operates to accentuate the sinuosity of 3D dunes, because h a t e segments of the lee face up-flow of scour pits generally migrate more slowly than linguoid segments at spurs. On a larger scale, non-uniform conditions of water depth and current speed between the axis of an estuarine channel and the adjacent bar crest or channel bank are also likely to produce oblique dune orientations, provided that the lateral extent (span) of individual crestlines is sufficiently large (Fig. 13-18). The orientation of the crestlines relative to the channel axis cannot be predicted in advance, as it depends on the precise, cross-flow distributions of depth, current speed, and dune height; both inward- and outward-facing orientations are possible. It is probable that the in-line inflections or kinks that are sometimes observed in the crests of larger dunes (Fig. 13-16; Dalrymple, 1984, fig. 3) are a result of along-crest changes in the relative influence of height and sediment discharge on migration speed (Fig. 13-18). If the lateral gradient of migration speed is too great, the crest may be stretched to the breaking point (Fig. 13-18E). Current-parallel lines where such breaks occur probably mark the boundaries of bedform fields with homogeneous or gradually varying morphological characteristics (e.g.,Langhorne, 1973; Aliotta and Perillo, 1987).

Discussion The two processes controlling dune orientation (variable current directions and non-uniform migration) may operate independently or simultaneously in an estuarine environment. The effects of non-uniform conditions are likely to dominate in straight channels because the divergence angle is usually nearly 180". On the other hand, the influence of variable transport directions may be important at channel bends and confluences, on bar crests, and near shoreline irregularities where divergence angles less than 180" are more common. Large and very large dunes which have a large span are more likely to show the influence of non-uniform flow conditions than are small to medium dunes (Dalrymple, 1984). This means that accurate determination of the transport direction using large bedforms may be subject to significant error. Rubin (1987b, pp. 10-12) discusses this problem and readers are referred there for additional information.

390 A

R.W. DALRYMPLE AND R.N. RHODES

I'

I

,

I

I

1

I

I

D

0

c

0

25

50

75

100

125

150

175

200

HORIZONTAL DISTANCE (m)

Fig. 13-18. Input data (A-C) and results (D, E) of numerical simulation of the influence of non-uniform conditions on bedform orientation. The cross-flow variations in mean current speed ( U ; part A) and water depth (B) were chosen to model the flank of an estuarine channel (zero distance at channel axis). The current speed at 200 m is approximately at the lower limit of the dune stability field. The distribution of dune height ( H ; part C) was derived from the water depth using eq. (13-l), but with a more rapid decrease beyond 100 m because of the decrease in current speed (cf. Figure 13-6D). (D, E) Relative migration distance ( U B )after 1 and 30 iterations, calculated as UB = KU3(sinq6)/H (cf. eq. (13-5)), starting with a straight, flow-transverse bedform (4 = 90"). The obliquity used at each point along the crest in subsequent steps is the obliquity obtained after the previous iteration. K is a coefficient incorporating the [p,(l - &)PI term of eq. (13-5) and the coefficient relating qs to U 3 . Because K is not easily determined, it has been assigned a value of 1 and the results are plotted without units. Thus, the distribution of U B approximates the shape of the dune crest but not the absolute migration distance or rate. Despite the monotonic distributions of current speed, depth, and dune height, the crestline becomes strongly kinked, due to the opposing influences of U 3 and H . The dashed portion of the crestline in (E) is greatly elongated (note the break in scale) and is unlikely to remain intact. Note that the segment between 125 m and 175 m faces outward toward the channel margin whereas adjacent segments face inward.

ESTUARINE DUNES AND BARS

391

SUPERIMPOSED DUNES

The superposition of smaller dunes on both the stoss and lee sides of larger ones (Figs. 13-2E-G; 13-12B; 13-16) is widespread in the tide-dominated, outer portion of estuaries. As discussed above, this is believed to represent a quasi-equilibrium superposition related to the development of an internal boundary layer on the stoss side of the larger dunes. For this process to operate, the larger form must be long enough to allow development of the internal boundary layer (Ashley, 1990). Judging by the lower limit on the size of compound dunes (sandwaves) given in previous studies (Boothroyd and Hubbard, 1975; Dalrymple et al., 1978; Dalrymple 1984), it would appear that the minimum wavelength is approximately 8-10 m. Because the internal boundary layer thickens as the near-bed currents accelerate up the larger stoss side (due to flow constriction over the main dune crest), one should expect to see progressive changes in the size and shape of the superimposed dunes between the trough and crest of the main dune. For instance, Dalrymple (1984) has noted that the superimposed dunes generally increase in wavelength and height toward the crest of the larger dune, with a progression from 2D dunes in the trough to 3D dunes at the crest (Fig. 13-2F, G). This is what would be expected if the near-bed flow conditions progressed from near the ripple-dune boundary in the trough toward the core of the dune stability field at the main crest (Figs. 13-2; 13-6). In cases where the near-bed flow remains attached down the lee side (cf., Richards and Taylor, 1981; Sweet and Kocurek, 1990), a reverse succession of superimposed forms should occur there. Some authors have reported, however, that the smallest superimposed dunes occur at the crest of the larger dune (Langhorne, 1973; Fenster et al., 1990). The reason for this is not clear, but three possibilities occur to us. 1) Current speeds at the main crest may be fast enough to push conditions into the upper half of the dune stability field where dune height (but not dune wavelength) is expected to decrease (Figs. 13-6; 13-7); however, because the wavelength of the superimposed forms also decreases this explanation is unlikely. 2) Sediment grain size typically becomes coarser toward the crest of large dunes (e.g., Dalrymple, 1984); thus, flow conditions at the main crest may be closer to the lower limit of the dune stability field than in finer sand lower on the profile (cf., Fig. 13-6B, D). 3) The small crestal dunes may have formed later and in a flow that was less intense than that which formed the larger superimposed dunes lower on the main stoss side. Such a weaker flow may have only been capable of remoulding the crestal part of the main dune, leaving the superimposed dunes in the trough unaltered. Because superimposed bedforms are relatively small, they lag behind changes in flow much less than the larger forms (see more below) and are observed to change their size, shape, and distribution markedly over relatively short time periods. For instance, superimposed dunes may be absent or smaller and less widely distributed at neap tides than at spring tides (Dalrymple, 1984) and may be obliterated by energetic storm waves (Langhorne, 1977). It is commonly observed that superimposed dunes are oriented at an oblique angle to the larger form (Figs. 13-2E-G; 13-12B; Terwindt, 1971; Langhorne, 1973; Boothroyd and Hubbard, 1975; Dalrymple, 1984; Aliotta and Perillo, 1987; Fenster

392

R.W. DALRYMPLE AND R.N. RHODES

et al., 1990; Bern6 et al., 1993), with the angular divergence reaching nearly 90" in some cases (30-60" divergences are more common). Given that the two sizes of dune are believed to respond to different portions of the flow, the most logical explanation is that the near-bed current which forms the superimposed dunes has a different direction than the main, external flow. This in turn is most likely due to deflection of the near-bed flow by the three-dimensional shape and oblique orientation of the larger dunes (e.g., Malikides et al., 1989; Sweet and Kocurek, 1990). In the case of relatively small compound dunes in Cobequid Bay (Bay of Fundy), on the other hand, Dalrymple (1984) has demonstrated that the smaller dunes are transverse to the main flow, while the larger forms are oblique to it. Here, the larger dunes are too small to affect the orientation of the near-bed flow significantly, but have a sufficiently large span that their orientation is skewed by along-crest differences in migration rate.

MORPHOLOGICAL RESPONSE TO UNSTEADY FLOW

Estuarine flow conditions (depth, speed, direction, temperature, etc.) are rarely steady for long. Thus, as already discussed in general terms, one or more aspects of dune morphology (size, plan shape, asymmetry, etc.) will also change in such a way as to move the bedform toward renewed equilibrium with the flow. Such morphological changes require the movement of a finite volume of sediment (Vr), the amount depending on the size of the bedform and the nature of the morphological change. The movement of this sediment in turn requires a finite amount of time (T,) which is called the lug time. Although it is rarely possible to determine the lag time accurately, it is evident that (13-6) where fi and f2 are unknown functions and qs is the sediment-transport rate. Thus, the lag time is directly proportional to dune size (as given by HD and L D ) and inversely proportional to the rate of sediment movement (Allen and Friend, 1976b; Bokuniewicz et al., 1977). Therefore, other things being constant, larger bedforms will lag behind changes in the flow more than smaller dunes, and dunes formed by slower currents (2D dunes) will lag more than dunes formed by faster currents (3D dunes) because of the strong positive relationship between current speed and sediment-transport rate. The effect of lag can be seen most easily by plotting the value of a morphological parameter (e.g., dune height or wavelength) against a flow parameter that is changing. If equilibrium were maintained at all times, all points would plot on a single line (Fig. 13-19, light lines). Because of lag, however, the extreme bedform size (maximum or minimum) achieved will not occur until after the flow has reached its extreme value. Thus, dune height, for example, will continue to increase after the flow depth has begun to decrease, with the maximum height lagging the maximum depth by a finite amount. As a result, the trajectory of points forms a hysteresis loop (Fig. 13-19; Allen et al., 1969; Allen, 1973b, 1976a). Ideally, the direction of

ESTUARINE DUNES AND BARS

393

--

5-

E

B

DUNES 2D

I

30

n

I I-

0

z W

2 -

W

>

s

1-

i'

I

I

01

2

W

02

03

DEPTH

04

(m)

06

08 -

08

SPEED

(m/s)

SPEED

(rn/s)

/ /

6

01

0.2

0 3 0 4

DEPTH

(m)

Fig. 13-19. Idealized representation of hysteresis loops expected for dune wavelength and height, due to variations in water depth and current speed. The light line in each section is the equilibrium relationship predicted from Fig. 13-6 (for a current speed of 0.7 m/s in A and C, and for a mean grain size of 0.5 mm in B and D). In A and C, all depths produce 3D dunes at the speed selected. Note the complex, double loop in D which develops because of the double-valued relationship between dune height and current speed (Fig. 13-6D). Current-speed changes which do not span the entire range will develop only part of this pattern: a counterclockwise trajectory at low speeds (i.e., where 2D dunes are present), a clockwise trajectory at high speeds (3D dunes present), or a complex mixture of both at intermediate speeds. Although the radius of curvature is shown as equal at both ends of each loop, a smaller radius (less lag) may occur at the upper end because the sediment discharge is higher.

movement around the loop will be counterczockwise if the morphological and flow parameters are directly related (Fig. 13-19A-C), but clockwise if the parameters are inversely related (e. g., dune height in the upper half of the stability field; Figs. 13-6D; 13-19D). Because the bedforms never reach equilibrium with the extreme conditions, the range of morphological parameters observed in an unsteady flow should be less than that predicted from equilibrium relationships (Allen, 1976a). Numerous studies have examined the response of estuarine dunes to variations in flow conditions over the neap-spring lunar cycle (e.g., Allen et al., 1969; Allen and Friend, 1976a; Tenvindt and Brouwer, 1986; Davis and Flemming, 1991; Rhodes, 1992). Although these studies generally show the effects of lag (Figs. 13-20; 13-21), the data are typically noisy and do not show the smooth hysteresis loops predicted theoretically (Fig. 13-19; Allen, 1974, 1976a). Furthermore, there is not always

R.W. DALRYMPLE AND R.N. RHODES

394

E

-64

I t

W

60

SPEED (m/s)

2D DUNES 056

048

040

064

072

080

SPEED (m/s) 036

-

016 -

+ I

(-7

014

W

I

-

0 12

2D DUNES 040

048

056

064

SPEED Im/s)

072

080

0 20

04

06

08

10

12

SPEED (m/s)

Fig. 13-20. Temporal variation of dune wavelength (A, B) and height (C, D) as a function of the maximum, dominant (flood) current speed (depth averaged) for 2D (A, C) and 3D (B, D) dunes in the Westerschelde estuary. The measurements span a complete neap-spring-neap cycle. After Tenvindt and Brouwer (1986, fig. 10).

agreement between studies. Factors which contribute to these inconsistent results include: 1) the collection of data at intervals that are too widely spaced in time, relative to the neap-spring cycle, to permit resolution of the loop (e.g., Davis and Flemming, 1991); 2) the measurement of only a small number of dunes (typically (10; Terwindt and Brouwer, 1986; Rhodes, 1992), thereby allowing the random behaviour of individuals to have a significant effect on the results; 3) irregular variations in flow conditions (e.g., Tenvindt and Brouwer, 1986) which cause the morphological response to deviate from a smooth trend; and 4) the occurrence of extraneous events (strong wave action or river floods) which reset the system to a new starting position. Future studies should attempt to minimize these factors. Because of the influence of lag, most studies of estuarine dunes find no statistically-significant relationship between any morphological parameter (height, wavelength, steepness or asymmetry) and either flow depth or current speed over one or more neap-spring cycles (Figs. 13-20; 13-21). Correlation coefficients are typically t0.5,and are commonly t O . l (Table 13-3; Tenvindt and Brouwer, 1986; Rhodes, 1992). This occurs despite the fact that all studies so far undertaken have examined small to medium dunes which should equilibrate with the flow relatively

395

ESTUARINE DUNES AND BARS

32

W

I 16

12

;::?p: 07

08

09

10

07

SPEED (m/s)

08

09

10

SPEED (m/s)

32

28

-

E

24-

I

c

-

I

p w

Id W -

3 --

20-

I

-

16 -

20

12

60

70

80

90

DEPTH (m)

Fig. 13-21. Temporal variation of mean, dune wavelength (A, C) and height (B, D) as a function of the maximum, dominant (ebb), current speed 100 cm above the bed (A, B) and high-tide water depth (C, D), which is a surrogate for tidal range, at a site in the Minas Basin, Bay of Fundy. The bedforms were small, 3D, simple dunes and the measurements span three neap-spring-neap cycles (only the second neap-tide interval is pronounced). Note the abrupt drop in mean wavelength and height shortly before peak spring tides, due to the creation of small, superimposed dunes. After Rhodes (1992).

Table 13-3 Correlation coefficients (Y) for the relationships between various morphological parameters and the current speed 100 cm above the bottom (for the dominant, ebb tide), for fields of 2D and 3D dunes in the Minas Basin, Bay of Fundy (after Rhodes, 1992, table 4.3). Morphological parameter

2D Dunes

3D Dunes

Height Wavelength MSI RI

0.18 0.03 0.03 0.22

0.50 0.41 0.05 0.10

MSI = modified symmetry index (Allen, 1980); RI = ripple index. Note: 3D dunes are more responsive to changes in flow speed (i.e., have higher correlation coefficients) than 2D dunes; height is the most strongly correlated parameter.

396

R.W. DALRYMPLE AND R.N. RHODES

rapidly. Nevertheless, the correlation coefficients are typically higher for 3D dunes (Table 13-3), as would be expected because of their occurrence in areas with higher sediment-transport rates. Despite of the scatter, the following points may be noted. The average wavelength of both 2D and 3D dunes commonly remains more or less unchanged (the variation is usually ~ 2 0 %of the mean value) over a neap-spring cycle (Fig. 13-20A, B; Tenvindt and Brouwer, 1986), in part because a decrease in the wavelength of one dune is offset by an increase in the lengths of its neighbours. When wavelength changes do occur, they take place abruptly, in a step-like fashion (Fig. 13-21A, C), due to the sudden increase (or decrease) in the number of dunes present. For example, Allen and Friend (1976a), Tenvindt and Brouwer (1986), and Rhodes (1992) all report a sharp decrease in average dune spacing at or just before the highest spring tides. The creation of smaller dunes at spring tides is contrary to the equilibrium relationship which predicts larger dunes in faster (and deeper) flow (Fig. 13-19A, B), but may be explained if the simple dunes which were present over most of the neap-spring cycle temporarily became compound dunes during spring tides due to intensification of flow in the internal boundary layer. The observation of Allen and Friend (1976a) that the new, small dunes occurred primarily near the crest of the larger dunes is consistent with this interpretation. Following the sudden decrease, wavelengths return slowly to near-original values (Fig. 13-21A7 C), presumably because the superimposed smaller dunes migrate faster than the major dunes and disappear by amalgamating with them. When hysteresis loops are observed in wavelength data (Figs. 13-20; 13-21), the trajectory is counterclockwise as predicted (Fig. 13-19A, B). Dune height is more variable than wavelength and there is a general tendency for it to increase as current speed (and depth) increase from neap to spring tides (and vice versa; Figs. 13-20; 13-21; Allen et al., 1969; Allen and Friend, 1976a; Tenvindt and Brouwer, 1986; Rhodes, 1992). However, instances are recorded where dune height decreases at the highest spring tides (Fig. 13-21B; Davis and Flemming, 1991; Rhodes, 1992), and Allen (1976b) notes a similar negative correlation with river discharge in the inner part of the Weser River estuary. Presumably flow conditions were in the upper half of the dune stability field so that the bedforms became smaller as the current speed increased; indeed, it would appear that all of these cases occurred in fields of 3D dunes as this explanation demands (Fig. 13-19D). (Note that 2D dunes show only a positive correlation between height and current speed.) Due to this complex response to changes in current speed, hysteresis loops are not always obvious in height data (Figs. 13-20; 13-21), even though the lag of dune height is generally short, being of the order of 1-3 tides (Tenvindt and Brouwer, 1986). A small number of studies have been done on the response of dunes to the longerterm changes in river discharge which characterize the inner part of estuaries. These studies, which are summarized and discussed by Allen (1973, 1976a, b), generally show that the response for wavelength and height are similar to those discussed above: height is more variable than wavelength and wavelength hysteresis loops have counterclockwise trajectories whereas height shows both counterclockwise and clockwise trajectories. Hysteresis loops are much more clearly developed than they are in tide-dominated settings, due to the slower rate-of-change of flow conditions.

ESTUARINE DUNES AND BARS

397

Fig. 13-22. Small, 2D, simple dunes with rounded and subdued profiles, as seen at neap tide in a flood-dominant channel, Arcachon Basin, France. No ebb cap is developed, and the trench shows no evidence of an angle-of-repose lee face at high tide. The tidal currents obviously do not reach the conditions necessary for dune maintenance, and the dunes are in the process of decaying to current ripples.

In several of the studies cited above (Allen and Friend, 1976a; Tenvindt and Brouwer, 1986) current speeds dropped into the ripple field, or even below the threshold of sediment motion, for several days during neap tides. However, little degradation of the dunes occurred in this brief period. Rhodes (1992) has examined a case where the longer-term reduction in current speeds associated with the passage from equinoctial spring tides in March to smaller spring tides in July and August caused flow conditions to fall into the ripple regime. At the beginning of the measurement period (early July), simple, 2D dunes with an average wavelength and height of 9 m and 0.26 cm were present. They exhibited subdued, rounded profiles and lacked an angle-of-repose lee face (cf., Fig. 13-22), suggesting that significant degradation had already occurred. Over the 35-day measurement period, both the height and wavelength decreased steadily, except for a brief increase in height during large spring tides when the maximum speeds briefly returned to the dune stability field. The dunes also became progressively more symmetrical. A projection of the trend of declining height suggests that another 4.5 months would be needed to obliterate the dunes. Although this value is subject to uncertainty and cannot be generalized, it indicates that the time needed to rework dunes into ripples is considerable, even for relatively small dunes. In estuarine environments current and dune reversal on a tidal period (usually semi-diurnal) is one of the most prominent aspects of flow unsteadiness. As discussed above, sediment is eroded from the former brink region with each reversal and

398

R.W. DALRYMPLE AND R.N. RHODES

A

B 15C

2 m

D Y W

zI-

1oc

d


-0.5-1 m), compound dunes, but a more reasonable, average slope for smaller dunes is 20-25” (Rhodes, 1992). For such dunes, S is closer to 0 . 7 5 L ~ thus, ; the results plotted in Fig. 13-23B underestimate the reversal time for small bedforms. Note also that this analysis assumes continuous transport in one direction; the actual reversal time may be considerably longer due to periods of no movement at slack water and transport in the opposite direction. Nevertheless, it is evident from Fig. 13-23B that dunes with heights of more than 0.3-0.5 m are very unlikely to reverse completely during a single, half tidal cycle. Clearly, the maximum height for which complete reversal is possible decreases as the discharge rate decreases, being closer to 0.1-0.2 m for the lowest transport rates plotted. The reversal of large to very large dunes requires a reversal in the direction of net transport for a period of weeks. In an estuarine setting, such transport and bedform reversals can occur as a result of seasonal changes in fluvial discharge (Bern6 et al., 1993).

DUNE MIGRATION RATES

The migration rate of estuarine dunes has been reported by numerous workers (Table 13-4). The values cover a very large range, varying greatly within and between areas. As predicted by eq. (13-5) above, the migration rate ( U B )generally decreases as the bedform height increases: average rates for small dunes ( 3 m) may have rates of only a few decimetres per year (Fenster et al., 1990). As a result of such differential migration rates, the small to medium dunes superimposed on large to very large compound dunes migrate faster than the larger form, moving up the stoss side and onto the lee face where they are partially or completely “absorbed by deposition as they migrate downward into areas with lower current speeds and sediment-transport rates (Dalrymple, 1984; Rubin, 1987b). Net sediment discharge, which depends on the speeds of the dominant and subordinate currents, also controls the dune migration rate [eq. (13-5)]. In unidirectional flow or in situations where there is little or no sediment transport by the subordinate current it is commonly suggested that (13-7a) where UD is the maximum or modal speed of the dominant current, although Salsman et al. (1966) suggest that U i gives a better correlation with dune migration rate. If both the dominant and subordinate currents transport appreciable amounts of sediment, eq. (13-7a) becomes

ue 0: qs VJ;

-

U,3)

(13-7b)

400

R.W. DALRYMPLE AND R.N. RHODES

Table 13-4 Compilation of reported dune migration rates Author

Location (dune type)

Salsman et al. (1966)

St. Andrews Bay, Florida: flood-tidal delta (2D simple dunes)

0.49-0.58

4.9

Chesapeake Bay: tidal inlet (compound dunes)

0.5-2.1

2-150

Thames estuary: outer sand bars (compound dunes)

1.5-8

Long Island Sound: “flood-tidal delta” (compound dunes?)

1

63;

0-125

Minas Basin-Cobequid Bay, Bay of Fundy: outer sand bars (compound dunes)

0.8

75;

7-220

Westerschelde estuary: middle estuary, intertidal shoal (2D simple dunes) (3D simple dunes)

0.15 0.26

120; 350;

Bahia Blanca estuary: subtidal channel (compound dunes)

3-4

33;

Fraser River: distributary channel (compound dunes)

0.3-2.1

Fenster et al. (1990)

Long Island Sound: “flood tidal delta” (compound dunes)

4.0-16.5

Rhodes (1992)

Minas Basin, Bay of Fundy: outer sand bars (2D simple dunes ) (3D simple dunes)

0.20 0.27

Ludwick (1972)

Langhorne(1973)

Bokuniewicz et al. ( 1977) Dalrymple (1984)

Terwindt and Brouwer (1986)

Aliotta and Perillo (1987) Kostaschuk et al. (1989)

Dune height (m)

Migration rate (m/Yr)

50-100 cm) to contain the scour. The amount of erosion by the subordinate current also increases from neap to spring tides; thus, erosional reactivation surfaces may be absent or only weakly developed at neap tides but become prominent, low-angle surfaces at spring tides (Fig. 13-24; de Mowbray and Visser, 1984; Rhodes, 1992). The angle of this surface in turn influences the succession of structures within the succeeding bundle. If the reactivation surface is steep, the first structures formed on the resumption of the dominant tide will be angular foreset laminae (Fig. 13-24A), but if the reactivation surface dips gently, the initial flow may remain attached and generate downslopemigrating current ripples (or concordant laminae if the lee-face flow is below the threshold of ripple; Figs. 13-24B, C). These are the reactivation structures of Kohsiek

403

ESTUARINE DUNES AND BARS

TIDAL BUNDLE COMPONENTS REACTIVATION ANGULAR

TANGENTIALKONCAVE

ubordinate Tide

activation Surface

Subordinate Tide

C

Fig. 13-24. Idealized succession of structures within tidal bundles formed by (A) relatively low current speeds and (B, C) high current speeds. In (A), the subordinate current has caused little erosion, leaving a steep lee face. The dominant current has not been strong enough to place significant amounts of sediment in suspension so that tangential toesets are hardly developed. In (B) and (C), stronger subordinate currents produce greater erosion and a lower-angle reactivation surface. Strong dominant currents place significant amounts of sediment in suspension and produce tangential toesets and concordant slackening structures. In ( C ) , strong currents during peak flow have caused a deepening of the scour pit, and truncation of the toe of the reactivation surface and its associated reactivation structures. This may accompany a transition to a more 3D morphology. The concordant laminae that are part of the reactivation structures form because grain size or near-bed current speeds are not in the ripple stability field. (A) and (B) modified after Kohsiek and Tenvindt (1981) and de Mowbray and Visser (1984); (C) after Rhodes (1992).

and Tenvindt (1981) and Tenvindt (1981). Only after some time has passed does the lee face steepen to the point where avalanching is re-established (de Mowbray and Visser, 1984; Rhodes, 1992). Initially the toesets are angular, due to the small amount of sediment in suspension. Concave foresets and slackening structures that consist of concordant laminae formed by suspension fallout (Fig. 13-24; Tenvindt, 1981; Kohsiek and Tenvindt, 1981) only form if the flow becomes strong enough to place large amounts of sediment into suspension, and become more prominent as current speeds increase toward spring tides. If the peak flow is sufficiently strong, scour in the

404

R.W. DALRYMPLE AND R.N. RHODES

70

1

I

1

I

I

I

1

1

60 h

E

2

S

50

v)

N

S

v)

40

Y

0

I I-

30

w

-1

kj!

20

3

m

10

n

0

10

20

30

40

50

60

70

80

90

100

110

120

130

BUNDLE NUMBER Fig. 13-25. Variation in the horizontal extent (thickness) of tidal bundles within a subtidal cross bed over several, neap-spring (N-S) lunar cycles. These thickness changes are caused by cyclic variations in tidal-current speed and dune migration rate. After Visser (1980).

trough may truncate the lower portion of the reactivation deposits (Fig. 13-24C). The presence of low-angle reactivation surfaces and draping laminae commonly impart a sigmoidal shape to the tidal bundles, leading some workers to use the term sigmoidal cross-bedding (e.g., Kreisa and Moiola, 1986). The deposits formed during slack water periods generally consist of drapes composed of mud (commonly pelleted) and/or organic material. The thickness of these drapes depends on the amount of sediment in suspension. This is greatest beneath the turbidity maximum during neap tides when very high suspended-sediment concentrations can occur at the bed (e.g., Allen et al., 1980). Either a single or double mud drape may be present (Fig. 13-24). Double drapes require mud deposition during both high and low slack water and a subordinate current which is strong enough to deposit a sand layer but not so strong as to erode the first drape. Thus, double drapes are less common than single drapes and are most abundant in, but not restricted to, the subtidal zone where mud deposition is more likely at low tide.

Compound dunes The larger size of compound dunes makes it less likely that their structures will show the effects of semi-diurnal or diurnal, tidal-current reversals and changes in speed. Instead, longer-term flow unsteadiness related to neap-spring cycles, changes in fluvial discharge, and storm events will be more clearly evident. The superimposed dunes which migrate over the crest and down the main lee face also have an significant influence (Dalrymple, 1984; Rubin, 198%). As a result, the lee-face

ESTUARINE DUNES AND BARS

405

A

Fig. 13-26. Schematic diagrams illustrating the variability of internal structures within compound dunes, associated with a decrease in the inclination of the lee face and master bedding planes (the Ez erosion surfaces). These diagrams are based on the assumptions that: the EZ surfaces are formed by erosion during the subordinate tide (periodic wave action would have a similar effect); the superimposed dunes are recreated in situ on the lee face at a later time; and superimposed dunes migrating over the crest have no effect on the lee-face structures. Because of this, the bases of the small, cross-bed sets (the E3 surfaces) in (C) and (D) have a very different geometry relative to the Ez surfaces than is observed by Dalrymple (1984) and modelled by Rubin (1987b). From Allen (1980).

structures of compound dunes are much more complicated than those formed by simple dunes. As illustrated by Allen (1980) and Dalrymple (1984), the structures may range from relatively simple foresets with few discontinuities (Fig. 13-26A, B), to complex cosets with cross-cutting, low-angle, master bedding planes (Fig. 13-26C, D; see also Berne et al., 1993). Subordinate-current cross beds are commonly present in the latter structures. The steepness of the master bedding planes reflects the inclination of the lee face which, as discussed above, is determined by a combination of factors including the relative strength of the dominant and subordinate tides (Allen, 1980), erosion by the superimposed dunes (Dalrymple, 1984), wave action (McCave, 1971), and the nature of lee-face flow as affected by the presence of

406

R.W. DALRYMPLE AND R.N. RHODES

superimposed dunes and the orientation of the main bedform (Sweet and Kocurek, 1990). In detail, the relationship of the smaller sets to the master bedding planes (the Ez surfaces in Fig. 13-26) will depend on whether the superimposed dunes are created in situ on a smooth lee face following an erosional event (Allen, 1980), or migrate over the main crest and (obliquely) down the lee face (Dalrymple, 1984; Rubin, 1987b). In the former case, the set bases of the smaller cross beds diverge from the master bedding planes (Fig. 13-26C), whereas in the latter case the set bases are the master bedding planes (cf. Figure 13-26B). Neap-spring cyclicity may be recorded within the small sets, and bioturbation and mud drapes may also be present. Additional examples of the complex structures which may be created by superimposed dunes are provided by Rubin (1987b). The nature of the bottomsets of compound dunes is poorly known, but a range of possibilities exists. If current speeds in the trough are high enough to sustain dunes, the bottomsets will be cross bedded, although probably with smaller set thicknesses than higher in the coset due to the lower current speeds in the trough. Bioturbation and mud drapes are likely to be more abundant than higher in the coset for the same reason. At the other extreme, current speeds in the trough may be below the threshold of sand movement most of the time, causing the bottomsets to consist of bioturbated, sandy mud. This is most likely near the limit of dune fields, where they pass outward into rippled sand flats, mudflats, or the lagoon floor (cf. Harris et al., 1992). In all cases the deposits of a single compound dune should coarsen upward because of the upward increase in current speed and wave action (McCave, 1971; Allen, 1980; Dalryrnple, 1984).

ESTUARINE BARFORMS

General characteristics and classification Like barforms in fluvial environments, estuarine barforms come in a bewildering array of sizes and shapes. Flow-transverse, oblique, and longitudinal orientations all occur, sometimes combined in a single, composite body. They may be more or less regularly repetitive in their spacing, either parallel or transverse to the flow, or occur as isolated individuals. As indicated at the beginning of this chapter, barforms are generally larger than dunes and commonly have dunes superimposed on them; however, an overlap in the range of possible sizes blurs the distinction between them. Typically, bars have flow-parallel spacings which are several times the channel width and a flow-transverse dimension which is a large fraction of the channel width. Thus, barforms are said to scale with flow width rather than flow depth as dunes do (see more below). This offers one of the few quantitative guidelines for distinguishing between dunes and bars in a modern estuary: if the feature has dimensions which are appreciably greater than those predicted from the water depth using eqs. (13-1) and (13-2), then it is probably a bar rather than a dune. For example, the bar segments which occur on elongate sand bars in Cobequid Bay, Bay of Fundy (Dalrymple et al., 1990) have many features in common with large and very large dunes,

ESTUARINE DUNES AND BARS

407

including quasi-regular spacing, a consistent asymmetry and migration direction, and superimposed, small to large dunes. However, their spacing (200-3000 m) and height (1-5 m) are significantly greater than would be expected of a dune (wavelength 18-72 m; height 0.5-2 m) in the effective water depths recorded (3-12 m). Thus, they are considered to be bars. A similar analysis indicates that the isolated “transverse bars” described by Boothroyd and Hubbard (1975) and Boothroyd (1985) are not dunes. Because of the diverse size and shape of barforms, a large number of terms have been used to define supposedly discrete types; for example, Smith (1975) has compiled more than 30 different names for fluvial bars. All of these types are probably present in estuaries, together with unique forms which result from reversing tidal flow. Because of our poor understanding of their genesis, there is little consensus on bar classification. Judging by the terms used to name fluvial bars, previous classifications have emphasised such attributes as: position in channel (mid-channel, bank-attached, and channel-junction bars); planform shape (linguoid, lunate, and elongate bars); orientation relative to flow (diagonal, transverse, and longitudinal bars); and hierarchical complexity (unit and compound bars). However, such descriptive terms do little to advance our understanding of their origin. Therefore, we have divided barforms into three broad categories which we believe to be genetically significant yet recognizable on observational grounds: 1) repetitive bars including alternate, point, and multiple, braid bars; 2) elongate tidal bars; and 3) isolated, delta-like bodies including spill-over lobes. In the following descriptions, only relatively simple, unmodified examples of each bar type will be considered. In many cases the superposition of the various bar types on each other, together with modification and dissection of the basic forms due to changing water levels, can produce very complex assemblages in which the separation of the individual elements is difficult. Comprehensive discussion of these complex forms is beyond the scope of this chapter.

Repetitive ba$orms Many bars exhibit a quasi-regular, repetitive spacing in the direction parallel to flow. The most abundant estuarine examples include the tidal and tidal-fluvial point bars and alternate, bank-attached bars which occur in the tidal channels and creeks of estuaries worldwide (e.g., Banvis, 1978; Arbouille et al., 1986; Ashley and Zeff, 1988; Dalrymple et al., 1990). Using an ingenious analogy with dunes, Yalin (1977) and Yalin and da Silva (1991, 1992) have argued that repetitive bars represent the imprint of horizontal turbulence (eddies shed from the banks), whereas dunes are the imprint of vertical turbulence (eddies shed from the bed). For relatively small values of the width (B)-to-depth ratio (less than approximately 100, the exact value depending on the relative roughness; Yalin and da Silva, 1992), alternate bars and meander point bars are produced. The spacing of these bars ( L B )is related to flow width in the same way that dune wavelength is related to flow depth (cf. eq. 13-1):

Lg = 6 B

(13-8)

408

R.W. DALRYMPLE AND R.N. RHODES

I

I

I

I

1.0

1.5

2.0

2.5

I

LOG CHANNEL WIDTH (log m) - B Fig. 13-27. Plot of along-channel spacing of alternate and point bars in tidal creeks in South Carolina (dots; Banvis, 1978) and the Salmon River estuary, Bay of Fundy (stars; Zaitlin, 1987). Note the close agreement with the predicted relationship (eq. (13-8)) over more than one order of magnitude. The larger starred dot is the mean value for multiple-row bars reported by Zaitlin (1987), with bar width plotted in place of channel width. This value clearly falls below the trend, indicating that multiple bars follow a different relationship.

For higher values of the width-to-depth ratio, the theory suggests that multiple rows of en echelon braid bars are produced, the number of rows increasing as the width-to-depth ratio increases. Numerous data on the spacing of alternate bars and meanders in rivers support eq. (13-8) (Yalin and da Silva, 1991, 1992) and the data plotted in Fig. 13-27 indicate that this relationship also holds in estuarine (tidal) environments. This suggests that the presence of reversing flow does not significantly alter the process by which this type of bar forms (as is also believed to be the case with dunes). The best morphological descriptions of alternate bars and meander point bars in estuaries are those provided by Barwis (1978) for tidal creeks in South Carolina and by Zaitlin (1987; see also Dalrymple et al., 1990, 1992) from the tidal-fluvial transition in the Cobequid Bay-Salmon River estuary, Bay of Fundy. They recognize a spectrum of bar shapes, the exact form depending on channel sinuosity. In nearly straight channels the alternate bars are completely welded to the bank, but as the sinuosity increases the tail of the bar (the upcurrent portion relative to the dominant current) becomes separated from the bank by one or more dead-end channels (Fig. 13-28) which are termed flood barbs if the dominant current is the flood. The resulting bar shape in plan forms half of a parabola, with a flow-transverse segment attached to the bank and a flow-parallel tail in the channel. Dissection of the bar by

ESTUARINE DUNES AND BARS

409

Fig. 13-28. Oblique aerial photograph of an alternate bar in the tidal-fluvial transition zone of the Cobequid Bay-Salmon River estuary, Bay of Fundy (Dalrymple et al., 1990, 1992). The dominant (flood) flow is to the left and the bar displays a flood asymmetry. The main crest of the bar is separated from the bank by a headward-terminating, shallow channel (a flood barb). The field of view is approximately 1 km wide.

small channels may isolate the tail as a mid-channel or diagonal bar. In longitudinal section these bars are asymmetric in the direction of local, net transport (Fig. 13-28; Dalrymple et al., 1990, 1992), with the crest located where the bar attaches to the bank. The steeper (lee) face may reach the angle of repose, but gentle slopes (t10") are more common. As the channel sinuosity increases further (radius of curvature t 2 . 5 times channel width; Banvis, 1978), more typical point bars are developed. They are asymmetrically disposed on the meander bend if there is one dominant flow direction (Banvis, 1978), but are symmetric, with ebb- and flood-dominated halves, if the two currents are more or less equal (Zaitlin, 1987; Dalrymple et al., 1990, 1992). Few descriptions exist of the multiple-row bar configurations which Yalin and da Silva (1991, 1992) predict should occur in areas with width to depth ratios greater than about 100, even though such situations are relatively common in sandy, tidal-flat environments. The multiple braid bars which occur in the broad, upper-flow-regime, sand-flat zone in the Cobequid Bay-Salmon River estuary (Zaitlin, 1987; Dalrymple et al., 1990) may be an example. Bar relief is low (0.3-1.5 m) and both flood- and ebbasymmetric forms occur, with lee-face slopes typically