DEVELOPMENTS IN SEDIMENTOLOGY 53
Geomorphology and Sedimentology of Estuaries
FURTHER TITLES IN THIS SERIES VOLUMES ...
480 downloads
2311 Views
11MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
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) “.<.a catch-all for estuaries not clearly included in the other three divisions”. He actually described in this category San Francisco Bay and its tributaries, the San Joaquin and Sacramento rivers. All of them formed by movements of the San Andreas’ fault system. Other examples of tectonic type of estuaries are the Valdivia river (Chile) (Pino et al., 1992) and Itamaraca (Brazil) (Medeiros and Kjerfve, 1993).
DEFINITIONS AND GEOMORPHOLOGIC CLASSIFICATIONS O F ESTUARIES
I/’ 41ni.t
31
MICROTIDAL
Fig. 2-8. Schematic diagrams representing the estuaries within the Hayes (1975) classification: A) microtidal, which are commonly associated to coastal lagoons; B) mesotidal; C ) macrotidal.
Although Pritchard called his classification Topographic or Physiographic, it is a genetic classification since he based the nomenclature on genetic attributes. Nonetheless all categories are quite broad and obviously inclusive, providing similar
32
G.M.E. PERILLO
characteristics to estuaries that are produced by the same process but have a totally different setting (e.g., rias and coastal-plain estuaries). Evidently geological processes may be similar on a regional or worldwide basis, but their dynamical and actual physiographic effects were not considered. For instance, estuaries located on terminal deltas or those not fully drowned (e.g., Amazon and de la Plata rivers) are not considered at all in the classification.
Classijication by tidal range (Hayes, 1975) Although based on a physical parameter, namely the tidal range, it is included as a geomorphologic classification because the tide is only employed as the leitmotiv for correlating several physiographic characteristics. Hayes (1975) in analyzing the morphology of sand deposits affiliated with estuaries recognized their different characteristics depending on tidal range. Following the coastal classification scheme given by Davies (1964), Hayes defined thus three types of estuaries: microtidaltidal range < 2 m; mesotidal-tidal range 2-4 m, and macrotidal-tidal range >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: <50 m3 s-l mean annual discharge; medium: 50-200 m3 s-l; high: >200 m3 s-l. Average water temperatures ("C)at or near the 200-m depth of fjord basins. Low: <1mm ax1 averaged over the entire fjord basin; medium: 1-10 mm a-l; high: 210 mm a-I.
of the major fjord coastlines. Each district has characteristics that are unique, with consequences to the record of sedimentation. Details beyond the review provided below can be found in Syvitski et al. (1987). Unfortunately each district has not been investigated with the same degree of thoroughness nor with the same tools and methods. Fjard coasts include the coast of Maine, USA, northeast of Portland (Embleton and King, 1970), the southern Swedish coast, and southern Oslofjord and the Skaggerak in Norway (Bird and Schwartz, 1985). Extensive parts of the Canadian coastline fall in this category, notably much of the coast of eastern Ungava Bay (Barry et al., 1977). The fjords of Greenland are among the world's largest (Funder, 1972). Most of these fjords drain the massive Greenland Ice Sheet, and variations in freshwater discharge are thus a result of air temperature fluctuations (Moller, 1984): meltwater contribution to the total freshwater input to the fjord may reach 90%. Glacial flour,
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS
121
often generated from the tidewater glaciers, is the principal source of sediment which can accumulate at > 1 m day-’ (Heling, 1974); terrestrial organic input is negligible. Greenland is the world’s largest producer of icebergs with individual fjords having typical ice production between 10 km3 a-l and 30 km3 a-l (Olsen and Reeh, 1969; Carbonnel and Bauer, 1968). Alaskan fjords are extremely rugged with hinterland mountains reaching up to 6,000 m. Freshwater discharge is a result of snow melt in the spring and rainstorms during the autumn; where glacier meltwater dominates, runoff is highest in late summer. Glacigenic sediments accumulate at very high rates, up to 9 m a-l in front of tidewater glaciers (Molnia, 1983; Hoskin and Burrell, 1972; Hoskin et al., 1976; Powell, 1991; Cowan and Powell, 1991). In this seismically active region, slumping is often triggered by earthquakes (von Heune, 1966). The island fjords of British Columbia receive little or no snowfall, and sediment input is directly related to rainstorm events. They receive limited detrital sediment input (small drainage basins, no icefields), relatively high terrestrial organic input (lush rainforest vegetation) and are sites of high primary production. Discharge into mainland British Columbia fjords often reflects spring melt of large winter snowfalls and highly turbid glacier-melt during the drier summer months. Subaqueous slope failures are a consequence of high sedimentation rates and moderate seismic activity. The fjords of Atlantic Canada mostly drain upland terrain and are strongly influenced by shelf storms; 10 m high waves may locally impact the outer fjord shoreline during winter storms. Much of the sediment distribution within these fjords reflects wave reworking of glacial (Pleistocene) sediments (Slatt, 1974; Piper et al., 1983). Exceptions include the fluvially-dominated Hamilton Inlet, Labrador, that drains a hinterland of 150,000 km2,and Saguenay Fjord, Quebec, with a hinterland of 100,000 km2, both with sizeable discharge (ranging from 1,000 to 5,000 m3 s-l ), where sedimentation rates may exceed 0.1 m a-l (Smith and Schafer, 1987). Nova Scotia inlets are better described as fjards, and seldom exceed 100 m in basin depth. Canada’s Arctic archipelago fjords are typically small and rugged (Lake and Walker, 1975), many receiving glacier melt. As a consequence of isostatic rebound of land, some sills have become subaerially exposed, resulting in the development of anoxic fjord lakes. In the extreme north, the fjords are permanently covered in sea ice: sediment deposits largely reflect warmer conditions of the early to mid Holocene (Stewart, 1991). An open water season exists further to the south, but the influx of shelf pack ice may extend the presence of sea-ice year round. Sedimentation events are very short and episodic (Gilbert, 1983; Lemmen, 1990; Syvitski and Hein, 1991). Wind blown sediment input is locally very important. These fjords are often sites of subaqueous slope failures, some related to large earthquakes. The mountainous Norwegian fjords contain waters warmed from the Gulf of Mexico, and the climate is not unlike the fjord coast of western North America. Shallow sills and low tidal range sometimes allow stagnant bottom water to form, intermittently or permanently (Stram, 1936). Sognefjord (200 km long and 1,300 m deep) and Hardangerfjord (180 km long and 900 m deep), are among the World’s largest and deepest fjords. Fluvial discharge is mainly governed by snow melting in
122
J.P.M. SYVITSKI AND J. SHAW
early spring and rainstorms in summer and autumn. Rivers are commonly filtered by numerous lakes, and most Norwegian fjords have low sedimentation rates (1 to 7 mm a-l), the exception being those that drain icefields. Sedimentation rates in the glacier-dominated fjords of Spitsbergen are considerably higher than mainland fjords (Elverhoi et al., 1983), even though Spitsbergen has near desert conditions. This fact underlines the importance of turbid meltwater. New Zealand’s South Island contains fjords that are short (15 and 45 km long) and narrow ( < 2 km) (Skerman, 1964), in a region of rugged topography. Glacier melt only affects the northern fjords, such as Milford Sound. Elsewhere, rainfall is extremely high throughout the year (up to 9 m; Pickard and Stanton, 1980) and there is little seasonality to discharge. The small drainage basins contain a lush water-saturated vegetation (Poole, 1951) that in turn yields little sediment (Stanton and Pickard, 1981). The fjord sediments are mostly organic mud with rarer layers of graded sand, possibly from side-wall slope failure (Pantin, 1964; Pickrill and Irwin, 1983; Pickrill, pers. commun., 1993). North of Milford Sound fjords are more glacier-influenced, and some have been completely filled with sediment (Adams, 1980; Pickrill et al., 1981). The fjords of Chile also drain rain-drenched mountains (Pickard, 1971). Fluvial discharge reflects the progressive melt of snow in the Andes, and meltwater contributions from glaciers are very turbid. Scottish fjords have been largely filled with sediment, and some have been converted to partially mixed estuaries (e.g., Firth of Tay; Cullingford, 1979). As a result they are small and shallow (<150 m). Hinterlands are mostly upland terrane with no permanent snow or icefields. Fluvial discharge is related to peaks in rainfall. Stagnant basin-water conditions seldom last longer than a few months and occur most frequently in summer (Milne, 1972). The fjords of Antarctica are for the most part either filled or partially filled with glacial ice. The ice termini are either floating or grounded, depending on the nature of the regional climate (polar conditions being favoured for ice terminus flotation with negligible ice surface melt: Anderson et al., 1991). As in northern Canada, some of the fjords are fenced in from the sea by thick ice shelves (Nichols, 1960). The relative sediment contribution from iceberg rafting is apparently much higher in Antarctica than in the Arctic (Elverhoi and Roaldset, 1983). The northern Antarctic Peninsula fjords are relatively warm and moist, with annual precipitation ( > 1 m) dependent on the passage of cyclonic storms. Sediment plumes emanating from glaciers and rivers are seldom seen, and aeolian transport of sediment and sediment gravity flows under floating glaciers is considered important (Griffith and Anderson, 1989).
SHORT-TERM DEPOSITIONAL PROCESSES
Most of the controlling parameters of fjords are variable, with scalar harmonics on the order of seconds to centuries (Table 5-1). Even historical “constants” such as sill depth can be variable on a geologic time scale. End-member fjords include: (1) ice-influenced fjords where glaciers, icebergs and sea ice control circulation and
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS A
PLANESECTIONS
123
CROSSSECTIONS
I .
FLUVIAL-DOMINATED FJORDS
WAVE OR TIDAL INFLUENCEDFJORDS
C
SEDIMENT GRAVITY FLOW FKJRDS
D SLIDE-SLUMP DOMINATED FJORDS
DEBRIS FLOW
E
Fig. 5-3. Spectrum and generalised features of end-member fjords, showing basic effects on sediment architecture.
sedimentation dynamics; (2) river-influenced fjords associated with high rates of particulate material settling out through a density-stabilized water column; ( 3 ) wellmixed fjords where circulation is dominated by tide and or wave action; (4) fjords subject to subaqueous slope failures and sediment gravity flows; ( 5 ) fjords containing anoxic basin waters and marked temporal and spatial biogeochemical gradients. The sedimentary architecture of a fjord depends on these primary controls on sedimentary processes (Fig. 5-3). Weak bottom currents may result in stratigraphy conformable to sea bottom topography; with stronger bottom currents the basin fill becomes more on-lapping in nature and even sea-bottom channel scouring by tidal action is possible. A fjord fill that largely reflects the deposition from sediment gravity flows contains a more ponded sedimentary architecture. The Coriolis effect on suspension plumes may allow for preferential sedimentation to one of the fjord margins (right in the northern hemisphere and left in the southern hemisphere). Irregular seafloor surfaces may result from side-wall or down-fjord slides and slumps.
124
J.P.M. SWITSKI AND J. SHAW
ICE-INFLUENCED FJORDS
All fjords, by definition, have been influenced by ice, sometimes for long periods (Syvitski et al., 1987). Glacimarine sediment forms the bulk of Quaternary filling in most fjords. Twenty-five percent of the World’s fjords remain under the influence of glaciers with marine termini. Sediment entering ice-influenced fjords is uniquely dispersed by (Syvitski, 1989): (1) ice-contact processes that influence the immediate deposition of sediment; (2) glacifluvial processes that discharge sediment and subsequently transport it within a fjord’s estuarine circulation; and (3) rafting by icebergs and sea ice. For development of the governing equations for many of the concepts outlined below see Syvitski (1989), Boulton (1990), Dowdeswell and Murray (1990) and Powell (1990).
Ice-front melt The rate of sediment deposition from ice front melt depends on (Syvitski, 1989): (1) the volume of ice being melted; (2) the distribution of sediment within the glacier; and (3) the movement of the glacier terminus. Ice melt is sensitive both to the temperature differential between the ice and the water and flow velocity along the face of the ice. For trunk glaciers occupying the main fjord valley (Fig. 5-3E), near-ice flow velocities are typically small and mostly from the vertical component (Syvitski, 1989). The ice front melt rate of a glacier in South Cape Fiord, Ellesmere Island, is approximately 0.1 m a-l (after Horne, 1985) which is negligible when compared to terminus retreat rates of 10 to 100 m ax1 observed in Canada’s Arctic fjords (Syvitski, 1989). This melt rate is equivalent to 0.1% of the annual flux of freshwater to the fjord. The melt rate for the ice front in Muir Inlet, Alaska, is estimated at 20 m axl (Syvitski, 1989), much higher than our Ellesmere Island example but still low compared to the Muir Glacier retreat rate of 400 m axl (Mackiewicz et al., 1984). Winddriven upwelling and downwelling near an ice terminus may further retard or accentuate the upward-rising melt velocity and thus the melt rate. Glaciers entering the margins of a fjord can experience proportionately higher melt rates if their ice termini protrude into the high currents of the established estuarine circulation (Fig. 5-4A). The deposition of englacial and supraglacial debris from melting at the ice front depends directly on the release rate and the area of the seafloor over which the released material is spread. The spread area is affected by the residual ice-margin velocity, which in turn is governed by the ice-flow rate and the terminus retreat rate (calving and melting). The two orders-of-magnitudedifference in melt rates between Muir Glacier and South Cape Glacier is counteracted by the ice front retreat rates in Alaska being nearly 100-fold greater than those of Arctic Canada, and thus the deposit thickness from terminus melt is quite similar (Syvitski, 1989).
GlacijluviaI processes The dynamics at the ice front are significantly influenced by the style of freshwater discharge into the denser marine waters. Near a main fjord glacier, the ambient basin
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS
125
-i
0.1-
wo3-
2
J
2
W05-
$07-
i
+09-
0
5
10
15
krn
Fig. 5-4. (A) Aerial photograph of McBeth Fjord, Baffin Island (1949: NAPL T219L105). (B) A r gun seismic reflection profile across the inner sill in Cambridge Fiord, Baffin Island. Note the exponential decrease in the sediment thickness seaward from the sill, a known stable ice terminus position for the period 7700 to 6700 year BP (Syvitski, 1989; Stravers and Syvitski, 1991).
water is often a low energy environment. There can be more than one discharge outlet for a given ice terminus and it is not uncommon to have supraglacial, englacial and subglacial discharge occurring simultaneously (Fig. 5-2B). The position of a discharge outlet may change with calving as new faces of the ice front are exposed. The location of a major discharge outlet may be associated with the position of a calving bay. For submarine discharge (i.e., Fig. 5-3E) the efflux is likely to be in the form of a jet whose behaviour depends on its buoyancy force and the jet’s momentum (inertial force). For most ice front discharge, the jet consists of a fluid less dense than the basin waters and thus is directed upwards immediately out from the ice tunnel (Fig. 5-2A). Near the ice tunnel, however, the discharge momentum is initially the controlling force. As the jet expands, basin water is entrained into the jet and a vertical buoyant “plume” driven by buoyancy forces is established (Pedersen, 1986). The vertical buoyant plume rises until it reaches a density level equal to the density
126
J.P.M. SYVITSKI AND J. SHAW
level within the plume, at which point it begins to flow horizontally down the fjord (Fig. 5-2B). The horizontal plume, driven by gravity, flows in a continually changing balance between buoyancy and frictional forces. The flow pattern of the jet can be divided into: (1) the zone of flow establishment, where there is a gradual transition from the flow dynamics within the ice tunnel to that of (2), the zone of established flow, where the jet expands linearly through entrainment of the ambient fluid into the plume. The spreading will continue until there is a complete loss in upward momentum or until the surface of the fjord is reached. Where the jet is sufficiently large to intersect the water surface, suspended sediment concentrations may be 60 times greater than for water found closer to the ice front (Syvitski, 1989). Suspended sediment concentrations greater than 30 kg mP3 are needed to overcome the density contrasts between the ice tunnel fluid (turbid but freshwater) and normal sea water, and thus cause the issuing plume to sink. Sediment concentrations are seldom this high, even during flood conditions (Syvitski et al., 1987), although j0kulhlaups (glacier outburst floods) are an exception. During the establishment of the jet flow, the plume will decelerate and release the coarse sand and gravel, thus allowing the buoyancy forces to change the direction of the plume. Thus even in rare conditions where a submarine discharge is initially directed downward, the plume will rise to within a few hundred metres of the ice front (Syvitski, 1989). Outwash muddy gravels are deposited immediately outside the subglacial discharge tunnel as a chaotic sequence of graded and welded layers (Powell, 1990; Rust, 1988). Texture may vary vertically and laterally due to the variation in discharge pulses and lateral movement of the jet effluent. Imbricate gravel is deposited closest to the tunnel, with sheet or weakly channelized gravels and sands deposited further out. Sedimentation of the suspended load will peak at the boundary between the zone of flow establishment and the zone of established flow, i.e. the position where the jet begins its upward curve. Sands and coarse silts sedimented out from the rising buoyant plume are often found mixed with remobilized tunnel mouth muddy gravels in the form of slumps and debris flows (Fig. 5-4B)(Syvitski, 1989). Much of the sediment load (70%) will initially be deposited within the first 500 m of an ice front (Syvitski, 1989; Cowan and Powell, 1991). Fan depocentres have been observed to accumulate at rates higher than lo6 m3 ax1 in some temperate glacial fjords, and may even aggrade to sea level to form ice-contact deltas (Powell, 1990). Supraglacial discharge is from the melt of snow on the surrounding hills and, more importantly, the hinterland icefields. Although they may contain a contribution from glacier melt, supraglacial streams are relatively non-turbid. Supraglacial discharge into the fjord may not occur if the glacier is highly crevassed near the tidewater position. In such cases, supraglacial streams enter the fjord as englacial discharge. Elsewhere supraglacial discharge will behave similarly to extraglacial (kame deltaic) or proglacial (normal deltaic) discharge. [Such discharge enters as a buoyant overflow, and will be discussed below under river-influenced fjords.] Supraglacial flow may also enter a fjord via waterfalls at the overhanging ice front. Here the establishment of the river plume differs from normal deltaic discharge in that much of the initial jet momentum is lost during the plunge into the stratified marine waters
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS
127
and the subsequent bounce back to the surface due to buoyancy forces. After the rebound to the fjord surface, the supraglacial-derived plume will flow seaward as a normal horizontally-spreading buoyant plume. Much of the coarse sediment load is lost during the initial plunge phase and thus deposited proximal to the ice front (Syvitski, 1989). Sedimentation rates under a buoyant plume exponentially (Fig. 5-4B) decrease down the fjord (Gorlich et al., 1987; Boulton, 1990; Powell, 1991): for most fjords over 50% of the suspended load is deposited within the zone of flow establishment, and nearly 80% of the load is released prior to plume spreading to the margins of the fjord. There is a much greater loss of hydraulic energy for ice front discharge when compared with normal fluvio-deltaic discharge into a fjord basin (Andrews and Syvitski, 1994). This relates to the energy loss within the initial discharge jet while either plunging (in the case of a supraglacial stream) or rising vertically (in the case of submarine discharge). Thus for a given level of discharge, coarser sediment particles are deposited closer to the discharge outlets in the ice front case, and transported much further seaward, riding on the buoyant surface plume, in the fluvio-deltaic case.
Iceberg calving and rafting Iceberg calving is important sedimentologicallybecause (1) supraglacial sediment is released during the calving process; (2) calving generates wave fronts; and (3) it is the major control on the position of the marine ice front. Icebergs are produced from a tidewater glacier by (Lliboutry, 1965): (1) subaerial jointing of ice blocks; (2) tidewater jointing of ice blocks around ice caves formed from submarine discharge or the tidewater indenture; (3) subaqueous jointing due to buoyancy forces; and (4) detachment of giant icebergs along transverse crevasse systems. The style of calving controls how much supraglacial material gets removed from an iceberg’s surface. Glaciers generally flow fastest in their centres. Calving of this central portion, particularly in polar and sub-polar regions, tends to be by a few giant icebergs (up to 0.2 km3), produced in such a manner that the old glacier surface on the iceberg is kept upright, and little supraglacial material is released within the fjord via rafting. Temperate glaciers, with their commonly fractured termini, do not produce icebergs of this kind. Bergs formed by calving modes (1) and (2) will generally lose their supraglacial material immediately at the ice front through icebergs overturning or plunging below the water line. All modes of iceberg calving will generate a solitary wave through the rapid displacement of water by the forward and downward, or even upward (style 3), motion of the iceberg. These impact waves are large and may be generated several times a day. Every few years particularly high waves will occur, with catastrophic effects as far as 50 to 100 km from the edge of the glacier (Petersen, 1977). The magnitude of the wave amplitude depends primarily upon the net potential energy of the calving berg, whereas the wave period depends primarily on the dimensions of the berg (Weigel, 1955). Only a very small fraction of the net potential energy is converted to wave energy (Weigel, 1955; Reeh, 1985). Iceberg overturning will similarly create large bow waves (Bass and Peters, 1985). A sedimentological
128
J.P.M. SWITSKI AND J. SHAW
consequence of iceberg-generated waves is to contribute ice-rafted material to shore deposits above the normal high tide line (Syvitski, 1989). Typically an iceberg maintains its position for several days while continuously releasing sediment from its submerged portion. The more rapid melting below the waterline eventually relocates the centre of gravity and the iceberg overturns and fragments. Subsequently, sediment that has been concentrated on the iceberg’s surface from meltout will be released (Ovenshine, 1970). As glacial sediment tends to be concentrated as basal or supraglacial sediment, it appears probable that most of the glacial debris is released within the fjord (Andrews and Matsch, 1983; Dowdeswell and Murray, 1990). Nevertheless, the contribution of iceberg-rafted sediment is highly variable for glacial fjords: in Kongsfjorden, Spitsbergen, the component is volumetrically minor (Elverhoi et al., 1983) as in Coronation Fiord, Baffin Island (Gilbert, 1980); in Greenland fjords the component is more significant (Berthois, 1969; Julian Dowdeswell, pers. comm. 1993). Andrews and Matsch (1983) give five controls on the rate of iceberg rafting: (1) the disposition of glacial debris within the icebergs; (2) the rate of iceberg production; (3) the rate of iceberg drift; (4) the temperature difference between the water and the iceberg; and ( 5 ) the amount of wave action. The residence time of a berg within the fjord may vary from less than one year to 10 years in some Greenland fjords. Large icebergs may be unable to exit the fjord because of a shallow sill (terminal moraine) at the fjord mouth (Vorndran and Sommerhoff, 1974; Blake, 1977). When icebergs impact upon the seafloor, they may create furrows with parallel berms of displaced sediment. The dimensions of these “ice scours” depend on the velocity of the iceberg at impact, the dimensions of the iceberg keel(s), and seafloor geotechnical properties (Syvitski et al., 1983a): widths typically vary from 10 to 30 m; depths range from 0.5 to 6 m; and slopes of berms from the furrow floor range from 6“ to 60”. The furrows may be straight or sinuous, continuous or a series of impact pits. Once grounded, an iceberg may contribute sediment to the seafloor, resulting in mounds (for example 15 m x 15 m) of coarser grained sediment (Syvitski et al., 1983b).
Ice-fi-ontmovement If the accumulation of winter snowfall exceeds the equivalent in snow and ice melt during the summer, then the ice sheet must increase its area of wastage by expanding along its lowest (altitudinal) perimeter. This effectively lowers the elevation of the position (equilibrium line altitude or ELA) that divides the ice sheet into zones of accumulation and ablation. The rate at which this mass redistribution takes place involves topographic influences on ice velocity and the thermomechanical properties of the ice. When an ice terminus enters the marine environment the mass balance equation changes (Powell, 1991). On land when there is an elevation drop in the ELA, there is a concomitant drop in the elevation of the ice terminus. However, the terminus of a tidewater glacier is always at sea level. Thus a drop in the ELA will result in a decrease in the vertical distance between the ELA and ice terminus (Mercer, 1961), engendering an advance of the ice terminus. The year-round melting
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS
129
Fig. 5-5. Stability of fronts during a glacier advance within a fjord (after Lliboutry, 1965).
of the submerged portion of a tidewater front is typically too small to contribute to the mass balance equation. The ablation area of the tidewater glacier will therefore be smaller and thus less effective than before, and once a marine ice terminus begins to advance it will try to continue its advance. Glaciers that advance into fjords in response to a drop in the ELA normally only reach standstill positions at the fjord mouth or at points of pronounced increase in fjord width (Fig. 5-5: Lliboutry, 1965), where the ice velocity is reduced and iceberg calving becomes more effective. If the fiord walls narrow, or if two tributary glaciers merge to occupy a single channel, the advance will be especially rapid (Fig. 5-5). Glaciers may advance at rates varying from tens of metres per year to 12 km a-', the surge rate measured on the Negri Glacier, Spitsbergen (Liestol, 1969). As a glacier advances into ever-increasing water depths, the calving rate will also increase (Brown et al. 1982; Pelto and Warren 1991). The ice terminus may eventually float as an ice shelf, given sufficiently deep water, ice velocity and structural support. Calving of icebergs, whether of a floating or grounded ice terminus, is the most effective way of limiting ice advance and ultimately initiating retreat of a marine ice terminus. Ice retreat may not necessarily alter the size of the ablation area, and once a marine ice terminus begins to retreat it will continue in that direction contrary to minor changes in the ELA (Mercer, 1961). If the fjord widens towards the head, the retreat may be rapid; between narrows the glacier terminus often maintains a tenuous stability. Retreat rates can vary from 0.01 km a-l (Spitsbergen: Lavrushin, 1968) to 5 km a-l recorded in Glacier Bay, Alaska (Powell, 1991). Emergence of the land is often associated with ice retreat. In Glacier Bay the maximum emergence rate is >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 (<1km).
Waveprocesses Wave-dominated fjords are rarely more than 100 m deep, the exception being the deeper basins of Newfoundland's fjords. Although most fjord coasts have a few examples, the fjord coasts of Nova Scotia, Newfoundland and Labrador are
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS
147
predominantly of this type (Piper et al., 1983). Wave-dominated fjords receive only 5 to 25% of their Holocene basin fill from rivers: almost all the remainder is derived from waves reworking older marine sediment or glacial till along the basin margins. Sediment accumulation rates range from 0.5 to 3 mm a-l (Piper et al., 1983; Piper and Keen, 1976). The highest sedimentation rates occur in deep basins adjacent to shore areas that receive the greatest wave intensity. The larger waves arrive as open-ocean swells or storm waves and in the case of the deeper Newfoundland fjords 10 m wave heights have been reported. Cliff retreat rates in some areas can exceed 1 m a-l near the fjord mouths decreasing to 25 cm a-l in exposed inner fjord areas (Piper et al., 1983). Severe storms are capable of erosion and resuspension of sediment found in water depths of many tens of metres, and of up to 100 m in Newfoundland bays (Slatt, 1974; Stehman, 1976). After a storm, suspended sediment concentrations remain high in a near bottom zone (Barrie and Piper, 1982; Pickrill, 1987). Fine sand and silt settle out rapidly, leading to the deposition of thinly-banded graded layers on the basin floors. The internal occurrence of sandy layers suggests that severe storms can recur at time intervals of the order of 20 to 50 years. Between these major storms, the basin margins accumulate a reservoir of sediment during less severe weather periods (Figs. 5-13B and 5-14C; Piper et al., 1983). Sediment texture in wave-influenced fjords correlates strongly with water depth; nearshore zones of sand and gravel pass into an offshore zone of lag sediment, till or bedrock (Deegan et al., 1973). The rate of downward fining depends on local bathymetry, fetch direction (coastal exposure) and the wave size (Exon, 1972). Commonly, wave resuspension results in onlapping basin-fill units (Fig. 5-14C). Severe storms may also generate strong currents capable of transporting sand (Exon, 1972). Along wave-dominated continental shelves, sands may be transported landward into the approaches to fjords and possibly into fjord basins themselves (Piper and Keen, 1976). An instructive example comes from Holyrood Pond, on the south coast of Newfoundland (Forbes, 1984; Shaw et al., 1990). A mixed sand and gravel barrier perched on the fjord sill is overwashed during storms, and sediment cascades into the fjord basin. The large sediment supply from eroding glacial bluffs ensures that the barrier remains stationary, effectively trapped on the sill.
FJORDS DOMINATED BY SLOPE FAILURE
Fjords are ideal environments for nearly every form of submarine slide and sediment gravity flow. Fjord sediments have a wide range of grain size and size-sorting. They often overlie very steep slopes where rates of sedimentation may far exceed rates of consolidation: fjord sediments may be more than 90% underconsolidated (Sangrey et al., 1979). These underconsolidated sediments may fail under their own weight or because of outside stimuli such as earthquakes. Four factors control slope failures in fjords: (1) topography; (2) supply of material; (3) physical properties of the sediment; and (4) the releasing mechanism.
148
J.P.M. SYVITSKI AND J. SHAW
Release mechanisms Release mechanisms are those natural or anthropogenic events that change one or more of the parameters that affect the strength of a sediment mass beyond some critical limit (i.e., cohesion, overburden stress, excess pore pressure, angle of internal friction). Release mechanisms include sediment loading, earthquakes, waves, sea-level fluctuations, changes in mass sediment properties, and human activities (cf. Kostaschuk and McCann, 1989). Loading of a sediment mass may result from: (a) long-term high rates of sedimentation, although usually combined with an additional trigger event; (b) short-term heavy sediment supply to delta areas, including exceptional storm-floods and jakulhlaups (Fig. 5-15A); (c) delta foresets advancing onto underconsolidated prodelta clays (Fig. 5-15A); (d) oversteepening of a depositional slope up to the critical angle of failure; (e) an advancing tidewater glacier; and ( f ) the impact of sea ice or an iceberg onto sloping seafloor environments. With the partial exception of (e) and ( f ) , these mechanisms are mostly responsible for failure in the areas offshore of river deltas. Side-entry river deltas tend to comprise relatively coarse-grained deposits which have accumulated on steep slopes (Fig. 5-15B). These deltas may advance through sediment failure on slopes near the angle of repose of the sediment mass. In undrained deposits a critical pore pressure may be developed due to earthquake shocks, high sedimentation rates, loss of buoyancy during an extreme low tide or the sudden change of hydrostatic pressure by wave action, and development of free gas within the sediment (Schwarz, 1982). On deltas, the overpressuring of the pore fluid permits the formation of very gently dipping slip planes (Mandl and Crans, 1981). Sudden pore-pressure disequilibrium within a sediment mass can also result in liquefaction, i.e. the sudden loss of sediment strength associated with the upward movement of pore fluid, temporarily allowing sediment particles to flow. During the 1964 Alaska earthquake, liquefaction-type slides incorporated sand layers exceeding 50 m in thickness (Andresen and Bjerrum, 1967). Earthquakes are also responsible for two types of giant waves in fjords: (1) tsunamis and (2) violent seiches. Both may result in subaerial and submarine slides (Murty, 1979). Tsunami waves are a result of seabed adjustment under the oceanic surface whereby seawater volume is displaced. The destructive effect of a tsunami will depend on: (1) the shoreline proximity and “line of sight” since tsunamis are highly directional; (2) the state of the tide (more destructive at high tide); ( 3 ) natural oscillations (i.e., the matching of the natural period of the fjord basin with the tsunami’s period); and (4) shoaling. If the fjord funnels the waves, the wave height may grow. The 1964 Alaska tsunamis, after travelling 1700 km, underwent wave amplification in Alberni Inlet, B.C., where the wave height grew to 7 m at the fjord head compared with 0.6 m at the fjord mouth (Thomson, 1981). Giant bow waves have also been produced in fjords after the sudden dislodgement of subaerial debris into the basin water. In 1958, 30 x lo6 m3 of rock plunged into Gilbert Inlet, Alaska, producing a giant wave whose waters reached 543 m above sea level and travelled down-fjord at between 156 and 209 km h-’ (Miller, 1960) - this
149
SEDIMENTOLOGY AND GEOMORPHOLOGY O F FJORDS
B
I
1500
1200
900
600
300
0
DISTANCE (metres) v.e. 24.9~
Fig. 5-15. (A) Interpretation of a very-high resolution seismic profile of the prodelta region of Itirbilung Fiord, showing the development of a complex slide geometry. The inset figure shows the largely terraced nature of the delta front a result of prograding sandy foresets over prodelta muds (after Syvitski and Hein, 1991). The labelled ponded slide mass is presumed to have resulting from a j0kulhlaup event initiated circa 1890 (after Syvitski and Hein, 1991). (B) Huntec seismic record of a sidewall debris flow deposit from Itirbilung Fiord (after Syvitski and Hein, 1991).
event is cited in the Guinness Book of World Records as the largest recorded wave! The impact of waves on slope stability results from the pressure changes on the seafloor associated with their passage. Waves impose an oscillatory motion on the seabed which, on a sloping area, leads to a mass transfer of sediment downslope. The collapse of bar-front sediment on the Adventfjord delta, Spitsbergen, is thought to result from wave-induced cyclic loading (Prior et al., 1981b). Giant waves, including those generated by submarine landslides, may in turn generate new subaqueous slope failures (Coulter and Migliaccio, 1966).
150
J.P.M. SYVITSKI AND J. SHAW
The frequency of slope failure is controlled by the rate of sediment deposition and the frequency of the releasing mechanism (Schwarz, 1982). For instance, occurrence of a slide frequently coincides with a period of normally low stability such as with a low tide (Terzaghi, 1956; Bjerrum, 1971). Earthquake-triggered slides may have a return frequency of the order of hundreds of years. Older deposits exposed by a recent slide are more highly consolidated than the removed sequences and, consequently, have a higher resistance to shear. Under similar stress conditions no further slope failures will occur until a new sedimentary cover of critical thickness develops (Schwarz, 1982).
Mass transport processes All forms of mass transport processes are believed to operate in fjords. These include mass transport with an elastic mechanical behaviour (rockfalls, slides, slumps, creep), transport with plastic deformation (debris and mud flows, and some grain flows), and viscous sediment flows (grain flows, fluidized and liquefied flows, turbidity currents). Prime sites of submarine failure in fjords include: (1) where sedimentation occurs over high slope “basement” topography (sidewall slopes, sills); (2) where tributary hanging valleys meet the central channel (cf. Aarseth et al., 1989); (3) areas of active tectonic faults; (4) near areas of high sedimentation rates (such as deltas) (Fig. 5-15A); and (5) fjord walls subject to rockfalls and avalanches (Fig. 5-15B). Gravitational deformation of cohesive sediment always commences with macroscopically perceptible creep movement. Creep is defined as continuous yielding of the soil particles under applied stress without brittle failure. Creep of stratified sediments causes contemporaneous deformation in the form of gentle folding: an upslope area of decreased sediment thickness is balanced by an increase of thickness in a downslope area. This increase in strain may eventually lead to elastic behaviour in the form of feather joints. Antithetic precursory faults grow, followed by the development of a main shear face, often with reorientation of secondary shear faces (Schwarz, 1982). The main shear face or glide path usually develops as a listric (curvilinear) fault. The main block slides downslope resulting in the formation of a tensional depression downslope of the scar zone, and a slide toe at the base of the slide where various fold forms and overthrust fault systems may result (Fig. 5-15B). The slide block may partially collapse into a convolute slide or slump, where internal stratification is chaotic (Fig. 5-15B). Since fjord slopes are relatively steep, acceleration of the slide mass usually results in at least partial collapse of the block system. Slide volumes in fjords range from less than 106 to lo9 m3 of sediment material. Slide velocities, usually calculated from the rate of rupture of underwater cables along the fjord length, range from 0.4 m s-l to 6.9 m s-’. Runout distances are wide ranging but may exceed 60 km (Syvitski and Farrow, 1989). Viscous flows tend to develop when the initial slide involves large quantities of saturated loosely deposited fine sand and silt. At the point of failure particles may begin to rearrange themselves into a viscous liquid and flow downslope. If the face of the slide scar is left unsupported it may in turn fail, and the slide will develop retrogressively slice by slice, widening the scar with each slice. Thus the dimensions
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS
151
Fig. 5-16. Schematic diagram of the 1975 and earlier slides off the Kitimat delta, British Columbia showing the diverse surface morphology (based on high resolution sidescan data; after Prior et al., 1983).
of the final slide may be disproportionately large compared with the size of the initial slide (Bjerrum, 1971). The retrogressive motion can occur at rates of fifty to several thousand metres per hour (Andresen and Bjerrum, 1967). The sediment flow will not cease until the slide scar reaches a more dense sand deposit, or another material type, or if the rate of sediment supply to the viscous flow is greater than can be transported away. The slide complex off the Kitimat Delta, British Columbia, involved more than lo8 m3 of material, caused a 26 m deepening of the seafloor at the slide head and resulted in downslope deposition on the fjord floor of as much as 30 m (Luternauer and Swan, 1978; Prior et al., 1982a, b; Bornhold, 1983; Fig. 5-16). A complex slide morphology resulted from a series of closely spaced events (i.e., 1971, 1974 and 1975).The slide geometry consists of three main segments: (1) a short and steep head on the delta front (4"-7"); (2) an intermediate blocky zone with 1"-to 2"-slopes; and (3) an elongate depositional area more than 3.2 km long with slopes tl"and marked by a steep toe. Side-entry deltas are sites of intermittent slope failures where the slide frequency is related to the development of oversteepened slopes (i.e., where the angle of internal friction is less than or equal to the slope of the delta front). The development of oversteepened slopes, in turn, can be linked to the slow or delayed switching of a subaerial distributary channel which allows a locally oversteepened slope to build. The 1955 slide off a side-entry fan delta at Woodfibre, Howe Sound, B.C., became a
152
J.P.M. SYVITSKI AND J. SHAW
classic example of slope instability involving coarse sand and gravel (Terzaghi, 1956; Prior et al., 1981a; Mandl and Crans, 1981). The set-up conditions included; (1) delta front slopes with very narrow limits (26“30’ to 28”); (2) increased silt content on the lateral margins of the delta front as a result of oversteepening of part of the delta lip; and (3) an extremely low tide. Lowered permeability associated with the siltation interfered with the rising and lowering of the water table and allowed excess pore pressure to be built up. The slide generated large-scale scarps, disturbed hummocky blocks of sediments and numerous discrete erosional chutes and channels. Small step-like terraces arranged roughly parallel to the bathymetric contours were also present (e.g., Fig. 5-15A). The channels were 3 to 5 m deep and 10 to 30 m wide, with levee-like features. Farther downslope the entire seafloor is composed of highly irregular and chaotically arranged blocks of sediment with a hummocky surface profile. Slides or slumps may develop into debris (units of poorly-sorted pebbly mud) or viscous sediment-gravity flows. If the flows become supercritical, they may progress further into turbulent turbidity currents. Acoustic evidence of this succession commonly shows slumps, slides and debris flow deposits, with channels eroded into their surfaces as the final stages of flow became turbulent (Syvitski et al., 1987). Lithological details of these deposits can be found in Syvitski et al. (1987), Syvitski and Farrow (1989), and Hein and Syvitski (1992).
DEEP-WATER RENEWALS AND ANOXIC FJORDS
One of the most discussed aspects of silled fjords is the periodic flushing or renewal of their basin waters with adjacent coastal or shelf waters (Gade and Edwards, 1980; Farmer and Freeland, 1983; Gade, 1970; Edwards and Edelsten, 1977). If the adjacent shelf water outside the sill is denser than the resident water of the fjord basin, and it can be “lifted” over the sill, then a density current will develop and attempt to replace the basin water (Fig. 5-17A). Density currents are seldom continuous but behave intermittently on a variety of time scales depending on circumstances.
Deep-water renewal Renewal events may be triggered by tidal motion, weather systems (land/sea breezes; atmospheric pressure), and seasonal events (monsoonal winds, runoff variations). Having entered the fjord, the heavier renewing water will sink towards the basin bottom as a turbulent density current or plume (Fig. 5-17). The downward acceleration can be related to the force of gravity but is modified by the density contrast, a pressure gradient due to the changing plume thickness, bottom drag and entrainment characteristics of the flow (Gade and Edwards, 1980). The plume density may be affected by any increase in density as a result of resuspended sediment particles and the entrainment of less dense resident basin water (Edwards et al., 1980).
153
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS
/
4ot DISSOLVED OXYGEN
AT 400m
NORDFJORD
I DEEPWATER RENEWAL
YEAR 1962
64
65
D
34
RUSSELL FJORD
Fig. 5-17. (A) Simple schematic of a complete fjord deepwater renewal (after Gade and Edwards, 1980). (B) Schematic of the low frequency cycling of renewal in Nordfjord, Norway, based on the dissolved oxygen content at 400 m (after Saelen, 1967). (C) Variation in salinity of different depths within Bonnefjord, Norway, showing both complete and partial renewal events (after Gade and Edwards, 1980). (D) Conceptualization showing dependence of depth of renewal on season in Russell Fjord (after Munech and Heggie, 1978).
Not all renewals result in complete replacement of the water masses within the fjord (Fig. 5-17C, D). The renewal may be considered partial under three conditions (Gade and Edwards, 1980): (1) the intruding shelf water, not sufficiently dense to replace the deepest water in the fjord, will sink to some more appropriate level, spread out, and begin to fill up the available space from that level up; (2) the intruding plume being thin may use up its available energy (for instance overcoming bottom friction) and soon come to rest without stirring up much of the resident basin water; (3) although all density conditions for complete replacement are met, the renewal event may be of such short duration that the fjord basin water is not completely overturned. In addition, those fjords with multiple basins tend not to have deep water renewal as a single process but as a series of events which may or may not affect all the basins within the fjord (e.g., Loch Etive: Edwards and Edelsten, 1977). The “source” water may use up much of its initial energy in deep water renewal of the outermost basin as a consequence of interlayer friction and dilution. Renewals that have a yearly return period are commonly related to the density structure offshore. The dissolved oxygen content of the basin water of Nordfjord, Norway shows a nine year period for complete renewal (Fig. 5-17B; Saelen, 1967). Biological activity over time will deplete the oxygen from the water until the next renewal event replenishes the fjord with oxygenated shelf water. Bonnefjord, Norway had a complete renewal in early 1963 and subsequently two partial renewals: the first at 40 m in 1964 and a deeper renewal (60 m) in 1965 (Fig. 5-17C; Gade and Edwards, 1980). Russell Fjord, Alaska, had a partial renewal to intermediate depths during the summer, a shallower renewal during the fall and a deep renewal during the spring (Fig. 5-17D; Muench and Heggie, 1978).
154
J.P.M. SYVITSKI AND J. SHAW
Renewal and sedimentation Little is known about the effect of deep-water renewals on fjord sedimentation. It is known that deep basin surficial sediments just landward of a sill are swept clean of silts and clays (Gade and Edwards, 1980; Tunnicliffe and Syvitski, 1983). In Loch Eil, Scotland, Edwards et al. (1980) found that little sedimentation occurred during neap tides. However during spring tides, a time of known deep-water renewals, local sedimentation increased by an order of magnitude. They considered that the increase in sedimentation was due to local resuspension during the renewal event, and to the sediment load carried by the renewal water. If the renewal occurs during the summer when the fjord waters are more likely to be turbid compared to adjacent waters, then a renewal event may result in the removal of a significant quantity of sediment from the fjord system. The opposite is considered true for the winter, when a deep-water renewal event may import resuspended storm sediment from the adjacent shelf at a time of low surface layer turbidity levels within the fjord (Winters and Syvitski, 1992).
Anoxia Fjord types span a wide spectrum from very high to very low carbon input and various modeling techniques have been applied to quantify the temporal and spatial distribution and flux of carbon (energy) within fjords (Farrow et al., 1983; Syvitski et al., 1990). The primary redox discontinuity is within the sediments but may occur in the water column, within or below the zone of photosynthetic activity (Skei, 1983). Migration of the redox zone, solubilization and precipitation of minor constituents and trace metals, cycling of carbon, nitrogen, trace metals and other mobile species, are processes associated with anoxic fjord environments. The oxygen regime in fjords depends on the periodicity of water exchange and the supply and rate of utilization of organic matter. In fjords with deep sills and frequent water exchange, oxygen saturation generally occurs above the pycnocline (sometimes in supersaturation due to primary production) and slight undersaturation at depth, due to organic matter breakdown and sluggish water movements. The oxygen content
Fig. 5-18. Depth profiles of idealized soluble transition ( T ) and type B cationic species concentrations in a fjord basin water column in the vicinity of the redox front (from Jacobs and Emerson, 1982).
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS
155
in the upper waters is also dependent on the extent of turbulence and mixing between river water and seawater. The residence time of the water is a critical factor with respect to oxygen consumption. In fjords with shallow sills and longer residence time, the basin water may turn anoxic. This will allow the reduction of nitrate and sulphate, and the creation of poisonous hydrogen sulfide (Fig. 5-18) (Richards, 1965). The levels of sulfide accumulated in the bottom water depend on the residence time of the water and the rate of sulphate reduction. Occasionally, vertically displaced anoxic water reaches the surface of the fjord, creating a massive fish kill (Hellefjord, Norway; Brogersma-Sanders, 1957). Such episodes have a definite impact on the bottom sediments of fjords with excess input of organic matter. Benthic biota are unable to survive prolonged periods of low oxygen concentrations, and there have been numerous reports of extinction of faunas within fjord basins due to the spread of anoxic conditions (e.g., Kitching et al., 1976; Rosenberg, 1980).
LONG-TERM DEPOSITIONAL TRENDS
Fjords are a dynamic coastal environments whose succession from formation to sediment infill can be related to the local glacial and paraglacial history of erosion and deposition (Forbes and Syvitski, 1995), superimposed on a well-developed preglacial geological setting. Much of the sediment accumulation within fjords relates to glacial/ proglacial sediment processes operating during and after the last major ice advance (Aarseth et al., 1975; Gilbert, 1985; Powell, 1991) or reflects several episodes of proglacial basin sedimentation. Basin deposits may include: (1) ice-contact deposits (lodgement or waterlain till, push and dump moraines); (2) proximal glacimarine sediments dominated by alternations of gravity flow deposits and hemipelagic layers; (3) distal glacimarine sediments that tend to be fine-grained and bioturbated; and (4) widely varying nonglacier influenced sediments that depend on local supply and energy conditions. Stages of fjord injilling
Five stages of fjord infilling have been recognized (Table 5-2) (Syvitski et al., 1987): (1) fjords are completely or nearly filled with glacier ice; (2) fjord sedimentation is influenced by the presence of one or more tidewater (or floating) glacier termini; (3) glaciers are solely land-based but still contribute substantial glacigenic sediment and meltwater to the fjord; (4) land-based glaciers are gone and fjord sedimentation is largely in response to precipitation events eroding older sedimentary deposits, with sediment redistribution by waves, tides and gravity (Syvitski and Praeg, 1989); and (5) fjord basins have become completely filled with sediment, or at least isolated from the sea. How close a particular fjord is to being infilled, depends on its present rate of fill and its volumetric storage capacity. Since the last glaciation, a large range of fill thickness have resulted (Table 5-3). Considering the present accumulation rates within fjords (e.g. Table 5-4) and the basin depth to be infilled (Table 5-3), few
J.P.M. SYVITSKI AND J. SHAW
156 Table 5-3 Quaternary sediment fill within fjords System
Max. thick. (m)
Basin depth (m)
Reference
Nuka Bay, Alaska Port Valdez, Alaska Glacier Bay fjords, Alaska Knight Inlet, B.C. Howe Sound, B.C. Douglas Channel, B.C. Mahone Bay, N.S. St. Margaret’s Bay, N.S. Makkovik Bay, Labrador Nain Bay, Labrador Lake Melville, Labrador Baffin Island fjords (10) S ~ n d r eStr~mfjord,Gr. Kangerdlugssuag Fjord, Gr. Loch Nevis, Scotland Maurangerfjord, Norway Hardangerfjord, Norway Oslofjord, Norway Sognefjord, Norway Arnafjord, Norway Lusterfjord, Norway Kongsfjorden, Spitsbergen Van Mijenfjorden, Spitsbergen Fournier Bay, Antarctica Maxwell Bay, Antarctica
250 800 100 600 750 600 80 40 50 112 400 20-200 400 500 30 70 140 130 300 180 200 100 20 80 70
300 280 300+ 540 250 420 50 80 160 75 250 320-800 ?? 900 40 280 900 340 1300 200 650 100 100 460 470
von Huene, 1966 von Huene et al., 1967 Powell, 1980 Syvitski, unpublished Syvitski and McDonald, 1982 Bornhold, 1983 Barnes and Piper, 1978 Piper and Keen, 1976 Barrie and Piper, 1979 Piper et al., 1975 Grant, 1975 Gilbert, 1985 Larsen, 1977 J.T. Andrews, pers. commun. Boulton, 1990 Cone et al., 1963 Cone et al., 1963 Richards, 1976 Aarseth, 1980 Aarseth, 1980 Aarseth, 1980 Elverhi et al., 1983 Elverhi et al., 1983 Griffith and Anderson, 1989 Griffith and Anderson, 1989
fjords will remain after the next million years, assuming no reglaciation occurs. Many of the New Zealand fjords have already filled with sediment and are now fronted by narrow outwash plains (Pickrill et al., 1981). The Mecatina coast of southern Quebec is fronted by a series of infilled fjords and overlapping deltas (Hein et al., 1993). In St. George’s Bay, southwest Newfoundland, a deep glacial trough has been completely filled by glacimarine and paraglacial sediments, late Holocene barrier platform deposits, and the prograded gravel beach ridges of an active barrier (Shaw and Forbes, 1992). The process of filling may be accelerated by man-made discharges of solid wastes, e.g., Jossingfjord, Norway, which was filled up to sill depth (60 m) from an initial basin depth of 96 m over 25 years of titanium mining.
Relative sea-level fluctuations During the growth of ice caps, water is removed from the World’s ocean and sea level falls, a process known as eustasy. Eustasy includes the effects of ocean-water volume changes, ocean-basin volume changes, and ocean-level distribution changes
157
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS Table 5-4 Accumulation rates within fjords System
Accumulation (mm a-l)
Reference
1. Northeast Alaska fjords (3) 2. Muir Inlet, Glacier Bay, Alaska 3. Boca de Quadra and Smeaton Bay, Alaska 4. Bute Inlet, B.C. 5. Douglas Channel, B.C. 6. Saguenay Fjord, Quebec 7. Bedford Basin, N.S. 8. Makkovik Bay, Labrador 9. Sondre Stromfjord, Greenland 10. Coronation Fiord, Baf. Is. 11. Spitsbergen Fjords (3) 12. Oslo Fjord, Norway 13. Korsfjorden, Norway 14. Norwegian Fjords (6) 15. Lock Striven, Scotland 16. New Zealand Fjords (3) 17. Cambridge, McBeth Fjord, Baf. Is.
450 to 2000 2000 to 9000 2.7 to 4.6 t300 0.2 to 2 0.4 to 23 3 0.7 to 3 25 to 35 39 0.1 to 100 <0.6 to 2.6 4 0.5 to 10 5 0.8 to 4
Molnia, 1983; von Heune, 1966 Powell, 1990 Sugai and Burrell, 1985 Farrow et al. 1983 MacDonald, 1983 Smith and Walton, 1980 Piper et al., 1983 Barrie and Piper, 1982 Larsen, 1977 Stravers and Syvitski, 1991 Elverhi et al., 1983 Richards, 1976 Aarseth et al., 1975 Skei, 1982 Deegan et al., 1973 Glasby, 1978; Pantin, 1964 Gilbert et al., 1990; Andrews, 1990
1
(Morner, 1980). With ice-cap melting, the eustatic effect results in a general rise in sea level. The meltwater contribution of the various ice sheets was most rapid between 14,000 and 8000 years before present, and was essentially complete 4000 years ago (Tushingham and Peltier, 1991; Fairbanks, 1989). However, waxing and waning of ice caps also causes a shift of the earth’s surface load and induces a mechanical response in the lithosphere, a process known as isostasy (Peltier and Andrews, 1976). The resultant changes in relative sea level due to these factors can be modelled. For example, the sea-level equation used by Lambeck (1991) comprised terms for: (1) eustatic sea level; (2) loading and unloading of the crust due to ice sheets; (3) loading and unloading of the crust due to meltwater added or removed from the oceans; and (4) gravitational attraction of ice and water. Considering these factors therefore, it is clear that relative sea-level fluctuations in a fjord will depend on the distance a fjord basin is from both the maximum extent of an ice sheet and the centre of the ice sheet, and the speed and timing of local deglaciation relative to general deglaciation of the ice sheet. There are also complications that relate to the growth of local hinterland ice caps. Generally, loading and unloading of the crust due to ice sheets has been a particularly important factor in fjords, due to their close proximity to former ice sheets. As a consequence the relative sea-level signatures of fjord regions are predominantly of zone 1 type (glaciated areas) in Clark’s classification (1980), so that many fjords have experienced emergence at exponentially declining rates following deglaciation. These are mainly fjords which became deglaciated late: they show land emergence following
158
J.P.M. SYVITSKI AND J. SHAW
attainment of the marine limit (Andrews et al., 1970; England, 1983), i.e. global sea level returned close to modern levels before the local land relaxed from the ice load. However, other fjords lie in the transition between zone 1 (glaciated areas) and zone 2 (the ice margin) in Clark's (1980) classification, and experienced initial emergence followed by later submergence. These fjords deglaciated early and experienced isostatic relaxation before a substantial amount of the ice sheets had melted, and thus before global sea levels had returned to present-day levels. Thus, a relative sea-level fall after the highstand was followed by a second period of marine transgression (Syvitski, 1992). Obviously, patterns of relative sea-level fluctuations between these extremes can occur. Furthermore, relative sea-level fluctuations related to ice sheet readvance (Goldthwait et al., 1966), or the relaxation of the peripheral bulge found seaward of the isostatically subsiding ice sheets (Quinlan and Beaumont, 1981), can further complicate relative sea-level patterns. Considering the advance phase of ice sheets, if the fjord is some distance from the ice sheet centre relative sea level initially falls because of eustasy. Such fjord basins may undergo initial isolation from the sea and possibly develop as proglacial lakes. The period of occupation by glacial ice, if it happens at all, would be short (at most a few thousand years). Eventually, however, relative sea level would rise due to crustal depression caused by the advancing ice sheet. For fjords closer to the ice centre, relative sea level rises initially as the crust is depressed to an even further extent under the newly formed ice load. Such proximal fjords could be covered by an ice sheet for up to tens of thousands of years, and sea level would thus adjust its position against the terminus of the glacier. Fjords proximal to the ice sheet centre would also typically show more glacial erosion. Turning to the retreat phase of ice sheets, if the local fjord-ice retreats earlier than the general ice sheet margin, or faster (i.e. >5 km a-' ), then because crustal response to glacier retreat is slow (<40 mm a-* ), the sea transgresses up the fjord as the ice retreats, for distances as great as 60 to 100 km up-valley. At the sea level maximum (i.e. marine limits up to 250 m above the present sea surface), sediment is deposited either conformably or disconformably on the glacial-proglacial sediments as shallow-water marine deposits.
Relative sea-level fall and its effects Following ice retreat, in most cases relative sea level falls rapidly due to crustal rebound. On the fjord coast of British Colombia, where the marine limit locally exceeds 200 m, most uplift occurred within 2000 years (Clague, 1983). In this and other settings deltaic and glaciomarine sediments deposited close to the ice front are elevated by subsequent rapid rebound. Matthews (1981) notes that fjords that undergo a rapid clearing of glacial ice will have continually changing circulation patterns. Land emergence, for instance, will affect sill depth and thus tidal velocities. The tidal velocities have increased by 50% in Glacier Bay, Alaska, since its deglaciation (Matthews, 1981). Another type of impact is the erosion of proglacial and earlier Holocene sediments as they emerge above sea level or come above the critical depth for reworking (Table 5-2 of Forbes and Syvitski, 1995). Piper et al. (1983) identified a class of
159
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS
--
---
-1
-STRAWBERRY ISLAND?
LOWER LI.MIT OF EROSION 10,000YR. 6.P PRESENT SEA LEVEL
PRESENT LOWER LIMIT OF EROSION CONFORMABLE COVER UNIT
/
h/
UPPER BASIN-FILL UNIT
Q ,
Fig. 5-19. Longitudinal profile of Makkovik Bay, Labrador, showing the facies development as a result of sea level lowering and increase in the effect of seafloor wave erosion (Barrie and Piper, 1982).
wave-dominated fjords on the Labrador coast that are presently emerging, that is, undergoing a marine regression. During the period of emergence more and more of the seafloor has been brought into the domain of marine erosion. (Fig. 5-19; Barrie and Piper, 1982). Three facies may be developed during this type of sedimentation history: (1) a lower basin-fill unit thought to be composed of coarse proglacial sediment; (2) a conformable cover unit deposited under high rates of hemipelagic sedimentation (ice-proximal); and (3) an upper basin-fill unit which results from increasing wave erosion concomitant with the land emergence. A common scenario involves a decrease in marine influence as a result of the relative lowering of sea level. As sea level drops, shallow sills will become exposed coincident with a change from marine to brackish to freshwater conditions (Fig. 5-20B). If the input of fresh water is large the trapped marine water will eventually become diluted and washed out of the basin. For example, the Hamilton Inlet sill that partially encloses Lake Melville was 80 m deep 7000 (radiocarbon) years ago, 50 m deep 5000 years ago, and is only 28 m deep at present. During this period the salinity of the fjord bottom has been reduced by 5%0 (Vilks and Mudie, 1983). The reduction in marine influence is carried to its extreme when basins become completely isolated from the sea. This sequence of events is reflected in biostratigraphic evidence from cores in coastal lakes (Hafsten, 1983): marine diatoms and dinoflagellate cysts (marine stage) are succeeded by brackish-water diatom assemblages and ultimately by freshwater diatoms and chlorophycae (freshwater lake stage). Radiocarbon dates from lakes with varying sill depths can be used to constrain relative sea-level curves (Hafsten, 1983). The magnificent freshwater fjord of Western Brook Pond, in western Newfoundland, a World Heritage Site (Berger et al., 1992), is a case in point. The fjord became isolated from the ocean by an emerged bedrock sill (Grant, 1989; Berger et al., 1992) which is now mantled with fresh-water peat.
160
J.P.M. SYVITSKI AND J. SHAW
aICI?,"Y4?NWPH PALE BROWN SILTY MUD
B
I(3)
TRONDHEIMSFJORD
..
.. .
. MARlNE
Fig. 5-20. (A) Schematic of a sediment core from the Bedford Basin, Nova Scotia (after Miller et al., 1982). High resolution biostratigraphic analysis has revealed a fjord influenced by a fluctuating sea level. (B) The effect of falling sea level as a result of isostatic rebound on the development of sediment at the head of Trondheimsfjord, Norway (after Kjemperud, 1981).
A meromictic lake, where the hypolimnion is occupied with old anoxic seawater, may form if the water depth preserved is deep and/or the drainage basin is small with limited fresh water supply. The bottom saline waters may become trapped for several thousand years (Northcote et al., 1964; Williams et al., 1961; Strom, 1957, 1961; Boyum, 1973; Bremmeng, 1974). During the final stage in the life of a fjord, a typical scenario might consist of (1) sediment filling by delta progradation, hemipelagic rain and sediment gravity flows, until the basin depth and sill depth are equal; (2) the development of salt-wedge or partially mixed estuarine hydraulics (depending on river and tidal conditions); ( 3 ) a dramatic increase in the supply of sediment to the once starved continental
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS
161
shelf; (4) depending on local wave conditions, the formation of a barrier that may partially enclose the estuary and develop a salt-marsh lagoon; (5) with further fluvial deposition, complete filling of the lagoon resulting in the formation of an open coastal delta; and (6) with the increased shelf sedimentation, bathymetric irregularities on the shelf would be smoothed out. A similar scenario was proposed for the Vancouver Island shelf/fjord complex by Carter (1973). Fjords may become isolated from the ocean by additional mechanisms, including: (1) cut-off by a prograding delta (Clague et al., 1983); (2) cut-off by a prograding barrier spit (Pickrill et al., (1981); (3) cut-off by a side-entry glacier (Gilbert et al., 1985); and (4) cut-off of a tributary fjord by a main trunk glacier (Goldthwait et al., 1966).
Relative sea-level rise Postglacial submergence in fjord regions typically occurs long after ice sheets melt, by which time the rate of crustal rebound has declined, unlike eustatic rise. Eustatic sea-level rise due to melting of the global ice sheets began to decelerate rapidly at ca. 8000 BP, and was largely completed only as late as ca. 4000 BP (fig. 5.12 in Tushingham and Peltier, 1991). In some regions submergence after that time can be attributed to continuing forebulge migration (Quinlan and Beaumont, 1981), as on the Atlantic coast of Nova Scotia. There, in a region of wave and tidedominated fjord inlets, the continuing Holocene transgression - with contemporary submergence rates of 0.36 m per century (Shaw and Forbes, 1990a) - has resulted in migration of the coastline and an associated erosional front through fields of drumlins (Orford et al., 1991). Sediment released by drumlin erosion has been redistributed, so that inlet infilling has kept pace with sea-level rise, even producing regressional signals within estuarine sediment sequences (Carter et al., 1989). The wave-cut platforms formed from erosion of till cliffs eventually become submerged by rising sea level (Barnes and Piper, 1978; Forbes and Syvitski, in press). The platforms undergo winnowing and by-passing, often forming gravel pavements (Stanley, 1968). Impacts of complex relative sea-level signatures Many fjords are in the transition zone between Clark's zones 1 and 2, and have experienced phases of both regression and transgression, largely as a result of the interplay between local isostatic effects and global eustatic change. On the fjord coast of northwest Scotland relative sea level dropped from a maximum of 60 to 80 m above the present level to slightly above it at 9000 BF', and subsequently rose to -+ 10 m at 6000 BP, before declining once more (Lambeck, 1991).The fjord coast of British Colombia emerged rapidly until ca. 10,000 BP, then submerged following a -12 m lowstand (Clague et al., 1982). The transgression triggered aggradation of the Fraser River floodplain (Clague, 1983). Bedford Basin, Nova Scotia, is a fjord which fluctuated from initial estuarine circulation (8000 radiocarbon year BP) to the present-day fjord condition (Fig. 5-20A, Miller et al., 1982). During a time when global sea level was falling as a result of a major ice-sheet advance, the coastal estuary, initially dominated by cold brackish water, turned into a freshwater lake. With the retreat of the ice sheet, relative sea
162
J.P.M. SYVITSKI AND J. SHAW
level rose locally and a deep salt-wedge estuary was produced. The trend continued until a deep-silled fjord formed. The transition from regression to transgression in fjord regions often results in a lowstand of relative sea level which can find morphological expression. Along the coast of southern Newfoundland a majority of fjords have shallow, submerged terraces at their heads. The comments of Flint (1940) regarding the origin of these have been proven correct by recent investigations: during the early Holocene relative sea level dropped to between 14 and 30 m below the present datum. Downcutting streams became incised into raised Late Wisconsinian ice-marginal deltas at fjord heads, forming Gilbert-style deltas graded to lowstand levels. They appear on seismic reflections records as prograded packages of clinoform reflections overlying glacial-marine stratified sediments (Shaw and Forbes, 1990b, 1992). The deltas were submerged during the subsequent late Holocene transgression. Turning to a region with a greater sediment supply, Thors and Boulton (1991) described events in Eyjafjordur, on the north coast of Iceland, where a large delta, up to 70 m thick, with well-developed foreset beds on boomer records, prograded into the fjord basin during a -40 m lowstand between 12,000 and 10,000 BP; continued sediment input during the later transgression formed deltas and spits at shallower depths. In terms of the availability of sediment in the littoral zone of fjords, St. George’s Bay, in southwest Newfoundland, is an extreme example. A glacially-deepened basin landward of a morainal sill has been almost completely infilled by up to 180 m of stratified glacimarine sediments (Shaw and Forbes, 1990b, 1992). Glacimarine muds deposited during and after a +44 m late-glacial highstand have been extensively winnowed in shallow areas, forming gravel pavements; fines have been transported into the deepest parts of the basin to accumulate as thick, gas-charged Holocene mud. A delta which formed by reworking of older, ice-marginal deposits during an early Holocene -25 m lowstand was subsequently enveloped by late Holocene prograded barrier-platform deposits (Shaw and Forbes, 1992; Forbes et al., 1993) released by wave-reworking of late-Wisconsinian glacial deposits exposed for 40 km along the coast. In addition to temporal variations in rates of sea-level change, fjord regions experience spatial variations in sea-level signal. On the fjord coast of southern Norway, relative sea level fell exponentially throughout postglacial time in the east, but the trend was reversed in the west, which experienced a marine transgression which culminated ca. 6000 BF! A notable example comes from the fjord coast of Labrador: Andrews (1989) showed that there had been up to -8 m of emergence along the inner coast since 2000 BF! Wave destruction of paleo-Eskimo sites suggested that the outer coast was at the limit of present day submergence.
Climate and sedimentation Since the last Pleistocene glaciation, the earth has undergone a warm climatic optimum called the Hypsithermal period: a regionally time transgressive period (13,000-8000 year BP for Alaska and 8000-5000 year BP for Baffin Island). It was a time of warm, dry summers with valley glaciers ablating rapidly. Major outwash
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS
163
deposits formed in front of the retreating ice (Church, 1978; Church and Ryder, 1972; Goldthwait et al., 1966). A cooling trend with minor oscillations that ended about 300 to 100 years ago followed this warm phase. Known as the Neoglacial, this period is marked by the alternation of cool moist conditions (and glacial advances) with warmer dry periods (growth of outwash deposits). The most recent of these ice advances, known as the Little Ice Age (circa 1700 AD), appears to be globally synchronous. Church and Ryder (1972) calculated that half of the annual sediment yield during the Neoglacial was derived from erosion of the isostatically uplifted Hypsithermal deposits. For example, in Cambridge Fiord, Baffin Island, basin-averaged (Hypsithermal) sedimentation rates between 8800 and 6700 years BP were 43 mm a-l (Stravers and Syvitski, 1991) which can be compared with rates in the postglacial Neoglacial period (6700 years BP to the present) of 1 mm a-' (Gilbert et al., 1990). This enormous drop in sedimentation rates, as effected by sea-level fluctuations, climate change and sediment supply, has been noted and modelled for other fjords (Syvitski et al., 1990; Syvitski and Hein, 1991; Andrews and Syvitski, 1994).
Numerical models Computer simulation models can be used to replicate sedimentary systems, so that insights into the character and origin of sedimentary deposits may be gained. A unified process-response numerical model attempts to simulate basinfilling through multiple transport pathways and depositional mechanisms. RIVERDELTA-GRAIN are examples of linked numerical models that, given climate data and river basin and fjord basin characteristics, simulate a sedimentary fill sequence from the growth of a prograding delta (Syvitski, 1993a, b; Syvitski and Daughney, 1992; Syvitslu and Alcott, 1993). The model simulates four mechanisms that affect the rate and style of basin-filling: (1) hemipelagic sedimentation of particles carried seaward by river plumes; (2) deltafront progradation as affected by bedload deposition at the river mouth; (3) proximal slope bypassing, primarily turbidity currents; and (4) downslope diffusive processes that work to smear previously deposited sediment into deeper water. The spatial distribution of seafloor texture is determined using: (1) a velocity distribution developed to simulate a buoyancy-dominated, free, two-dimensional jet flowing into a highly-stratified marine basin; (2) a particle-scavenging model that considers the biogeochemical affects of settling of particles within a marine environment (i.e. flocculation); and (3) contributions by sediment gravity flows. The temporal variations are affected by fluctuations in flow characteristics - the circulation factor, and changes affecting the position of the river mouth - the distance factor. Predictions of the seafloor slope, sediment texture and accumulation rate are at a seasonal (four times a year) or daily resolution. Sediment accumulation is predicted spatially from a parabolic partial differential equation that combines these four mechanisms for depositing sediment within a basin. The numerical solution employs a finite difference approximation solved by an explicit method. The models provide an opportunity to examine details of sedimenta-
SEDIMENTOLOGY AND GEOMORPHOLOGY O F FJORDS
"1)
165
A
500
4M)
300 200 100 0
"n
B
C 500
400 300 200 100
0 0 -
10
20
30
40
50
DISTANCE FROM ORIGINAL RIVER MOUTH (km)
Fig. 5-21. The Holocene record of seafloor positions (every 400 years) in Itirbilung Fiord, Baffin Island and as modelled by the DELTA5 program (cf. Syvitski and Daughney, 1992). (A) Two distinctive periods of sediment and water discharge conditions are represented (see text). (B) Same conditions as (A) but with a sea level rise in the second sedimentation period and current erosion increased by a factor of three. (C) Same conditions as (A) but the second sedimentation period segmented into two, with the latter half as warmer period providing increased glacier melt and associated sediment input.
J.P.M. SYVITSKI AND J . SHAW
166
2000
4000
6000
8OW
TIME (YR)
Fig. 5-22. Fraction of turbidites generated annually and averaged over a fifty year period (i.e. 0.7 would indicate that turbidites were generated at least once every year for 35 years of a fifty year period).
Progress has been made in the following ways: (1) We now have a better understanding of the processes of flocculation and agglomeration, particularly determination of in situ settling velocities and particle densities of suspended particles, and an understanding of the effect of “fine structure” (i.e., well mixed layers alternating with sharp interfaces) on the character of suspended matter. (2) Researchers have developed numerical models capable of predicting sedimentation rates and grain-size distributions of fjord sediments - the models are based on climate-driven (temperature, snowfall, rainfall, glacier melt) river discharge input. ( 3 ) We have seen the development of quantitative models that relate fjord basin processes to long term climatic variations and the effects of relative sea-level fluctuations. (4) Progress has been made in comprehending the dynamics of delta progradation, including detection of short-lived underflows and turbidity currents, the effect of internal waves breaking on the prodelta, and the effect of earthquake accelerations on large-scale failures. ( 5 ) A substantial amount of high-resolution seismic reflection data has been collected, together with accurately-positioned, long piston cores. These data aid construction of more rigorous bio-, seismo- and lithostratigraphic facies models of fjord deposits. (6) Progress has been made in determining the rate of sediment delivery to fjord basins by aeolian transport, sea-ice and iceberg rafting, and mountain slope colluvium. (7) We are better informed regarding the relative importance of the various styles of slope failure in the sedimentary record, and in particular the dynamics of release mechanisms and transport processes.
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS
167
(8) We appreciate more fully the effect of deep water renewals on the transport of suspended particulate matter in and out of fjords. (9) The role of ice-proximal sedimentation processes has become clearer, including the development of push moraines, ice-tunnel discharge processes, the effects of iceberg generated waves, jokulhlaup dynamics and deposits, and variations in the timing and location of ice terminus advances and retreats. This progress seems impressive, but many important questions remain to be answered. For example: (1) How does permafrost affect the development of arctic sandur? (2) What are the mechanical properties associated with rapid sedimentation of cohesive sediments? Overall, the score is quite good. Research papers cited post1987 have arguably increased our understanding of the sedimentology of fjords by at least an order of magnitude.
SUMMARY
In this chapter we have provided an overview of the range of geological issues which have been studied in fjords. Despite the good progress achieved in recent years, many research topics remain qualitative in their approach. However, we live in an era when financial and logistical support for marine geological research are becoming increasingly difficult to obtain, and demands to justify research cannot be ignored. The value of fjord research can be justified in terms of two important concerns which have rightly received much public attention - firstly the environment, or more particularly the human tendency to alter the environment in negative and deleterious ways, and secondly, the potential for global climate change. Fjord research is worthy in several respects. Fjord basins are especially susceptible to pollution on account of their special oceanographic setting. They can trap watercolumn pollutants rather than allow them to disperse onto the inner shelf; their complex configuration also favours the entrapment of oil spills, rather than their dispersal by wave action, and temperature inversions can allow atmospheric pollution to persist. Secondly, fjords are repositories of thick and complex sequences of sediments containing evidence of climatic change, in particular, climate change at the time of the Pleistocene-Holocene transition. Lastly, the high latitude glacier fjords, for example those of Greenland, are natural laboratories in which to examine processes which were active elsewhere during the last ice age; they are also sensitive indicators of changes in global climate.
ACKNOWLEDGEMENTS
We have had the fortune to have worked with or know many of the world’s fjord experts. They have directly contributed to our understanding of fjord science. This paper then reflects information we have received from John Anderson, John Andrews, Inge Aarseth, Bruno d’hglejan, Jim Bogen, Brian Bornhold, Joe Buckley, Geoffrey Boulton, Paul Carlson, Lionel Carter, Michael Church, Dave Burrell,
168
J.P.M. SYVITSKI AND J. SHAW
John Clague, Ellen Cowan, Ray Cranston, Gene Domack, Julian Dowdeswell, Anders Elverhoi, George Farrow, Don Forbes, Bob Gilbert, Kryzystof Giirlich, Al Grant, Monty Hampton, Alex Hay, Frances Hein, Hans Holtedahl, Ed Horne, Ray Kostaschuk, Al Lewis, Colin Levings, John Luternauer, Brian McCann, John Milliman, Peta Mudie, Jim Murray, George Pickard, Andrew Phillips, Dave Pickrill, David Piper, Ross Powell, David Prior, Jay Stravers, Charles Schafer, Jens Skei, Norm Smith, John Smith, Bob Taylor, Verina Tunnicliffe, Gus Vilks, and Gary Winters. Of course we have read and enjoyed the works of others, but these friends and colleagues are thanked for their support and kindness during our research. The quantity and quality of our data collected in fjords relates to the dedication and industry of many BIO technicians and staff, particularly Ken Asprey and Bill LeBlanc - thanks folks. This paper forms GSC contribution 24193.
REFERENCES Aarseth, I., Bjerkli, K., Bjorklund, R., Boe, D., Holm, J.P., Lorentzen-Styr, TJ., Myhre, L.A., Ugland, E.S. and Thiede, J., 1975. Late Quaternary sediments from Korsfjorden, West Norway. Sarsia, 58: 43-66. , and Giske~degaard,O., 1989. Submarine slides in glacirnarine sediments in some Aarseth, I., L ~ n n e0. western Norwegian fjords. Mar. Geol., 88: 1-21. Adams, J., 1980. Contemporary uplift and erosion of the Southern Alps, New Zealand. Geol. SOC.Am. Bull., Part 11, 91: 1-114. Anderson, J.B., Kennedy, D.S., Smith, M.J., and Domack, E.W., 1991. Sedimentary facies associated with Antarctica’s floating ice masses. Geol. SOC.Am. Spec. Pub., 261: 1-25. Andresen, A. and Bjerrum, L., 1967. Slides in subaqueous slopes in loose sand and silt. In: A.F. Richards (Editor), Marine Geotechnique. University Illinois Press, Urbana, pp. 221-239. Andrews, J.T., 1989. Postglacial emergence and submergence. In: R.J. Fulton (Editor), Quaternary Geology of Canada and Greenland. Geol. Surv. Can., Geology of Canada, 1, chapter 8. Andrews, J.T., 1990. Fiord to deep sea sediment transfers along the northeastern Canadian continental margin: models and data. Gtogr. Phys. Quat., 44: 55-70. Andrews, J.T. and Matsch, C.L., 1983. Glacial marine sediments and sedimentation: An annotated bibliography. Geo Abstracts No. 1, Short Run Press, Exeter. Andrews, J.T. and Syvitski, J.P.M., 1994. Sediment fluxes along high latitude glaciated continental margins: Northeast Canada and Eastern Greenland. In: Global Sedimentary Geofluxes. National Research Council and National Academy of Sciences Press, Washington, D.C., pp. 99-115. Andrews, J.T, Buckley, J.1: and England, J.H., 1970. Late-glacial chronology and glacioisostatic recovery, Home Bay, east Baffin Island. Geol. SOC.Am. Bull., 81: 1123-1148. Barnes, N.E. and Piper, D.J.W., 1978. Late Quaternary history of Mahone Bay, Nova Scotia. Can. J. Earth Sci., 15: 586-593. Barrie, C.Q. and Piper, D.J.W., 1982. Late Quaternary geology of Makkovik Bay, Labrador. Geol. Survey. Canada Pap., 81-17: 1-37. Barry, R.G., Crane, R.C., Locke, C.W., Locke 111, and Miller, G.H., 1977. The coastal environment of southern Baffin Island and northern Labrador-Ungava. Final Report, Project 136, to Imperial Oil Limited., Aquitaine Company of Canada, Ltd., and Canada-Cities Services Ltd. Institute of Arctic and Alpine Research, University of Colorado, Boulder, 165 pp. Bass and Peters, 1985. Dynamic analysis of unstable roll of icebergs. Proc. 8th Int. Conf. on Port and Ocean Engineering under Arctic Conditions, Narssarssuaq, Greenland, Sept. 7-14, 1985, pp. 966-979. Bell, L.M., 1975. Factors influencing the sedimentary environment of the Squamish River delta in southwestern British Columbia. Unpubl. M.A.Sc. thesis, Geol. Sci., University of British Columbia,
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS
169
Vancouver, 145 pp. Berger, A.R. (Co-ordinator), Bouchard, A., Brookes, LA., Grant, D.R., Hay, S.G., and Stevens, R.K., 1992. Geology, topography, and vegetation, Gros Morne National Park, Newfoundland. Geol. Sum. Can. Misc. Rep. 54, scale 1 : 150,000, Berthois, L., 1969. Contribution a 1’6tude stdimentologique du Kangerdlugssuag, cbte ouest du Gronland. Medd. Gronland, 187 pp. Bird, E.C. and Schwartz, M.L. (Editors), 1985. The World’s Coastline. Van Nostrand Reinhold Company, New York, 1071 pp. Bjerrum, L., 1971. Subaqueous slope failures in Norwegian fjords. Norweg. Geotech. Inst. Publ. 88, 8 PP. Blake Jr., W., 1977. Iceberg concentrations as an indicator of submarine moraines, eastern Queen Elizabeth Islands, District of Franklin. Geol. Surv. Canada Pap., 77-B: 281-286. Bogen, J., 1983. Morphology and sedimentology of deltas in fjord and fjord-valley lakes. Sediment. Geol. 36: 245-267. Bornhold, B.D., 1983. Sedimentation in Douglas Channel and Kitimat Fjord System. In: R.W. MacDonald (Editor), Proc. Workshop Kitimat Marine Environment. Can. Tech. Dept. of Hydrography and Ocean Science, No. 18: 88-114. Boulton, G.S., 1990. Sedimentary and sea level changes during glacial cycles and their control on glacimarine facies architecture. In: J.A. Dowdeswell and J.D. Scorse (Editors), Glaciomarine Environments: Processes and Sediments. Geol. SOC.London, Spec. Publ., 53: 15-52. Boulton, G.S. and Eyles, N., 1979. Sedimentation by valley glaciers; a model and genetic classification. In: C. Schliichter (Editor), Moraines and Varves. Proc. INQUA Symp. on Genesis and Lithology of Quaternary Deposits, 1978, pp. 11-25. Bouma, A.H., Hampton, M.A. and Orlando, R.C., 1977. Sand waves and other bedforms in lower Cook Inlet, Alaska. Mar. Geotechnol., 2: 291-308. Bouma, A.H., Hampton, M.A., Rappeport, M.L., Whitney, J.W., Teleki, P.G., Orlando, R.C. and Torreson, M.C., 1978. Movement of sand waves in lower Cook Inlet, Alaska. Offshore Technology Conf. Pap., 3311: 2271. Bowen, A.J., Normark, W.R. and Piper, D.J.W., 1984. Modelling of turbidity currents on Navy Submarine Fan, California Continental Borderland. Sedimentology, 31: 169-185. Beyurn, A., 1973. Salsvatn, a lake with old sea water. Swiss J. Hydrol., 35: 262-277. Bremmeng, G.S., 1974. Strandvatn, northern Norway, a lake with old seawater. Swiss J. Hydrol., 36: 351-356. Brown, C.S., Meier, M.F., and Post, A,, 1982. Calving speed of Alaska tidewater glaciers with applications to the Columbia Glacier, Alaska. U.S. Geol. Surv., Prof. Pap. 1258-C. Brogersma-Sanders, M., 1957. Mass mortality in the sea. In: J.W. Hedgepeth (Editor) Treatise on Marine Ecology and Paleoecology, Geol. SOC.Am. Mem., 67: 941-1010. Buckley, J.R. and Pond, S., 1976. Wind and the surface circulation of a fjord. J. Fish. Res. Board Can., 33: 2265-2271. Carbonnel, M. and Bauer, A., 1968. Exploitation des couvertures photographiques atriennes rtptttes du front des glaciers velant dans Disko Bugt et Umanak Fjord, juin-juillet 1964. Medd. Gronland Board, 173. Carlson, P.R., Powell, R.D. and Phillips, A.C., 1992. Submarine sedimentary features on a fjord delta front, Queen Inlet, Glacier Bay, Alaska. Can. J. Earth Sci., 29: 565-573. Carstens, T, 1970. Turbulent diffusion and entrainment in two-layer flow. Proc. Am. SOC.Civil. Eng. W W 97-104. Carter, L., 1973. Surficial sediments of Barkley Sound and the adjacent continental shelf, west coast Vancouver Island. Can. J. Earth Sci., 10: 441-459. Carter, R.W.G., Forbes, D.L., Jennings, S.C., Orford, J.D., Shaw, J. and Taylor, R.B., 1989. Barrier and lagoon coast evolution under differing relative sea-level regimes: examples from Ireland and Nova Scotia. Mar. Geol., 88: 221-242. Church, M., 1972. Baffin Island Sandurs: A study of arctic fluvial processes. Geol. Surv. Can. Bull., 216: 208 pp.
170
J.P.M. SYVITSKI AND J. SHAW
Church, M., 1978. Paleohydrological reconstructions form a Holocene valley fill. In: A.D. Miall (Editor), Fluvial Sedimentology. Can. SOC.Petrol. Geol. Mem., 5: 743-772. Church, M. and Ryder, R.M., 1972. Paraglacial sedimentation: a consideration of fluvial processes conditioned by glaciation. Geol. SOC.Am., 83: 3059-3072. Church, M. and Gilbert, R., 1975. Proglacial fluvial and lacustrine environments. In: A.V. Jopling and B.C. McDonald (Editors), Glaciofluvial and Glaciolacustrine Sedimentation. SEPM Spec. Publ., 23: 22-100. Clark, J.A., 1980. A numerical model of worldwide sea level changes on a viscoelastic Earth. In: N.-A. Morner (Editor), Earth Rheology, Isosytasy and Eustasy. John Wiley and Sons, London, pp. 525-534. Clague, J.J., 1983. Glacio-isostatic effects of the cordilleran ice sheet, British Colombia, Canada. In: D.E. Smith and A.G. Dawson (Editors), Shorelines and Isostasy. Academic Press, London, pp. 321-343. Clague, J.J., Luternauer, J.L. and Hebda, R.J., 1983. Sedimentary environments and post glacial history of the Fraser Delta and Lower Fraser Valley, British Columbia. Can. J. Earth Sci., 20: 1314-1326. Clague, J., Harper, J.R., Hebda, R.J., and Howes, D.E., 1982. Late Quaternary sea levels and crustal movements, coastal British Colombia. Can. J. Earth Sci., 19: 597-618. Cone, R.A., Neidell, N.S. and Kenyon, K.E., 1963. Studies of deep-water sediments with the continuous seismic profiler. The natural history of Hardangerfjord. Sarsia, 14: 61-78. Coulter, H.W. and Migliaccio, R.R., 1966. Effects of the earthquake of March 27, 1974 at Valdez, Alaska. U.S. Geol. Survey Prof. Pap. No. 542C, 35 pp. Cowan, E.A. and Powell, R.D., 1991. Ice-proximal sediment accumulation rates in a temperate glacial fjord, southeastern Alaska. In: J.B. Anderson and G.M. Ashley (Editors), Glacial Marine Sedimentation: Paleoclimatic Significance. Boulder Co., Geol. Soc. Am. Spec. Pap., 261: 61-73. Cullingford, R.A., 1979. Late glacial raised shorelines and deglaciation in the Earn-Tay area. In: J. M. Gray and J.J. Lowe (Editors), Studies in the Scottish Late glacial Environment. Pergamon Press, Oxford, pp. 15-32. D’Anglejan, B.F., and Smith, E.C., 1973. Distribution, transport and composition of suspended matter in the Saint Lawrence Estuary. Est. Coastal Mar. Sci., 4: 401-416. Deegan, C.E., Kirby, R. and Rae, I., 1973. The superficial deposits of the Firth of Clyde and its sea lochs. Inst. Geol. Sci. Rep. 73/9, 135 pp. Dowdeswell, E.K. and Andrews, J.T., 1985. The fiords of Baffin Island: Description and classification. In: J.T. Andrews (Editor), Quaternary Environments: Eastern Canadian Arctic, Baffin Bay and West Greenland. Allen and Unwin Publ., pp. 93-123. Dowdeswell, J.A., and Murray, T, 1990. Modelling rates of sedimentation from icebergs. In: J.A. Dowdeswell and J.D. Scorse (Editors), Glaciomarine Environments: Processes and Sediments. Geol. SOC.London, Spec. Pub., 53: 121-137. Ebbesmeyer, C.C., Barnes, C.A. and Langby, C., 1975. Application of an advective-diffuse equation to a water parcel observed in a fjord. Est. Coast. Mar. Sci., 3: 249-268. Edwards, A. and Edelsten, D.J., 1977. Deep water renewal of Loch Etive: A three basin Scottish fjord. Est. Coast. Mar. Sci., 5: 575-595. Edwards, A., Edelstein, D.J., Saunders, M.A. and Stanley, S.O., 1980. Renewal and entrainment in Loch Eil, a periodically ventilated Scottish fjord. In: H.J. Freeland, D.M. Farmer and C.D. Levings (Editors), Fjord Oceanography. Plenum Press, New York, pp. 523-530. Elverhoi, A. and Roaldset, E., 1983. Glaciomarine sediments and suspended particulate matter, Weddell Sea Shelf, Antarctica. Polar Res., 1: 1-21. Elverhoi, A,, Liestol, 0. and Nagy, J., 1980. Glacial erosion, sedimentation and microfauna in the inner part of Kongsfjorden, Spitsbergen. Norsk. Polarinst. Skr., 172: 33-58. Elverhoi, A,, Lonne, 0. and Seland, R., 1983. Glaciomarine sedimentation in a modern fjord environment, Spitsbergen. Polar Res., 1: 127-149. Embleton, C. and King, C.A.M., 1970. Glacial and Periglacial Geomorphology. Macmillan of Canada, Toronto, 608 pp. England, J., 1983. Isostatic adjustments in a full glacial sea. Can. J. Earth Sci., 20: 895-917.
SEDIMENTOLOGY AND GEOMORPHOLOGY O F FJORDS
171
Exon, N., 1972. Sedimentation in the outer Flensburg Fjord area, Baltic Sea, since the last glaciation. Meyniana, 22: 5-62. Fairbanks, R.G., 1989. A 17,000-year glacio-eustatic sea level record: influence of glacial melting rates on the Younger Dryas event and deep ocean circulation. Nature, 342: 637-642. Fairbridge, R.W., 1968. The Encyclopedia of Geomorphology. Reinhold, New York, 1295 pp. Farmer, D.M. and Osborne, TR., 1976. The influence of wind on the surface layer of a stratified inlet. Part I. Observations. J. Phys. Oceanogr., 6: 931-940. Farmer, D.M. and Freeland, H.J., 1983. The physical oceanography of fjords. Prog. Oceaonogr., 12: 147-220. Farrow, G.E., Syvitski, J.P.M. and Tunnicliffe, V., 1983. Suspended particulate loading on the macrobenthos in a highly turbid fjord: Knight Inlet, British Columbia. Can. J. Fish. Aquat. Sci., 40: 273-288. Feely, R.A. and Massoth, G.T, 1982. Sources, composition, and transport of suspended particulate matter in Lower Cook Inlet and Northwestern Shelikof Strait, Alaska. NOAA Tech. Rep. ERL 415 PMEL 34,28 pp. Flint, R.F., 1940. Late Quaternary changes of level in western and southern Newfoundland. Geol. SOC. Am. Bull., 51: 1757-1780. Forbes, D.L., 1984. Coastal geomorphology and sediments of Newfoundland. In: Current Research, Part B, Geol. Surv. Can., 84-1B: 11-24. Forbes, D.L. and Syvitski, J.P.M., 1995. Paraglacial coasts. In: C. Woodruffe and R.W.G. Carter (Editors), Coastal Evolution. Cambridge University Press, New York, pp. 373-424. Forbes, D.L., Shaw, J. and Eddy, B.G., 1993. Late-Quaternary sedimentation and the postglacial relative sea-level minimum in Port au Port Bay and vicinity, west Newfoundland. Atlantic Geol., 29: 1-26. Freeland, H.J., Farmer, D.M. and Levings, C.D., 1980. Fjord Oceanography. Plenum Press, New York, 715 pp. Funder, S . , 1972. Deglaciation of the Scoresby Sund fjord region, north-east Greenland. Polar Geomorphology, Inst. British Geogr. Spec. Pub., 4: 33-42. Gade, H.G., 1968. Horizontal and vertical exchanges and diffusion in the water masses of the Oslo Fjord. Helgol. Meer., 17: 462-475. Gade, H.G., 1970. Hydrographic investigations in the Oslofjord: A study of water circulation and exchange processes. Geophysical Inst. Rep. 24, Univ. of Bergen, Norway, 193 pp. Gade, H.G., 1976. Transport mechanism in fjords. In: S. Skreslet, R. Leinebo, J.B.L. Matthews and E. Sakshaug (Editors), Freshwater on the Sea. Assoc. Norweg. Oceanogr., Oslo, pp. 51-56. Gade, H.G. and Edwards, A,, 1980. Deep-water renewal in fjords. In: H.J. Freeland, D.M. Farmer and C.D. Levings (Editors), Fjord Oceanography. Plenum Press, New York, pp. 453-489. Gade, H.G., Lake, R.A., Lewis, E.L. and Walker, E.R., 1974. Oceanography of an Arctic Bay. Deep-sea Res., 21: 547-571. Gilbert, R., 1978. Observations on oceanography and sedimentation at Pangnirtung fiord, Baffin Island. Mar. Sedimentol., 14: 1-9. Gilbert, R., 1980. Observations on the sedimentary environments of fjords on Cumberland Peninsula, Baffin Island. In: H.J. Freeland, D.M. Farmer and C.D. Levings (Editors), Fjord Oceanography. Plenum Press, New York, pp. 633-638. Gilbert, R., 1982. Contemporary sedimentary environments on Baffin Island, N.W.T. Canada: Glaciomarine processes in fjords of eastern Cumberland Peninsula. Arc. Alp. Res., 14: 1-12. Gilbert, R., 1983. Sedimentary processes of Canadian arctic fjords. In: J.P.M. Syvitski and J.M. Skei (Editors), Sedimentology of Fjords. Sediment. Geol., 36: 147-175. Gilbert, R., 1985. Quaternary glacimarine sedimentation interpreted from seismic surveys of fiords on Baffin Island, N.W.7: Arctic, 38: 271-280. Gilbert, R., 1990. Rafting in glacimarine environments. In: J.A. Dowdeswell and J.D. Scorse (Editors), Glaciomarine Environments: Processes and Sediments. Geol. SOC.London, Spec. Pub., 53: 105-120. Gilbert, R., Syvitski, J.P.M. and Taylor, R.B., 1985. Reconnaissance study of proglacial Stewart Lakes, Baffin Island, District of Franklin. In: Current Research, Part A, Geol. Surv. Can., Pap. 85-1A: 505-510.
172
J.P.M. SWITSKI AND J. SHAW
Gilbert, R., Naldrett, D.L. and Horvath, V.V., 1990. Holocene sedimentary environments of Cambridge Fiord, Baffin Island, Northwest Territories. Can. J. Earth Sci., 27: 272-282. Glasby, G.P., 1978. Sedimentation and sediment geochemistry of Caswell, Nancy and Milford Sounds. In: G.P. Glasby (Editor), Fiord Studies: Caswell and Nancy Sounds/New Zealand. N.Z. Oceanog. Inst. Mem. 79: 19-38.. Glenne, B. and Simensen, T, 1963. Tidal current choking in the landlocked fjord of Nordasvannet. Sarsia, 11: 43-73. Goldthwait, R.P., Loewe, F., Ugolin, F.C., Decker, H.F., DeLong, D.M., Trantman, M.B., Good, E.E., Merrell, TR. and Rudolph, E.D., 1966. Soil Development and Ecological Succession in a Deglaciated Area of Muir Inlet; Southeast Alaska. Inst. Polar Studies, Rep. 20, 18 pp. Gorlich, K., Weslawski, J.M. and Zajaczkowski, M., 1987. Suspension settling effect on macrobenthos biomass distribution in the Hornsund fjord, Spitsbergen. Polar Res., 5: 175-192. Grant, A.C., 1975. Seismic reconnaissances of Lake Melville, Labrador. Can. J. Earth Sci., 12: 21032110. Grant, D.R., 1989. Quaternary geology of the Atlantic Appalachian region of Canada. In: R.J. Fulton (Editor), Quaternary Geology of Canada and Greenland, Geol. Surv. Can., Geology of Canada, 1, Chapter 5, pp. 393-440. Griffith, TW. and Anderson, J.B., 1989. Climatic controls on sedimentation in bays and fjords of northern Antarctic Peninsula. Mar. Geol., 85: 181-204. Gustavson, TC., 1975. Sedimentation and physical limnology in proglacial Malaspina Lake, southeastern Alaska. In: A.V. Jopling and B.C. McDonald (Editors), Glaciofluvial and Glaciolacustrine Sedimentation. SEPM Spec. Publ. 23, pp. 249-263. Hafsten, U., 1983. Biostratigraphical evidence for late Weichselian and Holocene sea-level changes in southern Norway. In: D.E. Smith and A.G. Dawson (Editors), Shorelines and Isostasy. Academic Press, London, pp. 161-181. Haselton, G.M., 1965. Glacial geology of Muir Inlet, southeastern Alaska. Inst. Polar Studies Rep., 18: 1-34. Hay, A.E., 1983. Acoustic remote sensing. In: J.P.M. Syvitski and C.P. Blakeney (Comp.), Sedimentology of Arctic Fjords Experiment. HU 82-031 Data Report, Vol. 1. Can. Data Rep. Hydrogr. Ocean. Sci., 12: 17-1-17-3. Hein, F.J. and Syvitski, J.P.M., 1992. Sedimentary environments and facies in an arctic basin, Itirbilung Fiord, Baffin Island, Canada. Sediment. Geol., 81: 1-29. Hein, F.J., Syvitski, J.P.M., Long, B.F. and Dredge, L., 1993. Quaternary sedimentation in the Sept-Illes area, Quebec North Shore: Potential for marine placer. Can. J. Earth Sci., 30: 553-574. Heling, D., 1974. Investigations on recent sediments in the Fiskenaesset region, Southern west Greenland. Rapp. Gronland Geol. Sunder., 65: 65-68. Holtedahl, H., 1965. Recent turbidites in the Hardangerfjord, Norway. Colston Res. SOC.Proc., Bristol, 17, 107-141. Holtedahl, H., 1975. The Geology of the Hardangerfjord, West Norway. Norges Geol. Sunder. Bull., 36: 1-87. Horne, E., 1985. Ice-induced vertical circulation in an arctic fiord. J. Geophys. Res., 90: 1078-1086. Hoskin, C.M. and Burrell, D.C., 1972. Sediment transport and accumulation in a fjord basin, Glacier Bay, Alaska. J. Geol., 80: 539-551. Hoskin, C.M., Burrell, D.C. and Freitag, G.R., 1976. Suspended sediment dynamics in Queen Inlet, Glacier Bay, Alaska. Mar. Sci. Commun., 2: 95-108. Hoskin, C.M., Burrell, D.C. and Freitag, G.R., 1978. Suspended sediment dynamics in Blue Fjord, Western Prince William Sound, Alaska. Est. Coast Mar. Sci., 7: 1-16. Huggett, W.S. and Wigen, S.O., 1983. Surface currents in the approaches to Kitimat. In: R.W. MacDonald (Editor), Proc. Workshop JSitimat Marine Environment. Can. Tech. Rep. Hydrogr. Ocean. Sci., 18: 84-65. Huppert, H.E., 1980. Topographic effects in stratified fluids. In: H.J. Freeland, D.M. Farmer and C.D. Levings (Editors), Fjord Oceanography. Plenum Press, New York, pp. 117-141. Jacobs, L., Emerson, S. and Skei, J., 1985. Partitioning and transport of metals across the 02/HzS interface in a permanently anoxic basin: Framvaren Fjord Norway. Geochim. Cosmochim. Acta, 49:
SEDIMENTOLOGY AND GEOMORPHOLOGY O F FJORDS
173
1433-1444. Kitching, J.A., Ebling, F.J., Gamble, J.C., Hoave, R., McLeod, A.A., and Norton, TA., 1976. The ecology of Lough h e . XIX. Seasonal changes in the western trough. J. h i m . Ecol., 45: 731-758. Kjemperud, A., 1981. A shoreline displacement investigation from Frosta in Trondheimsfjorden, Nord-Tr~ndelag,Norway. Norsk Geol. Tidsskr., 61: 1-15. Klinck, J.M., O’Brien, J. and Svendsen, H., 1981. A simple model of fjord and coastal circulation interaction. J. Phys. Oceanogr., 11: 1612-1626. Knight, R.J., 1971. Distributional trends in the recent marine sediments of Tasiujaq Cove of Ekalugad Fiord, Baffin Island, N.W.T Mar. Sedimentol., 7: 1-18. Kostaschuk, R.A. and McCann, S.B., 1983. Observations on delta-forming processes in a fjordhead delta, British Columbia. In: J. Syvitski and J. Skei (Editors), Sedimentology of Fjords. Sediment. Geol., 36: 269-288. Kostaschuk, R.A. and McCann, S.B., 1989. Submarine slope failure of a fjord delta: Bella Coola, British Columbia. Geogr. Phys. Quat., 43: 87-95. Lake, R.A. and Walker, E.R., 1975. Notes on the oceanography of d’Ilberville Fiord. Arctic, 26: 222-229. Lake, R.A. and Walker, E.R., 1976. A Canadian arctic fjord with some comparisons to fjords of the Western Americas. J. Fish. Res. Board Can., 33: 2277-2285. Lambeck, K., 1991. Glacial rebound and sea-level change in the British Isles. Terra Nova, 3: 379-389. Lavrushin, Y.A., 1968. Features of deposition and structure of the glacialmarine deposits under conditions of a fiord-coast. Translations from Lith. Econ. Min., 3: 63-79. Lemmen, D.S., 1990. Glaciomarine sedimentation in Disraeli Fiord, high arctic Canada. Mar. Geol., 94: 9-22. Lewis, E.L. and Perkin, R.G., 1982. Seasonal mixing processes in an Arctic fjord system. J. Phys. Oceanogr., 12: 74-83. Liestol, O., 1969. Glacial surges in West Spitsbergen. Can. J. Earth Sci., 6: 895-897. Lliboutry, L., 1965. Glaciers variations du c h a t sols gelts. In: Masson and Cie (Editors), Trait6 de Glaciologie Tomell, pp. 429-504, 517-520, 680-682, 763-764. Long, R.R., 1980. The fluid mechanical problem of fjord circulations. In: H.J. Freeland, D.M. Farmer and C.D. Levings (Editors), Fjord Oceanography, Plenum Press, New York, pp. 67-116. Luternauer, J.L. and Swan, D., 1978. Kitimat submarine slump deposit(s), a preliminary report. Current Research Part A, Geol. Surv. Can. Pap., 78-1A 327-332. Mackiewicz, N.E., Powell, R.D., Carlson, P.R. and Molnia, B.F., 1984. Interlaminated iceproximal glacimarine sediments in Muir Inlet, Alaska. Mar. Geol., 57: 113-147. Mandl, G. and Crans, W., 1981. Gravitational gliding in deltas. In: Thrust and Nappe Tectonics. Geol. SOC.London, pp. 41-54. Matthews, J.B., 1981. The seasonal circulation of the Glacier Bay, Alaska fjord system. Est. Coast. Shelf Sci., 12: 679-700. McAlister, W.B., Rattray, M. and Barnes, C.A., 1959. The dynamics of a fjord estuary: Silver Bay, Alaska. Dept. Oceanogr. Tech. Rep. 62, Univ. of Washington, Seattle, Washington. McCann, S.B., Dale, J.E. and Hale, P.B., 1981. Subarctic tidal flats in areas of large tidal range, southern Baffin Island, eastern Canada. Geogr. Phys. Quat., 35: 183-204. McClimans, TA., 1978a. Fronts in fjords. Geophys. Astrophys. Fluid Dyn., 11: 23-34. McClimans, TA., 1978b. On the energetics of tidal inlets to land-locked fjords. Mar. Sci. Commun., 4: 121-137. McClimans, TA., 1979. On the energetics of river plume entrainment. Geophys. Astrophys. Fluid Dyn., 13: 67-82. McKenna-Neuman, C. and Gilbert, R., 1986. Aeolian processes and landforms in glacifluvial environments in southeastern Baffin Island, N.W.T, Canada. In: W.G. Nickling (Editor), Aeolian Geomorphology. Proc. 17th Annu. Binghamton Geomorphology Symp. Geo Books, Nonvich, Chapter 12, pp. 213-235. Mercer, EH., 1961. The response of fjord glaciers to changes in the firm limit. J. Glaciol., 3: 850-858. Miller, A.A.L., Mudie, P.S. and Scott, D.B., 1982. Holocene history of Bedford Basin, Nova Scotia: foraminifera, dinoflagellate and pollen records. Can. J. Earth Sci., 19: 2342-2367.
174
J.P.M. SYVITSKI AND J. SHAW
Miller, D.J., 1960. Giant waves in Lituya Bay, Alaska. U S. Geol. Surv. Prof. Pap., 354C: 5186. Milliman, J.D. and Syvitski, J.P.M., 1992. Geomorphic/tectonic control of sediment discharge to the ocean: The importance of small mountainous rivers. J. Geol., 100: 525-544. Milne, Ph. H., 1972. Hydrography of Scottish west coast sea lochs. Mar. Res. Rep. No. 3, Dept. Agricult. Fish. Scotland, Edinburgh, 40 pp. Molnia, B.F., 1983. Subarctic glacial-marine sedimentation: a model. In: B.F. Molnia (Editor), GlacialMarine Sedimentation. Plenum Press, New York, pp. 95-114. Moller, J.S., 1984. Hydrodynamics of an arctic fjord. Field study. Affarlikassaa, West Greenland. Resonance of internal seiches and buoyancy driven circulation. Inst. Hydro-Dynamics and Hydraulic Engineering, Lyngby Denmark, Ser. Pap. 34, 197 pp. Molvaer, J., 1980. Deep-water renewals in the Frierfjord - an intermittently anoxic basin. In: H.J. Freeland, D.M. Farmer and C.D. Levings (Editors), Fjord Oceanography. Plenum Press, New York, pp. 531-539. Morner, N-A,, 1980. Eustasy and geoid changes as a function of Core/Mantle changes. In: N-A. Morner (Editor), Earth Rheology, Isostasy and Eustasy. John Wiley and Sons, London. pp. 535-553. Muench, R.D. and Heggie, D.T, 1978. Deep water exchange in Alaskan subarctic fjords. In: B. Kjerfve (Editor), Estuarine Transport Processes. Univ. S. Carolina Press, Columbia, S.C., pp. 239-268. Murty, TS., 1979. Submarine slide-generated water waves in Kitimat Inlet, British Columbia. J. Geophys. Res., 84: 7777-7779. Nichols, R.L., 1960. Geomorphology of Marguerite Bay area Palmer Peninsula, Antarctica. Geol. SOC. Am. Bull., 71: 1421-1450. Nihoul, J.C.J. (Editor), 1978. Hydrodynamics of Estuaries and Fjords. Elsevier Scientific Publishing Company, Amsterdam, 546 pp. Nordseth, K., 1976. Suspended and bed material load in Norwegian Rivers. In: S. Skreslet, R. Leinebo, J.B.L. Matthews and E. Sakshaug (Editors), Freshwater in the Sea, pp. 33-42. Northcote, TG., Wilson, M.S. and Hum, D.R., 1964. Some characteristics of Nitinat Lake, an inlet on Vancouver Island, British Columbia. J. Fish. Res. Board Can., 21: 1069-1081. Nye, J.F., 1976. Water flow in glaciers: jekolhlaups, tunnels and veins. J. Glaciol., 17: 181-207. Nyholm, N., Nielsen, TK. and Pedersen, K., 1983. Modeling heavy metals transport in an arctic fjord system polluted from mine tailings. ISEM Conf. on Modelling the rate and effects of toxic substances in the environment, Copenhagen. Olsen, 0. and Reeh, N., 1969. Preliminary report on glacier observations in Nordwestfjord, east Greenland. Rapp. Gronland. Geol. Sunder., 21: 41-53. Orford, J.D., Carter, R.W.G. and Jennings, S.C., 1991. Coarse clastic barrier environments: evolution and implications for Quaternary sea level interpretation. Quat. Int., 9: 87-104. Dstrem, G., 1975. Sediment transport in glacial meltwater streams. In: A.V. Jopling and B.C. McDonald (Editors), Glaciofluvial and Glaciolacustrine Sedimentation. SEPM Spec. Publ., 23: 101-122. Dstrem, G., Bridge, C.W. and Rannie, W.F., 1967. Glacio-hydrology, discharge and sediment transport in the Decade Glacier Area, Baffin Island, N.W.T Geog. Anna., 49A: 268-282. Dstrem, G., Riegler, T and Ekman, S.R., 1970. Slamtransportundersokelser i norske bre-elver 1969. Rep. 6/70, Hydrologisk Avdeling, Norges Vassdrags- og Elektrisitetsvesen, 68 pp. Ovenshine, A.T, 1970. Observations of iceberg rafting in Glacier Bay, Alaska and the identification of ancient ice-rafted deposits. Geol. SOC.Am. Bull., 81: 891-894. Pantin, H.M., 1964. Sedimentation in Milford Sound. In: TM. Skerman (Editor), Studies of a Southern Fiord. N. Zeal. Ocean. Inst. Mem., 17: 35-48. Pearson, TH. and Rosenberg, R., 1976. A comparative study of the effects on the marine environment of water from cellulose industries in Scotland and Sweden. Ambia, 5: 77-79. Pedersen, B., 1986. Environmental Hydraulics: Stratified Flows. VIII, Springer-Verlag, New York, 278 PP. Peltier, W.R. and Andrews, J.T, 1976. Glacial isostatic adjustment - I. The forward problem. Geophys. J. R. Astron. SOC.,46: 605-646. Pelto, M.S., and Warren, C.R., 1991. Relationship between tidewater glacier calving velocity and water depth at the calving front. Ann. Glaciol., 15: 115-118.
SEDIMENTOLOGY AND GEOMORPHOLOGY OF FJORDS
175
Petersen, G.H., 1977. Biological effects of sea-ice and icebergs in Greenland. In: M.J. Dunbar (Editor), Polar Oceans. Arctic Institute of North America, Calgary, pp. 319-329. Petersen, G.H., 1978. Life cycles and population dynamics of marine benthic bivalves from the Disko Bugt area of West Greenland. Ophelia, 17: 95-120. Phillips, A.C. and Smith, N.D., 1992. Delta slope processes and turbidity currents in prodeltaic submarine channels, Queen Inlet, Glacier Bay, Alaska. Can. J. Earth Sci., 29: 93-101. Phillips, A.C., Smith, N.D. and Powell, R.D., 1991. Laminated sediments in prodeltaic deposits, Glacier Bay, Alaska. In: J.B. Anderson and G.M. Ashley (Editors), Glacial Marine Sedimentation: Paleoclimatic Significance. Boulder Co., Geol. SOC.Am. Spec. Paper, 261: 51-60. Pickard, G.L., 1961. Oceanographic features of inlets in the British Columbia mainland coast. J. Fish. Res. Board Can., 18: 907-984. Pickard, G.L., 1971. Some physical oceanographic features of inlets of Chile. J. Fish. Res. Board Can., 28: 1077-1106. Pickard, G.L. and Stanton, B.R., 1980. Pacific fjords - a review of their water characteristics. In: H.J. Freeland, D.M. Farmer and C.D. Levings (Editors), Fjord Oceanography. Plenum Press, New York, pp. 1-51. Pickrill, R.A., 1987. Circulation and sedimentation of suspended particulate matter in New Zealand Fjords. Mar. Geol., 74: 21-39. Pickrill, R.A. and Irwin, J., 1983. Sedimentation in a deep glacier-fed lake - Lake Tekapo, New Zealand. Sedimentology, 30: 63-75. Pickrill, R.A., Irwin, J. and Shakespeare, B.S., 1981. Circulation and sedimentation in a tidal influenced fjord lake: Lake McKerrow, New Zealand. Est. Coastal Shelf Sci., 12: 23-37. Piper, D.J.W. and Keen, M.J., 1976. Geological Studies in St. Margarets Bay Nova Scotia. Geol. Surv. Can. Pap. 78-18, 18 pp. Piper, D.J.W., Letson, J.R.J., Delure, A.M. and Barrie, C.Q., 1983. Sediment accumulation in lowsedimentation, wave dominated, glacial inlets. Sediment. Geol., 36: 195-215. Poole, A.L., 1951. New Zealand-American Fiordland Expedition. N.Z. Dep. Sci. Ind. Res. Bull., 103, 99 PP. Powell, R.D., 1990. Glacimarine processes at grounding-line fans and their growth to ice-contact deltas. In: J.A. Dowdeswell and J.D. Scorse (Editors), Glaciomarine Environments: Processes and Sediments. Geol. SOC.London, Spec. Pub., 53: 53-73. Powell, R.D., 1991. Grounding-line systems as second-order controls on fluctuations of tidewater termini of temperate glaciers. In: J.B. Anderson and G.M. Ashley (Editors), Glacial Marine Sedimentation; Paleoclimatic Significance. Geol. SOC.Am. Spec. Pap., 261: 75-93. Prior, D.B., Wiseman Jr., W.J. and Gilbert, R., 1981a. Submarine slope processes on a fan delta, Howe Sound, British Columbia. Geo-Mar. Lett., 1: 85-90. Prior, D.B., Wiseman Jr., W.J. and Bryant, W.R., 1981b. Submarine chutes of the slopes of fjord deltas. Nature, 209: 326-328. Prior, D.B., Bornhold, B.D. and Coleman, J.M. and Bryant, W.R., 1982a. Morphology of a submarine slide, Kitimat Arm, British Columbia. Geology, 10: 588-592. Prior, D.B., Coleman, J.M. and Bornhold, B.D., 1982b. Results of a known sea floor instability event. Geo-Mar. Lett., 2: 117-122. Prior, D.B., Bornhold, B.D., and Coleman,, J.M., 1983. Geomorphology of a submarine landslide, Kitimat Arm, British Columbia. Geological Survey of Canada Open File Report 961. Prior, D. B., Bornhold, B.D., Wiseman Jr., W.J. and Lowe, D.R., 1987. Turbidity current activity in a British Columbia fjord. Science, 237: 1330-1333. Quinlan, G. and Beaumont, C., 1981. A comparison of observed and theoretical post glacial relative sea level in Atlantic Canada. Can. J. Earth Sci., 18: 1146-1163. Reeh, N., 1985. Long calving waves. Proc. 8th Int. Conf. on Port and Ocean Engineering under Arctic Conditions, Narssarssuaq, Greenland, Sept. 7-14, 1985, pp. 1310-1327. Relling, 0. and Nordseth, K., 1979. Sedimentation of a river suspension into a fjord basin. Gaupnefjord in Western Norway. Norsk Geogr. Tidsskr., 33: 187-203. Richards, F.A., 1965. Anoxic basins and fjords. In: J.P. Riley and G. Skirrow (Editors), Chemical Oceanography, Vol., 1. Academic Press, London, pp. 611-645.
176
J.P.M. SYVITSKI AND J. SHAW
Richards, A.F., 1976. Marine geotechnics of the Oslofjorden region. In: Contributions to Soil Mechanics, Laurits Bjerrum Memorial Volume. Oslo, Norweg. Geotech. Inst., pp. 41-63. Rosenberg, R., 1980. Effect of oxygen deficiency on benthic macrofauna in fjords. In: H.J. Freeland, D.M. Farmer and C.D. Levings (Editors), Fjord Oceanography. Plenum Press, New York. pp. 499-514. Rust, B.R., 1988. Ice proximal deposits of the Champlain Sea at South Gloucester near Ottawa, Canada. In: N.R. Gade (Editor), The Late Quaternary Development of the Champlain Sea Basin. Geol. Assoc. Can. Spec. Pap., 35: 37-45. Saelen, O.H., 1967. Some features on the hydrography of Norwegian fjords. In: G.H. Lauff (Editor), Estuaries. AAAS Publ. 83: 63-71. Sangrey, D.A., Cornell, U., Clukey, E.C. and Molnia, B.F., 1979. Geotechnical engineering analysis of underconsolidated sediments from Alaska coastal waters. 11th Offshore Technology Conf., Houston, 1: 677-682. Schafer, C.T, Cole, F.E. and Syvitski, J.P.M., 1989. Bio and lithofacies of modern sediments in Knight and Bute Inlets, B.C. Palaios, 4: 107-126. Schwarz, H.-U., 1982. Subaqueous Slope Failures - Experiments and Modern Occurrences. Contributions to Sedimentology 11, Elsevier, Stuttgart. Shaw, J., and Forbes, D.L., 1990a. Short- and long-term relative sea-level trends in Atlantic Canada. In: Proc. Canadian Coastal Conference, 1990, Kingston, Ont., pp. 291-305. Shaw, J. and Forbes, D.L., 1990b. Late Quaternary sedimentation in inner St. George’s Bay, southwest Newfoundland: acoustic stratigraphy and sea-bed deposits. Can. J. Earth Sci., 27: 964-983. Shaw, J. and Forbes, D.L., 1992. Barriers, barrier platforms, and spillover deposits in St. George’s Bay, Newfoundland: Paraglacial sedimentation on the flanks of a deep coastal basin. Mar. Geol., 105: 119-140. Shaw, J, Taylor, R.B., and Forbes, D.L., 1990. Coarse clastic barriers in eastern Canada: patterns of glaciogenic sediment dispersal with rising sea levels. J. Coastal Res., Spec. Issue, 9.: 160-200. Skei, J.M., 1982. Fluxes of pollutants to Norwegian fjord sediments based on lead-210. Proc. Symp. Mar. Chem. into the Eighties, Sidney, B.C., pp. 29-47. Skei, J.M., 1983. Geochemical and sedimentological considerations of a permanently anoxic fjord - Framvaren, South Norway. Sediment. Geol., 36: 131-145. Skei, J.M., Price, N.B. and Calvert, S.E., 1972. The distribution of heavy metals in sediments of Sorfjord, Norway. Water Air Soil Pollut., 1: 452-461. Skerman, TM., 1964. Studies of a southern fjord. New Zealand Oceanographic Institute Mem. No. 17, 102 pp. Slatt, R.M., 1974. Formation of palimpsest sediments, Conception Bay, southeastern Newfoundland. Geol. SOC.Am. Bull., 85: 821-826. Slatt, R.M. and Gardiner, W.W., 1976. Comparative petrology and source of sediments in Newfoundland fjords. Can. J. Earth Sci., 13: 1460-1465. Smith, J.N. and Walton, A., 1980. Sediment accumulation rates and geochronologies measured in the Saguenay Fjord using Pb-210 dating method. Geochirn. Cosrnochim. Acta, 44: 225-240. Smith, J.N. and Schafer, C.T, 1987. A 20th century record of seasonally-modulated sediment accumulation rates in a Canadian fjord. Quat. Res., 27: 232-247. Stanley, D.J., 1968. Reworking of glacial sediments in the N.W. Arm, a fjord-like inlet on the southeast coast of Nova Scotia. J. Sediment. Petrol., 38: 1224-1241. Stanton, B.R. and Pickard, G.L., 1981. Physical Oceanography of the New Zealand Fiords. New Zealand Oceanogr. Inst. Mem., 88: 1-36. Stehman, C.F., 1976. Pleistocene and Recent sediments of northern Placentia Bay, Newfoundland. Can. J. Earth Sci., 13: 1386-1392. Stewart, TG., 1991. Glacial marine sedimentation from tidewater glaciers in the Canadian High Arctic. In: J.B. Anderson and G.M. Ashley (Editors), Glacial Marine Sedimentation: Paleoclimatic Significance. Boulder Co., Geol. SOC.Am. Spec. Pap., 261: 95-107. Stravers, J.A. and Syvitski, J.P.M., 1991. Early Holocene land-sea correlations and deglacial evolution of the Cambridge Fiord basin, Northern Baffin Island. Quat. Res., 35: 7290.
SEDIMENTOLOGY AND GEOMORPHOLOGY O F FJORDS
177
Stream, K.M., 1936. Land-locked water. Hydrography and bottom deposits in badly ventilated Norwegian fjords with remarks upon sedimentation under anaerobic conditions. Skr. Norsk. Videnskap. Akad., Oslo, No. 7, 85 pp. Strcam, K.M., 1957. A lake with trapped sea water? Nature, 180: 982-983. Strcam, K.M., 1961. A second lake with old sea water at its bottom. Nature, 189: 913. Stromgren, T, 1974. The use of weighted arithmetic mean for describing the sediments of a landlocked basin (Borgenfjorden, Western Norway). Deep-sea Res., 21: 155-160. Suess, E., 1979. Mineral phases found in anoxic sediments by microbial decomposition of organic matter. Geochim. Cosmochim. Acta., 43: 339-352. Svendsen, H., 1977. A study of the circulation in a sill fjord on the west coast of Norway. Mar. Sci. Commun., 3: 151-209. Syvitski, J.P.M., 1989. On the deposition of sediment within glacier-influenced fjords: Oceanographic controls. Mar. Geol., 85: 301-329. Syvitski, J.P.M. (Editor), 1991. Principles, Methods and Applications of Particle Size Analysis. Cambridge University Press. New York, 368 pp. Syvitski, J.P.M., 1992. Marine geology of Baie Des Chaleurs. Geogr. Phys. Quat., 46: 331-348. Syvitski, J.P.M., 1993a. Fluvial sediments and marine interactions - a Canadian overview. Can. Water Resour. J. (in press). Syvitski, J.P.M., 1993b. Glaciomarine environments in Canada: an overview. Can. J. Earth Sci., 30: 354-371. Syvitski, J.P.M. and Murray, J.W., 1981. Particle interaction in fjord suspended sediment. Mar. Geol., 39: 215-242. Syvitski, J.P.M. and van Everdingen, D.A., 1981. A revaluation of the geologic phenomenon of sand flotation: A field and experimental approach. J. Sediment. Petrol., 51: 1315-1322. Syvitski, J.P.M. and MacDonald, R., 1982. Sediment character and provenance in a complex fjord; Howe Sound, British Columbia. Can. J. Earth Sci., 19: 1025-1044. Syvitski, J.P.M. and Farrow, G.E., 1983. Structures and processes in bayhead deltas: Knight and Bute Inlet, British Columbia. Sedim. Geol., 36: 217-244. Syvitski, J.P.M. and Farrow, G.E., 1989. Fjord sedimentation as an analogue for small hydrocarbonbearing submarine fans. In: M.K.G. Whateley and K.T. Pickering (Editors), Deltas: Sites and Traps for Fossil Fuels. Geol. SOC.London Spec. Publ., 41: 21-43. Syvitski, J.P.M. and Praeg, D.B., 1989. Quaternary sedimentation in the St. Lawrence Estuary and adjoining areas, Eastern Canada: An overview based on high resolution seismo-stratigraphy. Gtogr. Phys. Quat., 43: 291-310. Syvitski, J.P.M. and Hein, F.J., 1991. Sedimentology of an arctic basin: Itirbilung Fiord, Baffin Island, Canada. Geol. Surv. Can. Prof. Paper 91-11, 67 pp. Syvitski, J.P.M. and Daughney, S., 1992. DELTA2: Delta progradation and basin filling. Comput. Geosci., 18: 839-897. Syvitski, J.P.M. and Lewis, A.G., 1992. The seasonal distribution of suspended particles, and their Iron and Manganese loading, in a glacial runoff fjord. Geosci. Can., 19: 13-20. Syvitski, J.P.M. and Alcott, J.M., 1993. GRAlN2: Predictions of particle size seaward of river mouths. Comput. Geosci., 19: 399-446. Syvitski, J.P.M., Fader, G.B., Josenhans, H.W., MacLean, B. and Piper, D.J.W., 1983a. Seabed investigations of the Canadian east coast and arctic using Pisces IV. Geosci. Can., 10: 59-68. Syvitski, J.P.M., Silverberg, N., Ouellet, G. and Asprey, K.W., 1983b. First observations of benthos and seston from a submersible in the lower St. Lawrence estuary. Geog. Phys. Quat., 37: 227-240. Syvitski, J.P.M., Asprey, K.W., Clattenburg, D.A. and Hodge, G.D., 1985. The prodelta environment of a fjord: suspended particle dynamics. Sedimentology, 32: 83-107. Syvitski, J.P.M., Burrell, D.C. and Skei, J.M., 1987. Fjords: Processes and Products. Springer-Verlag, New York, 379 pp. Syvitski, J.P.M., Smith, J.N., Boudreau, B. and Calabrese, E.A., 1988. Basin sedimentation and the growth of prograding deltas. J. Geophys. Res., 93: 6895-6908. Syvitski, J.P.M., LeBlanc, K.W.G. and Cranston, R.E., 1990. The flux and preservation of organic carbon in Baffin Island fjords. In: J.A. Dowdeswell and J.D.Scorse (Editors), Glaciomarine Environments:
178
J.P.M. SWITSKI AND J. SHAW
Processes and Sediments. Geol. SOC.,London, Spec. Pub., 53: 217-239. Terzaghi, K., 1956. Varieties of submarine slope failures. Proc. 8th Texas Conf. Soil Mech. Foundation Eng., 41 pp. Thomson, R.E., 1981. Oceanography of the British Columbia Coast. Can. Fish. Aqua. Sci. Spec. Pub., 56: 1-291. Thors, K., and Boulton, G.S., 1991. Deltas, spits and littoral terraces associated with rising sea level: Late Quaternary examples from northern Iceland. Mar. Geol.,, 98: 99-112. Tomasson, H., 1991. Glaciofluvial sediment transport and erosion. In: J. Gjessing, J.O. Hagen, K.A. Hassel, K. Sand and B. Wold (Editors), Arctic Hydrology. Present and Future Tasks. Norwegian National Committee for Hydrology Report 23, Oslo, pp. 27-36. Thnnicliffe, V. and Syvitski, J.P.M., 1983. Corals assist boulder movement: an unusual mechanism of sediment movement. Limnol. Oceanography, 8: 564-568. Tushingham, A. M., and Peltier, W.R., 1991. Ice-3G: a new global model of late Pleistocene deglaciation based upon geophysical predictions of post-glacial relative sea level change. J. Geophys. Res., 96: 4497-4523. Vilks, G. and Mudie, P.J., 1983. Evidence for postglacial paleoceanographic and paleoclimatic changes in Lake Melville, Labrador, Canada. Arctic Alpine Res., 15: 307-320. von Huene, R., 1966. Glacial-marine geology of Nuka Bay, Alaska, and the adjacent continental shelf. Mar. Geol., 4 291-304. von Huene, R., Shor Jr., G.G. and Reimnitz, E., 1967. Geological interpretation of seismic profiles in Prince William Sound, Alaska. Geol. SOC.Am. Bull., 78: 259-268. Vorndran, V.G. and Sommerhoff, G., 1974. Glaziologisch-glazialmorphologische Untersuchungen im Gebiet des Qorqup-Austafsyle + Schers (Sudwest-Gronland). Polarforschung, 44: 137147. Weigel, R.L., 1955. Laboratory studies of gravity waves generated by the movement of a submerged body. Trans. AGU, 36: 759-774. Williams, EM., Mathews, W.H. and Pickard, G.L., 1961. A lake in British Columbia containing old seawater. Nature, 191: 830-832. Winters, G.V. and Syvitski, J.P.M., 1992. Suspended sediment character and distribution in McBeth Fiord, Baffin Island. Arctic, 45: 25-35. Yoshida, S., 1980. Mixing mechanisms of density current system at a river mouth. In: 'I: Carstens and T McClimans (Editors), 2nd Int. Symp. on Stratified Flows. Ziegler, T, 1973. Materialtransportunders kelser i norske bre-elver 1971. Rep. 4/73, Hydrologisk Avdeling Norges Vassdrags- og Electrisitetsvesen, 91 pp.
Geomurphulugy and Sedimeniology of Estuaries. Developmentsin Sedimeniology 53 edited by G.M.E. Perillo 0 1995 Elsevier Science B.V. All rights reserved.
179
Chapter 6
TIDE-DOMINATED ESTUARIES AND TIDAL RWERS JOHN T WELLS
INTRODUCTION
Most estuaries of the world are influenced, at least to some degree, by tides. Tidal energy serves as a mechanism for mixing both river-mouth and estuarine waters, resuspending and transporting sediments, creating bedforms, scouring channels, and redistributing pollutants. In extreme cases, tide is such an overwhelming source of energy within an estuary that it dominates virtually every aspect of physical forcing, morphologic response, and sedimentologic character. The role of tides is complex and less well studied than many other aspects of estuarine processes. We know that certain basins, such as the North Sea, the Yellow Sea, and the Bay of Fundy have especially large tide ranges because of basin geometry. We know that when tides converge in coastal embayments, such as estuaries, their energy is concentrated and the range increases as tides are forced through smaller crosssectional areas. We know that as tides move upstream through smaller cross-sectional areas, the tidal currents become progressively more asymmetric in both speed and duration. Yet, even when generalizations and simplifying assumptions are made, it is difficult to unravel the role of tides in sedimentation patterns and morphologic development because of the differences in source, distribution, and nature of both fine and coarse sediments found within estuaries. This chapter summarizes the important processes and attributes, as we presently understand them, in the class of estuaries referred to as tide-dominated estuaries or, in some cases, tidal rivers. Basic physical processes are reviewed in the first section of the chapter. Following that, the primary focus is on estuarine morphology and sedimentology. Many of the environments and habitats of tide-dominated estuaries, such as tidal flats and wetlands, are discussed in greater detail in separate chapters. Tidal sand ridges and smaller bedforms, which are discussed here insofar as they reflect sedimentation processes, are also given separate, more detailed treatment in another chapter (see Dalrymple and Rhodes, this volume). Holocene examples of both tide-dominated estuaries and tidal rivers, taken from the published literature, are discussed in the last two sections of this chapter.
The classification problem A logical set of questions is: what constitutes a tide-dominated estuary and how is it distinguished from a tidal river, how do we go about accentuating similarities and differences to allow formation of such a class of estuaries, and what was the basis for selecting the examples presented at the end of this chapter? Schubel and Pritchard
180
J.T. WELLS
(1990) state that “if classes are formed correctly (in this case for rivers and estuaries heavily influenced by tide), then what we know of one object in the class we know of the other objects” (quoted originally from Jevons, 1900). Clearly, the integration of two or more units into a single class requires that the units possess common characteristics (Dolan et al., 1972). Application of this conceptually simple idea to tide-dominated estuaries and tidal rivers is complicated by several factors. First, tide dominance cannot be determined solely on the basis of tide range. The classification of coastlines into microtidal ( < 2 m), mesotidal (2-4 m), and macrotidal ( > 4 m) environments (Davies, 1964; Hayes, 1975) has found widespread appeal in the geologic literature. However, some macrotidal river mouths, such as the Amazon, are not true tide-dominated estuaries because of high wave energy or extreme river discharge; in the case of the Amazon, river water meets and mixes with salt water on the open shelf (Nittrouer et al., 1986; Curtin and Legeckis, 1986), yet tides are important in the river mouth and their effects extend for hundreds of km upstream. Moreover, some microtidal estuaries, for example Chesapeake Bay entrance, display features such as mutually-evasive ebb and flood channels (Ludwick, 1974, 1975), which are characteristic of tide-dominated estuaries. A second complicating factor is that there are very few well-researched estuaries that are dominated by tides. We are biased by those few which have been studied in detail, most notably the Bristol ChanneUSevern Estuary (Kirby and Parker, 1983; Harris and Collins, 1985), Gironde Estuary (Allen, 1971; Nichols and Biggs, 1985), and parts of the Bay of Fundy (Dalrymple et al., 1990, 1992). Further, there is not a clear morphologic distinction between tide-dominated deltas and tide-dominated estuaries. Although deltas have been conveniently classified as river-, wave-, and tide-dominated systems (Coleman and Wright, 1971; Wright and Coleman, 1973; Galloway, 1975), some of the tide-dominated deltas, such as the Ord (Wright et al., 1975) and Hang (Coleman et al., 1970), might better be classified as estuaries. A third classification problem is the natural variability of estuaries. It is simply not possible to form mutually-exclusive classes of estuaries; there is a range of properties and processes, and some estuaries may fit into more than one class (the Keum Estuary of Korea is usually considered a ria-type estuary but is also dominated by tides; Wells and Park, 1992). Others vary considerably over an annual cycle, and still others have been altered by dredging or upstream engineering modifications. Because tide and wave energy are seldom quantified in the geologic literature (although some measure of their amplitudes is usually given), essentially all tide-dominated estuaries, including those described in this chapter, receive such status only on the basis of descriptive features. In simplest descriptive terms, tide-dominated estuaries are those in which tidal currents play the dominant role in determining the fate of the river-borne sediments. There is appreciable upstream transport of bedload sediment as a result of deformation of tide during propagation. Sediment is received from both the river and the shelf, yet most of the sediment received by the river ends up being transported and deposited by tidal currents. In extreme cases, there may be little or no density driven circulation. Tide-dominated estuaries usually have a funnel shape, are characterized
181
TIDE-DOMINATED ESTUARIES AND TIDAL RIVERS
by sand waves and linear subtidal sand ridges, and by intertidal mudflats, marshes, or mangrove swamps. As a subset of tide-dominated estuaries, tidal rivers are estuaries which have many of the same morphologic and sedimentologic features, but are associated with rivers that have very high discharge (e.g. Amazon and de la Plata in South America). Tidal influence usually extends substantially farther upstream than salt intrusion, and deltaic sedimentation occurs only in the subaqueous environment.
PHYSICAL PROCESSES IN TIDE-DOMINATED ESTUARIES
In the absence of strong tidal effects, rivers usually decelerate as flow expands into an estuarine section of greater cross-sectional area near the river mouth. Freshwater encounters, and may mix with, saltwater; sediment-transporting ability diminishes and sediments are deposited (summaries given in Dyer, 1973; Officer, 1976; Wright, 1977, 1985). In contrast, strong tidal effects destroy vertical stratification, create bidirectional currents, produce high shear stresses at the bed, and lead to time-velocity asymmetry in flow (Fig. 6-1). The periodic rise and fall of tide at the mouth of a river results in the temporary storage of large volumes of sea water in the estuary at high tide, followed by drainage at low tide. The volume of water exchanged by tide for a given estuary is known as the tidal prism. Tide-dominated estuaries typically have a tidal prism at least an order of magnitude greater than river discharge (Harris et al., 1993). The ratio between tidal
(.
2..
TIDE-DOMINATED ESTUARIES
.. .... .. ...... .,.
i.
'!': 'C
,.
PHYSlCAL PROCESSES
CHARACTERISTIC PHYSICAL PROCESSES 1 Bi-directional tidal currents and mutually-evasive flow 2 Broad region of strong turbulent mixing usually resulting in well-mixed estuary 3 Net upstream transport of coarse sediments from stronger flood currents 4 Rapid flow convergence and tidal wave deformation 5 Transient turbidity maximum developed by tides or combination of tides and density processes 6 Upstream limit of salt penetration 7 Upstream limit of tide
Fig. 6-1. Diagram showing the physical processes that are common to tide-dominated estuaries.
182
J.7: WELLS
transport and freshwater discharge provides a measure of the degree to which tidal mixing prevents vertical density stratification (Ippen and Harleman, 1966) which, in turn, affects lateral expansion of flow, estuarine circulation, and sediment dispersion (Wright, 1977). Because of significant changes in the tidal prism over a spring-neap cycle, tide-dominated estuaries often display large variations in mixing and therefore in density circulation. The most obvious effects of tide are to produce bi-direction currents and high shear stresses during peak flows. Although residual flow will be downstream, reversals in tide will periodically shift the fluvial-marine interface up and down the distributary channels. To a first approximation, maximum currents are determined by the crosssectional area of the estuary relative to the tide range. If frictional effects are neglected, constrictions will always accentuate currents. Because shear stress varies as the square of current speed, stresses at the sediment bed will also be accentuated by the presence of tide. In terms of sediment dynamics, one of the most important aspects of tidedominated estuaries is the way in which tide propagates upstream (Fig. 6-2). The speed at which tide moves up the axis of an estuary is governed by the equation for propagation of shallow water waves, and is therefore a direct function of water depth (see Shore Protection Manual, 1984). Because of this depth dependence, estuarine tides are deformed during upstream propagation as flood crests overtake ebb troughs. The result is twofold. First, flood velocities will exceed ebb velocities, but be of shorter duration. Second, the period of high-water slack will become longer than that for low-water slack. In many estuaries the degree of tidal asymmetry increases upstream, thereby magnifying the differences between ebb and flood velocities and slack-water durations. Tides also tend to increase in amplitude from upstream convergence. Decreases in depth and width resulting from the characteristic funnel-shaped geometry force the tidal wave through progressively smaller cross-sectional areas. However, frictional dissipation from the bottom and banks of the estuary occurs with convergence, and
4-
TIDE DEFORMATION ORD RIVER ESTUARY
3-
-4
;
0
6
12
24 30 TIME ( H R ) 4
18
36
42
48
Fig. 6-2. Deformation of tide in the Ord River from observations taken simultaneously at three stations (modified from Wright et al., 1975).
TIDE-DOMINATED ESTUARIES AND TIDAL RIVERS
183
tends to counteract it, thus decreasing the amplitude. If the effects of convergence exceed frictional dissipation, which is often the case in tide-dominated estuaries, then the system is referred to as hypersynchronous. The practical significance of hypersynchronous estuaries is that tidal currents actually reach their maximum speeds part way up the estuary; ultimately, however, tidal energy is diminished and both amplitude and speed approach zero in a region referred to as the tidal node.
Effects of tide on sediment dynamics The extension of tide into the lower course of a river has a substantial effect on both bed-load and suspended-load dynamics. Tides can affect sediment dynamics by: 1) producing bi-directional sediment transport, 2) creating mutually evasive transport pathways, 3) producing net landward bed-load transport, 4) prolonging flood-tide deposition of fine-grained sediments, 5) allowing formation of a tideinduced turbidity maximum, and 6) imparting a fortnightly rhythm that is important in the formation of fluid mud deposits. Although each of the above is controlled by tidal interactions, not all are equally well developed within a given estuary or from one estuary to another. The reversals in tide that produce bi-directional currents also produce patterns of bi-directional bed-load transport. Whereas much of the sediment that is transported and trapped in microtidal estuaries consists of organic-rich mud, many tide-dominated estuaries are characterized by sands derived from offshore and by bedforms that indicate high bed-load transport rates. Typical current speeds in tidedominated estuaries (1-3 m/s) are sufficient to keep sediments in motion for much of the tidal cycle, forcing channels to continuously adjust to erosion and deposition. This adjustment often leads to mutually-evasive currents and pathways of sediment transport, such that some channels may be dominated by ebb transport and others by flood transport. Tidal sand ridges that separate opposing transport pathways appear to have opposite-facing bedforms and complicated patterns of sediment movement (Swift, 1975; Twitchell, 1983). In response to a time-velocity asymmetry, there is net upstream sediment transport in most tide-dominated systems, with the largest bedforms migrating under the influence of flood currents. Examples of upstream sand transport include the Ord River estuary in northern Australia (Wright et al., 1973) and the Cobequid BaySalmon River estuary in eastern Canada (Dalrymple et al., 1990). Examples of net upstream transport of fine-grained sediments include the Gironde Estuary in France (Allen et al., 1980) and possibly the Severn Estuary in the United Kingdom (Murray and Hawkins, 1977). Net upstream transport continues (in the absence of densitydriven circulation) to the point where tide is damped out by frictional attenuation, and transport is controlled only by fluvial processes. In some estuaries, upstream flood transport of sand may be balanced by ebb transport in deeper channels where flow becomes concentrated during falling tide (Wright, 1977). Periods of slack water affect sedimentation by providing an opportunity for deposition of muds. Longer periods of slack water following flooding tides, as a result of tidal deformation, will favor deposition of sediment in the upstream reaches of an
184
J.T. WELLS
estuary. Over numerous tidal cycles processes of infilling are likely to be enhanced. The ability of ebbing tides to erode these deposits could be diminished by the fact that 1) there is a time delay between the moment at which flooding current of decreasing speed can no longer hold particles of a given size in suspension and the moment at which they reach the bottom and, 2) there is a difference between the maximum speed at which deposition of given particle sizes can occur and the minimum speed for the same material to be eroded by ebbing current. The above concepts, referred to as settling lag and scour lag, have been applied to tidal flat sediments since at least the 1950s (van Straaten and Kuenen, 1957; Postma, 1961).
Formation of a tidal turbidity maximum In terms of sediment dynamics, the most important feature in many estuaries is the broad zone of abnormally high suspended sediment concentrations known as the turbidity maximum. In partially-mixed estuaries, the turbidity maximum may affect rate of sediment accumulation on the bottom and serve as a distribution center for the rest of the estuary with respect to particle-reactive pollutants (Gelfenbaum, 1983; Wells and Kim, 1991). Although usually attributed to trapping of particles by the time-averaged (residual) flow at the landward limit of salt penetration (Schubel, 1971; Dobereiner and McManus, 1983; Nichols and Biggs, 1985) there is evidence that the turbidity maximum may also have a tidal origin (Swift and Pirie, 1970; Allen et al., 1980; Gelfenbaum, 1983; Dyer, 1986). Formation of a turbidity maximum in tide-dominated estuaries is governed by tidal resuspension in regions of high shear stress and by upstream trapping in the tidal node (Fig. 6-3). Strong bi-directional currents resuspend large amounts of fine-
SEAWARD
TURBID1T Y MAXIMUM
LANDWARD .
.
.:
TURBIDITY MAXIMUM
MAXIMUM
Fig. 6-3. Conceptual diagram showing the seasonal displacement of the Gironde Estuary turbidity maximum (modified from Allen et al., 1980). Whereas density processes appear to dominate the trapping at high river flow, tidal processes dominate at low river flow.
TIDE-DOMINATED ESTUARIES AND TIDAL RIVERS
185
grained sediment, including intertidal exchanges with large mudflats. Time-velocity asymmetry of tidal motion can lead to greater resuspension during flood than ebb, and to upstream transport. In hypersynchronous estuaries, the zone of maximum current (and sediment resuspension) is in the mid to upper reaches of the estuary. Resuspended sediment undergoes net upstream advection and, even in the absence of density circulation, can be trapped in the tidal node at the upstream limit of tides. Sediment concentrations in the turbidity maximum produced by these processes have been reported to reach 1-10 x lo3 mg/l (Allen et al., 1980) and are partly responsible for the formation of fluid mud deposits, discussed later in this chapter. Tidal turbidity maxima are transient features which can move along the axis of an estuary at semi-diurnal, fortnightly, and annual time scales (d'hglejan, 1981). Semi-diurnal excursions of 20 km in the Columbia Estuary and annual excursions of 40 km in the Gironde Estuary have been reported (Gelfenbaum, 1983; Allen et al., 1980). During periods of low river flow, a well-developed turbidity maximum may extend many kilometers upstream of the limit of salt intrusion. The development and dynamics of a turbidity maximum are complicated by the fact that most probably owe their origin to a combination of tidal transport and density circulation (Schubel, 1971).
MORPHOLOGIC AND SEDIMENTOLOGIC CHARACTER
Estuarine morphology The morphology of tide-dominated estuaries is characterized in plan view by a funnel-shaped geometry that has a high width-to-depth ratio (Fig. 6-4). The tidally-influenced distributary channels within the funnel usually have low sinuosity. Width of the estuarine section decreases rapidly upstream, often at an exponential rate. Extreme examples (such as the Fly in New Guinea: Harris et al., 1993), which are considered to be tidally-dominated end-members of a tripartite delta classification scheme (Miall, 1984), also represent estuarine end members with respect to morphology. Within and seaward of the estuarine funnel, bottom morphology consists of a wide variety of bedforms at scales ranging from centimeters to kilometers (Dalrymple et al., 1978; Harris, 1988). The location and distribution of bedform types reflects the patterns of sediment transport (Allen and Collinson, 1974). Tidal sand ridges with superimposed sand waves are perhaps the most pronounced morphologic features, and are considered to be diagnostic of tide-dominated estuaries. Formed by strong tidal currents, tidal sand ridges (also termed tidal current ridges, subtidal sand bars, and linear sand banks) are oriented approximately parallel to the axis of main tidal flow (Off, 1963; Kenyon et al., 1981; Belderson et al., 1982; Fig. 6-5). Morphologically and stratigraphically tidal sand ridges are marine sand bodies that replace the distributary mouth bar in fluvially-dominated systems. Bedforms in tide-dominated estuaries can reverse their direction of migration under alternating ebb and flood currents (Clifton, 1982). The frequency and magnitude
186
J.T. WELLS
Fig. 6-4. Two-dimensional morphologic variability in 8 tide-dominated estuaries. Most are funnel shaped, but entrance widths and rates of upstream width reduction vary greatly.
of reversals depends on bedform scale, strength of main and subordinate currents, and sediment texture and lithology. In the Severn Estuary, megaripple crests can be rebuilt within a tidal cycle, resulting in amplitude changes by as much as 2 m (Harris and Collins, 1985). However, larger bedforms such as sand waves are often stable, even at spring-neap tidal frequency. Long-term asymmetry of bedforms is produced and maintained by tidal flows in which one phase dominates over the other. The classic work by Off (1963), which delineated the largest scale of bedforms in coastal
TIDE-DOMINATED ESTUARIES AND TIDAL RIVERS
.
/:._..... . . ..:.. . .. . . .
a. -
187
TIDE- DOMINATED ESTUARIES
. . CHARACTERISTIC MORPHOLOGIC FEATURES 1
I
MORPHOLOGY
Receiving basin with coastal or open marine morphology formed by combined tides and waves 2 Tidal sand ridges elongated parallel to bidirectional currents with superimposed sand waves and ripples 3 Mutually-evasive channels with opposing transport directions 4 Funnel shape with large width-to-depth ratio 5 Bay-head delta occasionally formed when sediment input is high 6 Flanking wetlands dominated by marsh or mangrove vegetation with incised tidal channels 7 Meandering river channel near upstream end of funnel shape 8 Straight river channel within alluvial valley
Fig. 6-5. Diagram showing the morphologic features that are common to tide-dominated estuaries
embayments (including several tide-dominated estuaries), showed that some were of sufficient scale and stability to appear as regular features on bathymetric charts (length 8-65 km,height 8-30 m). Further details of bedform morphology and dynamics are given elsewhere in this volume (see Dalrymple and Rhodes, this volume). As one progresses up the estuary, it becomes apparent that the funnel shape gives way to an upstream pattern of sinuous channels (Woodroffe et al., 1989; Dalrymple et al., 1992; Fig. 6-5). The region of intense meandering usually occurs as the upper limit of tide influence is approached. Within the meander zone, the long-term transport by flood currents is balanced by river flow (Dalrymple et al., 1992). Active point bar migration is common in this section of the estuarine system, and meander scars indicate that high sinuosity was also present in the past (Wright et al., 1975; BartschWinkler and Schmoll, 1984). Farther upstream, beyond the limit of tidal influence, the channel again straightens thus completing the “straight-meandering-straight’’ morphologic model proposed by Woodroffe et al. (1989) and Dalrymple et al. (1992). In cross section, the estuarine funnel may show tidal ridges that have been exposed above mean sea level and colonized by vegetation to form linear islands (e.g. GangesBrahmaputra River, Coleman, 1969; Hang River, Coleman et al., 1970; and the Fly River, Harris et al., 1993). It is in these systems, which all require large sediment delivery, that the distinction between estuaries and deltas becomes blurred. If sufficient fluvial sediment is delivered, then in addition to tidal sand ridges in the funnel, deltaic deposits may form near the lower end of the fluvial system. Tidal flats, incised by tidal creeks and colonized by mangrove swamp or marsh grass, usually flank the main fluvial
188
J.T. WELLS
channel (Fig. 6-5). Further details of tidal flats (see Amos, this volume) and associated wetlands (see Luternauer et al.; Augustinus; this volume) are given elsewhere.
Estuarine sedimentology Tide-dominated estuaries are composed of sediments derived from both fluvial and offshore sources; the grain sizes range from gravel to mud. Sediments can be comprised of terrigenous materials (Allen, 1991), reworked glacial deposits (Amos, 1978), volcanic debris (Bartsch-Winkler, 1988), and shelf sands and muds (Collins, 1983). Whereas finer sands and muds are from fluvial sources, coarser sands are derived from the shelf or from erosion of the shoreline (Harris, 1988). Surficial sedimentary deposits reflect the interaction between fluvial processes, which decrease in strength seaward, and marine processes, which decrease in strength landward (Fig. 6-6). Although rigorous sediment budgets have yet to be determined for any of the tide-dominated estuaries, most appear to be sites of active deposition (Larsonneur, 1975; Parker and Kirby, 1982; Bartsch-Winkler and Ovenshine, 1984; Woodroffe et al., 1989; Dalrymple et al., 1990). Few sedimentary deposits are as well defined as those originating from tidal processes (Reineck and Singh, 1980; Nio and Yang, 1991). Estuarine tidal deposits include: channel sands (bedforms) that are characterized by cross bedding (Dalrymple, 1984), ripple laminations (Clifton, 1982), herringbone structures (Dalrymple et al., 1992), and reactivation surfaces (Klein, 1970; Visser, 1980); and, finer-grained
TIDE-DOMINATED ESTUARIES
\
1 Open shelf or gulf deposits, influenced by marine
2 Platform of sandy sediments transported from
.. . .. . .
3
:. . .
I
I
4
I
SEDIMENTOLOGY
I
5 6
7
offshore Tidal sand ridges aligned approximately with flow and characterized by cross bedding Time-varying fluid mud deposits, controlled by spring-neap tidal signal Tidal flats and associated marsh or mangrove vegetation Coarse to fine fluvlal or tidal sediments Point bar sands
Fig. 6-6. Diagram showing the sedimentologic features that are common to tide-dominated estuaries.
I
I
TIDE-DOMINATED ESTUARIES AND TIDAL RIVERS
189
flanking tidal flat deposits that are characterized by tidal bedding (Dalrymple and Makino, 1989), tidal rhythmites (Dalrymple et al., 1991; Martino and Sanderson, 1993), and flaser, wavy, and lenticular bedding (Reineck and Singh, 1980; Weimer et al., 1982). Although each estuary is sedimentologically unique, the above structures occur within a framework that shares the following attributes: 1) a broad range of sedimentary structures that may be interlayered in sharply contrasting strata, 2) a dominance of sand in the estuarine channel, 3) a dominance of silt and clay in adjacent tidal flat deposits, 4) a diminished or absent tripartite facies distribution that is typical in microtidal or wave dominated estuaries, and, 5 ) a coupling between high suspended sediment concentrations and fluid mud deposits. Tidal channel sands, the dominant facies, are coarsest near the mouth and near the head of the estuary (Fig. 6-6). According to Dalrymple et al. (1992), lateral shifting of channel bedforms produces an upward-fining sequence since currents are stronger at greater depths and weaker over bar crests. Subtidal sand ridges, capped by muds, form islands in the Fly, Klang, and parts of the Ord estuaries (Coleman et al., 1970; Wright, 1985; Harris et al., 1993). Some estuaries have broad sand flats that occur landward of the tidal sand ridges in a region of enhanced tidal flow (Dalrymple et al., 1992). The muddy lagoonal facies, typical of microtidal and wave-dominated estuaries (Nichols et al., 1991), is diminished or absent. However, mud drapes and fluid mud deposits are a common feature because tidal flow is variable; muds can accumulate over bedforms and in troughs during slack water or during weak flows. Moreover, it is now recognized that erosion and deposition of fine-grained sediments are not mutually exclusive processes, but can operate continuously in estuaries (Sanford and Halka, 1993). Finally, the tidal flat facies which forms the margins of the funnel and the sinuous upstream channel, represent a low-energy environment that is dominated by fine-grained sediments, incised by tidal creeks and migrating tidal channels, and capped by marsh, mangrove, or evaporite deposits (Fig. 6-6). Fluid-mud deposits Although the literature on tide-dominated estuaries is biased towards coarsegrained systems, many tide-dominated estuaries contain regions of gel-like fluid mud (e.g. Thames, Severn, Gironde, Loire, and parts of the Fly). An understanding of fluid mud distribution and dynamics is critical because of the well established connection between the transport of fine-grained sediments and the transport of toxic and non-toxic substances (Goldberg et al., 1977, 1978). Mechanisms that concentrate sediments as fluid mud deposits may also concentrate heavy metals, pesticides, and other pollutants that are adsorbed onto surfaces of the particles, at least until such time as they are released diagenetically or re-enter the system through resuspension (Wells, 1989). Often the same development pressures and activities that lead to higher levels of pollutants also lead to increased sediment loads. Fluid muds form in estuaries from high suspended sediment concentrations, usually in association with lutoclines or regions of sharp concentration gradients (Wolanski et al., 1988; Ross and Mehta, 1989). For convenience, fluid muds have been defined by sedimentologists on the basis of bulk density or by acoustic detectability
190
5.7: WELLS
CONCENTRATION, C (g/l 1
vEmcln,
u knlh)
Fig. 6-7. Diagram of instantaneous velocity and sediment concentration profiles (modified from Ross and Mehta, 1989) showing steps (layering) within the suspension. Mobile fluid muds can move freely throughout the tidal cycle, whereas stationary suspensions are most likely to be fully re-entrained only during spring tides. Gradual consolidation and incomplete re-entrainment during neap tides may allow an increment of the stationary suspension to become a permanent part of the bed, thus gradually infilling the estuary.
on echo sounders (Wells, 1983; Wells and Kemp, 1986). However, fluid muds occur in several different states of consolidation (Fig. 6-7) and, in fact, engineers base their definitions on rheological properties and processes (Ross et al., 1987). Based on decade-long measurements in the Severn Estuary and inner Bristol Channel, Kirby and Parker (1983) identified muds as forming settled mud deposits, stationary suspensions, and mobile suspensions, in accordance with their behavioral characteristics. Whereas settled mud forms the consolidated bed of the estuary and is resistant to erosion, stationary suspensions are deposits that have a lifespan measured in hours or days, and mobile suspensions are actually dense layers of moving mud. Kirby and Parker (1983) concluded that, when compared to other estuaries (Gallenne, 1974; Allen, 1990), fluid mud behavior in the Severn was typical rather than exceptional. In the Severn and Gironde estuaries, the formation of fluid mud is coupled to the fortnightly spring-neap tidal cycle (Allen et al., 1977; Kirby and Parker, 1983). As the range in tide gradually decreases from spring to neap, mud deposits that have formed from settling at slack tide become stationary and are able to resist resuspension fur-
TIDE-DOMINATED ESTUARIES AND TIDAL RIVERS
191
ther into the next tide. This happens because muds deposited at slack water have a greater opportunity to consolidate before currents reach strengths sufficient to erode them. Eventually, fluid mud deposits may be able to resist resuspension and remain on the bottom for an entire tidal cycle. If subsequent resuspension is not complete during the increase in tide range from neap to the next spring tide, then a layer of mud may remain on the bottom and become part of the sedimentary record. Laminations in settled muds which have been identified in cores are thought to represent gradual accumulation from the above process (Kirby and Parker, 1983; Dyer, 1986; Allen, 1990).
Estuarine infilling A fundamental problem in understanding estuarine infilling is that sediments do not simply “accumulate” on the bottom by slow, steady-state processes of deposition. Rather, they undergo exchange with the overlying water column (as described above for muds) on time scales which range from hours to years, and perhaps even decades. Details of how this cycling operates are important not only in addressing water quality issues (since particle-reactive pollutants adhere to fine-particle surfaces and share their immediate transport dynamics), but also in assessing rates of infilling. It is reasonable to expect that different estuaries will have different infilling histories. Inherent variations in sediment type, rate of delivery, subsidence and local sea level rise, estuarine depth and length of marine zone, and intensity of transport processes are so great that a single infilling model will not suffice. However, most tide-dominated estuaries will probably share several attributes that are common to the infilling process. First, infilling from a marine transgression will cause a landward shift in facies but will occur only if sediment supply exceeds the rate of local sea level rise. Second, in the absence of a marine transgression, a tide-dominated estuary will fill from both the seaward and landward ends. High rates of sediment delivery may result in progradation as a delta. Third, to a first approximation, the maximum thickness of estuarine fill is determined by the depth of the basin. Fourth, the sedimentary record may be incomplete as a result of erosional truncation and removal of fine-grained deposits under high energy conditions. Using Bristol Channel, Moreton Bay (Australia), and the Thames Estuary as examples, Harris (1988) has proposed a 3-stage infilling process that begins with deposition of intertidal flats and scour of offshore sediments. As infilling progresses, patches of sand are replaced by bedform sequences which evolve in a systematic manner to fill the estuary mouth. In the final stages, intertidal flats and sand banks merge to complete the replacement of water with sediment. A fining-upward vertical sequence is produced that includes scoured bedrock, gravel and sand lag deposits, coarse sand bank deposits that fine upwards to intertidal muds, and finally to supratidal deposits. Harris (1988) concludes that a transgressive vertical sequence is reversed and coarsens upwards. Similarly, Dalrymple et al.( 1992) has proposed an infilling sequence that, if complete, would fine upwards during transgression from fluvial sands and gravels to interbedded sands and muds of the inner estuary. An overall coarsening upwards section would then be produced with the deposition of elongate sand bars. Finally,
192
J.T WELLS
at maximum transgression, the estuary would infill from a progradational sequence (as has begun in the Cobequid Bay-Salmon River system) in which sand bars are overlain by sand flats, mud flats, and eventually capped by marsh deposits.
HOLOCENE EXAMPLES: TIDE-DOMINATED ESTUARIES
Gironde Estualy The Gironde Estuary is an excellent example of a mixed sand and mud system with a moderate tide range (average 4 m). As shown in Fig. 6-4, it has a long, relatively smooth funnel shape that tapers from a width of 15 km at its opening to only 1 km at a distance 100 km upstream, the limit of tidal influence. Approximately 3 x lo6 tons of fluvial sediments are delivered to the estuary per year, of which about 75% is silt and clay and 25% is sand. According to Castaing and Allen (1981), 60% of the suspended load leaves the estuary and accumulates on the shelf, but all of the bedload is retained. Ruch et al. (1993) found that 2-3% of the Gironde load moves across the shelf and down the Cap Ferret Canyon in a bottom nepheloid layer. One of the most significant features of the Gironde is the pronounced relationship between spring-neap tides (and river discharge), the turbidity maximum and fluid mud deposits, and the degree of mixing (well mixed at spring to partially mixed or well stratified at neap). Figure 6-3, which was based on a model for the Gironde, shows the highly transitory nature of the turbidity maximum (relative to river discharge) and the concentrated source of mud it provides. In addition to the seasonal suspended-sediment signal, lower current velocities and longer slack-water durations lead to net mud accumulation during neap tides and thus alternating sand and mud laminae in the estuarine deposits. The processes of sedimentation and resulting sedimentary facies of the Gironde Estuary have been described in considerable detail and are relatively well understood (Allen, 1971; Allen et al., 1977, 1980; Castaing and Allen, 1981; Nichols and Biggs, 1985; Allen, 1991). Grain size along the axis of the estuary decreases downstream to the lower funnel, then increases to the inlet opening. According to Allen (1991), the zone of grain size minimum corresponds to the downstream limit of fluvial sand, which displays a fining-seaward trend because of strong flood currents which impede downstream transport of coarser sediment. The central section of estuarine mud within the funnel is particularly characteristic of the wave-dominated estuaries described by Dalrymple et al. (1992); however, the transport and deposition of finegrained sediment within the Gironde are processes that are clearly controlled by tides (Allen et al., 1980), and this provides the rationale for classifying the Gironde as a tide-dominated estuary. The sedimentary facies of an axial stratigraphic column are shown in Fig. 6-8. In the fluvial section, the meandering channel produces point bars composed of sand and gravel, backed by levees and overbank deposits. The fluvial-to-tidal transition, located where upstream tide reversals first occur (at high river stage), marks a region where point bars become more estuarine in nature with crossbedded sand
193
TIDE-DOMINATED ESTUARIES AND TIDAL RIVERS TIDAL BARS fPROQRADINQ AND COALESCINQ
INLET SHOALS --
COARSE SIND AND GRAVEL WITH MIXED TIDE AND WAVE
k(FAc'Es I N LET
I
ESTUARINE
I MEANDERING CHANNELS
Fig. 6-8. Stratigraphic section of the Gironde Estuary showing transgressive ( T ) and regressive units in idealized form (modified from Allen, 1991). Although the sediments show a tripartite facies distribution similar to wave-dominated estuaries, the dominant processes are tidally forced.
and interbedded mud and clay flasers. Tidal estuarine point bars differ from fluvial point bars by their rhythmic sand and mud alternations, mud drapes and interbeds, and lack of overbank deposits (Allen, 1991). Within the estuarine funnel, tidal sand bars, prograding over estuarine muds (Fig. 6 4 , form a 6- to 7-m thick section (Allen, 1991). Their sedimentary structures are tidally produced. At the lower end of the estuarine funnel, coarse sand and gravel are derived from adjacent beaches and deposited in an inlet shoal (Fig. 6-8). Although the channel is scoured by tidal currents, sediments of the lower estuary are also reworked by storm waves.
Severn Estualy The Severn Estuary-Bristol Channel is a mixed sand, mud, and gravel system with a high tide range (8 m average, increasing to as much as 14 m on spring tides). It is an excellent example of a sediment-starved estuary, still in the earliest stages of infilling (Harris, 1988) with exposed bedrock covering extensive sections of the bottom (Fig. 6-9). The sediment load from fluvial sources is estimated to be only 1.6 x lo6 tons/yr (Collins, 1983). The estuarine system is funnel shaped with a major bend separating the seaward Bristol Channel from the interior Severn Estuary (Fig. 6-4). Along its lower course, depths exceed 30 m. The most characteristic feature of the Severn Estuary and Bristol Channel is the complexity of bedforms and facies distribution patterns (Fig. 6-9). The primary bedforms, which have a highly uneven distribution, are megaripples, sand waves, and sand ribbons (Harris and Collins, 1985). Whereas megaripples are common, sand waves are generally isolated at the mouth, and sand ribbons are patchy throughout.
194
J.T. WELLS
I
BRISTOL CHANNEL-SEVERN ESTUARY
COO'
3-06
Fig. 6-9. The distribution of sediments and bedforms in the Bristol Channel and Severn Estuary (modified from Harris and Collins, 1985). Arrows indicate locations of sand waves and their orientation, bars indicate sand ribbon locations and orientations, and stars represent regions of extensive megaripples.
Because of the complexity, patterns of sediment transport have been examined on the basis of heavy minerals (Barrie, 1980), foraminifera1 assemblages (Culver and Banner, 1979), bedform asymmetries (Kenyon and Stride, 1970), numerical models (Uncles, 1983), satellite imagery (Collins, 1983), and water circulation (Hamilton et al., 1980). These and other studies indicate that sand transport pathways are mutually evasive in a very broad sense, with ebb dominance carrying estuarine sands offshore in Bristol Channel and flood dominance transporting sediments onshore along the north and south margins. Divergence of bedload transport implies that the Celtic Sea is both a source and a sink. Suspended sediment is thought to have a net seaward movement, supplying offshore mud deposits (Collins, 1983). However, one of the interesting aspects of the sediment budget is the large amount of mobile sediment within the system at any time. Resuspension of bottom sediment may increase the mobile sediment population to values equivalent to 3-8 yr of fluvial input (Collins, 1983). Earlier contradictory models, which showed onshore transport of fine-grained sediments (e.g. Culver, 1980), are now taken as support of the hypothesis of upstream flow along the coastal margins. Within the Severn where much of the estuarine margin is comprised of marsh, recent studies have shown that vertical accretion rates may have increased over the last few hundred years and may be as rapid as 10-100 mm/yr (Allen and Rae, 1988).
Ord Estuary The Ord is an example of a sand-rich, moderate-tide-range (4.3 m) estuary that has generally been classified as a tide-dominated delta. The lower course of the river
TIDE-DOMINATED ESTUARIES AND TIDAL RIVERS
195
has a funnel shape that decreases in width from 9 km at the entrance to 90 m at a location 65 km upstream (Wright et al., 1975). Sediment discharge is approximately 22 x lo6 tons/yr and apparently includes an appreciable amount of bedload sand as well as large quantities of suspended silt and clay. As a tide-dominated estuary, the Ord is significant in three ways: it is very shallow (average depth 4 m); it accommodates tide as a standing wave (amplitudes decrease upstream; Fig. 6-2); and, there is extreme annual variability in river discharge (Wright et al., 1975). Much of what might be termed deltaic sedimentation in the Ord occurs as tidal flat deposits. Suspended sediments are spread across shoreline margins at high tide and, because of the arid climate, are capped by evaporite deposits (Thom et al., 1975). Mangroves fringe the distributaries. However, the primary morphologic features are the tidal sand ridges which choke the river channel and extend through the Cambridge Gulf north to the Timor Sea. These bedforms average 2 km in length, range in height from 10 to 22 m, and collectively account for 5 x lo6 m3 of sand (Wright et al., 1975). Smaller-scale bedforms are also present throughout the estuarine funnel. Whereas bedforms on the linear sand ridges have opposite migration directions on opposite sides of the ridges, the smallest scale of bedforms within the Ord channel reverse their direction of migration with tide. Figure 6-10 shows an idealized composite stratigraphic column for infilling of the Ord, based on a delta model proposed by Coleman (1976). The lowermost unit represents fine sediments that have escaped the estuary to accumulate in a marine shelf environment. Most of the sediments in the major sand unit are transported and deposited by tidal currents. Bi-directional cross bedding is common. The overbank splay deposits form within smaller channels that cut across the major distributary channels; they form deposits that accumulate in the adjacent basins. The tidal channel sands, which may scour into underlying deposits, are from infilling of meandering tidal channels. At the top of the sequence, tidal flats and evaporites contain alternating sand, silt, and clay laminations.
Cobequid Bay-Salmon River Estualy The Cobequid Bay-Salmon River Estuary is a relatively small sand-dominated system with an extreme tide range (average 12 m with maximum spring range 16 m). It is one of the most intensively-studied macrotidal estuaries in the world; previous research has involved flow dynamics (Greenberg and Amos, 1983), estuarine sedimentology (Dalrymple et al., 1990), and associated tidal flat deposition (Dalrymple et al., 1991). The estuary is funnel shaped but only 40 km long (Fig. 6-11). Because of the exceedingly small freshwater discharge relative to the tidal prism, the Cobequid Bay-Salmon River system is in many ways more of a macrotidal embayment than it is an estuary. The two most prominent features which characterize this estuary are 1) very well-defined headward transport of sand, and 2) very well-defined facies zones that can be explained in terms of physical processes. Observations by Dalrymple et al. (1990) show that sediment which is derived seaward of the estuary is accumulating faster than local sea level rise and that the system is therefore progradational. The
196
u)
J.T. WELLS COMPOSITE STRATIGRAPHIC SECTION: ORD
t
ALTERNATINO SAND SILT AND CLAY I BU1ROWINO AND Nub CRdCKS
NEDIUM SORTED SAND WITH SCOUR BASE ; BIDIRECTIONAL CROSS BEDDINO
ALTERNATINC SAND, SILT. AND CLAY; MUD CRACKS AND BURROWINO W l T W MUDS
ALTERNATINO SILTS AND CLAYS; SCATTERED SnELLSi MORE INTENSE BURROWINO NEAR BASE
THN OEDDED MARINE MUD WITH SILT AND SAND STRINQERS. HKIHLY BURROWED NEAR BASAL Z b E S
Fig. 6-10. Composite stratigraphic column of the Ord (modified from Coleman, 1976). Although produced as a hypothetical column for the Ord in the context of delta models, the salient features should remain in the model when the Ord is considered as a tide-dominated estuary.
facies zones include tidal sand bars in the lower half of the funnel, high energy sand flats with braided channels in the middle region, and a tidal-fluvial transition near the head of the estuary (Fig. 6-11). The tidal sand bars are composed of medium to coarse sand, whereas the inner two zones are composed of fine to very fine sand. The landward decrease in grain size, as in the Gironde Estuary, results from the inability of river flow to transport coarse bedload sediments against currents that are flood dominated. Extensive research on bedforms in the Cobequid Bay-Salmon River Estuary has been undertaken since the mid-1970s and is summarized in Dalrymple et al. (1990). Extensive marsh and tidal flat deposits form the shoreline margins in the middle and inner parts of the system (Fig. 6-11). These deposits record both diurnal and fortnightly inequality in the tides, and can actually be recognized in the inner estuary where the sedimentation rates are high and bioturbation is low (Dalrymple et al., 1991). Annual deposits can be identified in the middle estuary. Short-term
197
TIDE-DOMINATED ESTUARIES AND TIDAL RIVERS
I
I
COBEQ~IDBAY-SALMON RIVER ES~UARY
ELONGATE SAND
~
-
D S A N D FLATS
MIXED FLATS, MUD FLATS
-4515'
SALT MARSH
,
63-35, I
,
63VS I
I
Fig. 6-11. Facies distribution map of the Cobequid Bay-Salmon River Estuary (modified from Dalrymple et al., 1991). Because of the small freshwater discharge, the estuary has characteristics of a macrotidal embayment.
sedimentation rates vary from essentially zero at the mouth to 5 to 30 cm/yr in the middle section to more than 1m per year in the tidal-fluvial transition (Dalrymple et al., 1991). The Cobequid Bay-Salmon River Estuary owes its existence to marine transgression, yet the volume of sediment entering the system suggests it is regressive under present sedimentation and sea-level conditions. Infilling of the Cobequid BaySalmon River Estuary should produce, if a complete section were available, 20 m of fill. Dalrymple et al. (1990) suggest that the section would contain gravel lag at the base, followed by fining-upward sands from the sand bar to sand flat zones. At the top of the sequence mixed flats, mudflats, and marsh would complete the overall fining-upward sequence. Cross bedding in the lower sands of the tidal sand bars would give way to parallel-laminated fine sands of the sand flat and to tidal bedding in the upper tidal flat sediments.
HOLOCENE EXAMPLES: TIDAL RIVERS
Rio de la Plata The Rio de la Plata is an example of a tidal river in a microtidal setting (range <1 m). Located on the border between Uruguay and Argentina, the estuary is a large, funnel-shaped embayment that receives water from the Parana and Uruguay Rivers which, together, form the second largest drainage basin in South America (Depetris and Griffin, 1968). The average annual suspended sediment load from
198
5.7: WELLS
both rivers, 79.8 x l o 6 tons/yr, contains 75% coarse to medium silt, 15% fine to very fine silt, and 10% clay (Urien, 1972). Although small quantities of fine sand are apparently transported as bedload from the rivers, most of the estuarine sand is either relict or derived locally from erosion along the northern shore of the estuary. The Rio de la Plata is significant because of its size; it is so large that more than one tidal wavelength can be accommodated in the estuarine funnel (width 35-230 km; length 270 km). As tides enter the estuary, they become asymmetric with shorter, and presumably stronger, flows during the flood phase of the cycle. Tidal currents, which can reach speeds of 1.5 m/s, display a slight tendency towards counterclockwise circulation (Urien, 1972). Circulation patterns, together with bottom morphology, suggest mutually-evasive flow around several large sand banks that characterize parts of the interior of the estuary. As is typical of tidal rivers, penetration of salt in the Rio de la Plata is significantly limited by the influence of freshwater discharge; more than half the length of the estuarine funnel is fresh. Thus the trapping of suspended sediment particles at the landward limit of salt penetration, where elevated sediment concentrations have in fact been observed (Urien, 1972), is displaced considerably seawards of the uppermost region of tidal influence. Reversing tidal currents of sufficient strength to resuspend and transport bottom sediments are prevalent throughout the freshwater section of the estuary. Figure 6-12 shows the relationship between sediment distribution, salinity, and offshore bathymetry, as modified from Urien (1972). Much of the floor of the inner estuary is actually a subaqueous delta that extends from distributaries on the lower course of the Parana River. Here, in a fashion typical of delta front progradation, elongate distributary mouth bars of sand grade first into silts, then into clays in the intermediate and outer estuary. Ortiz Bank, English Bank and Plata Shoals are relict sandy features that were formed during the Holocene transgression, but are now in an estuarine environment dominated by silt and clay. These sandy features, especially in the outer estuary, are thought to have been produced by the advancing shoreline. The resulting pattern of facies development was an onlap marine facies of sand. According to Urien (1972), when sea level stabilized about 2,000 yr ago, the advancing marine environments were progressively replaced by an offlap fluvial facies that was silty. Future infilling of the Rio de la Plata will produce a coarsening-upward sequence of sediments in the inner and intermediate estuary (from subaqueous deltaic progradation in the narrow part of the funnel) and a fining-upward sequence in the outer estuary as fluvial muds are deposited over transgressive sands.
Amazon River The lower course of the Amazon River in Brazil is an example of a tidal river in a macrotidal setting (range about 4-8 m at the mouth). The Amazon is the largest river in the world in terms of both sediment discharge and drainage basin size (drainage basin approximately 2 x larger than Rio de la Plata (Depetris and Griffin, 1968), but discharge is 15x larger). Because of the large discharge (12 x 108 tons/yr; Meade et al., 1985), seawater never enters the river mouth, even during low river stage, and
TIDE-DOMINATED ESTUARIES AND TIDAL RIVERS
199
Fig. 6-12. The percentage of sand, salinity range, and offshore bathymetry of the Rio de la Plata Estuary (modified from Urien, 1972). Much of the estuarine funnel is freshwater with bottom sediments deposited from subaqueous delta progradation.
the Amazon thus serves as a classic example of a system where mixing and estuarine processes occur on the shelf. Because tide range is so large, the Amazon also serves as a good example of the extent to which tidal influence can be felt upstream (800 km), and the degree to which penetration of salt and tide can be decoupled from each other in large river systems. Most of the sediments transported by the Amazon River are silt- and clay-sized particles (85-95 %) and are carried in suspension. Suspended sediment concentrations are on the order of 100-400 mg/l, values that are typical of those on the Mississippi River. Sand and small amounts of fine gravel occur on the bed of the main stem; however, unlike many other large rivers, there is no significant decrease in grain size or increase in degree of sorting with distance along the river channel towards the mouth (Nordin et al., 1980). On the lower Amazon River, suspended sediments are stored during rising river stage and resuspended during falling stage, thereby damping out extremes in sediment discharge (Meade et al., 1985). This seasonal pattern probably overrides shorter-term tidal processes along a significant part of the lower reach of the river. Seaward of the river mouth where landward transport of water produces an estuarine-like circulation pattern, the sediments can be generally divided into wellsorted, negatively-skewed outer shelf sands, poorly-sorted, positively-skewed inner
200
J.T. WELLS
shelf muds, and a corridor extending from the river mouth across the inner shelf in which mud is interbedded with sand (Nittrouer et al., 1986). Accumulation rates on the shelf range from essentially zero to 10 cm/yr, depending on sediment supply and intensity of physical processes (Nittrouer et al., 1986). Roughly half of the sediment discharged by the Amazon accumulates on the shelf (Nittrouer et al., 1986). Far less is known about sediments and sedimentation processes near the river mouth than on the adjacent shelf. Figure 6-13 shows a region of modern sands that extends from the river mouth as a corridor of bedload transport, but also shows large shoals, presumably comprised of sand, where the distribution of sediments has not been mapped. Sands near the river mouth are very well sorted and are characterized by cross laminations, scour and fill structures, rip-up clasts, and convolute laminations (Kuehl et al., 1986). Tidal ellipses have a cross-shelf orientation and are strongest nearshore and near the river mouth, where tidally-induced shear stresses are almost certainly capable of producing all of these sedimentary structures.
SUMMARY
Physical and biological processes in nearly all estuaries are influenced by tides. The degree of influence is governed by estuarine morphology, tide range, water and sediment discharge, winds, and shelf processes. Tide dominated estuaries are those in which tidal currents play the dominant role in the fate of river-borne sediments, resulting in appreciable upstream transport of bedload sediment and, in extreme cases, little or no density-driven circulation. Tidal rivers, which have many of the same morphologic and sedimentologic features, are estuaries that occur in the lower reaches of large rivers where the penetration of tide extends farther than, and is decoupled from, the upstream penetration of salt. Here, subaqueous deltaic sedimentation is common. Most tide dominated estuaries and tidal rivers have a funnel shape, bidirectional sediment transport, mutually-evasive transport pathways, a tide- or density-induced turbidity maximum, and extensive regions of fine-grained sediment deposition, often in the form of fluid mud. Bottom sediments range from mud to gravel. As tides move upstream through smaller cross-sectional areas, the tidal currents become progressively more asymmetric in both speed and direction. In many cases, this leads to net landward transport of the bedload sediments. Characteristic bedforms include tidal sand ridges, large sand waves, and megaripples; characteristic sedimentary structures include cross bedding, tidal bedding, reactivation surfaces, and flaser, wavy, and lenticular bedding. Tidal flats, mangrove swamp, or marsh grass usually form the margins of the estuaries. Examples of tide-dominated estuaries and tidal rivers can be found in a wide variety of settings: for example, in the Rio de la Plata and Amazon tidal rivers, where respective tide ranges are less than 1 m and 4-8 m, much of the sedimentation occurs in the form of subaqueous deltas that have built over and around transgressive sands; in the Gironde, where tide range is range 4 m, a highly transitory turbidity maximum
TIDE-DOMINATED ESTUARIES AND TIDAL RIVERS
RIVER ZONE I SETTLING I
201
INTERACTION ZONES (COAGULATIONI
I HIGH DIATOM PROD.)
C UPWELLING 1 1 SUBJECT TO WIND I
(WAVE SHOALlNG/TIDAL STIRWNGI n
DIVERGENCE
CONMRQENCE
DIVERGENCE
20 krn
Fig. 6-13. Mouth of the Amazon River showing bathymetry and sedimentary facies in lower panel (modified from Kuehl et al., 1986) and hypothetical section across river mouth in upper panel (modified from Curtin and Legeckis, 1986). Because of the large volume of water discharge, mixing and estuarinelike circulation occur on the shelf within several discrete interaction zones. Lengths of arrows indicate relative velocity magnitude.
characterizes the estuary yet 60% of the suspended sediment leaves the estuary and accumulates on the shelf; in the Severn, where tide range is 8 m, sedimentation rates are so low that exposed bedrock covers extensive sections of the estuary; and, in the Cobequid Bay-Salmon River, where tide range can exceed 12 m, extensive progradation is occurring from sands derived seaward of the estuary.
J.T. WELLS REFERENCES Allen, G.P., 1971. Relationship between grain size parameter distribution and current patterns in the Gironde Estuary (France). J. Sediment. Petrol., 41: 74-88. Allen, G.P., 1991. Sedimentary processes and facies in the Gironde estuary: a recent model for macrotidal estuarine systems. In: D.G. Smith, G.E. Reinson, B.A. Zaitlin and R.A. Rahmani (Editors), Clastic Tidal Sedimentology. Can. SOC.Pet. Geol. Mem., 16: 29-39. Allen, G.P., Salomon, J.C., Bassoullet, P., Du Penhoat, Y. and De Grandpre, C., 1980. Effects of tides on mixing and suspended sediment transport in macrotidal estuaries. Sediment. Geol., 26: 69-90. Allen, G.P., Sauzay, G., Castaing, S.P. and Jouanneau, J.M., 1977. Xansport and deposition of suspended sediment in the Gironde estuary, France. In: M. Wiley, (Editor), Estuarine Processes, Vol. 2. Academic Press, New York, pp. 63-81. Allen, J.R.L, 1990. The Severn Estuary in southwest Britian: its retreat under marine transgression, and fine sediment regime. Sediment. Geol., 66: 13-28. Allen, J.R.L. and Collinson, J.D., 1974. The superimposition and classification of dunes formed by unidirectional aqueous flows. Sediment. Geol., 12: 169-178. Allen, J.R.L. and Rae, J.E., 1988. Vertical salt marsh accretion since the Roman Period in the Severn Estuary, southwest Britian. Mar. Geol., 83: 225-235. Amos, C.L., 1978. The post glacial evolution of the Minas Basin, N.S.: a sedimentological interpretation. J. Sediment. Petrol., 48: 965-982. Barrie, J.V., 1980. Heavy mineral distribution in bottom sediments of the Bristol Channel, U.K. Est. Coastal Mar. Sci., 11: 369-381. Bartsch-Winkler, S., 1988. Cycle of earthquake-induced aggradation and related tidal channel shifting, upper Turnagain Arm, Alaska, USA. Sedimentology, 35: 621-628. Bartsch-Winkler, S. and Ovenshine, A.T, 1984. Macrotidal subarctic environment of Turnagain and Knik Arms, Upper Cook Inlet, Alaska: sedimentology of the intertidal zone. J. Sediment. Petrol., 54: 1221-1238. Bartsch-Winkler, S. and Schmoll, H.R., 1984. Bedding types in Holocene tidal channel sequences, Knik Arm, Upper Cook Inlet, Alaska. J. Sediment. Petrol., 54: 1239-1250. Belderson, R.H., Johnson, M.A. and Kenyon, N.H., 1982. Bedforms. In: A.H. Stride (Editor), Offshore Tidal Sands. Chapman and Hall, London, pp. 27-57. Castaing, P. and Allen, G.P., 1981. Mechanisms controlling seaward escape of suspended sediment from the Gironde; a macrotidal estuary in France. Mar. Geol., 40: 101-118. Clifton, H.E., 1982. Estuarine deposits. In: PA. Scholle and D. Spearing (Editors), Sandstone Depositional Environments, AAPG, Tulsa, pp. 179-189. Coleman, J.M., 1969. Brahmaputra River: channel processes and sedimentation. Sediment. Geol., 3: 129-239. Coleman, J.M., 1976. Deltas: processes of deposition and models for exploration. Continuing Education Publ. Co., Inc., Champaign, IL., 102 pp. Coleman, J.M. and Wright, L.D., 1971. Analysis of major river systems and their deltas, procedures and rationale, with two examples. Louisiana State Univ., Coastal Studies Inst., Tech. Rep. 95, 125 pp. Coleman, J.M., Gagliano, S.M. and Smith W.G., 1970. Sedimentation in a Malaysian high tide tropical delta. In: J.P. Morgan (Editor), Deltaic Sedimentation. SEPM Spec. Publ. 15, Tulsa, pp. 185-197. Collins, M.B., 1983. Supply, distribution, and transport of suspended sediment in a macrotidal environment: Bristol Channel, U.K. Can. J. Fish. Aquat. Sci., 40 (suppl. 1): 44-59. Culver, S.J., 1980. Differential two-way sediment transport in the Bristol Channel and Severn Estuary, U.K. Mar. Geol., 34: M39-M43. Culver, S.J. and Banner, F.T., 1979. The significance of derived pre-Quaternary foraminifera in Holocene sediments of the north-central Bristol Channel. Mar. Geol., 29: 187-207. Curtin, TB. and Legeckis, R.V., 1986. Physical observations in the plume region of the Amazon River during peak discharge - I. Surface variability. Cont. Shelf Res., 6: 31-51. Dalrymple, R.W., 1984. Morphology and internal structure of sand waves in the Bay of Fundy. Sedimentology, 31: 365-382.
TIDE-DOMINATED ESTUARIES AND TIDAL RIVERS
203
Dalrymple, R.W. and Y. Makino, 1989. Description and genesis of tidal bedding in the Cobequid BaySalmon River Estuary, Bay of Fundy, Canada. In: A. Taira and F. Masuda (Editors), Sedimentary Facies in the Active Plate Margin. Terra Scientific Publishing, Tokyo, pp. 151-177. Dalrymple, R.W., Knight, R.J. and Lambiase, J.J., 1978. Bedforms and their hydraulic stability relationships in a tidal environment, Bay of Fundy, Canada. Nature, 275: 100-104. Dalrymple, R.W., Knight, R.J., Zaitlin, B.A. and Middleton, G.V., 1990. Dynamics and facies model of a macrotidal sand-bar complex, Cobequid Bay-Salmon River Estuary (Bay of Fundy). Sedimentology, 37: 577-612. Dalrymple, R.W., Makino, Y. and Zaitlin, B.A., 1991. Temporal and spatial patterns of rhythmite deposition on mud flats in the macrotidal Cobequid Bay-Salmon River estuary, Bay of Fundy, Canada. In: D.G. Smith, G.E. Reinson, B.A. Zaitlin and R.A. Rahmani (Editors), Clastic Tidal Sedimentology. Can. SOC.Petroleum Geol. Mem., 16: 137-160. Dalrymple, R.W., Zaitlin, B.A. and Boyd, R., 1992. Estuarine facies models: conceptual basis and stratigraphic implications. J. Sediment. Petrol., 62: 1130-1146. d’Anglejan, B., 1981. On the advection of turbidity in the Saint Lawrence middle estuary. Estuaries, 4: 2-15. Davies, J.H.L., 1964. A morphogenic approach to world shorelines. Z. Geomorph., 8: 127-142. Depetris, P.J. and Griffin, J.J., 1968. Suspended load in the Rio de la Plata drainage basin. Sedimentology, 11: 53-60. Dobereiner, C. and McManus, J., 1983. Turbidity maximum migration and harbor siltationin the Tay Estuary. Can. J. Fish. Aquat. Sci., 40 (suppl. 1): 117-129. Dolan, R., Hayden, B., Hornberger, G., Zieman, J. and Vincent, M., 1972. Classification of the coastal environments of the world. Part 1. The Americas. Univ. Virginia, Tech. Rep. 1 (Office of Naval Res.), 163 pp. Dyer, K.R., 1973. Estuaries: A Physical Introduction. Wiley, London, 140 pp. Dyer, K.R., 1986. Coastal and Estuarine Sediment Dynamics. John Wiley and Sons, New York, 342 pp. Gallenne, B., 1974. Study of fine material in suspension in the estuary of the Loire and its dynamic grading. Est. Coastal Mar. Sci., 2: 261-272. Galloway, W.E., 1975. Process framework for describing the morphologic and stratigraphic evolution of deltaic depositional systems. In: M.L. Broussard (Editor), Deltas: Models for Exploration. Houston Geol. SOC.,pp. 87-98. Gelfenbaum, G., 1983. Suspended-sediment response to semidiurnal and fortnightly tidal variations in a mesotidal estuary: Columbia River., U.S.A. Mar. Geol., 52: 39-57. Goldberg, E.D., Gamble, E., Griffin, J.J. and Koide, M., 1977. Pollution history of Narragansett Bay as recorded in its sediments. Est. Coastal Mar. Sci., 5: 549-561. Goldberg, E.D., Hodge, V, Koide, M., Griffin, J., Gamble, E., Bricker, O.P., Matisoff, G., Holdren, G.R., Jr. and Braun R., 1978. A pollution history of Chesapeake Bay. Geochim. Cosmochim. Acta, 42: 1413-1426. Greenberg, D.A. and Amos, C.L., 1983. Suspended sediment transport and deposition modeling in the Bay of Fundy, Nova Scotia - a region of potential tidal power development. Can. J. Fish. Aquat. Sci., 40 (suppl. 1): 20-34. Hamilton, D.J., Sommerville, J.J. and Stanford, P.N., 1980. Bottom currents and shelf sediments, southwest of Britian. Sediment. Geol., 26: 115-138. Harris, P.T., 1988. Large-scale bedforms as indicators of mutually evasive sand transport and the sequential infilling of wide-mouthed estuaries. Sediment. Geol., 57: 273-298. Harris, P.T. and Collins, M.B., 1985. Bedform distributions and sediment transport paths in the Bristol Channel and Severn Estuary, U.K. Mar. Geol., 62: 153-166. 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., 1975. Morphology of sand accumulations in estuaries. In: L.E. Cronin (Editor), Estuarine Research, Vol. 2. Academic Press, New York, pp. 3-22. Ippen, A.T. and Harleman, D.R.F., 1966. Tidal dynamics in estuaries. In: A.?: Ippen (Editor), Estuary and Coastline Hydrodynamics. McGraw-Hill Book Co., Inc., New York, pp. 493-522. Jevons, W.S., 1900. The Principles of Science. 2nd ed., The Macmillan Co., New York, pp. 673-674.
204
J.T. WELLS
Kenyon, N.H. and Stride, A.H., 1970. The tide-swept continental shelf sediments between the Shetland Isles and France. Sedimentology, 14: 159-173. Kenyon, N.H., Belderson, R.H., Stride, A.H. and Johnson, M.A., 1981. Offshore tidal sand banks as indicators of net sand transport and as potential deposits. Int. Assoc. Sedimentol., Spec. Publ., 5: 257-268. Kirby, R. and Parker, W.R., 1983. Distribution and behavior of tine sediment in the Severn Estuary and Inner Bristol Channel, U.K. Can. J. Fish. Aquat. Sci., 40 (suppl. 1): 83-95. Klein, G. D., 1970. Depositional and dispersal dynamics of intertidal sand bars. J. Sediment. Petrol., 40: 1095-1127. Kuehl, S.A., Nittrouer, C.A. and DeMaster, D.J., 1986. Distribution of sedimentary structures in the Amazon subaqueous delta. Cont. Shelf Res., 6: 311-336. Larsonneur, C., 1975. Tidal deposits, Mont Saint-Michel Bay, France. In: R.N. Ginsburg (Editor), Tidal Deposits: A Casebook of Recent Example and Fossil Counterparts. Springer-Verlag, pp. 21-30. Ludwick, J.C., 1974. Tidal currents and zig-zag shoals in a wide estuary entrance. Geol. SOC.Am. Bull., 85: 717-726. Ludwick, J.C., 1975. Tidal currents, sediment transport and sand banks in Chesapeake Bay entrance. In: L.E. Cronin (Editor), Estuarine Research, Vol. 2. Academic Press, New York, pp. 365-380. Martino, R.L. and Sanderson, D.D., 1993. Fourier and autocorrelation analysis of estuarine tidal rhythmites, Lower Breathitt Formation (Pennsylvanian), eastern Kentucky, USA. J. Sediment. Petrol., 63: 105-119. Meade, R.H., Dunne, T, Richey, J.E., Santos, U.de M. and Salati, E., 1985. Storage and remobilization of suspended sediment in the lower Amazon River of Brazil. Science, 228: 488-490. Miall, A.D., 1984. Deltas. In: R.G. Walker (Editor), Facies Models. Geol. Assoc. Can., pp. 106-118. Murray, J.W. and Hawkins, A.B., 1977. Sediment transport in the Severn Estuary during the past 8000-9000 years. J. Geol. SOC.London, 132: 385-398. Nichols, M.M. and Biggs, R.B., 1985. Estuaries. In: R.A. Davis, Jr. (Editor), Coastal Sedimentary Environments. Springer-Verlag, New York, pp. 77-186. Nichols, M.M., Johnson, G.H. and Peebles, P.C., 1991. Modern sediments and facies model for a microtidal coastal plain estuary, the James Estuary Virginia. J . Sediment. Petrol., 61: 883-899. Nio, S.D. and Yang, C.S., 1991. Diagnostic attributes of clastic tidal deposits: a review. In: D.G. Smith, G.E. Reinson, B.A. Zaitlin and R.A. Rahmani (Editors), Clastic Tidal Sedimentology. Can. SOC. Petrol. Geol. Mem., 16: 3-28. Nittrouer, C.A., Kuehl, S.A., DeMaster, D.J. and Kowsmann, R.O., 1986. The deltaic nature of Amazon shelf sedimentation. Geol. SOC.Am. Bull., 97: 444-458. Nordin, C.F., Jr., Meade, R.H., Curtis, W.F., Bosio, N.J. and Landim, P.M.B., 1980. Size distribution of Amazon River bed sediment. Nature, 286: 52-53. Off, T, 1963. Rhythmic linear sand bodies caused by tidal currents. AAPG Bull., 47: 324-341. Officer, C.B., 1976. Physical Oceanography of Estuaries (and Associated Coastal Waters). John Wiley, New York, 465 pp. . Parker, W.R. and Kirby, R., 1982. Sources and transport patterns of sediment in the inner Bristol Channel and Severn Estuary. In: Institute of Civil Eng., Severn Barrage, Thomas Telford, pp. 181-194. Postma, H., 1961. Transport and accumulation of suspended matter in the Dutch Wadden Sea. Neth. J. Sea Res., 1: 148-190. Reineck, H.E. and Singh, I.B., 1980. Depositional Sedimentary Environments. Springer-Verlag, New York, 549 pp. Ross, M.A., Lin, C.P., and Mehta, A.J., 1987. On the definition of fluid mud. In: R.M. Ragan (Editor), Proc. 1987 Nat. Conf. on Hydraulic Engineering., Am. SOC.Civil Eng., New York, pp. 231-236. Ross, M.A. and Mehta, A.J., 1989. On the mechanics of lutoclines and fluid mud. J. Coastal Res., SI5: 51-61. Ruch, P., Mirmand, M., Jouanneau, J.-M., and Latouch, C., 1993. Sediment budget and transfer of suspended sediment from the Gironde estuary to Cap Ferret Canyon. Mar. Geol., 111: 109-119. Sanford, L.P. and Halka, J.P., 1993. Assessing the paradigm of mutually exclusive erosion and deposition of mud, with examples from upper Chesapeake Bay. Mar. Geol., 114: 37-57.
TIDE-DOMINATED ESTUARIES AND TIDAL RIVERS
205
Schubel, J.R., 1971. Estuarine circulation and sedimentation. In: J.R. Schubel (Editor), The Estuarine Environment: Estuaries and Estuarine Sedimentation. Am. Geol. Inst., Short Course Lecture Notes, Washington, D.C. Schubel, J.R. and Pritchard, D.W., 1990. Great Lakes Estuaries-phooey. Estuaries, 13: 508-509. Shore Protection Manual, 1984. U.S. Army Corps of Engineers, Coastal Engineering Res. Center. Swift, D.J.P., 1975. Tidal sand ridges and shoal retreat massifs. Mar. Geol., 18: 105-134. Swift, D.J.P. and Pirie, R.G., 1970. Fine-sediment dispersal in the Gulf of San Miguel, western Gulf of Panama: a reconnaissance. J. Mar. Res., 28: 69-95. Thom, B.G., Wright, L.D. and Coleman, J.M., 1975. Mangrove ecology and deltaic estuarine geomorphology: Cambridge Gulf-Ord River delta, W.A. J. Ecol., 63: 203-232. Twichell, D.C., 1983. Bedform distribution and inferred sand transport on Georges Bank, United States Atlantic continental shelf. Sedimentology, 30: 695-710. Uncles, R.J., 1983. Modeling tidal stress, circulation, and mixing in the Bristol Channel as a prerequisite for ecosystem studies. Can. J. Fish. Aquatic Sci., 40 (suppl. 1): 8-19. Urien, C.M., 1972. Rio de la Plata Estuary Environments. Geol. SOC.Am. Mem., 133: 213-234. van Straaten, L.M.J.U. and Kuenen, Ph.H., 1957. Accumulation of fine grained sediments in the Dutch Wadden Sea. Geol. Mijnb., 19: 329-354. Visser, M.J., 1980. Neap-spring cycles reflected in Holocene subtidal large-scale bedform deposits: a preliminary note. Geology, 8: 543-546. Weimer, R.J., Howard, J.D., and Lindsay, D.R., 1982. Tidal flats and associated tidal channels. In: PA. Scholle and D. Spearing (Editors), Sandstone Depositional Environments. Am. Assoc. Pet. Geol., Tulsa, pp. 191-245. Wells, J.T, 1983. Dynamics of coastal fluid muds in low-, moderate-, and high-tide-range environments. Can. J. Fish. Aquat. Sci., 40 (supp. 1): 130-142. Wells, J.T., 1989. A scoping study of the distribution, composition, and dynamics of water-column and bottom sediments: Albemarle-Pamlico estuarine system. Albemarle-Pamlico Estuarine Study, Rep. No. 89-05, 39 pp. Wells, J.T. and Kemp, G.P., 1986. Interaction of surface waves and cohesive sediments: field observations and geologic significance. In: A.J. Mehta (Editor), Estuarine Cohesive Sediment Dynamics, Series on Coastal and Estuarine Studies. Springer-Verlag, New York, pp. 43-65. Wells, J.T. and Kim, S.Y., 1991. Trapping and escape of fine-grained sediments: Neuse River Estuary, N.C. Proc. Coastal Sediment. '91, ASCE, Seattle, pp. 775-788. Wells, J.T. and Park, Y.A., 1992. Observations on shelf and subtidal channel flow: implications of sediment dispersal seaward of the Keum River estuary, Korea. Est. Coastal Shelf Sci., 34: 365-379. Wolanski, E., Chappell, J., Ridd, P. and Verbessy, R., 1988. Fluidization of mud in estuaries. J. Geophys. Res., 93: 2351-2361. Woodroffe, C.D., Chappell, J., Thom, B.G. and Wallensky, E., 1989. Depositional model of a macrotidal estuary and floodplain, South Alligator River, Northern Australia. Sedimentology, 36: 737-756. Wright, L.D., 1977. Sediment transport and deposition at river mouths: a synthesis. Geol. SOC.Am. Bull., 88: 857-868. Wright, L.D., 1985. River Deltas. In: R.A. Davis, Jr. (Editor), Coastal Sedimentary Environments. Springer-Verlag, New York, pp. 1-76. Wright, L.D. and Coleman, J.M., 1973. Variations in morphology of major river deltas as functions of ocean wave and river discharge regimes. AAPG Bull., 57: 370-398. Wright, L.D., Coleman, J.M. and Thom, B.G., 1973. Processes of channel development in a high-tiderange environment: Cambridge Gulf-Ord River delta. J. Geol., 81: 15-41. Wright, L.D., Coleman, J.M. and Thom, B.G., 1975. Sediment transport and deposition in a macrotidal river channel, Ord River, Western Australia. In: L.E. Cronin (Editor), Estuarine Research, Vol. 2. Academic Press, New York, pp. 309-322.
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.
207
Chapter 7 DELTA FRONT ESTUARIES BRUCE S. HART
INTRODUCTION
Deltas are shoreline protuberances formed where river supplied sediment has accumulated in standing bodies of water faster than it can be redistributed by basinal processes such as waves, currents, tides or submarine failures (adapted from Elliot, 1986; and Wright, 1985). Deltaic environments and deltaic sediments are of considerable applied interest. Modern deltas are sites of much human activity. Large urban areas (including the cities of New Orleans and Cairo), productive agricultural lands and important transportation corridors are found in deltaic settings. Similarly, deltas can have substantial ecological importance as nurseries, migration routes and habitat for wildlife. Ancient deltaic sediments can be host to important accumulations of fossil fuels such as hydrocarbons and coal. As such, deltaic processes, morphology, sediments and stratigraphy have all formed the object of considerable study (for recent syntheses and examples see reviews by Wright, 1985; Elliot, 1986; and volumes edited by Whately and Pickering, 1989; Oti and Postma, 1995). Deltas form in a large number of settings and on a broad range of spatial scales. Small streams entering bays or lakes can form deltas a few square metres in breadth. Conversely, large rivers entering the sea have built subaerial delta plains many tens of thousands of kilometres in area. Just as small deltas can form as sub-environments of larger sedimentary systems (e.g., in lakes, bays, estuaries), larger deltas are composed of a myriad of smaller sub-environments. Many portions of deltas are affected by tidal dynamics and mixing of fresh and saline waters. These are the “delta front” estuaries which form the subject of this chapter. The character of the stratigraphic record produced by deltas has been dealt with previously by many authors (e.g., Coleman and Wright, 1975; Elliot, 1986; Postma, 1990) and will only be briefly touched upon here. Numerical treatments of sediment transport in estuarine settings form the subject of another chapter (see Dyer, this volume).
DELTA MORPHOLOGY AND GROWTH
The controls on delta form and size are many, and include characteristics of the source basin which affect sediment calibre and supply rate (catchment area, precipitation, source lithologies, etc.) and the characteristics of the receiving basin which affect the redistribution of that sediment (wave climate, tides, basin morphology, etc.). Following Elliot (1986), a conceptual framework (Fig. 7-1) can be established in which the fundamental controls on all these factors are climate (precipitation, tem-
208
B.S. HART
feeder+^^^^^^+
Receiving Basin
Tectonics Fig. 7- 1. Conceptual framework for understanding interactions of climate and tectonics in deltaic settings.
perature, etc.) tectonics (drainage basin relief, receiving basin geometry, subsidence) or the interaction of these two variables (e.g., lithology and climate of the hinterland determining sediment calibre, yield). Many attempts have been made to portray deltaic morphology andlor facies patterns as a function of the interplay of a few controlling variables. Perhaps the best known of these is Galloway’s (1975) ternary classification of delta morphology based on the relative importance of fluvial, tidal and wave influences (Fig. 7-2). This diagram has since been adapted to incorporate expected sand isopach patterns RIVER
WAVE
TIDE
Fig. 7-2. Ternary diagram showing classification of deltas with respect to relative importance of wave, river, and tidal energies. Location of some deltas shown for illustrative purposes. Adapted from Galloway (1975). Although attractive, this classification system cannot deal with factors such as relative sea level history, sediment grain size, or offshore slope.
DELTA FRONT ESTUARIES
209
(Wright, 1985) and expanded to classify all coastal environments (Boyd et al., 1992). Despite the apparent success and simplicity of this ternary classification, it is apparent that the three factors which form its basis can not explain all the variability to be found between deltas. Other factors, such as the character of the fluvial feeder system, sediment grain size, water depth of the receiving basin and relative sea level changes can also play important roles (e.g., Postma, 1990; Orton and Reading, 1993). The following two sections present a review of the characteristics of the alluvial feeder system and receiving basin which most influence delta morphology. Alluvial feeder systems The alluvial valley acts as a conduit which supplies most of the water and sediment (sourced in the drainage basin) to the delta. A complete discussion of the sedimentology and geomorphology of fluvial systems is beyond the scope of this chapter, however some generalities will be noted which are most of interest to deltaic studies. Various authors (e.g., Postma, 1990) have examined the way in which characteristics of alluvial feeder systems affect the morphology and sedimentary facies development of deltas. The size, stability and number of channels in the fluvial system determine whether sediment is supplied as a line source to the coast, or at a single point. Sediment yield of the drainage basin helps determine whether the delta will prograde, aggrade or be transgressed as a function of relative sea level change, and is itself a function of many variables, such as drainage basin size, relief, geology, and climate. Milliman and Meade (1983) reviewed the delivery of sediment to the oceans by rivers. They found that smaller rivers tend to have higher sediment yields per km2 of the drainage basin than larger rivers, and suggested that this reflects the inability of smaller drainage basins to store sediment. Some humid tropical deltas in areas of high relief can have enormous sediment discharges for the size of their drainage basin. For example, the Fly River has an estimated sediment discharge of 85 million tonnes per year, but drains an area of only 76,000 km2 (Harris et al., 1993). Deforestation of the drainage basin increases sediment yields, as do poor agricultural practices. Milliman and Meade (1983) suggested that poor agricultural practices in the drainage basins of the Ganges-Brahmaputra and Yellow rivers contribute to the enormous suspended sediment discharges (in excess of lo9 tonnedyear) of those two systems. Lakes along a river’s course act as sediment traps, decreasing sediment yield at the river mouth. Similarly, the construction of dams has led to major declines in sediment supply to some deltaic areas. For example, dam construction on the Nile and Ebro Rivers has cut off most of the fluvial sediment supply to the deltas (Milliman and Meade, 1983; GuillCn and Palanques, 1992), and in the case of Nile Delta has resulted in high rates of erosion at the delta front (e.g., Lotfy and Frihy, 1993). Dredging removes much of the sandy bedload from the lower part of the Fraser River as it crosses the delta (Milliman, 1980). There are few major river systems which have not been affected to some extent by mankind’s activities, and care must be taken when examining the relationships between current processes and existing delta morphologies.
210
B.S. HART
Sediment grain size characteristics affect channel morphology and river mouth processes. Like sediment yield, grain size characteristics are influenced by drainage basin physiography, relief, geology, and climate. According to Schumm (1977), the ratio of suspended load to bedload exerts a primary control on channel morphology. He related channel types to sediment transport processes, with “suspended load” channels (0-3% bedload) being meandering, “mixed load” (3-11% bedload) channels also typically meandering and “bed load” channels (> 11% bedload) being either straight or braided. Bridge (1993) however summarized existing literature on the relation between channel geometry, water flow and sediment transportldeposition in fluvial systems (with emphasis on braided systems) and concluded that the associations presented by Schumm (1977) are not universally valid. Fan deltas and braidplain deltas, composed of coarse-grained sediments, form where alluvial fans and braided channel systems with high bedload concentrations prograde into the sea along coasts adjacent to areas of high relief (e.g., Ethridge and Wescott, 1984; Galloway, 1976). Much sand is supplied by alluvial systems in high latitude, formerly glaciated areas with pronounced seasonal discharge variations (e.g., Fraser River, Milliman 1980; Outardes River, Hart and Long, 1990) and also by rivers draining arid tropical basins (e.g., Burdekin River, Coleman and Wright, 1975). In humid tropical areas, weathering rates are intense and large volumes of fine-grained sediments are supplied by alluvial systems (e.g., Harris et al., 1993; Staub and Esterle, 1993). Mud-dominated deltas have formed at the mouth of most long rivers (Amazon, Ganges-Brahmaputra, Mississippi) reflecting the combined effects of storage/deposition of coarse sediments within the alluvial channel systems and chemical and physical breakdown of sediment grains down the axis of the rivers. River discharge characteristics, including seasonality and absolute discharge, affect channel morphology (but see Bridge 1993), sedimentation rate, morphology of interchannel areas, and river mouth processes. Discharge characteristics are a function of many variables, but climate and drainage basin parameters such as length of the river system play major roles. Schumm (1977) suggested that the primary influence of river discharge is in controlling the size of stream channels, and the amplitude and wavelength of meanders (where present). Pronounced seasonality tends to prevent stable channel systems from developing. Smaller basins will tend to have greater seasonality, since tributaries in larger basins generally drain a variety of climatic and geomorphologic zones and hydrograph peaks from the tributaries often destructively interfere. Changes in discharge characteristics due to dam construction led to morphological changes in the estuary of the Outardes Delta (Hart and Long, 1990). Diversion of flow for irrigation reduces the discharge of some rivers; the Colorado River is a notable example.
Receiving basin characteristics The roles of tectonic and climatic variables are also manifest in the basins into which deltas have formed (Fig. 7-1). These variables control the size and morphology of the basin, and combine to produce the processes which redistribute sediment supplied by the alluvial systems.
DELTA FRONT ESTUARIES
211
The tectonic setting (e.g., active/passive margins, intracratonic basins) plays a fundamental role in determining morphological characteristics of the receiving basin such as depositional slope, subsidence rates and basin size and shape. Depositional slope influences progradation rates (for any given combination of sediment supply, relative sea level change and distributary network pattern (see below), progradation will be faster across flat platforms than across steeply dipping platforms), sediment dispersal processes related to slope instability on the subaqueous portions of the delta (e.g., Prior and Coleman, 1984; Hart et al., 1992) and exposure of the delta front to wave action. Factors such as the shape, orientation and size of the basin interact with the gravitational attraction of the sun and moon to determine the tidal regime of the basin. Microtidal regimes are those with tides less than 2 m (e.g., Nile, Mississippi, Mackenzie deltas), mesotidal regimes have tidal ranges between 2 and 4 m (e.g., Ganges-Brahmaputra, Mekong deltas) and macrotidal regimes are those with tidal ranges greater than 4 m (e.g., Amazon). Tidal currents in the receiving basin can be either perpendicular to the delta front, as in the case of deltas formed at the end of narrow basins (e.g., Tigris-Euphrates Delta of the Persian Gulf), or parallel to the delta front, as in the case of deltas which have grown along the margins of narrow straits (e.g., Fraser Delta; Hart et al., 1992). Changes in sea level have components related to tectonic and climatic controls. Eustatic (i.e. global) sea level changes reflect tectonic processes (e.g., mid-ocean ridge formation) operating on time scales of millions of years or more, and climatic changes (principally related to the waxing and waning of major ice sheets). Local subsidence or uplift combine with the eustatic sea level changes to produce relative sea level changes. The Mississippi Delta has grown on a passive margin, with compaction of the thick (several km) underlying sedimentary deposits accounting for approximately 90% of the observed 1-2 cm/yr relative sea level rise in that area (Ramsay et al., 1991). In formerly glaciated regions, isostatic uplift of the crust has produced deltas which have grown during conditions of falling relative sea level (e.g., Outardes Delta; Hart and Long, 1990). Dominguez et al. (1987) suggested that short term (200-300 years) small-scale (2-3 m) oscillations of sea level have profoundly affected the morphology and stratigraphy of the S5o Francisco Delta. The growth of sequence stratigraphic modeling and concepts has prompted a renewed interest in how eustatic sea level fluctuations might affect fluvial processes in the drainage basin (e.g., Schumm, 1993).
DELTAIC ENVIRONMENTS
Deltas can be divided into a number of discrete physiographic zones shown schematically in Fig. 7-3. As discussed below, the size of each zone and the type of geomorphologic element present within each can vary significantly from delta to delta. Below the alluvial feeder system, the channel enters the delta plain, comprised of the subaerial and intertidal portion of the delta. A wide variety of morphological components can be present here, with feeder system character, relative sea level history, climate and other factors playing important roles. In a general sense, the
212
B.S. HART
C-
Alluvial
ProdeltdDelta Slope
Fig. 7-3. Physiographic zones of deltas (after Coleman and Wright, 1975).
delta plain consists of channels flanked by low-lying interchannel areas comprised of forests, swamps, bays, lakes, salt flats, beach ridges or other morphological features. Commonly, two sub-divisions are made: a) the upper delta plain which is the older portion of the delta at some distance from the coast and is beyond the limit of tidal inundation or other type of marine incursion (e.g., storm surges), and b) the lower delta plain which is directly affected by tidal inundation or other marine incursion. The delta front is that area seaward of the delta plain where fluvial and basinal processes interact, and much of the distributaries’ coarse sediment load is deposited. Further offshore is the prodelta area, dominated by fine-grained suspended sediment deposition. A delta slope environment, a relatively steeply dipping portion of the delta below the influence of most surface gravity waves (i.e. below the delta front), may be present between the delta front and prodelta where deltas are prograding into deep water (e.g., Fraser Delta, Hart et al., 1992). This section will examine the estuarine processes, morphology and sedimentology of channels, river mouths and interchannel areas of deltas.
Channels Two principal types of channels can be recognized, i) distributary channels which transport the fluvial discharge seaward, and ii) tidal channels which drain tidally inundated areas of the lower delta plain. Deltaic distributary channels resemble alluvial channels in many respects, but are affected by periodic stage fluctuations associated with tides or other sea level changes (e.g., storm surges) in the receiving basin and, in places, flow reversals. Coleman and Wright (1975) suggested that 3 main types of distributary channel networks are observed on delta plains and related these distributary patterns to subsidence rates, offshore slopes, wave and tidal action, and sediment calibre. The first type consists of bifurcating channel systems (Fig. 7-4a) such as those found on the Mississippi delta. The river begins to bifurcate as it enters the delta plain, leading to the development of many river mouths (although the discharges carried by each are not equal). On that delta, active distributaries range
DELTA FRONT ESTUARIES
213
Fig. 7-4. End-member distributary channel patterns (after Coleman and Wright, 1975).
from 2-3 m wide and 1 m deep to 1 km wide and 30 m deep (Coleman, 1981). The Mahakam Delta also is of this type (Allen et al., 1979). There, the channel bifurcates upon entering the delta plain and then each distributary subsequently bifurcates approximately every 5-10 km further downstream, producing about 15 river mouths. In the Mahakam Delta, the Niger Delta (Allen, 1970) and the Rajang Delta (Staub and Esterle, 1993), channels tend to shallow downstream as they bifurcate, reflecting the decrease in discharge carried by any one distributary channel segment. Rejoining channel systems (Fig. 7-4b) show complicated patterns of bifurcation and rejoining of channel segments. Fewer active river mouths are present in these systems. Coleman and Wright (1975) suggested that the GangedBrahmaputra Deltas shows this type of distributary pattern. The main channel there begins to bifurcate about 160 km downstream from the confluence of the Ganges and Brahmaputra rivers, with subsequent bifurcations and rejoinings leading to a series of more or less minor, shallow channels separated by migrating sand bars (Coleman 1969). Barua (1990) studied the distributary channels of this delta and found that flood and ebbdominated channels were present, with aspect ratios (depth/width) less than 1 being typical of flood dominated channels and aspect ratios greater than 1 being associated with ebb-dominated channels. Finally, single channel systems (Fig. 7-4c) have one or a few distributaries which originate from a single point near the apex of the delta. The SCo Francisco (Dominguez et al., 1987) and Fraser deltas have channel networks of this type. The fluvial discharge is typically not spread evenly among the distributaries. For example, in the Fraser River, the Main Arm transports about 80% of the river’s discharge (Milliman, 1980). River planforms can be quite variable. Meandering channel forms are typical of distributary systems having high concentrations of fine-grained suspended sediment and more or less constant fluvial discharges. The distributaries of the Niger River are
214
B.S. HART
meandering (Allen, 1970) as were the principal distributaries of the Nile prior to river modification (Sestini, 1989). Tornqvist et al. (1993) used shallow borings to suggest that accelerated rates of base level rise caused the distributary system of the Middle Holocene Rhine-Meuse delta to change from a meandering to an anastomosing channel pattern. The Mississippi Delta, despite the high suspended sediment load of the river, is characterized by high sinuosity channels which become straighter towards the river mouths. In the Mahakam distributary system, the thalwegs of the channels migrate within the bounds of straight channel systems (Allen et al., 1979). Where fine-grained suspended sediment levels are high, levees develop along channels during overbank flooding events, most often in response to river flood discharges. Flow expansion outside of the confines of the channel leads to a decrease in competence and sediment deposition directly adjacent to the channel. The height of the levees is not greater than the maximum flood stage, and Coleman (1981) noted that the height of the levees decreased towards the river mouths. Levees up to 2 m in height are found on the Mississippi Delta (Coleman, 1981) and Niger Delta (Allen, 1970), and levees are present on the Nile (Sestini, 1989) and Orinoco (van Andel, 1967) delta plains. No levees are present along the distributaries of the Mahakam Delta, apparently because the relative importance of fluvial processes is reduced in this delta (Allen et al., 1979). Rivers with flashy discharges tend to have wide, shallow, braided distributary systems. The Burdekin River of Australia drains an arid area which supplies abundant coarse-grained sediments to the delta. Channels bifurcate, are separated by shoals and are choked by sandy sediments (Coleman and Wright, 1975). These characteristics are similar to those of sub-boreal deltaic estuaries of the north shore of the St. Lawrence where highest discharges occur during snowmelt derived freshets (Hart and Long, 1990). Lateral migration of these channels is the rule, and levees do not develop. Sediment within distributaries is transported either in suspension (very fine sand to clay) or as bedload (fine sand and coarser). Quantification of the former tends to be simpler, as calculation of suspended sediment fluxes is easier than measurement of bedload transport rates (particularly in tidal settings). Suspended sediment load carried at any given moment by a particular river is generally a function of variations in rainfall or snowmelt, although in some settings events such as landslides may contribute to pulses of sediment transport (e.g., Harris et al., 1993). Milliman (1980) showed that in the Fraser River, a temporal lag or hysteresis was present between the times of peak suspended sediment discharge (rising limb of hydrograph) and peak water discharge. Absolute differences in suspended loads between river systems reflect drainage basin characteristics such as lithology and climate. Suspended sediment profiles in distributaries show a great degree of temporal and spatial variability. In salt wedge estuaries a turbidity maximum can develop, such as in the Handil Distributary of the Mahakam Delta (Allen et al., 1979). In the Main Arm of the Fraser River, a salt wedge is present at higher discharges, but the flow becomes moderately stratified at low discharges according to Kostaschuk et al. (1992). Those authors attributed a seaward decrease in the suspended sediment concentrations of the upper layer to deposition (due to flocculation and a decrease
DELTA FRONT ESTUARIES
215
in turbulence as the flow accelerates over the salt wedge and becomes stratified) and dilution of the turbid river flow by entrainment of the underlying cleaner marine waters. Staub and Esterle (1993) noted that the distributary mouths of the Rajang River include salt wedge, partially mixed and fully mixed types. They too noted that the suspended sediment concentrations within the distributaries tended to decrease in a seaward direction, and suggested that dilution of turbid river waters by “black water discharge” from interdistributary mangrove swamps was primarily responsible. In these distributaries, a vertical stratification was present, with the less turbid black water at the surface; this stratification was thought to cause sediment by-passing of the delta plain, and thus favoured peat formation. In some arctic deltas, the occurrence of peak fluvial discharges before shore ice break-up can lead to the phenomenon known as overflow, where highly turbid waters spill over the ice and suspended sediments are subsequently ice rafted (e.g., Naidu and Mowatt, 1975). Transport of bedload sediments is, in general, less well documented. Although bedload discharge of rivers is commonly ignored or considered minor with respect to the suspended load (e.g., Milliman and Meade, 1983), sands and gravels can form an important and even dominant proportion of the total sediment discharge on fan deltas and braidplain deltas where the delta is in close proximity to orogenic belts or glacial deposits (e.g., Burdekin Delta, Coleman and Wright, 1975; Copper Delta, Galloway, 1976; Fraser Delta, Milliman, 1980). Bedload sediments are typically sandy, but gravels may be present in some cases (e.g., Ethridge and Wescott, 1984). Bedforms of various sizes and types have been reported from distributary channels (e.g., Allen et al., 1979; Guillen and Palanques, 1992; Hart and Long, 1990). Allen et al. (1979) and Guillkn and Palanques (1992) noted a downstream decrease in bedform size in their study areas. Kostaschuk et al. (1989) used bathymetric profiling to monitor bedform geometry, dimensions and movement in the Main Arm of the Fraser River, and derived estimates of bedload transport from their observations. Bedload transport can only continue downstream to the limit of the intrusion of the salt wedge. In some rivers the salt wedge can be completely pushed out of the lower courses of the estuary during peak fluvial discharges and at such times sand can be discharged into the sea. Tidally generated bedforms are present near the mouths of some distributaries (e.g., Burdekin Delta, Coleman and Wright, 1975; Outardes Delta, Hart and Long, 1990), and a fuller discussion of river mouth processes will be presented later. Channel switching is well documented in deltaic settings. Historical records of channel switching or migration have been presented for the GangedBrahmaputra Delta (Coleman, 1969), Nile Delta (Sestini, 1989), Yellow River Delta (Xue, 1993) and Fraser Delta (Clague et al., 1983; Fig. 7-5), to name a few examples. Channel switching/avulsion is typical of deltaic settings of mud-dominated systems. There, distributaries lengthen until they are no longer hydraulically efficient, and the main flow diverts itself to another, shorter course. Coleman and Wright (1975) presented three types of channel switching, and related each to factors such as distributary channel type, shelf slope, and wave and tidal energies. Suter and Berryhill (1985) suggested that channel switching of late Quaternary shelf margin deltas of the Gulf of Mexico was sometimes the result of diapiric movements of underlying salts,
216
B.S. HART
Fig. 7-5. Historical changes in position of the Main Arm of the Fraser River prior to jetty construction in the early/mid-1900s. Note that some former positions of the Main Arm are currently occupied by tidal channels which cross the broad intertidal flats. From Clague et al. (1983).
and tectonic movements are known to have led to channel migration in other settings. The relative channel stability of fine-sediment dominated systems leads to the development of linear, laterally discontinuous sandbodies. In the Mississippi Delta however, distributaries, once abandoned, are filled with fine-grained sediments deposited during overbank flooding (Coleman, 1981). Channel migration is more typical of sand-dominated distributary systems with erratic fluvial discharges. Sandbodies tend to have higher widthldepth ratios. Monahan et al. (1993) showed that migration of the distributaries of the Fraser Delta (principally on the lower delta plain) has led to the generation of a nearly continuous sheet sand beneath the delta plain. In this case, the upper part of the delta front can be removed from the stratigraphic record during progradation. Allen et al. (1979) reported both channel migration and avulsive behaviour for distributaries of the Mahakam Delta plain. Tidal channels drain portions of delta plains which are inundated by tides (lower delta plain). Water surface gradients are commonly negligible in these channels, but an upstream shallowing is generally present (e.g., Coleman and Wright, 1977). In tropical settings, mangrove swamps or peat bogs of the lower delta plain tend to be drained by highly meandering tidal channels which have depths ranging from a few metres up to 30 m (e.g., Allen et al., 1979; Allen, 1970; Staub and Esterle, 1993). In such heavily vegetated areas, channel positions tend to be relatively stable. Elsewhere, in places where the delta plain is not vegetated, or the substrates are sandier, channels are shallower and lateral migration of tidal channels is facilitated. Tidal creeks across tidal flats are discussed in the chapter by Amos (this volume). Bedload sediment transport in tidal channels is bi-directional, and bedforms within the channels may change directions with the ebb and flood of the tides. In some
DELTA FRONT ESTUARIES
217
environments, such as the Mahakam Delta (Allen et al., 1979), only muds are deposited, except near the mouth of the tidal channel where basinal processes lead to an influx of sands and shell material. Black water discharges, devoid of significant concentrations of suspended sediment, are typically reported from tidal creeks draining mangrove swamps and peat bogs (e.g., Staub and Esterle, 1993). Oomkens (1974) suggested that shifting of the tidal inlets of the Niger Delta in response to longshore currents has led the Late Quaternary portion of the section to be dominated by tidal channel sands. On most deltas which enter the ocean, distributary channels are affected to some degree by tidal/basinal processes, and a spectrum exists between “pure” distributary and tidal channels. In general, the relative importance of tidal processes increases in a channel as the distributary is abandoned. On the lower Fraser Delta plain, it can be shown that some tidal creeks occupy previous distributary channel positions (Fig. 7-5) and it is possible that some tidal creeks may have a similar origin in other deltaic settings.
River mouths River mouths are the place where the degree of interaction between basinal and fluvial processes is greatest. Flow expansion as the river discharge leaves the confines of the distributary channel leads to loss of sediment transport competence and deposition. Wright (1977; 1985) presented a framework discussing the interaction between form and process in this setting which has served as a basis for much further work. The variables which control river mouth morphology are approximately the same as those which control the overall form of the delta, and they include: i) density contrasts between the river flow and the water of the receiving basin, ii) relative strengths of fluvial discharge and tidal flows, tidal range and the orientation of the tidal currents with respect to the shoreline, iii) wave power, and orientation of the wave attack with respect to the shoreline, and iv) depth and morphology of the river and delta front or delta slope. Orton and Reading (1993) argued that the grain size of the sediments delivered by the fluvial system also plays an important role. Where rivers enter marine basins, the density of the fresh waters is almost always less than that of the saline receiving basin, regardless of temperature or sediment load (but see Wright et al., 1986). When the river mouth is shallow, such as at the mouths of rivers with erratic discharges and/or high bedloads, friction between the fluvial discharge and the sediment/water interface causes rapid flow expansion and deceleration. Sediment is deposited rapidly and “middle ground bars” are formed (Wright, 1977; Fig. 7-6a). The mouths of some distributaries of the Mississippi which have built out into shallow interdistributary bays (e.g., Atchafalaya distributary) are dominated by friction effects, and middle ground bars have been recognized on many other deltas (e.g., Outardes Delta Estuary, Hart and Long, 1990). In the simplest case, when the velocity of the river discharge is relatively high and the river mouth area is sufficiently deep, a buoyant effluent develops (Wright, 1977). Here, vertical stratification between the effluent and the ambient sea waters is pronounced, the river discharge does not directly “feel” the sediment water interface
218
B.S. HART
Coarsest Sand
Sand and Mud
_____
isobaths
J waterlevel
Fig. 7-6. Examples of river mouth bar types. (A) Middle ground bar typical of friction dominated river mouths; (B) simple subaqueous bar formed at buoyancy dominated river mouths; ( C ) strong waves of shore normal incidence cause mouth bar to be arcuate. Shoaling waves drive sand back toward beach as swash bars. River mouth can be deflected alongshore where waves approach the shoreline obliquely. Adapted from Wright (1977).
219
DELTA FRONT ESTUARIES
and the fresh water spreads out as a buoyant surface plume. Plumes are prominent on the Mississippi and Fraser Deltas, and sedimentation from a plume has been modeled by Sparks et al. (1991). The “type” morphology produced at this type of river mouth (but not present at the Fraser Delta, see below) consists of a single subaqueous river mouth bar (Fig. 7-6b). The mouths of the “birdfoot” delta of the Mississippi are of this type (Wright and Coleman, 1974). Strong waves arriving at the river mouth can augment deposition by breaking down vertical stratification and slowing down the effluent’s outward movement (Wright, 1977; Wright et al., 1986). In shallow water, shoaling waves can redirect sands landward as swash bars (Fig. 7-6c), and when the angle of incidence of the approaching waves is high, the longshore sediment transport can divert the location of the river mouth. The Senegal Delta (Coleman and Wright, 1975) has had its river mouth deflected several tens of kilometres alongshore in this way. Once a river mouth is abandoned, the sandy sediments deposited there become available for reworking by waves. Barrier island complexes can be constructed this way, and the formation of such complexes has been noted on the Niger (Allen, 1970; Oomkens, 1974), Mississippi (Penland et al., 1988; Fig. 7-7) and Copper (Galloway, 1976) Deltas. These barrier complexes protect the delta front landward of the islands from the effects of wave attack, permitting the development of quiet bays or swamp areas on high energy coasts.
TRANSGRESSIVE MISSISSIPPI DELTA BARRIER MODEL
r
REGRESSIVE ENVIRONMENTS
<
a Fresh Marsh
(i)
Distributary
f l Beach Ridge
$
tDrdlstrlbvtarY
TRANSGRESSIVE ENVIRONMENTS
A
REOCCUPATION
SUBMERGENCE
w
0Subaerial
Barrier Sands Subqqueous Barrier Sands U Sand Sheet Salt Marsh --- Recurved Spit ;k Shell Reef t Tidal Channel
Fig. 7-7. Schematic representation of the growth and transgression of barrier island complexes on the Mississippi Delta. From Penland et al. (1988), reproduced with permission of SEPM (Society for Sedimentary Geology).
220
B.S. HART
a 142'E t
Kerema
Gulf of Papua
++ +
+ .o
,b TIDE DOMINATED ESTUARY
WAVE D O M I N A T E D DELTA
Island Stabilised by Vegetation
Pleistocene
Fig. 7-8. a. Map view showing elongate tidal ridges of the Fly Delta estuary. Note classic funnel shape. b. Cross-section (location shown in Fig. 7-8a) shows some tidal ridges are cored by resistant, earlier deposits. From Harris et al. (1993), reproduced with permission of Pergammon Press Ltd.
Where tidal currents are high, the river mouth flares and assumes the funnel shape characteristic of these environments (Fig. 7-2). The Fly Delta (Harris et al., 1993) and Ganges-Brahmaputra Delta (Barua, 1990) are examples of river mouths subjected to strong tidal flows. Elongate tidal ridges form in the river mouth, and sandy bedforms can be present in channel axes (Fig. 7-8). Harris et al. (1993) noted however that fluid muds could be present near the bottom of some tidal channels of the Fly Delta. There, some of the subtidal sand banks become stabilized through trapping of muds by vegetation (seagrasses, mangroves, etc.) and eventually become emergent. According to Harris et al. (1993), these islands (some "anchored" by resistant pre-
DELTA FRONT ESTUARIES
221
Holocene deposits) are prone to rapid lateral migration. Coleman (1969) similarly noted the tendency of tidal ridges to shift in the mouth of the Ganges-Brahmaputra River. Inhabited low-lying islands of this delta are sometimes inundated by river floods or storm surges, and much loss of life has occurred. Sandy deltas which form in areas of significant tidal range typically have large sandy shoals which become exposed at low tide. Vegetation is not able to colonize and stabilize these banks. Examples include the Burdekin (Coleman and Wright, 1975), Outardes (Hart and Long, 1990) and Copper (Galloway, 1976) deltas. Within any one large delta, variations in the importance of river discharge, wave action and tides can vary between distributary mouths. For example, in the Mahakam Delta, river dominated and tide dominated river mouths can be recognized, with other river mouths showing combinations of effects (Allen et al., 1979). Furthermore, temporal changes at any one river mouth are to be expected, at least on longer time periods (hundreds, thousands of years), as the importance of the associated distributary waxes and wanes. Finally, where deltas have formed into deep water, mouth bar formation may be inhibited by the initiation and growth of submarine channel systems. On the Fraser Delta, which is prograding into water over 300 m deep and has gradients in excess of 6” on the upper delta slope, a major submarine channel system heads at the river mouth where the development of a mouth bar might be expected (Hart et al., 1992). Repeated submarine failures at the river mouth have been documented there using comparative bathymetric surveying by McKenna et al. (1992).
Interchannel areas Estuarine environments of the lower delta plain include those interchannel areas which are inundated at times by marine waters. These can include tidal flats, lagoons, bays and swamps or marshes, and details of the sedimentology and geomorphology of these environments are discussed in other chapters. Even in areas of very low tidal range such as the Ebro Delta, tides and minor sea level changes due to atmospheric perturbations can lead to inundation of low-lying areas by marine waters (Maldonado, 1975). One of the most characteristic morphological components of deltaic systems are the crevasse splays which form at high river stages when levee breaching occurs (Fig. 7-9). Sediment-laden river waters spread out into the interdistributary areas forming “microdeltas” of relatively coarse-grained sediment. In the Mississippi Delta, these splay deposits can cover areas up to 160 km2, with thicknesses of up to 15 m (Coleman and Wright, 1975). In areas of high subsidence, growth of the crevasse splay creates land areas which become inundated (through compaction and other causes of subsidence) following healing of the levee breach (Fig. 7-10). Away from the crevasse splays, sedimentation is dominated by deposition of fine-grained sediments from suspension during river floods, high tides or storm surges. In humid areas, swamps can develop, leading to thick accumulations of peat (e.g., Staub and Esterle, 1993). In areas of seasonally high river discharges, large areas can be flooded during high river stages, leading at times to loss of life and property. For example, in
222
B.S. HART
Fig. 7-9. Growth of a crevasse splay on the Mississippi Delta. From Coleman (1988).
1955, approximately 38% of the land area of Bangladesh - then East Pakistan - adjacent to the Ganges-Brahmaputra River were inundated (Coleman, 1969). These floods provide rich soils, and low lying interdistributary areas are therefore
DELTA FRONT ESTUARIES
223
Years after Subaerial Land
Fig. 7-10. Graphic representation on the growth and inundation of crevasse splays on the Mississippi Delta (from Coleman, 1988).
prime sites for dyking and cultivation (e.g., Nile Delta, Fraser Delta). Paradoxically, dyke construction (and dam construction) prevents the nutrients in the soils from being replenished, and lower yields result.
SUMMARY
This chapter presented an overview of the factors which control the geomorphology and physical development of delta front estuaries. The growth of large deltas occurs on geologic time scales, in response to allocyclic phenomena such as changes in sea level and sediment supply which typically occur on time periods of lo5 years or longer. These factors are responsible for establishing whether a delta will prograde or be transgressed (e.g., Boyd et al., 1992). Autocyclic phenomena such as lobe switching or channel avulsion occur on shorter time periods (hundreds to thousands of years) and control the patterns of sedimentation and erosion within the overall transgressive or regressive framework. Historical records are long enough in some places so that changes which have occurred in on time scales of centuries in the organization of distributary networks or coastal configuration can be documented (e.g., Ganges-Brahmaputra, Coleman, 1969; Nile, Sestini, 1989; Yellow, Xue, 1993). Coastal changes over periods of several years to decades have been documented from deltaic settings (e.g., Mississippi example) and can sometimes be attributed to humankind’s activities (e.g., Lotfy and Frihy, 1993; Hart and Long, 1990). Although two dimensional modeling of deltas can provide insights into how deltaic systems respond to allocyclic phenomena (e.g., Syvitski and Daughney, 1992), it is generally the geologic record which provides the evidence used to reconstruct (empirically) patterns and processes of delta growth of any given delta. This knowledge can then be used to help predict the future evolution of the delta.
224
B.S. HART
Estuarine areas form sub-environments of deltas. Distributary channels are similar to alluvial channels in many respects, but are affected by tidal processes and the meeting of saline and fresh waters in their lower portions. River mouths are the places where the interaction between fluvial and marine processes is greatest, and patterns of sedimentation there are governed by the interaction of these processes together with the modifying influence of waves. Most marine deltas show the effects of waves, tides and river processes to some extent. Interdistributary areas are host to a wide variety of estuarine environments including tidal flats, mashes and swamps, and lagoons. The effects of mankind’s activities in deltaic settings are being recognized as including coastal erosion, habitat loss, loss of fertility of deltaic soils or more simply as changes in morphology or sedimentation patterns. Deltaic areas, like many coastal areas, can be sites where conflicting demands on natural resources need to be reconciled. For example, dyking of distributary channels in order to cultivate interdistributary areas prevents those same areas from having nutrients replenished by floods and, in areas of high subsidence, could lead to greater flood risk as the elevation of the areas decreases. Achieving the objectives of reconciling our conflicting interests and preserving essential natural ecosystems will require us to broaden or understanding of the complex interactions between process and form in the physical environment, and to prioritize our desired uses of these environments.
REFERENCES Allen, G.P., Laurier, D. and Thouvenin, J., 1979. Etude Stdimentologique du Delta de la Mahakam. Compagnie Franqaise des Pttroles, Notes et Mtmoires, 15, 115 pp. Allen, J.R.L., 1970. Sediments in the modern Niger Delta. A summary and review. In: J.P. Morgan (Editor), Deltaic Sedimentation: Modern and Ancient. SEPM Spec. Pub., 15: 138-151. Barua, D.K., 1990. Suspended sediment movement in the estuary of the Ganges-Brahmaputra-Meghna river system. Mar. Geol., 91: 243-253. Boyd, R., Dalrymple, R. and Zaitlin, B.A., 1992. Classification of coastal depositional environments. Sediment. Geol., 80: 139-150. Bridge, J.S., 1993. The interaction between channel geometry, water flow, sediment transport and deposition in braided rivers. In: J.L. Best and C.S. Bristow (Editors), Braided Rivers. Geol. SOC., Am., Spec. Pub., 75: 13-71. Clague, J.J., Luternauer, J.L., and Hebda, R.J., 1983. Sedimentary environments and postglacial history of the Fraser River and lower Fraser Valley, British Columbia. Can. J. Earth Sci., 20: 1314-1320. Coleman, J.M., 1969. Brahmaputra River: channel processes and sedimentation. Sediment. Geol., 3: 131-239. Coleman, J.M., 1981. Deltas, Processes of Deposition and Models for Exploration (2nd ed.). Burgess Publishing Co., Minneapolis, 124 pp. Coleman, J.M., 1988. Dynamic changes and processes in the Mississippi River delta. Geol. SOC.Am. Bull., 100: 999-1015. Coleman, J.M., and Wright, L.D., 1975. Modern river deltas: variability of processes and sand bodies. In: M.L. Broussard (Editor), Deltas - Models for Exploration. Houston Geol. SOC.,Houston, Texas, pp. 99-149. Coleman, J.M., and Wright, L.D., 1977. Sedimentation in an arid macrotidal alluvial river system: Ord River, western Australia. J. Geol., 86: 621-642. Dominguez, J.M.L., Martin, L. and Bittencourt, A.C.S.P., 1987. Sea-level history and Quaternary
DELTA FRONT ESTUARIES
225
evolution of river mouth-associated beach-ridge plains along the east-southeast Brazilian coast: a summary. In: D. Nummedal, O.H. Pilkey and J.D. Howard (Editors), Sea Level Fluctuation and Coastal Evolution. SEPM Spec. Pub., 41: 115-127. Elliot, T, 1986. Deltas. In: Reading H.G. (Editor), Sedimentary Environments and Facies. Blackwell Scientific Publications, Oxford, pp. 113-154. Ethridge, F.G. and Wescott, W.A., 1984. Tectonic setting, recognition and hydrocarbon reservoir potential of fan-delta deposits. In: E.H. Kosters and R.J. Steel (Editors), Sedimentology of Gravels and Conglomerates. Can. SOC.Pet. Geol. Mem., 10: 217-235. Galloway, W.E., 1975. Process framework for describing the morphologic and stratigraphic evolution of deltaic depositional systems. In: M.L. Broussard (Editor), Deltas - Models for Exploration. Houston Geol. SOC.,Houston, Texas, pp. 87-98. Galloway, W.E., 1976. Sediments and stratigraphic framework of the Copper River fan delta, Alaska. J. Sediment. Petrol., 46: 726-737. Guilltn, J. and Palanques, A,, 1992. Sediment dynamics and hydrodynamics in the lower course of a river highly regulated by dams. Sedimentology, 39: 567-579. 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 Papau. Cont. Shelf Res., 13: 441-472. Hart, B.S. and Long, B.F., 1990. Recent evolution of the Outardes Estuary, Quebec, Canada: consequences of dam construction on the river. Sedimentology, 37: 495-507. Hart, B.S., Prior, D.B., Barrie, J.V., Currie, R.A., and Luternauer, J.L., 1992. A river mouth submarine landslide and channel complex, Fraser Delta, Canada. Sediment. Geol., 81: 73-87. Kostaschuk, R.A., Church, M.A. and Luternauer, J.L., 1989. Bedforms, bed material, and bedload transport in a salt-wedge estuary: Fraser River, British Columbia. Can. J. Earth. Sci., 26: 1440-1452. Kostaschuk, R.A., Church, M.A. and Luternauer, J.L., 1992. Sediment transport over salt-wedge intrusions: Fraser River estuary, Canada. Sedimentology, 39: 305-317. Lotfy, M.F., and Frihy, O.E., 1993. Sediment balance in the nearshore zone of the Nile Delta coast, Egypt. J. Coastal Res., 9: 654-662. Maldonado, A,, 1975. Sedimentation, stratigraphy, and development of the Ebro Delta, Spain. In: M.L. Broussard (Editor), Deltas - Models for Exploration. Houston Geol. SOC.,Houston, Texas, pp. 311-338. McKenna, G.T., Luternauer, J.L. and Kostaschuk, R.A., 1992, Large-scale mass-wasting events on the Fraser River delta front near Sand Heads, British Columbia. Can. Geotech. J., 29: 151-156. Milliman, J.D., 1980. Sedimentation in the Fraser River and its estuary, southwestern British Columbia (Canada). Est. Coast. Mar. Sci., 10: 609-633. Milliman, J.D. and Meade, R.H., 1983. World-wide delivery of river sediment to the oceans. J. Geol., 91: 1-21. Monahan, P.A., Luternauer, J.L., and Barrie, J.V., 1993. A delta topset sheet sand and modern sedimentary processes in the Fraser River delta, British Columbia. Geol. Surv. Can. Paper, 93-1A 263-272. Naidu, AS., and Mowatt, TC., 1975. Depositional environments and sediment characteristics of the Colville and adjacent deltas, Northern arctic Alaska. In: M.L Broussard (Editor), Deltas - Models for Exploration. Houston Geol. SOC.,Houston, Texas, pp. 283-309. Oomkens, E., 1974. Lithofacies relations in the Late Quaternary Niger Delta complex. Sedimentology, 21: 195-222. Orton, G.J., and Reading, H.G., 1993. Variability of deltaic processes in terms of sediment supply, with particular emphasis on grain size. Sedimentology, 40: 475-512. Oti, M., and Postma, G. (Editors), 1995. The Geology of Deltas. A.A. Balkema Publishers, in press. Penland, S., Boyd, R. and Suter, J.R., 1988. Transgressive depositional systems of the Mississippi Delta plain: a model for barrier shoreline and shelf sand development. J. Sediment. Petrol., 58: 932-949. Postma, G., 1990, Depositional architecture and facies of river and fan deltas: a synthesis. In: A. Collela and D.B. Prior (Editors), Coarse-Grained Deltas. Int. Assoc. Sediment. Spec. Publ., 10: 13-27. Prior, D.B. and Coleman, J.M., 1984. Submarine slope instability. In: D. Brunsden and D.B. Prior (Editors), Slope Instability. John Wiley and Sons Ltd., pp. 419-455. Ramsay, K.E., Penland, S., and Roberts, H.H., 1991. Implications of accelerated sea-level rise on
226
B.S. HART
Louisiana coastal environments. In: N.C. Krause, K.J. Gingerich and D.L. Kriebel (Editors), Coastal Sediments '91. ASCE, Seattle, pp. 1207-1222. Schumm, S.A., 1977. The Fluvial System. John Wiley and Sons, Toronto, 338 pp. Schumm, S.A., 1993. River response to baselevel change: implications for sequence stratigraphy. J. Geol., 101: 279-294. Sestini, G. 1989. Nile Delta: a review of depositional environments and geological history. In: M.K.G. Whateley and K.T. Pickering (Editors), Deltas, Sites and Traps for Fossil Fuels. Geol. SOC.Spec. Pub., 41: 99-127. Sparks, R.S.J., Carey, S.N. and Sigurdsson, H., 1991. Sedimentation from gravity currents generated by turbulent plumes. Sedimentology, 38: 839-856. Staub, J.R. and Esterle, J.S., 1993. Provenance and sediment dispersal in the Rajang River delta/coastal plain system, Sarawak, East Malasia. Sediment. Geol., 85: 191-201. Suter, J.R. and Berryhill, H.L., 1985. Late Quaternary shelf margin deltas, northwestern Gulf of Mexico. AAPG Bull., 69: 77-91. Syvitski, J.P.M. and Daughney, S., 1992. DELTA2: delta progradation and basin filling. Comp. Geosci., 18: 839-895. Tornqvist, TE., van Ree, M.H.M and Faessen, E.L.J.H., 1993. Longitudinal facies architectural changes of a Middle Holocene anastomosing distributary system (Rhine-Meuse delta, central Netherlands). Sediment. Geol., 85: 203-219. Whateley, M.K.G. and Pickering, K.T (Editors) 1989. Deltas, Sites and Traps for Fossil Fuels, Geol. SOC.Spec. Pub., 41. van Andel, TJ.H., 1967. The Orinoco Delta. J. Sediment. Petrol., 37: 297-310. Wright, L.D., 1977, Sediment transport and deposition at river mouths: a synthesis. Geol. SOC.Am. Bull., 88: 857-868. Wright, L.D., 1985. River Deltas. In: R.A. Davis, Jr. (Editor), Coastal Sedimentary Environments (2nd ed.). Springer-Verlag, New York, pp. 1-76. Wright, L.D. and Coleman, J.M., 1974. Mississippi River mouth processes: effluent dynamics and morphologic development. J. Geol., 82: 751-778. Wright, L.D., Yang, Z.-S, Bornhold, B.D., Keller, G.H., Prior, D.B., and Wiseman, W.J. Jr., 1986. Hyperpycnal plumes and plume fronts over the Huanghe (Yellow River) delta front. Geo-Mar. Lett., 6: 97-105. Xue, C., 1993. Historical changes in the Yellow River delta, China. Mar. Geol., 113: 321-329.
Geomorphology and Sedimentology of Estuaries. Developments in Sedimentology 53 edited by G.M.E. Perillo 0 1995 Elsevier Science B.V. All rights reserved.
227
Chapter 8
STRUCTURAL ESTUARIES MARIO P I N 0 QUIVIRA
INTRODUCTION
The first associations between tectonics and coastal sedimentary environments were established over a century ago following the development of the geosynclinal theory (Dana, 1873). Later on, and during decades, the relationships between tectonics and sedimentation were restricted preferably to the analysis and interpretation of sedimentary facies (Pettijohn, 1949; Krumbein and Sloss, 1963). Knowledge of the continental drift was complemented by the measurements of paleolatitudes and the discovery of the sea-floor spreading (Dietz, 1961; Hess, 1962) resulted, at the end of the ~O’S,in the development of new concepts related to plate tectonics (McKenzie and Morgan, 1969; LePichon, 1968; Isacks et al., 1968). Therefore, the different types of tectonic zones recognized on the continents (extensional, contractional and horizontal shear zones) were readily associated with the respective plate contacts (divergent, convergent and transform) on the ocean. At the divergent contacts, the plates move in opposing directions meanwhile the expansion of the oceanic floor accretes new lithosphere, accompanied by the formation of volcanoes, normal faults and the intrusion of dykes. Convergent plates imply that one of them moves below the other and part of the lithosphere is consumed, with the development of faulting and thrusting. Finally, at the transform contact, the plates move laterally between them and there is no recycling of lithosphere, but producing large scale transcurrent faults. On a global scale, at the convergence zones (Pacific type active margins) or marginal seas (back-arc basins) practically all the events involved in coastal sedimentation are connected with structural processes, specially tectonics. This connection is produced either from the point of view of the origin of the sediments (i.e., material coming from volcanic arcs), the relationship between block uplift and coastal physiography, or due to changes in sea level no related to typical isostatic events. Inclusive, at a local scale, it is possible to observe, for instance, that many rivers and glaciers that originated estuaries are controlled by faults that determined the movement direction and subsequent erosion by water or ice. Along the coast of the continents there have been many changes of the sea level during the geological time scale. Most variations have been eustatic in origin, however, they have deep roots in the plate tectonic movements and local vertical movements. The later may have also occurred by isostatic effects produced by loading and downloading of ice, crustal expansion, etc. Within the present context we consider that structural processes that occurred within the Quaternary and produced vertical changes in the earth’s crust are termed Neotectonism.
228
M. P I N 0
The Neotectonic component in estuaries has been recognized long time ago. Pritchard (1960) introduced the idea of “Tectonic Estuaries,” or the idea of “Compound Estuaries” by Fairbridge (1980). Even in the definition of “Dynamic Environmental Factors” where the concepts of geotectonic and neotectonics are included, are based on the ideas exposed by Jennings and Bird (1967). Unfortunately, there is a limited bibliography referring to the structural processes acting during the Quaternary on river mouths and their effect in the formation of estuaries. Therefore, the objective of the present chapter is to provide a basic review of the processes that form the structural estuaries as defined by Perillo (this volume) and a description of some characteristic estuaries of the type. Because of the limited references available, this review does not attempt to be extensive, but rather focusing on the processes themselves.
GENERAL CLASSIFICATIONS OF STRUCTURAL ESTUARIES
According to Pritchard (1960, 1967), estuaries originated by tectonic processes constitute a group that contains all types of estuaries that were no possible to classify within his three classical categories (drowned river valleys, fjords and bar-built estuaries). Schubel (1972) indicates that tectonic estuaries correspond to those that fill basins formed by faulting, folding, or other diastrophic movements. However, in both cases there is no discrimination among the different characteristics. More recently, Hume and Herdendorf (1988) introduced a general classification of estuaries, primarily for New Zealand ones, based on the primary mechanism that shaped the basin prior to any modification produced by Holocene depositional processes. Within the estuaries of interest here, they defined two categories formed by vulcanism (volcanic embayments) and tectonism (fault-defined embayments and diastrophic embayments). The basic characteristics that can be observed Table 8-1 Caracteristics of tectonic and volcanic estuaries according to Hume and Herdendorf (1988) Type
Name
Origin mechanism
Main features
Tectonic
Fault-defined embayments
Faults
Parallel shores, structurally defined Rectangular embayment with wide deep rocky inlet Inlet width 1 2 km Well mixed
Diastrophic embayments
Diastrophic processes other than faults
Inlet width >5 km Well mixed
Volcanic embayments
Explosion craters
Circular in plan, defined by crater rim Very small catchment Little freshwater inflow Tide-dominated hydrology Narrow inlet
Volcanic
STRUCTURAL ESTUARIES
229
A
I
N
-e;
Fig. 8-1. Typical examples of structural estuaries: (A) fault-define embayment (Manaia Harbour); (B) diastrophic embayment (Hawke Bay), and (C) volcanic embayment (Panmure Crater) (modified from Hume and Herdendorf, 1988).
in each type are given in Table 8-1. Examples of them are presented in Fig. 8-1. Fault-defined estuaries are elongated features with subparallel shores guided by faults having small inlets (normally less than 2 km) and dominated by tidal currents. Examples are the Manaia (Fig. 8-1A) and Te Kouma Harbours, and San Francisco Bay (Fig. 8-2). On the other hand, diastrophic embayments are much larger coastal bodies originally controlled by folding and other deformations. Since they have wide mouths, the circulation within the estuary is affected by wind and oceanic forcing such as coastal currents. New Zealand examples are Hawke (Fig. 8-1B) and Tasman Bays (Fig. 8-2). Volcanic embayments are estuaries developed in drowned explosion craters. The rim was breached by the sea forming a narrow inlet. Freshwater input is rather low
230
M. PINO
Fig. 8-2. Location map of the rivers and estuaries mentioned in the text: I = rias of Galicia, EspaBa; 2 = ria de Gallegos, Argentina; 3 = Itamaraci estuary, Brazil; 4 = San Francisco Bay, USA; 5 = Yangtze-Kiang, Yellow River, China; 6 = Orakei Basin, Panmure Crater, Manaia and Te Kouma Harbors, New Zealand; 7 = Hawke Bay, New Zealand; 8 = Tasman Bay, New Zealand; 9 = Bahia Magdalena, Mexico; 10 = Bahia San Quintin, Mexico; I 1 = Queule River Estuary, Chile; 12 = Maullin River Estuary, Chile; 13 = Pudeto, Mechaico and Quetalmahue River Estuaries, Chilot Island, Chile; 14 = Upper Cook Inlet, Portage, USA.
since the rim of the crater marks the drainage divide. Panmure Basin (Fig. 8-lC) and Lyttelton Harbour are examples from the New Zealand coast (Fig. 8-2).
Morpho-tectonic classification As plate tectonics influence is always present on the origin and evolution of the structural estuaries considering a macroscale criteria, some researchers such as Fairbridge (1980) have preferred to use the term “morpho-tectonic factors” to indicate the processes that controlled the formation of these coastal features. At a local scale, Fairbridge (1980) subdivided the continental border in hard-rock and softrock coasts. In this case, the association is made between tectonically positive coasts, generally characterized by hard rocks and a long-term history of net emergence, in opposition to tectonically negative coasts, subject to a net submergence and the presence of soft rocks. Both the emergence and submergence are considered long-term processes and not necessarily they are active today. In hard rocks the most characteristic type of estuary is the ria (von Richtofen, 1886; Davis, 1915). Rias (see also Castaing and Guilcher, this volume) are the most important feature of the Galician coast landscape (Fig. 8-2). Their origin is related
STRUCTURAL ESTUARIES
231
to the perpendicular Hercinic faulting of the coast (Vidal Romani, 1983) which produced a continental uplift in blocks. However, Pannekoek (1966) considered them as rift valleys. Next some deep fractures were formed which were occupied first by the rivers and later on invaded by the sea. The history of the rias in the geological time is only comparable with the many decades dedicated to their description and interpretation. The geological history starts during the Lower Mesozoic, when the first separation by rifting is produced between the Canadian shelf and the Galician coast. This rifting ends during the Upper Jurassic-Lower Cretaceous, with the deposition of shales over evaporitic sediments on the new continental slope. After the descent of the continent on the Cretaceous-Tertiary boundary, a short period of normal, long-term subsidence of the passive margin was superposed. This process was interrupted by a rifting phase (Eocene) that created the rias, and it was stressed during the post-orogenic phase (Vanney et al., 1979). The term ria is also employed in other latitudes, for instance in the Argentinean Patagonic coast, where the European immigrant recognized physiographic characteristics of the estuaries that were similar to those in Galicia. An example of this type is Rio Gallegos (Fig. 8-2) which, even though has always been considered a typical ria (Piccolo and Perillo, in press). The work by Pastor and Bonilla (1966) shows that modern faulting processes gave origin to the present structure. It is important to mention that this important work passed unnoticed until late 1993. In this case, a marine sedimentation platform was formed during the Eocene, followed by a regression with deposition of pyroclastic sediments and a smooth folding. During the Pleistocene a fracture normal to the coast lifted the northern block. The fracture defined the position of the estuary axis, emphasising its depth by the erosive action of rivers and glaciers on the sunken block (Pastor and Bonilla, 1966). Faulted estuaries are also common in northeastern Brazil (passive margin). A mountain range (Serra do Mar) leans along the coast, which is characterized by along shore faulting and transversal offsets. Wherever the ocean penetrates the zones weakened by the faults, large rocky blocks become isolated forming islands. River runoff transforms these channels in estuaries, which are relatively shallow and have two well-defined inlets (Medeiros and Kjerfve, 1993). The Itamarach estuarine system (Fig. 8-2) is an example among many estuaries originated by faulting and without this regard may be considered as rias; nevertheless, they are clearly structural estuaries. The Itamaraca estuary has the shape of elongated U, approximately 20 km in length. Because it is within a microtidal, tropical coast, its hydrodynamics is strongly influenced by alternation of dry and rainy seasons (Medeiros and Kjerfve, 1993). San Francisco Bay (Fig. 8-2) corresponds partially to this class of estuaries, since the Golden Gate acts in a similar way as other structural estuaries formed by faulting and resembling a ria. Nevertheless, the bay proper is on top of a subsidence area due to the development of a graben and a downwarping process (Fairbridge, 1980). Another example also described by Fairbridge (1980) is the estuaries formed on the Dalmatian-type coasts where the river systems developed a superposed drainage structure across folds and subsequent rivers eliminated the softer sediments in the
232
M. PIN0
axes of the folds. The estuaries actually formed by a slow subsidence of the coast due to crustal cooling amplified by the last post-glacial sea level rise. The coast in Dalmatia is rather young and there are many “false cliff” because the sea had no time to actually modify its original structure (Guilcher, 1957). In soft-rock coasts (Fairbridge, 1980) many type estuaries can be grouped together associated to large rivers. They are observed in tectonic environments of the Atlantic, Pacific or Back Arc Basin types, with a relief marked by such features as downwarps of the crust, fault-bounded grabens and sedimentary infilling. Within the Atlantic environment, the rifted borders tend to form a zig-zag thalweg related to the intersection of triple-point junctures, and in each of them a graben is developed initially. This structure locks in the locus of major rivers and their deltas (e.g., Niger river, Weber, 1971; Weber and Daukoru, 1975; Fairbrigde, 1980). According to the latter author, the Pacific type border and on the external part of the Back Arc Basin type, the rivers have small catchment areas, high discharge regimes and short channels due to the high relief. Estuaries formed in these environments are small, unless there are graben neotectonics as is the case of San Francisco Bay (Schubel, 1972; Cloern and Nichols, 1985). The inner sectors of the Back Arc Basin are in contact with the older parts of the continents; therefore there are the conditions for the development of large river basins (i.e., Yangtze-kiang and Yellow rivers). On the tectonic environments of the Pacific type of the Mexican coast, most estuaries correspond to the coastal lagoon type. Lankford (1976) analyzed the origin of 83 coastal lagoons from which only four correspond to tectonic originated basins. They are associated to uplift-downwarp of blocks (tectonic lagoon) or to old volcanoes (volcanic lagoons). The possibility that many estuaries had this origin cannot be discarded, however, evidences may have been blurred because of superposed sedimentation and/or erosion processes. The former (i.e., Magdalena Bay, Pacific border of Baja California, Fig. 8-2) is characterized by a barrier formed by uplifted rocks which give shape to an irregular coast with very deep inlets. Nearshore processes tie rocky barriers with shoals or beaches. On the other hand, volcanic lagoons have a central depression with a single opening and formed by volcanoes/lava flows independent of the sea level history. Their shape and bathymetry are highly variable, and salinity variability is mostly related to climatic conditions (i.e., San Quintin Bay, Baja California, Fig. 8-2). In both cases, Back Arc and Pacific type margins just described (New Zealand and Mexico), faulting, diastrophism and volcanism are different expressions of the structural conditions that originate the basins where estuaries developed.
NEOTECTONIC INFLUENCE ON THE FORMATION OF ESTUARIES
All the types of structural estuaries (including those of volcanic origin) described so far have in common that they have developed on any tectonic macroenvironment. Also the structural feature that initiated the original fluvial or coastal channel had no influence on the hydrodynamics of the system, except for the characteristics of the inlet. This is specially true because normally estuarine inlets formed by structural
STRUCTURAL ESTUARIES
233
processes are well defined, deep and always are open. Most probably the situation is preserved because the structural process was not limited to the coast itself but also involved the adjacent continental shelf. There are, nevertheless, evidences from a group of estuaries on Pacific type borders in which the neotectonics have a constant influence over the hydrodynamic and sedimentation patterns. Subsidence and uplift produced by plate convergence are probably the most important factors in the evolution of coasts related to subduction zones. Such coasts may fall or rise in few seconds during earthquakes (coseismic subsidence and uplift), and/or along centuries during interseismic intervals. When earthquakes of magnitude (Mw) > 8 are produced at the interface between overriding and subducting plates (i.e., South American and Nazca plates), the region nearest the subduction trench is frequently uplifted. Simultaneously, a parallel zone in an arcward direction (which corresponds to the external part of the estuaries) may subside (Plafker, 1972, Thatcher, 1984). According to Plafker (1988) vertical and horizontal movements related to plate convergence at Pacific type margins are the sum of coseismic, interseismic and transient deformations of a complete earthquake cycle, on time scale of tens to thousands years. Coseismic deformations result from seaward thrusting of the upper plate and depend upon many features of the megathrust (i.e., dip, displacement and subsidiary faults). Interseismic movements are function of elastic and permanent deformations, and they are in turn a function of the coupling across the megathrust interfaces between the two colliding plates, and the direction, rate, and duration of relative plate movement. Finally, transient postseismic displacements may occur that result from relatively slow elastic strain release or creep deformation following an earthquake. Two zones of the Pacific border of the American Continent present the best evidences in this sense: the coast of the IX and X Regions in southern Chile (Fig. 8-3) and the coast of the Washington and Alaska states in the USA. In the 1960 the Chilean earthquake ( M w = 9.5) coastal deformation occurred for about 1,000 km in a line parallel to the arc over an area of about 80,000 km2. At the shoreline, vertical displacements between -2.3 and +5.7m were detected (Plafker and Savage, 1970; Plafker, 1972, 1988). That seismic event released more energy than any other earthquake in the instrumental record (Kanamori, 1977; Cifuentes, 1989). Unfortunately the only information available for the evaluation of the changes produced by the coseismic deformation corresponds to some slide pictures taken just before the seism (Weischet, pers. commun., 1993). The subsidence effect on the Valdivia area (including the Valdivia River estuary), where the downward movement was of the order of 2 m, has been described by several authors (Weischet, 1960; Tazzief, 1960, 1961; Wright and Mella, 1963). It is important to note that seisms of similar magnitude (if descriptions of historical reporters are compared) occurred in 1575, 1737 and 1837 (Lomitz, 1970). During the last decade, research on the historical evidences of seisms has been reactivated in the USA. Also new evidences have been obtained from seisms that happened in prehistoric epochs based on the interpretation of fine estuarine deposits or sandy laminae accumulated by tsunamis over salt marshes (Nelson et al., 1987; Nelson, 1988; Atwater, 1987, 1988). The 1960 tsunami, in fact, deposited sand over
234
M. P I N 0
Fig. 8-3. Plate-tectonic setting in south central Chile (Regions 1X and X). Contours (m) show level change of the continent after the 1960 earthquake. Barbed line represents seaward end of subduction zone. Paired lines show Chilean spreading ridge. I = ChiloC Island. (Modified after Plafker and Savage, 1970.)
the salt marshes. Recent studies in the USA have demonstrated that the only ways delicate vegetal structures may be preserved are: a) rapid burial; b) burial as the product of a rapid subsidence, and c) sharp contact between the salt marsh deposit and the sandy or mud laminae that underlies it. Stratigraphic studies have been done in different localities in the south of Chile corresponding to the southern extreme of the area affected by the 1960 seism (Maulin and Chiloe Island, 41-43"s) (Bartsch-Winkler and Schmoll, 1990, 1993; Atwater et al., 1992). In this area, repetitions of events have been recorded due to the rapid burial of the salt marsh deposits. However, they have not been space- and temporally correlated so to allow the recognition of past large earthquakes similar to the one in 1960. Even though the coseismic subsidence, it is possible to observe a net emergence during the Holocene (about 2.5 m in the last 4,000 yr). Meaning that besides the continental recuperation between two successive seisms, regionally, there is a net emergency due to, probably, a monotonic uplift from postglacial loading of the Pacific floor. On the time scale between two successive seisms (i.e., no more than few centuries), estuaries suffer deep changes. Figure 8-4 shows the variations produced by the Queule River estuary due to the 1960 seism. This estuary is an adequate representant of the many microtidal estuaries of the south of Chile. Subsidence induced that previous supratidal meadows that bordered the estuary became either inter or subtidal beds (Fig. 8-5). The surface of the outer and middle estuary doubled after the earthquake.
STRUCTURAL ESTUARIES
235
Fig. 8-4. Variation due to neotectonics in the physiography of the Queule River Estuary, obtained from slides taken immediately before and after May, 1960 seism (courtesy of Prof. Dr. W. Weischet). The width of the estuary in the middle section increased approximately twice with a subsidence of the order of 1.8 m.
The sedimentologic map (Fig. 8-6) made after 1980 show that such areas correspond, presently, to muddy zones of the estuary (Pino and Mulsow, 1983; Rojas, 1986) because they are out of the main channel. The litter of some houses deposited over the upper parts of the original meadows and the sandy laminae due to the tsunami can be used as datum level, and allowed to figure out that after the seism and tsunami the salt marshes have grown at an average rate of 1 cm/yr. In the Queule Estuary there is no evidence of postseismic emergence. Informations provided by the local inhabitants suggest that the ebb tidal delta has grown narrowing and making shallower the entrance channel. Also on the intertidal flats accumulation is quite active, because of the reduction of the exchange with the open sea increasing the residence time of sediments and pollutants. In general, the increase in the accumulation tendency within the estuary agrees with the expected process for a transgression. For the case of the Chiloe Island, the seism caused the transformation
M. P I N 0
236
Fig. 8-5. Bathymetry of the Queule River Estuary (1992). The positions of the old channel and the small islands that were submerged are observed (see Fig. 8-4).
GRAVEL
MUD 9 1
11
1 9 SAND
Fig. 8-6. Map of the sedimentary facies of Queule River Estuary (1986). On both shores intertidal sedimentation (specially muddy sand) is observed, and on the zone formerly occupied by the meadows (Fig. 8-4) muddy sand and mud were deposited (now they are subtidal bottoms).
of several rivers in estuaries (i.e., Pudeto, Mechaico and Quetalmahue estuaries) (Bartsch-Winkler and Schmoll, 1993). A seism of MW = 9.2 intensity was registered in 1964 on the Alaska region. The seism caused a deformation parallel to the arc along 950 km and covering a surface of 140,000 km2. The vertical movements in the area fluctuated between + 11.3 and -2.3 m, and a tsunami also occurred afterward (Plafker, 1972, 1988). There are no specific studies about changes of the bottom sediment composition within estuaries, neither
STRUCTURAL ESTUARIES
237
of their hydrography. At the town of Portage, (Upper Cook Inlet) the subsidence was 2.4 m divided in 1.6 m due directly to the crust movement, and the rest product of the compaction of 300 m of sediments. The downward movement affected a surface of 18 km2 where over 20 x lo6 m3 of intertidal silts were deposited (Ovenshine et al., 1976). Core records and the convolute bedding observed in the zone (approximately 8,000 yr of stratigraphic register) indicate that the rapid deposition of silt is an event that repeats itself along the stratigraphic section. They can be interpreted in terms of the recurrence of the coastal subsidence at approximately 600-yr intervals (Bartsch-Winkler et al, 1983; Bartsch-Winkler and Schmoll, 1993). A similar time interval is suggested for the major earthquakes in the south of Chile (Barrientos and Ward, 1991).
SUMMARY
Within the structural estuaries, it is possible to differentiate two major groups: a) those where the global tectonic characteristics (in any type of margin) have determined the origin of the fluvial or coastal channel, and b) those where strong neotectonic impacts are observed on the sediments deposited after the csseismic subsidence. The effect of the subsidence on the estuarine circulation has not been studied at all, as well as the affect of the vertical movement that occurs between two successive major earthquakes. Study of cores coming from this type of estuaries, specially those cores taken from sub and intertidal areas, are prone to provide new indications. Evidences obtained from the Pacific coast of the Americas (south of Chile and Oregon, Washington and Alaska) suggest that the ideas proposed by Fairbridge (1980) in relation with the morpho-tectonic factor must be re-evaluated. In the same zone and with short time scales (days to thousands of years) it is possible to present abrupt sinking (inclusive with the creation of new estuaries), and sustain uplift in interseismic and transient periods. The tectonics does give, in this case, the opportunity to study intense processes of creation and changes in the estuaries during human-scale time period.
REFERENCES Atwater, B.F., 1987. Evidence of great Holocene earthquakes along the outer coast of Washington State. Science, 236: 942-944. Atwater, B.F., 1988. Buried Holocene wetlands along the Johns River, southwest Washington. Holocene Subduction in the Pacific Northwest, A Symposium. Quaternary Research Center, Univ. of Washington, 4 pp. Atwater, B.F., Jimenez, H. and Vita-Finzi, C., 1992. Net late Holocene emergence despite earthquakeinduced submergence, south central Chile. In: N. Rutter (Editor), Impacts of Tectonic on Quaternary Coastal Evolution. Quater. Int., 15/16: 77-85. Barrientos, S.E. and Ward, S.M., 1991. Chile earthquake; inversion for slip distribution from surface deformation. Geophys. J. Int., 103: 589-598. Bartsch-Winkler, S. and Schmoll, H.R., 1990. Stratigraphy of late Holocene intertidal deposits, Isla
238
M. P I N 0
Grande de ChiloC Region, southern Chile - Evidence for relative sea level change. 13th Int. Sedimentol. Congr., Nottingham, England, pp. 39-40. Bartsch-Winkler, S. and Schmoll, H.R., 1993. Evidence for late Holocene sea-level fall from reconnaissance stratigraphical studies in an area of earthquake-subsided intertidal deposits, Isla ChiloC, souther Chile. Int. Assoc. Sedimentol., Spec. Pub., 20: 91-109. Bartsch-Winkler, S., Ovenshine, A.T and Kachadoorian, R., 1983. Holocene history of an estuarine area surrounding Portage, Alaska as recorded in a 93 m core. Can. J. Earth Sci., 20: 802-820. Cifuentes, I.L., 1989. The 1960 Chilean earthquake. J. Geophys. Res., 94: 665-680. Cloern, J.E. and Nichols, F.H. (Editors), 1985. Temporal dynamics of an estuary: San Francisco Bay. Dr. W. Junk Publ., Dordrecht, 237 pp. Dana, J.D., 1873. On some results of the earth’s contraction from cooling, including a discussion of the origin of mountains and the nature of the earth’s interior. Am. J. Sci., 6: 161-171. Davis, W.M., 1915. The principles of geographic description. Ann. Assoc. Am. Geogr., 5: 61-105. Dietz, R.S., 1961. Continental and ocean basin evolution by spreading of the sea floor. Nature, 190: 854. Fairbrigde, 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., pp. 2-35. Guilcher, A., 1957. Morfologia Litoral y Submarina. Ed. Omega, Barcelona, 262 pp. Hess, H.H., 1962. History of ocean basins. In: A.E. Engel et al. (Editors), Petrologic Studies: A Volumen to Honour A.F. Budddington. Geol. SOC.Am., pp. 559-580. Hume TH. and Herdendorf C.H., 1988. A geomorphic classification of estuaries and its application to coastal resource management - a New Zealand example. Ocean Shoreline Manag., 11: 249-274. Isacks, B., Oliver, J. and Sykes, L.R., 1968. Seismology and the new global tectonics. J. Geophys. Res., 73: 5855-5862.. Jennings, J.N. and Bird, E.C.F., 1967. Regional geomorphological characteristics of some Australian estuaries. In: G.H. Lauff (Editor), Estuaries. AAAS Washington, DC, Publ. 83, pp. 121-128. Kanamori, H., 1977. The energy release in great earthquakes. J. Geophys. Res., 82: 2981-2987. Krumbein W.C. and Sloss, L.L., 1963. Stratigraphy and Sedimentation. W.H. Freeman, San Francisco, 600 pp. Lankford, R.R., 1976. Coastal lagoons of Mexico. Their origin and classification. In: M. Wiley (Editor), Estuarine Processes, Vol 11, Academic Press, New York, pp. 182-215. LePichon, X., 1968. Sea-floor spreading and continental drift. J. Geophys. Res., 73: 3661-3675. Lomitz, C., 1970. Major earthquakes and tsunamis in Chile during the period 1535 to 1955. Geol. Rundsch., 59: 938-960. Mckenzie, D.P. and Morgan, W.J., 1969. Evolution of triple junctions. Nature, 224 125-126. Medeiros, C. and Kjerfve, B., 1993. Hydrology of a tropical estuarine system: Itamarac6 Brazil. Est. Coastal Shelf Sci., 36: 495-515. Nelson, A.R., 1988. Implications of late Holocene salt-marsh stratigraphy for great earthquake recurrence along the coast of south-central Oregon. Holocene Subduction in the Pacific Northwest, A Symposium. Quaternary Research Center, Univ. of Washington, 12 pp. Nelson, A.R., Atwater, B.F. and Grant, W., 1987. Estuarine record of Holocene subduction earthquakes in coastal Oregon and Washington, USA. 12th Int. Congr., Int. Union Quat. Res., Ottawa, p. 231. Ovenshine, A.T., Lawson, D.E. and Bartsch-Winkler, S., 1976. The Placer River Silt - an intertidal deposit caused by the 1964 Alaska earthquake. J. Res. U.S. Geol. SUIT., 4: 151-162. Pannekoek, A.J., 1966. The ria problem. Tijdschr. K. Ned. Aardr. Genootsch., 83: 289-297. Pastor, J.M. and Bonilla, J., 1966. Estudio para la formulaci6n del plan de desarrollo fisico de la ciudad de Rio Gallegos. In: Municipalidad de Rio Gallegos (Editor), Bases para el Desarrollo, Vol. 2A, pp. 27-5 1. Pettijohn, F.J., 1949. Sedimentary Rocks. Harper and Row, 526 pp. Piccolo, M.C. and Perillo, G.M.E., in press. Geomorfologia e hidrografia de 10s estuarios de la Repdblica Argentina. In: INIDEP (Editor), El Mar Argentino y sus Recursos Pesqueros. Pino, M. and Mulsow, S., 1983. Distribuci6n de facies granulomttricas en el estuario del rio Queule, IX Regibn; un analisis de componentes principales. Rev. Geol. Chile, 18: 77-85.
STRUCTURAL ESTUARIES
239
Plafker, G., 1972. Alaskan earthquake of 1964 and Chilean earthquake of 1960 - Implications for arc tectonics. J. Geophys. Res., 77: 901-925. Plafker, G., 1988. Tectonic deformation related to great subduction zone earthquakes. Holocene Subduction in the Pacific Northwest. A Symposium. Quaternary Research Center, Univ. of Washington, 3 PP. Plafker, G. and Savage, J.C., 1970. Mechanism of the Chilean earthquakes of May 21-22, 1960. Geol. SOC.Am. Bull., 81: 1001-1030. 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. AAAS Washington, DC. Pub. 83, 3-5. Rojas, C., 1986. El estuario del rio Queule: un ambiente sedimentario en el sur de Chile. An. Inst. Cienc. Mar Limnol. Univ. Nac. Auton. Mexico, 13: 231-240. Schubel, J.R., 1972. Classification according to mode of basin formation. In: J.R. Schubel (Editor), The Estuarine Environmcnt: Estuaries and Estuarine Sedimentation. Am. Geol. Inst., Washington, DC, pp. 11-2-8. Tazzieff, H., 1960. propos de la signification tectonique des important glissements de terrain provoquts par le grand stisme du Chili en 1960. C.R. Acad. Sci. Paris, 251: 2204-2206. Tazzief, H., 1961. Interprttation des glissements de terrain accompagnant le grand stisme du Chili. Bull. SOC.Belge Gtol., 69: 1-11. Thatcher, W., 1984. The earthquake deformation cycle, recurrence and the time-predictable model. J. Geophys. Res., 89: 5674-5680. Vanney, J.R., Axietre, J.L. and Dunand, J.P., 1979. Geomorphic provinces and the evolution of the northwestern Iberian continental margin. Ann. Inst. Ocean. Paris, 55: 5-20. Vidal Romani, J.R., 1983. Origen y evoluci6n de las rias de Muros y Noia. GGQ/GETC VI Reunidn de Cuaternario. Excursi6n N2, Univ. de Vigo, pp. 9-41. von Richtofen, F., 1886. Fuhrer fur Forschungsreisende, Janecke, Hannover, 734 pp. Weber, K.J., 1971. Sedimentological aspects of oilfields of the Niger delta. Geol. Mijnb. 50: 559-576. Weber, K.J.and Daukoru, E., 1975. Petroleum geology of the Niger delta. Proc. 9th World Petrol. Conf., pp. 209-221. Weischet, W., 1960. Contribuciones al estudio de las transformaciones geogrificas en la parte septentrional del sur de Chile por efecto del sismo del 22 de mayo de 1960. An. Fac. Cienc. Fis. Mat., Univ. Chile, Publ. 15, Vol. 17, 23 pp. Wright, C and Mella, A,, 1963. Modifications to the soil patterns of south-central Chile resulting from seismic and associated phenomena during the period May to August 1960. In: P. Saint-Amand (Editor), Oceanographic, Geologic and Engineering Studies of the Chilean Earthquakes of May 1960. Bull. Seismol. SOC.Am., Spec. Issue, 83: 1367-1402.
This Page Intentionally Left Blank
Geomorphology and Sedimentology of Estuuries. Developments in Sedimentology 53 edited by G.M.E.Perillo 0 1995 Elsevier Science B.V. All rights reserved.
241
Chapter 9
COASTAL LAGOONS FEDERICO I. ISLA
INTRODUCTION
Although there are many definitions and descriptions of coastal lagoons in the literature, most of them share a series of basic features: 1) they are shallow estuarine environments where salt and fresh water interact; 2) the water mass is impounded by some type of sedimentary barrier, and 3) the connection with the open sea is restricted to one to several openings (inlets) in the barrier maintained mainly due to the action of tidal currents against the effect of wave-generated longshore currents. Because of the barrier conditions the free communication with the ocean, many authors call coastal lagoons as “bar-built estuaries” (i.e., Pritchard, 1960; Fairbridge, 1980). The latter implicitly suppose that the water body was enclosed by the growing of a barrier spit. While such a process may be true for many coastal lagoons, others were created by inundation of low lying plains behind pre-existent barriers either sedimentary, biogenic or formed by hard rocks. Reineck and Singh (1983) proposed the following definition of coastal lagoons: “... are shallow water bodies, running parallel to the coast, and connected to the open sea with an outlet”. Later Kjerfve (1986) and Kjerfve and Magill (1989) have advanced further over the previous definition and indicated that they are “...inland water bodies usually oriented parallel to the coast separated from the ocean by a barrier, connected to the ocean by one or more restricted inlets, and having depths that seldom exceed a couple of meters”. Both definitions attach themselves to the widespread criteria given previously and, in general, can be considered as suitable to describe the coastal features subject of the present chapter; nevertheless, Kjerfve’s definition adds a very important phrase to it “...usually oriented parallel...” Although most coastal lagoons follow a shoreline-parallel pattern, there are many exceptions to this rule [i.e., Mobile Bay (USA); San Antonio Bay (Argentina); Laguna de Rocha (Uruguay)]. Coastal lagoons are parallel to the coast when they originated by a wave-built barrier. Those that occupy a submerged river valley can be positioned normal to the shoreline (Phleger, 1981). From an evolutionary standpoint, coastal lagoons are related to longshore drift or sea-level variations. Some coastal lagoons are associated to ancient depressions flooded by the postglacial transgression, some are connected to former (Upper Pleistocene) barriers,and others are recently conceived by the longshore growing of spits and barriers. As most coastal environments, coastal lagoons developed during the last 7,000 years, when the sea level diminished its rise, became stable or began to drop (the Holocene quasi-stillstand).
242
F.I. ISLA
Sea-level variations have extreme importance in the development of coastal lagoons. For tectonically stable coasts, submergence led to the flooding of the previous landscape (accommodation regime, in the sense of Swift and Thorne, 1991). In these conditions, development of lagoons may have depended on the reworking of ancient barrier islands. On emergence coasts, river input and longshore drift dominate (supply regime). Coastal lagoons are suitable environments for man’s activities as occupation, tourism, fishing, aquaculture and transportation. However, it is not yet properly known the impacts of man intervention in estuarine environments. The experiences of Grevelingen estuary (The Netherlands) converted into a lake (Bannink et al., 1984), or the San Francisco estuary (Nichols et al., 1986) resume the consequences of these interventions. The ecology of an estuary depends upon its geological stage, and the rate and direction of its natural changes is a yardstick to assess impacts induced by man (Roy, 1984). Due to the influence of coastal lagoons in a wide range of coastal processes such as environmental (i.e., pollution control), fisheries (i.e., artisanal fisheries, aquaculture, nursery), human development (i.e., local and regional economies, harbours), coastal management, etc., to have a detailed knowledge of the geomorphology and sedimentology of these features becomes crucial. Therefore, the objective of the present chapter is to describe the general characteristics of coastal lagoons, their main dynamical factors and evolution.
ORIGIN OF COASTAL LAGOONS
Most coastal lagoons are related to barrier developments. In fact, Phleger (1969) suggests that “The formation and development of the lagoon barrier are the critical element in the history of a lagoon”. There was much controversy about the origin of barrier islands-coastal lagoons systems. Barriers were repeatedly proposed as emergent features (de Beaumont, 1845; Johnson, 1919; Leontyev and Nikiforov, 1966; Zenkovitch, 1969; Otvos, 1970). Hoyt (1967,1970) suggested that submergence could also lead to reworked barriers, while Fischer (1968) discussed Hoyt’s idea in the sense that complex longshore spits could lead to barrier islands without submergence. Zenkovitch (1969) gave detailed examples of Russian lagoons and barriers caused either by emergence or submergence. Both positions are today locally probed and a multiple causality (equifinality) accepted (Schwartz, 1971). Another discussion focused on the timing of the barrier submergence. The “shoreward retreat model” considers a continuous landward migration of the barrier (Swift, 1975). The example of the New York shelf, where a rapid (episodic) rise of the sea level overstepped the barrier, is based mainly on 14C ages of backbarrier and marsh peats deposits on the shelf. As a result, an in-place drowning of an ancient barrier was contemplated (Sanders and Kumar, 1975; Rampino and Sanders, 1981) as the possible way for the present status of that coastal area. On the coast of Nova Scotia, an episodic overstepping was historically documented (Forbes et al., 1991; Fig. 9-1). Also on the shelf of Massachusetts (Jeffreys Ledge), a drowned barrier spit
COASTAL LAGOONS
243
1954
2
-6
m
0
Distancio
(m)
Fig. 9-1. Gravel barrier overstepped at Story Head, Nova Scotia (after Forbes et al., 1991).
is indicating a Holocene stillstand preserved by a rapid sea-level rise (Oldale, 1985). On the Mississippi River delta, Penland et al. (1985) clearly described how coastal lagoon deposits could be the response of the erosional shoreface retreat of an abandoned delta lobe reworked by waves while subsiding. A barrier island developed during this reworking allowed the preservation of the deposits as the transgression proceeded (Fig. 9-2).
GEOMORPHOLOGY
Coastal lagoons occur along low-lying coasts and they are commonly related to past shoreline deposits with availability of sediments. Coastal plains are therefore reworked to enable the formation of barriers, spits, beaches, cheniers, marshes and mudflats.
F.I. ISLA
244 DELTI
PLAIN
ENVIRONMENTS
Dtstributary Levee Marsh
Wench Ridge%
BARRIER ENViRONMENTS
0
S uboerlol Reworked Deltoic Sands
tLT]
Subaquaous Reworked D e l t ~ t c Sonds
R e c u r v e d Spit
I @@
I N N E R S H E L F SHOAL
I
EARRIER i s i.iN%
ARC
T i d a l l n i e t Channel
Shell Reef6
1
Fig. 9-2. Idealized coastal evolution of the Mississippi delta in relation to the occupation and subsidence (modified after Penland et al., 1985).
Other environments are easily recognized within or related to coastal lagoons. The sand barriers are accumulations of eolian origin related to sand availability and a significant wave energy during the initial stages of the transgression. Short (1988) has explained the activity of sand ramps during the initial stages of the Holocene transgression while their disactivation is related to sand consumption. Present or former directions of dominant winds are responsible for dune patterns and barrier dynamics. Tidal inlets are connections between the sea and lagoons maintained mainly by tidal currents. Several subenvironments can be discriminated close to tidal inlets as many processes can occur simultaneously or closely. Tidal deltas are sand accumulations produced by the action of flood or ebb currents. Tidal range and wave action are the main processes responsible for the shape of the inlet and related sand banks. Tidal flats are accumulations of sand, mud or both (mixed flats) along the coast of estuaries. Their extension is related to tidal range and shoreline slope. Marshes are mudflats that become vegetated by specific plants in a position related to spring tides or storm surges. Tidal flats (Amos, this volume) and marshes (Luternauer et al., this volume) are specifically described in other chapters of this book.
SEDIMENTOLOGY
Template coastal lagoons are usually dominated by seasonal laminated muds. If the lagoon is shallow enough to be easily modified by wind-induced waves, the bottom is continuously resuspended. Where bioturbation is significant, lamination is
COASTAL LAGOONS
24.5
altered. At tropical hyperarid coastal lagoons, evaporite precipitation dominates over physical and biogenic processes. Usually towards the inlet, there is more sand availability and higher dynamics. Transverse bedforms as ripples, megaripples, dunes or sand waves may occur in relation to the magnitude of flood or ebb currents (Boothroyd, 198.5). These bedforms vary in response to the stage of the tide to maintain a higher bottom stress (Bruun, 1969). Within big coastal lagoons, the formation of micro- and macroflocs plays a significant role in their sedimentation. Pejrup (1988) proposed a hydrodynamic classification of estuarine sediments using the percentage of the flocculated grain-size population in the mud fraction. The coarsest sediment in a coastal lagoon is usually close to the inlets where there is higher energy. On the other hand, the finest sediment is at tidal flats or at ponds or lagoons distant from the river source. Some lagoons have very high sediment discharges. Dos Patos Lagoon outputs sands, silts and clays that can be recognized offshore at depths of 22 m in front of the inlet (Calliari and Fachin, 1993). Within Dos Patos Lagoon, there are some grain-size anomalies characterizing small area depressions (up to 19 m depth) where clay is settling (Herz and Mascarenhas, 1993).
CONDITIONING FACTORS FOR THE DEVELOPMENT OF COASTAL LAGOONS
Although coastal lagoons are found all over the world, covering over 13% of its coastline (Nichols and Allen, 1981), they require some particular conditions to develop. Regarding the dominance between advective and diffusive transport in flushing the lagoon, Kjerfve and Magill (1989) recognized choked lagoons (Fig. 9-3A) where diffusion dominates and leaky lagoons (Fig. 9-3B) where advection dominates. Restricted lagoons are transitional between them (Fig. 9-3C). At the same time and in relation to the type of saline stratification, they also recognized between coastal lagoons with estuarine circulation (surface outflow and inflow by the bottom) and hypersaline lagoons for those with anti-estuarine circulation (surface inflow and outflow by the bottom). COASTAL LAGOONS
Fig. 9-3. Classification of coastal lagoons: A) choked; B) restricted, and C) leaky (after Kjerve and Magill, 1989).
246
F.I. ISLA
I SOLAR
IA
/\
RADIATION
B
P R EC I PITAT I0N 8 EVAPORATION (60 140
120 100
..
80 60 40
2:vl N 6O0
I
I
I
30"
,
,
I
,
I
00
EROSION R A T E S ON L A N D
I
30°
,
I
60°
s
C
D
Fig. 9-4. Latitudinal factors such as A) solar radiation, B) precipitation-evaporation, C) land erosion rates, and D) type of sediment availability affect coastal lagoon behaviours (after Nichols and Allen, 1981; Hayes, 1980).
Climate effects Climate conditions the amount and type of sediment supplied to the coast (Hayes, 1980). Latitude also conditions the amount of light (indirectly primary productivity, Fig. 9-4A), temperature, precipitation (evaporation, weathering, Fig. 94B), vegetation cover (transpiration) and winds and storms frequency. Furthermore, there are notable differences in the sedimentological composition of coastal lagoons as a function of the latitude (Fig. 9-4D). Climate effects can be readily associated to latitudinal distribution, and Nichols and Allen (1981) compiled differences among coastal lagoons belonging to high, medium and low latitudes:
._
COASTAL LAGOONS
247
Seepage
0 0 Transitional channel /seepage
00 0 1
0 01 Discharge
01
c u m e c s / l o o 0 rn l e n q t h
1
/1W m
10
width borrler
Fig. 9-5. Discharge vs. potential head at SE Ireland coastal lagoons. These factors control processes of seepage and channelization (after Carter et al., 1984).
High-latitude coastal lagoons have dynamics restricted to gravel dispersals (Fig. 9-4D) or snow cover. Most of the year, ice dynamics (freezing, breakup and rafting) dominate. During the unfrozen months, runoff, winds, waves, tides and organisms play a more significant role. In low-lying coasts (tidal flats), seasonal frost and permafrost produce features as mounds, polygonal patterns, non-sorted circles, boulders and thermokarst depressions (Dionne, 1989a). The shore ice provides either protection, erosion or sedimentation. However, protection was previously overestimated because, during breakup, ice blocks are usually lifted up by spring tides tearing out pieces of mudflat or marsh (Dionne, 1989b). The sedimentation effects are related to the amount of sediment supply; however, they are mostly input by the rivers rather than shore erosion (Fig. 9-4C). Boulders are usually the more conspicuous features, but there may be also a significant fine sediment load (Dionne, 1989a; Fig. 9-4D). On partly closed lagoons with gravel barriers, seepage can play a significant role if the tidal range is large (Fig. 9-5) and the barrier is long enough to delay the ebb flow. These processes have been recognized in Ireland (Carter et al., 1984) and Tierra del Fuego (Isla et al., 1991). Mid-latitude coastal lagoons are related to sand-duned barriers. The directions of dominant winds and storm frequency contribute to migrating trends and evolution. Potential obstruction of tidal inlets is related to longshore currents. Precipitation and vegetation cover are significant factors for dune stability and backbarrier dynamics. The availability of sand in relation to the Holocene quasi-stillstand (rising or dropping) is also a matter of the wave energy and river inputs. High-energy beaches allow sand ramps to feed up the foredunes. On low-energy beaches, the dunes are easily stabilized by vegetation and they induce beach progradation (Short, 1988). Low-latitude coastal lagoons are dominated by evaporation (sabkhas or arid coastal lagoons), siltation (related to tropical deltas), biogenic activity (coral lagoons, mangrove swamps) or episodic events (hurricanes, monsoons). When evaporation exceeds freshwater input, salts begin to precipitate as carbonates, chlorides or sulphates. If winds are strong enough, oolites and skeletal sands dominate the shallower portions. At the Abu Dhabi sabkha, algal mats, fecal pellets and burrows
248
F.I. ISLA
are interlayered with gypsum, anhydrite, calcite and dolomite (Evans et al., 1969). Dry seasons in low-latitude lagoons can cause that the salinity maximum in the lagoons becomes larger than coastal salinity. This process causes an inverse estuarine circulation, with surficial inflow and an outflow of denser saline water through the bottom. As coastal lagoons are usually shallow, wind effects mix water and the salinity difference can be recognized only across the channel (Kjerfve, 1990). Some tropical lagoons are dominated by siltation in response to high input of mud (Fig. 9-4D). Generally, lagoons are the longshore-current reworking of a river delta. The Keta lagoon is related to the Volta river, Ghana (Ntiamoa-Baidu and Hollis, 1992; Blivi, 1993); Nokoue lagoon is associated to the OuCme river, Benin (OyCdC and Lang, 1993). Manzala, Idku and Maryut coastal lagoons are related to abandoned branches of the Nile delta (Coutellier and Stanley, 1987). Same processes occurred on the Atlantic coast of Nicaragua: Ulang river-Pahara lagoon, and Grande river-Perlas lagoon (Owens and Roberts, 1978), and Brazilian coast, e.g., Doce River (Martin and Suguio, 1992). Lagoons dominated by biogenic activity comprised either coral lagoons or those associated to mangrove swamps are described below. Specific coastal lagoons behave in response to episodic events. Hurricanes produce either new tidal entrances or their obstructions. Corpus Christi Pass is a typical hurricane modified tidal inlet. In between storms (Carla in 1961, or Beulah in 1967) it was closed by littoral drift and became open for short time immediately after these storms (Davis et al., 1973). Washovers are common features associated to episodes as in the coastal lagoons of Texas, Louisiana, Alabama and Florida. The effects of these storms also affect the inner shelf. Hurricane Beulah (1967) influenced the coast of northern Mexico and southern Texas, causing storms channels and wedge-shaped sand bodies at the coastal lagoons (Scott et al., 1969). Some years later, on the coasts of Louisiana and Alabama, hurricane Frederic (September, 1979) originated washovers, shore erosion and reopening of inlets (Nummedal et al., 1980). Hurricane Allen produced similar processes at Laguna Madre coastal lagoon, Texas (Maynard and Sutter, 1983). Analogous episodic events surely affected other tropical coastal areas, but they are not well recorded (Ntiamoa-Baidu and Hollis, 1992). Storm surge effects are well recognized for the coast related to the Mississippi delta and a typical dynamic process can be described as follows: previous fair-weather waves try to attach a tidal berm to the barrier (Fig. 9-6A). During the storm, washovers transported sand to the lagoon while sand sheets sparse on the inner shelf (Fig. 9-6B). After the storm, the beach attempts to recover but the washovers and the inner-shelf deposits remain as a record of the episode (Fig. 9-6C; Penland et al., 1985). Vegetation recovery studies from washovers of Texas coast indicate that species richness was restored within a year; although there was a significant change in species composition and dominance (Lonard and Judd, 1993). Precipitation and the drainage regime caused the segmentation of Los Patos Lagoon, although this segmentation is not affecting today its southeastern coast (Toldo, 1991). For the Holocene lagoonal system of Rio Grande do Sul, meteorological deltas (lagoonal “tidal” deltas in the sense of Tomazelli and Wilwock, 1991) were described between interconnected lagoons. These deltas resulted from different effects on the
249
COASTAL LAGOONS
A. FAIR WEATHER
m
Deposition Erosion
a
Resultont Storm Deposlt
-Sediment
Dispersal
OVERWASH
6 STORM IMPACT
rp
Resultant Storm Deposlt
C. RECOVERY WASHOVER DEPOSIT \
Fig. 9-6. Effect of storms and beach recovery at the Mississippi delta coast (modified after Penland et al., 1985).
lagoon levels. They can be induced by winds, different drainage discharge or meteorological tides (Tomazelli and Wilwock, 1991). On the other hand, at Terminos lagoon (Mexico), fish capture is related to river discharge (Yafiez-Arancibia et al., 1992).
Tectoniceffects Although coastal lagoons and barriers develop along different coastal conditions, it is statistically recognized that they are preferentially developed on trailing-edge coasts (49%) compared to collision margins (24%) or marginal seas (27%; Glaesser, 1978). At Willapa Bay (Washington, USA), it was probed that six episodic subsidence phenomena during the last 7,000 yr were associated to earthquakes. Peat layers signify the submergence of Triglochin maritima salt marshes, while sand sheets may be deposited by related tsunamis (Atwater, 1987). In a similar way, convolute bedding and slump structures characterized the 3,500-yr sequence along the tidal channels of
250
F.I. ISLA
the Knight Arm (Cook Inlet, Alaska). These structures are undoubtedly related to seismicity (Bartsch-Winkler and Schmoll, 1984). The tectonic architecture for estuarine deposits varied along the recent past as the sea level went up and down. When the sea level was 100-120 m below present, 15,000 yr ago, estuaries (s. str.) would have been connected to canyon headlands. However, during the postglacial sea-level rise the flat, outer half of the shelf should have favoured the formation of coastal lagoons (Emery, 1967). When the sea level drowned shelf channels, estuaries would have formed in some places but most of them must have characteristics of coastal lagoons. At the maximum reaches of the transgression the sea level encountered a steeper topography (shoreface) favouring the formation of estuaries (Emery, 1967).
Biogenic effects In some coastal lagoons as Langebaan, S. Africa, there is a significant effect of biogenic structures on some environments as the main channel, intertidal and subtidal flats and marshes. Upper intertidal flats have more diverse biological activity while lower intertidal flats are strongly and more uniformly bioturbated (Flemming, 1977). Within the Georgia estuaries, on the backbarrier environments, characteristic lebensspuren (traces of biogenic activity) help to recognize the activity of polychaeta, decapods, rays, amphipods, ophiuroids, bivalves and anemones. The distribution of bioturbation features is controlled by sediment type, its accumulation and reworking rates, and salinity. Bioturbation increases seaward while lower-energy environments have more potential of preservation (Howard and Frey, 1985). In Mar Chiquita Lagoon, Pezzani and Obenat (1988) have described reef-like structures produced by the Ficopomatus enigmaticus polychaeta which constitute a sediment trap and contribute to the silting of the inner lagoon. Coral-reef coasts have a completely different behaviour in the sense that they are dominated by biogenic processes. A study at Davis Reef (Great Barrier Reef, Queensland) proposed two sedimentation models depending on the relative protection to physical action. Leeward margins are characterized by bioturbated muddy sands that grade upwards to reefs (microatolls), storm layers and bioclastic beaches (cay sands) with aragonitic cement (Fig. 9-7A). Windward-margin sequences, on the other hand, initiate with basal coral gravel, overlaid by heavily bioturbated muddy sands (subtidal biogenic mounds) that grade to similar intertidal and supratidal deposits with more gravel content (Tudhope, 1989). Washovers are preserved as shingle ramparts (Fig. 9-7B). The shape and evolution of these microatoll lagoons are conditioned by dominant wind direction.
Wind-waveeffects Some lagoons are large enough and therefore subject to internal wind and wave effects. If lagoons are shallow and the sediment fine, winds can cause frequent sediment resuspension and turbidity may restrict light penetration. When winds blow
COASTAL LAGOONS
25 1
Leeward margin woter Depth
A Srdimmf ACCUrn.
Rote
mcr
1 II
'1
I
Windward margin
1/01
fa*)
i B
Fig. 9-7. Schematic sequences of the evolution of A) leeward and B) windward margins of coral reef lagoons (after Tudhope, 1989).
parallel to elongated lagoons, stationary waves can produce septation (Zenkovich, 1959; Fig. 9-8). For instance, the action of dominant winds from the north and south counteracts the fluvial effect of the Guaiba Complex (Brazil) and causes a regular oscillation along the Itapua Spit (Toldo, 1989). An analogous process has been recognized by Kjerfve (1990) for some Texas coastal lagoons. For the septation of Dos Patos Lagoon, Toldo (1991) measured the growth of points at a rate of 59.5 m/yr. However, he distinguished that those points on the western margin are growing while those on the southern half of the southern margin are eroding (Fig. 9-8B). Again, winds from opposing directions can cause that the spits grow (Fig. 9-8C) originating small lagoons on the western shore of the Dos Patos Lagoon (Fig. 9-9).
252
F.I. ISLA
C
Fig. 9-8. Theoretical model of septation stages of coastal lagoons due to wind effects (modified after Zenkovitch, 1959).
Wind also conditions barrier and tidal channel migration rates. On the Frisian Islands (Germany), the eastward displacement of the backbarrier drainage divides has altered the inlet drainage networks and produced hook-shaped main-inlet channels (Fitzgerald et al., 1984). Comparatively, the tidal channels carved on the San Sebastian Bay mudflats suffer a straight lateral migration due to the dominant westerlies (Isla et al., 1991).
Tidal and wave effects Coastal lagoons are estuaries (sensu Zuto) maintained by tidal currents (via tidal inlets) with a vertical mixing. The effects of tides control the development of different sand bodies (Hayes, 1975; Fig. 9-10). Tidal inlets have a minimum flow area in relation to the lagoon tidal prism (O’Brien, 1969; Jarrett, 1976; Fig. 9-11). Although, fluctuations in mean sea level should correlate with variations in the inlet’s throat (scour or deposition), some inlets as Price Inlet do not present this pattern (Fitzgerald and Fitzgerald, 1977). The minimum flow area of Mar Chiquita inlet, for example, correlates fairly good with precipitation in the region but having a lag of 30 days (Isla, 1986). Based on laboratory tests, Mayor-Mora (1977) conclude that waves superimposed on tides reduce 40% of small inlets cross-sectional area, suggesting that waves cause a net transport toward the inlet that reduces its area. Tides also sustain the presence and geometry of tidal deltas. Microtidal inlets have flood tidal deltas, while at meso- and macrotidal coasts ebb deltas and sand ridges are more important (Hubbard et al., 1979; Fig. 9-12). Maximum currents occur at the inlets and channels. At the inlet there is a segregation of flows: the ebb channel is deep while two or more flood-dominated channels are shallower; while on the
COASTAL LAGOONS
253
Fig. 9-9. Septation effects in Laguna dos Patos (modified after Toldo, 1991).
inner mouth, the reverse is true for a major flood channel and lateral ebb-channels. Sediment concentration varies in relation to the tidal current velocity (Fig. 9-13).
Longshore-driji effects The amount of longshore drift is a major constrain for coastal lagoons since it can produce their complete obstruction from the sea. It was largely indicated that waves create inlets, tides maintained them and longshore currents close the inlets (Lucke, 1934). There are many case studies about inlet migration rate and temporal closure. Occasionally, the longshore drift led to inlets narrowing instead of their migration (Fitzgerald et al., 1984; Fig. 9-14). Preferably in microtidal coasts, the drift volume conditions the inlet stability. Ecological and commercial purposes require inlets operable and therefore lead to the proposal of several hydraulic stability criteria. Bypassing plants are usually established to maintain operable inlets on alluvial shores (Bruun, 1981). Further
254
F.I. ISLA TIDAL RANGE M
0
N
B
L
I
I
mtcrolidol
mesolidol
+
m
I
macrotidal
description of the relation between the stability of tidal inlets against littoral drift will be provided in the Tidal Inlets section. The modifications suffered by the tidal deltas due to this process will be described in the corresponding section.
RELATED ENVIRONMENTS
The general morphology and sediment distribution of the lagoon is highly dependent on the barrier conditions, number and distribution of inlets, tidal range and sediment supply. Hereinafter, a description of the environments related to coastal lagoon geomorphology and evolution is presented.
Tidal inlets Tidal entrances are the key environments in the behaviour of coastal lagoons, since inlets condition sediment, salinity, nutrients, pollutant and organism dispersals between the lagoon and the ocean. When they facilitate navigation, coastal-lagoon areas may become good harbours and developed regions. On alluvial shores, tidal inlets owe their existence to the tidal flow that flushes the sediment carried into the throat by littoral currents on both sides of the entry (Bruun, 1966). In summary, a tidal inlet is a restricted, relatively narrow channel developed across a barrier where tidal currents are accelerated in a jet-like fashion. Bruun (1969) recognized that channel bottoms change in relation to mean maximum velocity. He realized that for several tidal inlets, maximum velocity ranges from
COASTAL LAGOONS
Io2
255
10'
lo5
106
Ac(ft2)
Fig. 9-11. Tidal prism vs. minimum flow areas for inlets with one or two groynes (modified after Jarrett, 1976). In Mar Chiquita inlet, tidal prism is a wedge that affects only the channel.
Fig. 9-12. Morphological differences between A) tide-dominated, B) wave-dominated, and C) transitional tidal inlets for the Georgia embayment (after Hubbard et al., 1979).
0.9 to 1.05 m/s when they were subjected to heavy littoral drift, meanwhile at inlets controlled by jetties the velocities increased to 1-1.1 m/s. These mean maximum velocities originate shear stresses in the range of 34-39 dynes/cm2 for exposed inlets and 49-54 dynes/cm2 for those controlled with jetties. The indicated shear stresses (considering typical fine sands 0.125-0.250 mm in diameter) correspond to the transition from flat crested dunes to plane-bed configuration. Furthermore, Mehta et al. (1976) confirmed that inertia effects are dominant during slack water while bed-friction effects are important during most of the flood and ebb periods.
256
F.I. ISLA CS (rn/sec)
1
1
lo
I 1
I 1
140-
-05
0-
Fig. 9-13. Variations in tidal current velocity (CS) and suspended sediment concentrations (CSS) during a semidiurnal tide in Chesapeake Bay.
Oo0
r"
1 TOTAL BARRIER
60 000 \
F
L
2
40 000
h
1
' ''.\\,
TOTAL DRAINAGE AREA
a, c 4
20 000
0I 1650
-
3 1750
0 1860
0 1960
Fig. 9-14. Variations in the length of barriers, drainage area and inlet width along the Frisian Island coastline (Fitzgerald et al., 1984).
A major issue comprising tidal inlets is the degree of stability of the feature. For ecological and economical purposes, a stable, deep inlet driving into a lagoon is required thus leading to several hydraulic stability criteria. Most studies of inlet stability considered the original O'Brien (1931) approach assuming that the minimum cross-sectional area A c (ft2),measured below the mean sea level (MSL) is a function of the tidal prism P (ft3) corresponding to the semidiurnal spring tide. Escoffier (1940) sustained that a stable inlet cross-section is maintained by the balance between the amount of sediment scoured by tidal currents and the amount supplied by the longshore drift. In the relationship between minimum flow area and tidal prism established by O'Brien (1969), A c can decrease when a jetty stabilized the inlet (Kieslich, 1981). Jarrett (1976) considered both these effects and the differentiation between inlets without jetty, with one jetty and with two jetties. = 5.37 x 10-6~1.07
(9-1)
This expression is used by Isla (1986) to locate the Mar Chiquita Inlet (Fig. 9-11).
COASTAL LAGOONS
257
On the other hand, Bruun (1966,1981) proposed the relation between tidal prism ( P ) and drift amount ( M ) to determine roughly the stability of tidal inlets. Bruun (1978) identified the following limits: P / M > 150 conditions are relatively good, little bar and good flushing; 100 < P / M < 150 conditions become less satisfactory, and offshore bar becomes more pronounced; 50 < P / M < 100 entrance bar may be rather large, but there is usually a channel through the bar; 20 < P / M < 50 all inlets are typical “bar-bypassers”;waves break over the bar during the storms, and the reason why the inlets “stay alive” is because of freshwater input due to the rain; P / M < 20 entrances became unstable rather than permanent inlets. Based on these criteria, McBride (1987) recognized, for the Florida coast tidal inlets, wave-dominated (t50), transitional and tide-dominated (>150) conditions (see Fig. 9-12 for the geomorphologic shape associated to each particular domain). O’Brien (1980) proposed another inlet closure criterion (ICC): the relation of tidal prism power to wave power,
where a() is the tidal amplitude; g is the acceleration of gravity; b is the channel width; T, is the wave period; Tt is the tidal period, and HO is the wave height. This closure coefficient was used to classify inlets as wave, transitional and tide dominated (Hubbard et al., 1979; Fig. 9-12). In an attempt to approximate closure feasibility, it is useful to determine shoal accumulation. Therefore, several authors have monitored sand volumes along the coast of Florida (Marino and Mehta, 1986), South Carolina (Fitzgerald and Fitzgerald, 1977), North Carolina (Jarrett, 1976), and along the east coast of USA (Walton and Adams, 1976). Other possible solution against inlet closure could be to open another inlet in a different situation (drift amount, wave or wind effects). Using the stability criteria proposed by Escoffier (1940), van de Kreeke (1985) suggested stability fields to predict whether a two-inlet system would operate or one of them can be subject to scour or shoaling (Fig. 9-15). Dealing with tidal inlet morphology, significant differences were recognized between the dominance of waves, tides or both (Hubbard et al., 1979). The morphology is related to the time-velocity asymmetry of tidal currents (Fig. 9-13). Usually, maximum ebb currents occur close to low water and can continue flowing seawards along the main channel while the tide is coming up (Hayes, 1980).
Tidal deltas Tidal deltas are sand accumulations produced by the action of flood or ebb currents at the respective entrances of tidal inlets. Tidal range and wave action are
F.I. ISLA
Fig. 9-15. Equilibrium stability curves for a coastal lagoon system with two tidal inlets (after van de Kreeke, 1985).
Fig. 9-16. Idealized bedform distribution for an ebb-tidal delta (after Hayes, 1980).
the main processes responsible for the shape of the deltas, although the ebb currents are more important for the case of ebb tidal deltas, which form seaward of the inlet. Flood-tidal deltas accumulate inland of the inlet gorge and their development is mostly produced by the action of flood currents. Ebb-tidal deltas are more common on mesotidal coasts and consist of a main ebb channel flanked by two banks with marginal flood channels and swash bars, forming sand ramps (Hayes, 1980; Fig. 9-16). On the ebb-tidal delta, wave swash impedes that ebb currents increase the net sand transport in the direction of flood-tidal currents and forming swash bars (Fitzgerald et al., 1976). At the Essex River ebb-tidal delta, the period between swash bar formation in the terminal lobe and their eventual attachment to the beaches is of approximately 5-7 yr (Smith and Fitzgerald, 1994). Estimates of sediment transport rates indicated that the volume transported over
259
COASTAL LAGOONS rn MARSH
\
C L A M / MUD F L A l
EBB
2
ENVlRONMENl SWASH BAR
EBB SHIELD FLOOD - T I D A L DELTA
-
DOMINANTLY FLOOD ORIENTED SAND WAVES
SHALLOW CHANNEL
FLOOD C H A N N E L DOMINANTLY EBB - ORIENTED S A N D WAVES
EBB CHANNEL
DEEP CHANNEL
_c ~
SEA
Fig. 9-17. Idealized stratigraphic sequences for the regression of a flood-tidal delta and an ebb-tidal delta (after Hayes, 1980).
the swash platforms is an order of magnitude greater than the onshore transport via bar complexes (Smith and Fitzgerald, 1994). The segregation of flows produces a longer duration of the flood and higher mean ebb-tidal currents at the main channel. Thus, the flushing ability of the main channel denotes the stability of mesotidal inlets where there is a significant longshore transport (Fitzgerald et al., 1976). The terminal lobe constitutes the inertia deceleration of the ebb flow. Fripp Inlet (South Carolina) and San Antonio Inlet (Argentina) are good examples of ebb-tidal deltas. The ebb-tidal delta sedimentary sequence consists of sand waves oriented seawards (deep channel) with shells or mudballs as lag deposits, overlaid by megaripples or dunes (shallow channels). The top of the sequence is dominated by sand waves or dunes oriented towards the inlet and caused by the action of waves (swash bars; Hayes, 1980; Figs. 9-16,9-17). The sand ramp of flood-tidal deltas is landwards of the inlet. The flood channel usually ends in an ebb shield and flanked by sand banks (ebb spits; Hayes, 1980). The Essex estuary (Hayes, 1980) and the Mar Chiquita inlet (Buenos Ares) are good examples of flood-tidal deltas. In the flood-tidal delta sequence, the bottom can be made up by transverse bedforms oriented in the direction of the ebb flow (ebb channel). Above, the sand ramp is preferentially dominated by megaripples or dunes oriented landwards. This flood-tidal delta is capped by muds and peats representing mudflats and marshes (Hayes, 1980; Fig. 9-17). At the Keurbooms macrotidal inlet (S. Africa), Reddering (1983) differentiated the depositional sequences between the tide- and wave-dominated zones. Within the tide-dominated zone, crossbeds related to the flood channel is recognized. Whereas in the wave-dominated zones, the ebb-tidal delta deposits are characterized by higher energy and thin crossbeds oriented to the sea (Reddering, 1983). Dealing with the environments of Mar Chiquita microtidal inlet, Isla Mendy (1989) compares the sequences of structures related to an inlet (of 2 m depth) and the beach (Fig. 9-18). At a depth of approximately one meter, the inner rough
260
F.I. ISLA
Fig. 9-18. Compared evolution model for a beach and a tidal inlet under a microtidal regime (modified after Isla Mendy, 1989).
facies of the beach (sensu Clifton et al., 1971) are significantly different from the ebb-oriented facies of the ebb channel or from sets dipping towards the lagoon that were produced in a similar way as the migration of a berm crest on a spit. At the highest positions of the sequence, thick strata with heavy-mineral concentrations are characteristic of the backshore (Isla Mendy, 1989). For the mesotidal inlets of the Dutch coast, Boersma and Terwindt (1981) described bundles due to the effect of tides. For the flood-tidal delta of the Westerschelde estuary, they recognized different stages in the effect of the flood. They also introduced different concepts in the interpretations of tidal sequences. The bundles are separated by pause planes representing either erosive, non-erosive or depositional (mud drapes) periods (Boersma and Tenvindt, 1981). Within the bundles, reactivation
COASTAL LAGOONS
261
sui$aces correspond to the acceleration period. Full-vortex structures reflect the full stage of the tide, while slackening structures represent the deceleration period. However, on sandy shoals several kilometres landwards of the Eastern Scheldt inlet, these tidal-bundle cycles do not strictly define the sedimentation process. At Galgeplaat Shoal, bedload transport is absent and transport during neap tides is insignificant. Erosion takes place during storms and sedimentation occurs during spring tides, fair weather conditions (Kohsiek et al., 1990). Barriers Barriers may form, in stable sea-level conditions, by the longshore growing of complex spits (Fischer, 1968). Inlets migrations vary from 2 to 90 m/year at the east coast of USA (Hayes, 1980). Inlets between the Frisian Islands (Germany) migrate due to westerlies and other processes at a rate of 20-46 m/year (Fitzgerald and Penland, 1987). In relation to the degree of blocking of spitsbarriers, coastal lagoons are classified into estuarine-lagoons, “open” lagoons, partly closed lagoons and closed lagoons (Nichols and Allen, 1981; Fig. 9-19). In a similar way and regarding the shoreline continuity index and headland interactions, spits and barriers were classified into fringing beaches, flanking spits, flying spits and linking barriers (Carter et al., 1987; Fig. 9-20). It was proposed that ebb-tidal deltas deposits associated to migrating inlets provide the initial sand body for the development of shoreface-attached sand ridges. Linearity and obliquity of these sand ridges are a function of shoreline transgression, lateral inlet migration and wave reworking (McBride and Moslow, 1991). Although it is not associated to a coastal lagoon, a series of shoreface-connected sand ridges (G6mez and Perillo, 1992) have developed offshore Bahia Blanca Estuary as a continuation of ebb deltas highly modified. Not only longshore growth of barriedspits affects inlets, the rate of landward growth conditions directly the infilling rate of the lagoons parallel to the shore. Onshore persistent winds produce the migration of sand dunes and backbarrier accumulations in their transition to marshes. To counteract coastal dune erosion, sand fencing has been proposed in several places. Parallel sand fencing probed to be a good way to fix dunes even in coastal areas where sand is deficient and with subsidence (Mendelssohn et al., 1991). Sand barriers can be built of ridges or dunes; in gravel barriers each ridge is clearly recognized. Sand or shell ridges are not part of the backshore but they are closely related to it. As the beach ridges are produced by storms, they generally exhibit little relief (Leatherman, 1982). In the transition to barrier dunes, dune ridges are beach ridges capped by wind-blown sand. Leatherman (1982) pointed to a biogeochemical mechanism for building beach ridges: As a spit grows into open water, an above-water platform originates due to the transport of waves and currents. Washovers transport beach grass fragments and seeds to the crest of this platform. Plants then begin to grow up while algae and other plants provide nutrients, and they soon become large enough to trap aeolian and subaqueous transported sand (Leatherman, 1982). ASthe
F.I. ISLA
262 4
WDYF
OCtlO"
r
Fig. 9-19. Classification of coastal lagoons in relation to tides, waves and longshore currents (modified after Nichols and Allen, 1981).
beach ridge plain grows, grasses characterize the dune ridges, while wetland species cover the interdune flats. Paired and arcuate sand ridges mark the location of former infilled inlets (Leatherman, 1982). The formation of duned barriers depends on the availability of sand and high wave and wind energy (Short, 1988). The heights of the dunes are controlled by sand size and wind velocity (Leatherman, 1982). There is a "wind limit" for a given grain size above which sand cannot be retarded by vegetation. As in temperate areas dune migration patterns are rather difficult that prevail without vegetated patches, it is common to find wind action concentrated in lower areas known as blowouts (Leatherman, 1982). Sand barriers grew during the initial stages of the Holocene
COASTAL LAGOONS
263
0.
.
--..-
1
SECTION VIEW B
. ... . . .R, . . .-... -
B
HEADLAND INTERACTION B / R
Fig. 9-20. Classification of barriers and spits in relation to the interaction between headlands and the shoreline continuity index (after Carter et al., 1987).
transgression. For the Australian coast, sand was abundant during the postglacial transgression (10,000-6,000 yr) and became exhausted by mid-Holocene (Short, 1988).
Tidal flats Tidal flats (see also Amos, this volume) are environments of net accumulation within coastal lagoons, and may influence their dynamics, heat exchange and nutrient inputs. Tidal flats can also play an important role in the heat balance of coastal lagoons with significant tidal ranges (van Boxel, 1986). Furthermore, heat exchange may induce evaporation effects and bioturbation of the sediments modifying the erodability of the material (Piccolo et al., 1993). Sand, mixed and mudflats are discriminated in relation to the sediment availability and dynamics (waves, winds). Another factor that conditions the physics of tidal flat sediments is the subaerial exposure (number of days exposed during neap-spring cycles; Amos et al., 1988). Tidal flats can be dominated by episodes (storms, episodic sediment inputs) leading to a chenier plain, or by cycles. Common cycles are flood-ebb, neap-spring and summer-winter. These three cycles were particularly recognized and described in Cobequid Bay-Salmon River estuary (Dalrymple et al., 1990a, b; Fig. 9-21).
Marshes Marshes (see also Luternauer et al., this volume) are very specified vegetated environments subject to sedimentation. Sedimentation rates increase in relation to tidal ranges (Harrison and Bloom, 1977). Their production of particulate organic carbon (POC) made them very important areas for recruitment. They have also been referred as natural tertiary treatment plants because of their assimilation ability to inorganic nutrient excess (Valiela and Teal, 1979; Sherr and Payne, 1981). Some beneficial aspects of tidal marshes are usually disregarded when they are intensively
264
F.I. ISLA MOUTH OF, ESTJARY
HEAC
OF
’ESTJARY
Cobeauid boy mlnas basin
Powerline section
BlGCk rock section r
JUNE-DECEMBER
_ -_-_--_------- - MARCH- MAY
Bridge section
: ,---7 5-
c
c
NEAP-SPRING CYCLE
SPRING NEAP
SPRING
ll
Biotdrboted mJd
H
Ice- p r o d x e d deformation
H
Tidal beddiig
E3
I c e - r a f t e d pebbles
Dl
Escape b u w w
Fig. 9-21. Schematic record of the summer-winter and neap-spring cycles along the estuary of Cobequid Bay-Salmon River (Dalrymple et al., 1990b).
occupied or separated from the tidal influence by the construction of roads. At high latitudes, annual frost and permafrost are the determined factors for tidal flats (Dionne, 1989a). Marsh clumps can be transported by ice floes ground during low tides (Dionne, 1989b). During winter, Salicomia can survive below the ice (Isla et al., 1991). In some coastal areas, salt marshes are threatened by sea-level rise. In order to test if modern accretion rates could stand present rates of compaction, sea-level rise and submergence, ecological models were proposed (Chmura et al., 1992). Other factors are frequency and duration of floodings, suspension concentration, tidal range, storm frequency, exposure to wave attack, vegetation characteristics and biodegradation (Oenema and de Laune, 1988). For the marshes within the Eastern Scheldt (The Netherlands), mean accretion rates vary between 0.4 and 1.5 cm/yr. Vertical accretion exceeds sediment losses induced by the retreat of the marsh cliffs by a factor of 10-20 (Oenema and de Laune, 1988).
Mangroves Mangroves (see also Augustinus, this volume) are halophytic (salt tolerant), woody, seed-bearing plants ranging from trees to shrubs (Snedaker and Getter, 1985). These environments characterize subtropical and tropical lagoons as mangroves require minimum wave activity, salt and freshwater (the hydroperiod or seasonal flooding), nutrients and oxygen (Snedaker and Getter, 1985). Wind waves have minor importance, as mangroves grow where waves are absent (Kjerfve, 1990). Muddy sediments with high silt content are optimum for mangrove’s growth. Leaves deposition acidify the soil and thus favour mangrove dominance. Six species (4 genera) dominate America’s mangroves while 70 species (45 genera) are found in the Indo-Pacific mangroves (Kjerfve, 1990). Their productivity span between 30 and 2,000 gC/m2/year (Mann, 1982). In a physiographic attempt of classification of Florida mangroves (Lug0 and Snedaker, 1974), only fringe and ovenvash forests would be related to coastal lagoons.
COASTAL LAGOONS
265
COASTAL LAGOON EVOLUTION
Geographic or engineering approximations to coastal lagoon behaviour fail to recognize their evolutionary stage. Relative sea-level trends and tectonics should be considered when recognizing long-term coastal lagoon behaviour. Bedform classifications restricted to areas of sea level rising (most of Northern Hemisphere) are seldom useful to uplifting coasts or where sea level is stable or dropping (Southern Hemisphere). Coastal lagoons evolve in relation to the accumulation rate and the sea-level behaviour. Coastal-lagoon infilling is an expected process in regressive coasts. However, in transgressive phases, their evolution depend on the sedimentation rate: wether it exceeds sea-level rise (“surplus”) or this rate is under the transgressive rate (“deficit”; Nichols et al., 1986). Surplus lagoons are, therefore, characterized by shallow depths, low water capacity and deltaic processes. In deficit lagoons, the bathymetry is controlled by antecedent topography with drowned valleys; bay-head deltas are limited (Nichols et al., 1986). Many coastal lagoons from Argentina, Australia, Brazil, Mexico and USA, are controlled by littoral deposits related to the last interglacial. Most Holocene barriers are reworking Sangamon or Eemian barriers, and showing surprising similarities in the surmounted patterns of lagoons and barriers. It has been stated that the rolling topography with hills and basins of coralline lagoons may have resulted from extensive limestone solution during Pleistocene low sea levels (Shepard, 1970). Although inlets’ competence and closure are common phenomena in coastal lagoons, in the Southern Hemisphere there are evidences of other related environmental changes during the Holocene regression. Long-term beach drift reversals have been suggested for the Nayarit coast, Mexico (Curray et al., 1969), Mar Chiquita Lagoon, Argentina (Isla Mendy, 1989) and Doce River inlet, Brazil (Martin and Suguio, 1992). In the sense of Herz and Mascarenhas (1993), the cellular structure of Dos Patos Lagoon proves a high maturity degree of the system, where connections between cells facilitate an interchange of water and matter. At the Gippsland Lakes, Bird (1978) explained that where ecological conditions, specially sediment supply, are suitable for reed swamps, shrinkage of lagoons would dominate. On the other hand, where waves and currents rework lagoon’s shorelines, segmentation would be dominant. Other environmental natural changes during Late Holocene have critic effects on coastal lagoons. In Mar Chiquita, mass mortalities were related to a replacement of the infaunal communities with epifaunal assemblages (Isla and Rivero D’Andrea, 1994). In Langebaan Lagoon, South Africa, and in relation to the Holocene quasistillstand, there was a significant increase in sedimentation for the last 2,000 yr (Flemming, 1988). Many other factors condition coastal lagoon evolution. They are either inherited (bedrock, morphology) or physical factors (tidal currents, river and wave dynamics). In the Australian coast, sea level has been approximately stable for the last 6,000 yr (Roy, 1984). In the NSW coast, Roy (1984) proposed an estuarine classification considering: drowned river valley estuaries (ria-like), barrier estuaries (coastal la-
266
F.I. ISLA
I SALINITY ran-
and variabtlity
1. DROWNED RIVER VALLEY
I
ESTUARY
more
WIND WAVES water mixing: bottom TURBIDITY
waves t river flaw
HABITAT +SUBSTRATE diversity MANGROVES and SALT MARSH SEAGRASSES
species diverrlty
Fig. 9-22. Evolution model - including water quality and biota - for three estuary types from New South Wales (modified after Roy, 1984). There is no correlation between estuaries of different type.
goons) and saline coastal lakes (blind coastal lagoons). Roy attempted to describe the environmental or biology changes related to the infilling of estuaries with a stable sea level. Three stages were proposed for these three estuary types (Fig. 9-22). In the youthful stages (A), all estuaries are deep with organic-rich muds supporting the infauna communities (high density and stable population). Salinity is consistently high and waters relatively well oxygenated, except in closed lagoons where the fauna is limited to salinity (Fig. 9-22). At intermediate stages (B) of infilling, intertidal and shallow subtidal environments expanded. Fauna densities and specie diversity increase as shorelines are more complex and sediment types diversify (Fig. 9-22). Increase in tidal ranges favour mangroves but more turbidity affects seagrasses and intolerant fauna. Excessive turbidity and nutrient availability could lead to algal blooms that may cause oxygen depletion (Roy, 1984). In mature stages, plants and animals decline their distribution, population and species diversity. Flood plains expand in replacement of intertidal environments. Significant variations and salinity and turbidity restrict faunal population and diversity (Roy, 1984).
SUMMARY
Coastal lagoons are environments of different origins. Sea level either rising, dropping or stable could lead to a coastal lagoon. Depending on the sediment input rate, Southern Hemisphere lagoons would be filling while in the Northern Hemisphere flooding would be expected. Biochemical processes control low-latitude
COASTAL LAGOONS
267
coastal lagoons while physical processes are more significant in high latitudes. There, the sediment availability (sand and gravel) and wave regime control the longshore growing of sand barriers and gravel spits. Tidal inlets are critic in coastal-lagoon behaviour and evolution. They control the sediment and water exchange but also condition the ability of the lagoons for navigation, fisheries (recruitment and harvesting) and aquaculture. Tidal deltas and tidal flats are very important sedimentation areas because they can record (with different energy conditions and sedimentation rates) each tide, each neap-spring cycle and each summer-winter season.
ACKNOWLEDGEMENTS
E. Toldo (CECO, Porto Alegre) provided unpublished information about Laguna dos Patos (“mais grande do mundo”). M. Farenga, M. J. Bo, M. Tom& and M. V. Bernasconi did the draftings. G.M.E. Perillo (IADO, Bahia Blanca) made useful comments to the first draft.
REFERENCES Amos, C.L., van Wagoner, N.A. and Daborn, G.R., 1988. The influence of subaerial exposure on the bulk properties of fine-grained intertidal sediments from Minas Basin, Bay of Fundy. Est. Coastal Shelf Sci., 27: 1-13. Atwater, B.F., 1987. Evidence for great Holocene earthquakes along the outer coast of Washington. Science, 236: 942-944. Bannink, B.A., van der Meulen, J.H.M. and Nienhuis, P.H., 1984. Lake Grevelingen: from an estuary to a saline lake: an Introduction. Neth. J. Sea Res., 18: 179-190. Bartsch-Winkler, S. and Schmoll, H.R., 1984. Bedding types in Holocene tidal channel sequences, Knik Arm, Upper Cook Inlet, Alaska. J. Sediment. Petrol., 54: 1239-1250. Blivi, A., 1993. The geomorphological evolution of the Volta Delta during the Holocene. Int. Symp. on Quater. Coastal evolution: Models, Processes and Local to Global Factors. Extended abstracts, IGCP 274 Final Meeting, Oostduinkerke, Belgium, pp. 11-14. Boersma, J.R. and Tenvindt, J.H.J., 1981. Neap-spring tide sequences on intertidal shoal deposits in a mesotidal estuary. Sedimentology, 28: 151-170. Bird, E.F.C., 1978. The geomorphology of the Gippsland Lakes Regions. Ministry for Conservation, Australia, Pub. 186, 158 pp. Boothroyd, J., 1985. Tidal inlets and tidal deltas. In: R.A. Davis (Editor), Coastal Sedimentary Environments. Springer-Verlag, New York, pp. 445-532. Bruun, P. 1966. Tidal Inlets and Littoral Drift. Oslo Univ. Book Co., 193 pp. Bruun, P., 1969. Tidal inlets on alluvial shores. In: A. Ayala-Castafiares and F.B. Phleger (Editors), Lagunas Costeras, un Simposio. UNAM-UNESCO, Mexico, pp. 349-366. Bruun, l?, 1978. Stability of Coastal Inlets: Theory and Engineering. Elsevier Sci. Publ. Co., Amsterdam, 509 pp. Bruun, P., 1981. Port Engineering. Gulf Publ. Co., Houston, TX, 800 pp. Calliari, L.J. and Fachin, S., 1993. Laguna dos Patos: influtncia nos depdsitos lamiticos costeiros. Pesquisas, 20: 57-69. Carter, R.W.G., Johnston, TW. and Orford, J.D., 1984. Stream outlets through mixed sand and gravel coastal barriers: examples from southeast Ireland. Z. Geomorphol., 28: 427-442.
268
F.I. ISLA
Carter, R.W.G., Orford, J.D., Forbes, D.L. and Taylor, R.B., 1987. Gravel barriers, headlands and lagoons: an evolutionary model. Coastal Sediments ’87, ASCE, New Orleans, LA, pp. 1776-1792. Clifton, H.E., Hunter, R.E. and Phillips, R.L., 1971. Depositional structures and processes in the non-barred high-energy nearshore. J. Sediment. Petrol. 41: 651-670. Coutellier, V. and Stanley, D.J., 1987. Late Quaternary stratigraphy and paleogeography of the Eastern Nile delta, Egypt. Mar. Geol., 77: 257-275. Curray, J.R., Emmel, F.J. and Crampton, P.J.S., 1969. Holocene history of a strand plain, lagoonal coast, Nayarit, Mexico. In: A. Ayala-Castanares and F.B. Phleger (Editors), Lagunas Costeras, un Simposio. UNAM-UNESCO, Mexico, pp. 63-100. Chmura, G.L., Costanza, R., and Kosters, E., 1992. Modelling coastal marsh stability in response to sea level rise: a case study in coastal Louisiana, USA. Ecol. Model., 64: 47-64. Dalrymple, R.W., Knight, R.J., Zaitlin, B.A. and Middleton, G.V., 1990a.Dynamics and facies model of a macrotidal sand-bar complex, Cobequid Bay-Salmon River estuary (Bay of Fundy). Sedimentology, 37: 577-612.
Dalrymple, R.W., Makino, Y. and Zaitlin, B.A., 1990b. Temporal and spatial patterns of rhythmite depositon on mudlfats in the macrotidal Cobequid Bay-Salmon River estuary, Bay of Fundy, Canada. In: D.G. Smith, G.E. Reinson, B.A. Zaitlin and R.A. Rahmani (Editors), Clastic Tidal Sedimentology. Can. SOC.Petrol. Geol. Mem., 16: 137-160. Davis, R.A., Fingleton, W.G., Allen, G.R., Crealese, Ch.D., Johanns, W.M., O’Sullivan, J.A., Reive, C.L., Scheetz, S.J. and Stranaly, G. L., 1973. Corpus Christie Pass: A hurricane modified tidal inlet on Mustang Island, Texas. Contributions in Mar. Sci., 17: 123-131. de Beaumont, E.L., 1845. LeGons de GCologie Pratique. 7me LeGon - Levtes de Sables et Galets. P. Bertrand, Paris, pp. 221-252. Dionne, J. C., 1989a. The role of ice and frost in tidal marsh development. A review with particular reference to Quebec, Canada. Essener Geogr. Arbeiten., 18: 171-210. Dionne, J.C., 1989b. An estimate of shore ice action in a Spartina tidal marsh, St. Lawrence estuary, QuCbec, Canada. J. Coastal Res., 5: 281-293. Emery, K.O., 1967. Estuaries and lagoons in relation to continental shelves. In: Lauff, J. (Editor), Estuaries. AAAS Washington DC, pp. 9-11. Escoffier, F.F., 1940. The stability of tidal inlets. Shore and Beach, 8(4), 114-115. Evans, G., Schmidt, V., Bush, P. and Nelson, H., 1969. Stratigraphy and geologic history of the sabkha Abu Dhabi, Persian Gulf. Sedimentology, 12: 145-159. Fairbridge, R.W., 1980. The estuary: its definition and geodynamic cycle. In: E. Olausson and I. Cat0 (Editors), Chemistry and Biogeochemistry of Estuaries. Wiley, NY, pp. 1-35. Fisher, J.J., 1968. Barrier island formation: discussion. Geol. SOC.Am. Bull., 79: 1421-1426. Fitzgerald, D.M., Fico, C. and Hayes, M. O., 1979. Effects of the Charleston Harbor, S.C. jetty construction on local accretion and erosion. Proc. Coastal Structures Conf., ASCE, Alexandria, Virginia, pp. 641-664. Fitzgerald, D.M. and Fitzgerald, S.A., 1977. Factors influencing tidal inlet throat geometry. Coastal Sediments ’77. ASCE, Charleston, pp. 563-581. Fitzgerald, D.M. and Penland, S., 1987. Backbarrier dynamics of the East Freisian Islands. J. Sediment. Petrol., 57: 746-754. Fitzgerald, D.M., Penland, S. and Nummedal, D., 1984. Control of barrier island shape by inlet sediment bypassing: East Frisian Islands, West Germany. Mar. Geol., 60: 355-376. Flemming, B.W., 1977. Langebaan Lagoon: a mixed carbonate-siliciclastic tidal environment in a semi-arid climate. Sediment. Geol., 18: 61-95. Flemming, B.W., 1988. Process and pattern of sediment mixing in a macrotidal coastal lagoon along the west coast of South Africa. In: P.L. de Boer (Editor), Tide-Influenced Sedimentary Environments and Facies. D. Reidel Publ. Co., pp. 275-288. Forbes, D.L., Taykir, R.B., Orford, J.D., Carter, R.WG. and Shaw, J., 1991. Gravel-barrier migration and overstepping. Mar. Geol., 97: 305-313. Glaesser, J.G., 1978. Globaldistribution of barrier islands in terms of tetonic setting. J. Geol., 86: 283-297.
COASTAL LAGOONS
269
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). Harrison, E.Z. and Bloom, A.L., 1977. Sedimentation rates on tidal salt marshes in Connecticut. J. Sediment. Petrol., 47: 4. 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. Hayes, M.O., 1980. General morphology and sediment patterns in tidal inlets. Sediment. Geol., 26: 139-156. Herz, R.and Mascarenhas, S., 1993. Patos Lagoon. In: J. Sorensen, F. Gable and F. Bandarin (Editors), The Management of Coastal Lagoons and Enclosed Bays. Coastal Zone '93, ASCE, New Orleans, LA, pp. 52-64. Howard, J.D. and Frey, R.W., 1985. Physical and biogenic aspects of backbarrier sedimentary sequences, Georgia Coast, USA. Mar. Geol., 63: 77-127. Hoyt, J.H., 1967. Barrier island formation. Geol. SOC.Am. Bull., 78: 1125-1136. Hoyt, J.H., 1970. Development and migration of barrier islands, Northern Gulf of Mexico: discussion. Geol. SOC.Am. Bull., 81: 3779-3782. Hubbard, D.K., Oertel, G. and Nummedal, D., 1979. The role of waves and tidal currents in the development of tidal-inlet sedimentary structures and sand body geometry: examples from North Carolina, South Carolina and Georgia. J. Sediment. Petrol., 4 9 1073-1092. Isla, EI., 1986. Dinimica sedimentaria de la zona de la desembocadura de la laguna Mar Chiquita, provincia de Buenos Aires. Unpublished thesis, UNLP, La Plata, 257 pp. Isla, F.I. and Rivero D'Andrea, I. 1994. Procesos retrotafon6micos en secuencias de aguas someras. Pesquisas, 20: 90-95. Isla, F.I., Vilas, F.E., Bujalesky, G.G., Ferrero, M., Bonorino, G.G. and Miralles, A.A., 1991. Gravel drift and wind effects on the macrotidal San Sebastian Bay, Tierra del Fuego, Argentina. Mar. Geol., 97: 211-224. Isla Mendy, F.I., 1989. Modelo evolutivo comparado de playa y boca de micromareas: Mar Chiquita, Argentina. XI1 Congr. Espafiol de Sedimentologia. Com., Leioa, Bilbao, pp. 121-124. Jarrett, J.T., 1976. Tidal prism-inlet area relationships. GITI Rep. 3, US Army CERC, Ft. Belvoir, Va., pp. 32. Johnson, D.W., 1919. Shore Processes and Shoreline Development. J. Wiley and Sons, New York, 584 pp. Kieslich, J.M., 1981. Tidal inlet response to jetty construction. GITI Rep. 19, US Army CERC, Ft. Belvoir, Va., pp. 63. Kjerfve, B., 1986. Comparative oceanography of coastal lagoons. In: D.A. Wolfe (Editor), Estuarine Variability. Academic Press, New York, pp. 63-81. Kjerfve, B., 1990. Manual for investigation of hydrological proceses in Mangrove ecosystems. Baruch Inst. Mar. Biol.. Marine Sci. Prog., 35 pp. Kjerfve, B. and Magill, K.E., 1989. Geographic and hydrodynamic characteristics of shallow coastal lagoons. Mar. Geol., 88: 187-199. Kohsiek, L.H.M., Buist, H.J., Bloks, P.,Van der Berg, J.H. and Visser, J., 1988. Sedimentary processes on a sandy shoal in a mesotidal estuary (Oosterschelde, The Netherlands). In: P.L. de Boer (Editor), Tide-Influenced Sedimentary Environments and Facies. D. Reidel Publ. Co., 201-214. Leatherman, S.P., 1982. Barrier Island Handbook. Univ. of Maryland, Baltimore, 109 pp. Leontyev, O.K.K. and Nikiforov, L.G., 1966. An approach to the problem of the origin of barrier bars. Int. Oceanogr. Congr., 2nd Abs., pp. 221-222. Lonard, R.I. and Judd, F.W., 1993. Recovery of vegetation of barrier island washover zones. Coastal Zone '93. ASCE, New Orleans, LA, pp. 2332-2342. Lucke, J.B., 1934. A theory of evolution of lagoon deposits on shorelines of emergence. J. Geol., 42: 561-584. Lugo, A.E. and Snedaker, S.C., 1974. The ecology of mangroves. Annu. Rev. Ecol. Syst., 5: 39-64. McBride, R.A., 1987. Tidal inlet history, morphology and stability, eastern coast of Florida, USA. Coastal Sediments '87, ASCE, New Orleans, LA, pp. 1592-1607.
270
F.I. ISLA
McBride, R.A. and Moslow, T E , 1991. Origin, evolution and distribution of shoreface sand ridges, Atlantic inner shelf, USA. Mar. Geol., 97: 57-85. Mann, K.H., 1982. Ecology of Coastal Waters: A System Approach. Univ. California Press, Berkeley, 322 pp. Marino, J.N. and Mehta, A.J., 1986. Sediment volumes around Florida’s East Coast tidal inlets. Interim Rep., Coastal and Ocean. Eng. Dept., Univ. Florida, Gainesville, 71 pp. Martin, L. and Suguio, K., 1992. Variation of coastal dynamics during the las 7000 years recorded in beach-ridge plains associated with river mouths: example from the central Brazilian coast. Palaeogeogr. Palaeoclimatol. Palaeoecol., 99: 119-140. Maynard, A.K. and Suter, J.R., 1983. Regional varaibility of washover deposits on the South Texas coast. Trans. Gulf Coast Assoc. Geol. Sci., 33: 339-346. Mayor-Mora, R.E., 1977. Laboratory investigation of tidal inlets on sandy coasts. US Army Corps of Eng., GITI Rep. 11, 106 pp. Mendelssohn, LA., Hester, M. W., Monteferrante, F.J. and Talbot, F., 1991. Experimental dune building and vegetative stabilization in a sand-deficient barrier island setting on the Lousiana coast, USA. J. Coastal Res., 7: 137-149. Mehta, A.J., Byrne, R.J. and De Alteris, J.T, 1976. Measurements of bed friction in tidal inlets. Proc. 5th Coastal Eng. Conf., Honolulu, pp. 1701-1720. Nichols, M.M. and Allen, G., 1981. Sedimentary processes in coastal lagoons. In Coastal lagoon research, present and future. UNESCO Tech. Pap. Mar. Sci., 33: 27-80. Nichols, F.H., Cloern, J.E., Luoma, S.N. and Peterson, D.H., 1986. The modification of an estuary. Science, 231: 567-573. Ntiamoa-Baidu, Y.and Hollis, G.E., 1992. Planning the management of coastal lagoons in Ghana. In: E. Maltby, P.J. Dugan and J.C. Lefeuvre (Editors), Conservation and Development: The Sustainable Use of Wetlands Resources. Proc. 3rd. Int. Wetland Conf., Rennes, IUCN, Gland, Switzerland, pp. 113-121. Nummedal, D., Penland, S., Gerdes, R., Schramm, W., Kahn, J. and Roberts, H., 1980. Geologic response to hurricane impact on low-profile Gulf coast barrier. Trans. Gulf Coast Assoc. Geol. Sci., 30: 183-195. O’Brien, M.P., 1931. Estuary tidal prism related to entrance areas. Civil Eng., 1: 738 O’Brien, M.P., 1969. Equilibrium flow areas of inlets on sandy coasts. ASCE J. Waterw. Harbours Div., 95: 43-52. O’Brien, M.P., 1980. Comments on tidal entrances on sandy coasts. Proc. 17th. Coast. Eng. Conf., 111, pp. 2504-2516. Oenema, 0. and De Laune, R.D., 1988. Accretion rates in salt marshes in the Eastern Scheldt, south-west Netherlands. Estuar. Coastal Shelf Sci., 26: 379-394. Oldale, R.N., 1985. A drowned Holocene barrier spit off Cape Ann, Massachesetts. Geology, 13: 375-377. Otvos, E.G., 1970. Development and migration of barrier islands, northern Gulf of Mexico. Geol. SOC. Am. Bull., 81: 241-246. Owens, E.H. and Roberts, H.H., 1978. Variations of wave-energy levels and coastal sedimentation, Eastern Nicaragua. Proc. Coastal Eng. Conf., ASCE, Hamburg, pp. 1195-1214. OyCdC, L.M. and Lang, J., 1993. Breaks in environmental equilibria and sedimentary discontinuities in a margino-littoral geosystem inherited from the Holocene transgression: The OuCme delta and “Lake” NokouC (Benin, West Africa). Int. Symp. on Quater. Coastal Evolution: Models, Processes and Local to Global Factors. Extended abstracts, IGCP 274 Final Meeting, Oostduinkerke, Belgium, pp. 107-110. Penland, S., Suter, J.R. and Boyd, R., 1985. Barrier island arcs along abandoned Mississippi River deltas. Mar. Geol., 63: 197-233. Pejrup, M., 1988. The triangular diagram used for classification of estuarine sediments: a new approach. In: de Boer, P. L. et al. (Editors), Tide-influenced sedimentary environments and facies. Reidel Publ. Co., pp. 289-300. Pezzani, S. and Obenat, S., 1988. Estudio integrado de la lagna costera Mar Chiquita, Provincia de Buenos Aires, Argentina. I. Caracteristicas de la poblacidn de Ficopomatus enigmaticus. Inf.
COASTAL LAGOONS
27 1
UNESCO Cienc. Mar, 47: 101-102. Phleger, F.B., 1969. Some general features of coastal lagoons. In: A. Ayala-Castafiares and F.B. Phleger (Editors), Lagunas Costeras, un Simposio. UNAM-UNESCO, Mexico, pp. 5-26. Phleger, F.B., 1981. A review of some general features of coastal lagoons. In Coastal lagoon research, present and future. UNESCO Tech. Pap. Mar. Sci., 33: 7-14. Piccolo, M.C., Perillo, G.M.E. and Daborn, G.R., 1993. Soil temperature variations on a tidal flat in Minas Basin, Bay of Fundy, Canada. Est., Coast. Shelf Sci., 35: 345-357. Pritchard, D.W., 1960. Lectures on estuarine oceanography. B. Kinsman (Editor), J. Hopkins Univ. 154 pp. Rampino, M.R. and Sanders, J.E., 1981. Evolution of the barrier islands of southern Long Island, New York. Sedimentology, 28: 37-48. Reddering, J.S.V., 1983. An inlet sequence produced by migration of a small microtidal inlet against longshore drift: the Keurbooms Inlet, South Africa. Sedimentology, 30: 201-218. Reineck, H.-E. and Singh, I.B., 1980. Depositional Sedimentary Environments. Springer-Verlag, Berlin, 549 pp. Roy, P.S., 1984. New South Wales estuaries: their origin and evolution. In: B.G. Thom (ed.), Coastal Geomorphology in Australia, Academic Press, Australia, pp. 99-121. Sanders, J.E. and Kumar, N., 1975. Evidence of shoreface retreat and in-place “drowning” during Holocene submergence of barriers, shelf off Fire Island, New York. Geol. SOC.Am. Bull., 86: 65-76. Schwartz, M.L., 1971. The multiple causality of barrier islands. J. Geol., 79: 91-94. Scott, A.J., Hoover, R.A. and McGowen, J.H., 1969. Effects of hurricane “Beulah”, 1967, on Texas coastal lagoons and barriers. In: A. Ayala-Castafiares and F.B. Phleger (Editors), Lagunas Costeras, un Simposio. UNAM-UNESCO, Mexico, pp. 221-236. Shepard, F.P., 1970. Lagoonal topography of Caroline and Marshall Islands. Geol. SOC.Am. Bull., 81: 1905-1914. Sherr, B.F. and Payne, W.J., 1981. The effect of sewage sludge on salt-marsh denitrifying bacteria. Estuaries, 4: 146-149. Short, A.D., 1988. Holocene coastal dune formation in Southern Australia: A case study. Sediment. Geol. 55, 121-142. Smith, J.B. and Fitzgerald, D.M., 1994. Sediment transport pattern at the Essex River Inlet ebb-tidal delta, Massachusetts, USA. J. Coastal Res., 10: 752-774. Snedaker, S.C. and Getter, Ch.D., 1985. Coastal Resources Management Guides. Coastal Management Publ. 2, National Park Service, Columbia SC, 205 pp. Swift, D.J.P., 1975. Barrier island genesis: evidence from central Atlantic shelf eastern USA. Sediment. Geol. 14: 1-43. Swift, D.J.P. and Thorne, J.A., 1991. Sedimentation on continental margins. I: A general model of shelf sedimentation. IAS, Spec. Publ., 14: 3-31. Toldo, E.E., 1989. 0 s efeitos do transporte sedimentar na distribucfio dos tamanhos de gr5o e morfodinamica da Lagoa dos Patos. Unpubl. Thesis, Univ. Fed. Rio Grande do Sul, Porto Alegre, 143 pp. Toldo, E.E., 1991. Morfodindmica da Laguna dos Patos, Rio Grande do Sul. Pesquisas, 18: 58-63. Tomazelli, L.J. and Wilwock, J.A., 1991. Geologia do sistema lagunar HolocCnico do litoral norte do Rio Grande do Sul. Pesquisas, 18: 13-24. Tudhope, A.H., 1989. Shallowing-upwards sedimentation in a coral reef lagoon, Great Barier Reef of Australia. J. Sediment. Petrol., 59: 1036-1051. Valiela, I. and Teal, J.M., 1979. The nitrogen budget of a salt marsh ecosystem. Nature, 280: 652-656. van Boxel, J., 1986. Heat Balance Investigations in Tidal Areas. Centre for Mathematics and Computer Science, Vrije Universiteit te Amsterdam, 135 pp. van de Kreeke, J., 1985. Stability of tidal inlets: Pass Cavallo, Texas. Estuar. Coastal Shelf Sci., 21: 33-43. Walton, TL. and Adams, W.D., 1976. Capacity of inlet outer bars to store sand. Proc. 15 Coastal Eng. Conf., pp. 1919-1937. Yaiiez-Arancibia, A,, Aguirre-Leh, A. and Soberon-Chavez, G., 1992. Estuarine-related fisheries in Terminos lagoon and adjacent continental shelf (Southern Gulf of Mexico). In: E. Maltby,
272
F.I. ISLA
P.J. Dugan and J.C. Lefeuvre (Editors), Conservation and Development: the Sustainable Use of Wetlands Resources. Proc. 3rd Int. Wetland Conf., Rennes, IUCN, Gland, Switzerland, pp. 145-153. Zenkovich, V.P., 1959. On the genesis of cuspate spits along lagoon shores. J. Geol., 67: 269-277. Zenkovitch, V.P., 1969. Origin of barrier beaches and lagoon coast. In: A. Ayala-Castafiares and F.B. Phleger (Editors), Lagunas Costeras, un Sirnposio. UNAM-UNESCO, Mexico, pp. 27-38.
Geomorphologyand Sedimentology of Estuaries. Developments in Sedimentology 53 edited by G.M.E. Perillo 0 1995 Elsevier Science B.V. All rights reserved.
273
Chapter 10
SILICICLASTIC TIDAL FLATS CARL L. AMOS
THE CLASSIFICATION OF TIDAL FLATS
Hans-Erich Reineck (1972), in an article titled “Tidal Flats”, defined tidal flats as “...sandy to muddy or marshy flats emerging during low tide and submerging during high tide...”. This definition, while useful, can be ambiguous as the positions of low and high tides are time-variable (Fig. 10.1). The frequency of tidal flat inundation, examined over long time-scales (months to years) shows that approximately the lower 25% of the extreme tidal range is intermittently exposed, while the higher 25% is intermittently inundated, leaving only the central 50% consistently within Reineck’s definition. Consequently, some divergences in classification of tidal flats have resulted, particularly at the two extremes of exposure. Klein (1985) and Wang and Eisma (1988) proposed a tripartite subdivision of the intertidal (littoral) region, bound between the subtidal (sublittoral) and supratidal (supralittoral) regions. Figure 10-1 shows this classification: their higher tidal flats occur between mean high water spring tides (MHWST) and mean high water neap tides (MHWNT) and are intermittently inundated; their middle tidal flats occur between MHWNT and mean low water neap tides (MLWNT) and are inundated by every tide; the lower tidal flat is situated below MLWNT and is only intermittently exposed. The supratidal region is rarely inundated (the exception being storm surges) and the subtidal region rarely exposed. The genesis and character of sediments deposited within each of these three zones differ. Reineck (1972) observed that these sediments “...form a wedge-shaped body which is elongated parallel to the shore line, but may be intersected by channels and river estuaries.”. His definition embraced the supratidal salt marsh as did Evans’ (1965) classification, whereas Knight and Dalrymple (1975) and Klein (1985) excluded it. Klein (1985) defined tidal flats as “...low relief environments containing unconsolidated and unvegetated sediments that accumulate within the intertidal range, including the supratidal zone. They are present where salt marshes are absent or between the marsh and the subtidal environment.”. It is sediment associated with this latter definition that is described in this paper. Reineck (1978) and Weimer et al. (1980) considered tidal flats to be part of a genetically-related group collectively termed tidal deposits. Such deposits form largely on: (1) intertidal flats where vertical and lateral accretion are primarily influenced by waves and to a lesser extent by tidal currents; and (2) subtidal flats, where lateral accretion is dominant and tidal flow dominates over wave motion. For present purposes, tidal flats are equated with intertidal flats, although it is recognised that tidal influences may extend well below the level of extreme low water.
274
C.L. AMOS
1.o
M.H.W.S.T.
0.9
. r h
5
0.8
Z
0.7
s>
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
.-
\<
,SLOPE=-0.29
-\ ,SLOPE=
102
d
22 9
FLOODING WATER MASS-
l0-
-
-0.16
\ ."
(OUTER ESTUARINE WATER MASS
)
EBBING WATER MASS
-\
I -
CUMBERLAND BASIN _ _ _ _ SHEPODY BAY ~
1
I
1-
zGNECTO 1
BAY I
1
period of two years. These observations clearly demonstrated a gradient of SSC that conformed in all cases to an exponential decay function of the form: (10-3) where SSCOis the concentration in the turbidity maximum and k is the decay constant which varied with season between -0.023 and -0.091 about a mean (storm-free) value of -0.049 (Fig. 10-4; Amos and Tee, 1989). The existence of the SSC-gradient yielded the means of accurate estimation of the total suspended mass in the Basin through integration of the product of SSC and cross-sectional area along the basin length. It was thus found that, for the two years that Keizer and co-workers carried out their measurements, the total suspended mass in Cumberland Basin remained remarkably constant at lo5 metric tonnes. This despite fluctuations in SSC in the inner bay (normalized to the tidal limit) of 0.6 g/l to 15 g/l. Similar gradients with similar decay constants were found by Amos and Tee (1989) in other embayments (of similar tidal range) of the Bay of Fundy. This finding points to the existence of an equilibrium SSC-gradient that is independent of local sediment supply, wave exposure, or basin geometry, and supports the notion of an equilibrium capacity of a restricted tidal water mass which is continually being upset, and which the system is continually striving to maintain. So, the residual transport of sediments may be viewed as a diffusive rather than an advective process, and appears to be the product of a dynamic balance between landward transport due to tidal asymmetry and seaward dispersion due to the resulting SSC-gradient as suggested by Groen (1967). The nature of the
286
C.L. AMOS
dynamic balance in the SSC-gradient becomes apparent when it is understood that the SSC-gradient in the Bay of Fundy is largely controlled by the balances between the sources and sinks of sediment along the Bay. During periods of ice break-up and wave activity (resuspension from the tidal flats) the equilibrium gradient is exceeded through sediment input at its upper end with a consequent export of material in suspension (Amos and Tee, 1989). The mechanism of import, in the case of an abundance of sediment supply, appears to be governed by the capacity of the tidal flats to accommodate new material from suspension; precisely the concept proposed by Evans (1965). If there is no accommodation space for suspended solids, then export must balance import irrespective of settling or scour lag; the system is accommodation space limited. Where accommodation space is available, then import will take place either up to the rate of maximum mass deposition on the higher flats, or to the rate governed by supply. This latter case we may term supply limited. Within these two extremes, the headward sediment flux is most sensitive to SSC; the higher the SSC the greater is the flux. A sensitivity analysis of the parameters influencing this residual flux under conditions of zero SSC-gradient was carried out by Amos and Tee (1989). The parameters tested were: SSC; the critical threshold for deposition, the critical threshold for erosion; the erosion rate; and the settling rate. This analysis showed that variations in SSC (over observed concentration ranges) have a much larger effect on the residual flux of suspended solids than do variations in the thresholds for erosion and deposition, or indeed the particle settling rate. Changing the erosion threshold (over reasonable limits), the basis of Postma’s (1961) hypothesis, did little to alter the residual horizontal flux. The tidal region thus behaves somewhat like a bellows, constantly adjusting the sediment flux at its source to maintain a constant sediment capuciy. This being so, then it follows that the residual flux across the tidal flats is a dynamic balance between the ability of the flats to accept sediments and thus prograde, and the supply of that sediment. Returning to the tidal flats of the Wash, we see that the reclamation umbra of Kestner (1975) provides the accommodation space for an increase in sediment import, and that this umbra extends half way across the flats (Amos, 1974). In the absence of such an umbra, sediment import would be reduced in order to be lock-stepped to natural progradational processes as defined by Evans (1965). Evans (1965) proposed that the accretion and associated zonation of tidal deposits were the result of a gradual reduction in capacity and competency of the inundating tidal flows: “...The gradually decreasing velocity of the tidal currents as they move in over the intertidal flats causes a reduction in the capacity and competency of the waters and results in a gradual differentiation of the load.”
This concept of competency is visualized by reference to Fig. 10-5. In this figure we’ see six (6) thresholds respecting sediment behaviour through which passes an hypothetical plot of asymmetrical tidal current speed. These thresholds are: (1) the threshold for mud deposition also approximately equal to the transport threshold of very fine sand; (2) the threshold for bedload motion of medium sand; (3) the threshold for the erosion of mud flat sediments; (4) the suspension threshold for
287
SILICICLASTIC TIDAL FLATS
10
n
P
3
2rn
r
rn C
3
e
1 I
zi
1.0
E
ZONE OF M U D ACCUMULATION SALT MARSH INNER M U D FLATS
MIXED FLATS
(I)
0.5 W
U IW
I 0.0-
Fig. 10-5. A schematic illustration of the change in competency of the tidal flows over a tidal flat. Six thresholds are defined that appear to adequately explain the zonation of sediments in the Wash: ( I ) the threshold for deposition of fine-grained sediment; ( 2 ) the threshold for motion of fine/medium sand; (3) the threshold for erosion of mud flats, and the suspension of fine sand; ( 4 ) the suspension threshold of fine/medium sand; (5) the traction threshold of gravel; ( 6 ) the suspension threshold medium/coarse sand.
fine/medium sand; ( 5 ) the traction threshold for coarse sand and gravel; and (6) the suspension threshold of medium/coarse sand. Following each of these threshold lines upwards through the curve of tidal elevation, thence across to the tidal flat profile yields the relative elevations above which each of the thresholds is not exceeded by the inundating tidal flows. The regions between respective thresholds provide the spatial range in energetics within which discrete sedimentary sub-environments are potentially formed. In this example we see that only the region landward of (above) threshold (3) will accommodated mud flats and that the region between thresholds (3) and (4)will host the fine sand of the sand flats. The landward gradient in peak flow occurs only on the flats above approximate MSL. Below this level, the entire flats are subject to the peak in tidal flow, and so no zonation of the bed on the basis of peak tidal energy is possible. The notion of capacity, particulary in the Wash, is less easy to define than is competency. The papers of Evans and Collins (1975, 1987) clearly show that the Wash is well below its capacity and that many of the turbid events that bring a large part of the sediment to its flats are unrelated to local conditions of weather or sea state, and are probably transported alongshore with the residual current. In order to maintain the integrity of these turbid events the settling rate of the constituent material must be extremely low. As a consequence, the relative importance of settling and scour lag is much reduced (Groen, 1967). Yet, as we have seen, deposition on the mud flats of the Wash continues to take place.
288
C.L. AMOS
A graphic example of mud flat deposition can be seen in the Avon River estuary, Minas Basin. A solid-fill causeway constructed across that estuary resulted in the rapid development of a mud flat within the sedimentation umbra to its seaward side. The deposition rate and net accretion of this mud flat were found to conform to the trend inherent in eq. (10-2) and was still in excess of 360 mm/a nine years after causeway construction. It has now reached phase 2 of mud flat development (Dalrymple et al., 1991). This mud flat is colonised by halophytic plants and is no longer actively accreting. It is, however, prograding seawards over the tidal sand flats at its seaward edge. Recent surveys undertaken by Vaughan Engineering Associates Ltd. (1993) show net accretion of 1-2 m (in the 20 years between 1972 and 1992; 50-100 mm/a) some 5 km seawards of the causeway. Thus the umbra of the causeway appears to be propagating down the estuary at an approximate rate of 200 m/a. The rate of mass settling ( S M / S t )and the net deposition (D) on this tidal mud flat were calculated using the following equations of Krone (1962):
SM
-=
St
ssc w,(1 -
2)
(10-4)
and Net deposition (D) =
1SSC(t) W, (1
-
$)Sl
(10-5)
where to is the ambient shear stress at time t , and rd is the critical shear stress for deposition (Fig. 10-6), given as 0.12 N/m2 by Creutzberg and Postma (1979). Particle settling rates (W s )of the material in suspension over the Avon River mud flats varies m/s and the SSC of the inundating water mass between 1.2 x 10W4 and 3.3 x is circa 100 mg/l (Amos and Mosher, 1985). The predicted mass settling [using eqs. (10-4) and (10-5)] and that observed on the causeway mud flat were within 20% (Amos and Mosher, 1985). The inferences of this are that (1) sedimentation within the umbra is purely the result of a reduction in flow speed, and (2) that the pattern of deposition may be estimated with reasonable accuracy provided in situ measures are made of the free parameters and that waves are unimportant. In situ monitoring of mass settling rate ensures that the appropriate mode of settling is used (floccule settling, mass settling or hindered settling; see Dyer, 1986 for review).
Mud jlut erosion The erosion of mud flat sediments takes place in two ways: as Type I erosion the erosion rate quickly reaches a maximum and then decreases with time in an exponentially-decaying fashion; and as Type I1 erosion - where the erosion remains constant with time. The first type of mud flat erosion, also called surface erosion by Mehta et al. (1982), we term as benign as the process is self-limiting and short-lived. This pattern of erosion is equated with the breakdown of weak primary bonds of surface organic aggregates and pellets under hydrodynamically turbulent smooth flow. The erosion rate peaks within 30 seconds of application of the bed shear stress and is order lop4 kg/m2/s. After attaining this peak it quickly drops back - where
SILICICLASTIC TIDAL FLATS
289
-
(HOURS) c--NO DEPOSITION (To > T d )
NO DEPOSITION (To >Td)
Fig. 10-6. A schematic diagram of the tidal inundation of the Avon River mud flat and associated current speed. The bed shear stress for deposition (td) is also plotted. Notice that deposition was possible for only 1.4 hours about high tide. Once the mud flat reaches an elevation whereby the inundation period is less than 1.4 hours the mud flat becomes accrerion restricted.
to zero within 2 to 3 minutes. This erosion type takes place at relatively low bed shear stresses (0.2 to 2.5 Pa). Type I1 erosion, also called bulk erosion by Mehta et al. (1982), we term as chronic as much higher values of SSC are the possible result. The peak erosion rates are comparable with those of Type I erosion, however, the erosion continues unabated. This pattern of erosion occurs at bed shear stresses in excess of 4 Pa; that is, under turbulent rough conditions of flow. Under such conditions, excavation of roughness elements (through spatially-varying hydrodynamic pressure distribution) can take place with failure along planes of weakness defined by the microfabric of the sediment. Much of the evidence for the above comes from in situ observations made by Amos et al. (1992). They found 5-fold variations in mud flat strength (equated with the critical shear stress for surface erosion) over 20 days of observations, and spatial variations of the same magnitude. Also, the rate of bed erosion showed no relationship to the absolute bed shear stress, but was strongly correlated to the excess bed shear stress in the exponential form: (10-6) where E is the erosion rate, to is the applied bed shear stress, and t e ( z )is the critical bed shear stress for erosion at depth z below the original (un-eroded) sediment surface. Recent evidence with the Sea Carousel (unpublished data, 1992) shows that the erosion rates and threshold vary considerably with location. Consequently,
290
C.L. AMOS
accurate predictions of the responses of mud flats to applied stresses without prior in situ measurements are unlikely in the near future. There is, however, some hope in predicting the fate of newly-deposited sediments where the consolidation and stress histories are known.
Sand flat stability and the transport of non-cohesive sediment The evolution and stability of the fine-grained sand flats (as distinct from the bars, channels and banks of the Bay of Fundy) has had less attention than the mud flats. The detailed measurements of Collins et al. (1981), however, shed insight into the stability of these sand flats. They detected considerable amounts of sand in suspension across the flats, the majority of which was in the fine sand range. The effects of peak tidal flows observed by Collins et al. (1981) and by Amos (1974) along the transect of the suspension measurements, are plotted against relative intertidal elevation in Fig. 10-7. The competency of the peak flow across the flats is expressed in terms of the bed shear stress (to), which decreases landwards across the sand flats in a linear fashion. The potential effect of the flow on seabed material [of mean diameters ranging from fine sand (D5o = 100 microns) to gravel (Dso = 2500 microns)] is expressed in the mode of transport (no motion, bedload or suspension).
-BEDLOAD-BEDLOAD-
1.o
--
VERY FINE SAND (100 M/CRONS) FINE SAND (200 MICRONS) MEDIUM SAND (300 MICRONS) MEDIUM/COARSE SAND (500 MICRONS) COARSE SAND (750 MICRONS) VERY COARSE SAND (1200 MICRONS) GRAVEL (2500 MICRONS)
SUSPENSION SUSPENSION SUSPENSION SUSPENSIONBEDLOAD -SUSPENSIONBEDLOAD BEDLOAD
EROSION THRESHOLDS FOR LABORATORY MUDS
I I :
0.9
,.*I\
1
MUDFLATS
1
SANDFLATS
CHANNELSANDS a BARS
I L
>
0.2
E 0.1 J
WAVE-FORMED RIPPLES AND W A V E S 4 CURRENT RIPPLES 3-D MEGARIPPLES
1
o-o.',
'. ..\ ,
2
.4
.6
.8 1.0
2
4
6
1
1
1
1
810
1
I
I
20
40
60 80
PEAK BED SHEAR STRESS (Pa)
Fig. 10-7. A comparison between the peak bed shear stresses across the tidal flats of the Wash (a), and Minas Basin (0). Notice that the flows are comparable over the sand flats and in the low water tidal channel, but diverge over the mud flats and marshes. The higher nearshore current speeds in the Wash is reflected in the higher relative elevation of the mud flats and mixed flats. Also shown in the figure are the range of thresholds for erosion of laboratory muds (taken from Amos and Mosher, 1984), and the modes of transport (no motion, bedload or suspension) of sediments ranging in mean diameter from very fine sand (100 microns) to gravel (2500 microns).
291
SILICICLASTIC TIDAL FLATS
Also shown is the potential range of bedforms (ripples, megaripples or sand waves) across the flats based purely on peak bed shear stress. The threshold for traction is based on a solution of the modified Shields parameter 0 (after Yalin, 1972): (10-7)
(10-8)
so
0 = 0.1-
v u*D50
for
u*D50 < 2.3
(10-10)
V
where U*D50/v is the grain Reynolds number, U, is the friction velocity, v is the kinematic viscosity, tocrit is the threshold bed shear stress, and ( p s - p0)g is the sediment buoyant unit weight. The suspension threshold is based on the suspension criterion of Bagnold (1966): (10-11) The range of possible bedforms is based on thresholds defined by Allen (1982) for wave and current ripples, and Dalrymple et al. (1978) for large-scale bedforms. The results of the above analyses are given in Tables 10-1 and 10-2 for the Wash and Minas Basin, respectively. Note that the peak flows have the competency to move gravel as bedload to the approximate position of MLWNT while coarse sand could Table 10-1 The Wash - peak velocity (Urn,,), peak bed shear stress (rrnax) and sediment transport mode as a function of relative elevation ( h / R )a, for a range of grain sizes found on tidal flats ( D in metres) hlR
umnx
rmax
D1 = 0.0001
0.95 0.82 0.67 0.58 0.51 0.22
0.05 0.15 0.32 0.45 0.48 0.98
0.01 0.12 0.38 0.65 0.86 3.40
Dz = 0.0002
D3 = 0,0005
D4 = 0.001
-
-
susp susp susp susp susp
-
-
-
susp susp susp susp
-
-
bed bed susp
bed bed
0 5 = 0.002 -
-
bed
h is height above extreme low water, R is extreme tidal range. D1: fine sand; D2: fine/medium sand; D3: medium/coarse sand; D4: coarse sand; D5:gravel; -: no motion; bed: bedload; susp: suspension. a
292
C.L. AMOS
Table 10-2 and sediment transport Minas Basin, Bay of Fundy - peak velocity (Urnax),peak bed shear stress (rmax) mode as a function of relative elevation ( h / R )a, for a range of grain sizes found on intertidal flats h/R 0.60 0.55 0.48 0.42 0.26 0.0 a
Urn, 0.20 0.27 0.35 0.50 0.90 2.4
rmax 0.12 0.49 0.77 0.93 3.7 17.7
D1 = 0.0001 susp susp susp susp susp
D2
= 0.0002
susp susp susp susp
D3 = 0.0005
D4 = 0.001
D5
bed bed bed susp
bed bed susp
-
= 0.002
-
bed bed
h is height above extreme low water, R is extreme tidal range.
D1:fine sand; Dz:fine/medium sand; D3:rnedium/coarse sand; D4:coarse sand; Dg:gravel.
-:
no
motion; bed: bedload; susp: suspension.
be moved as bedload to the seaward limit of the mud flat (MLWNT). Medium sand and finer grades could be moved in suspension across the entire width of the sand flats. The mode of transport is most sensitive to grain size changes over the medium o 300-500 microns). Fine sand (D50 = 100 microns) and finer sand range ( D ~ = material would move largely in suspension across the entire flats of both regions. Applying the concept of competency to the development of the sand flats, we would expect to see a gravel fraction in its lower part, a very coarse sand fraction on the central flats, and a coarse sand component on the inner sand flats. This is not the case. Well-sorted medium sand dominates the lower sand flats, well-sorted fine sand prevails on the central flats, and fine to very fine sand typifies the inner sand flat. The two major gradients in the size of bottom sediments, the mud flat/sand flat boundary, and the sand fldchannel sand boundary, correspond to the thresholds for suspension of fine sand and medium/coarse sand respectively. Thus sand (in large quantities) appears not to be found landward of its threshold for suspension. Thus the distribution of sizes conforms more closely to the bedload/suspension transition than to the threshold for incipient motion of sand; but why? It seems that for sediment to occupy a position on the tidal flat it must arrive onto the tidal flats in suspension in order to move up the steep landward flank of the low water tidal channel. This is perhaps demonstrated by Collins et al. (1982) who observed that measurable quantities of fine sand were suspended over the sand flats (circa 100 mg/l) and even over the mud flats (10 mg/l) during “quiet” conditions. It is to be marvelled that the fine sand remains on the tidal flat and does not disperse seawards under storms. The fact that over the long-term it does not, underlines the importance of the tidal asymmetry and consequent residual transport over a tidal inundation. This tendency now presents us with a conundrum: as sand possesses no cohesion it cannot be subject to consolidation effect in the scour lag concept, which is one of the supposed main agents responsible for the headward residual motion of fines. Given that even coarse sand moves headwards regardless of its lack of cohesion why are we to believe that fines would not do the same irrespective of scour lag? The next section attempts to address this question.
293
SILICICLASTICTIDAL FLATS ,,U ,,
FOR SAND TRANSPORT
4 7
PEAK TIDAL VELOCITY ( m l s ) 0.4 0.6 0.8 1.0 1.2
0.2
1.4
I
ZONATION
I
SIZE
1
THE WASH
(PROGRADATIONAL SEQUENCE)
I
1
Fig. 10-8. A synthesis of the Wash tidal flat sedimentary character, peak tidal flow and exposure relative to elevation, taken from Amos (1974). The slope of each zone is given in brackets.
A model for sediment accretionlerosion on the tidal pats of the Wash and Minas Basin The above concepts of tidal flat sedimentation take into account only the peak tidal-flows. The development of a tidal flat and its lithology, on the other hand, is the time-integrated effect of the total tidal inundation. Due to landward decreases in both the duration of inundation and the peak bed shear stress, the total energy expenditure at the bed will decrease landwards across the flats in a non-linear fashion. Also the composition of the bed is likely to be the product of the size and quantity of material deposited versus that eroded, and so the seaward edge of the mud flat should be at the position on the tidal flat where the net erosion and net deposition of fine-grained sediment is equal. A synthesis of the zonation, lithology, and peak tidal flow across the tidal flats of the Wash is given in Fig. 10-8. Notice that the transition from a mud flat to a sand flat occurs at the level of MHWNT The peak tidal current speed shows a linear decrease landwards across the flats. It is greatest in the low water tidal channel (1 m/s), ranges from 0.3 to 0.7 m/s over the sand flats, and is generally less than 0.3 mls over the mud flats and marsh. The simulation of tidal inundation ( H ) and tidal current speed ( U ) over the Wash tidal flats was calculated following the method of Doodson and Warburg (1941) using the first four dominant tidal constituents (M2, S2, K1, and 01): H(t) = Al(sinwt
-81) + A2(sinwt
-
B z ) + A3(sinwt - 83) + A4(sinwt
- 84)
(10-12)
294
C.L. AMOS
and U(t) = Ul(C0S wt
-
81) + Uz(C0SWt - 82) + U3(COSWt - 83) + U4(COS wt - 84) (10-13)
where A1 to A4 are the elevation amplitudes of each constituent (3.15, 1.00, 0.14, and 0.18 m), respectively, and U1 to U4 are the amplitudes of the current speeds of each of the four constituents (0.42, 0.13, 0.02, and 0.03 m/s). Also w = 2 n / T , where T is the tidal constituent period (M2 = 12.42 h; Sz = 12.00 h; K1 = 23.94 h; and 01 = 25.82 h), and are the phase lags (6.33, 7.70, 20.01, and 10.25 h). Tide height and current speed were determined at 30-minute time-steps for 993 hours or eighty M2 tidal cycles. For each time-step, bed shear stress was evaluated adopting the quadratic stress law: TO =
CdPO U ( t ) 2
(10-14)
Also assigned were: the critical shear stress for deposition of fines r d = 0.1 Pa; the critical shear stress for erosion re = 0.5 Pa; the mass settling rate W, = 0.00027 m/s; and the drag coefficient Cd = 0.003. Mass settling rate was determined using eq. (10-5). Erosion rate was computed using eq. (10-6). Continuity of mass in suspension was determined assuming no lateral or longitudinal advection (a closed system). The starting SSC was set in turns to 10, 100, 1000 and 10,000 mg/l and was assumed to be constant across the flats. In all cases, the net balance in sedimentation was determined for each of 200 elevations spaced equally between extreme low and high water levels. For each elevation the following parameters were calculated: (1) the time of inundation; (2) the time series of water level; (3) the instantaneous bed shear stress; (4) the cumulative deposited mass; (5) the cumulative eroded mass; (6) the SSC; and (7) the mean (time-averaged) bed shear stress. The total sediment deposition and erosion of the entire flats (integrated over the eighty tidal cycles) was also determined. The time-series of results for the Wash for a starting SSC of 100 mg/l is shown in Fig. 10-9. The figure shows a clear 20-day modulation of the tidal elevation (Fig. 10-9a) and current speed (Fig. 10-9b) for a position at MLWST Notice that the total predicted (across flat) deposition and erosion show complex time-variability (Figs. 10-9c and d). For present purposes of demonstration, we have assumed an infinite supply of all sediment sizes across the flats. Peak erosion appears to exceed peak deposition during spring tides; during neap tides the converse is evident. The net predicted result is one of long-term erosion of the flats and an overall increase in SSC that is modulated by the spring-neap cycle. Notice that net deposition is predicted to be relatively steady with time (i.e. insensitive to the peak tidal current speed), whereas erosion is highly sensitive to tidal current speed and appears absent during neap phases of the tide. The net deposition and erosion (integrated over the eighty tidal cycles shown in Fig. 10-9a) is shown against elevation across the Wash tidal flat in Fig. 10.10. In this case, predicted net deposition and potential erosion are shown for a starting SSC of 1000 mg/l. Notice the asymptotic decrease in mean bed shear stress across the tidal flats. Also notice that net erosion follows this general trend decreasing in an
296
h
E
v
z
z
C.L. AMOS
;I ssc
3
2
0 1 -
0.001
A
=
supratidal region DEPOSITION salt marsh
1
NO CHANGE
L -- t
100 mg/L
MEAN BED SHEAR STRESS (Pa)
0 NET EROSION (kp)
NET DEPOSITION (kgl
0.01
0.1
1
10
100
:
Arenicola sand flats
lower s a n d flats
1 1
__________1 channel sands
Fig. 10-10. The predicted mean bed shear stress, potential erosion, and net deposition for the Wash, U.K. plotted against tidal flat elevation for a starting suspended sediment concentration of 100 mg/l. Notice that the elevation where deposition and erosion intersect corresponds closely with the seaward limit of the mud flats.
hyperbolic fashion with elevation with an asymptote at the level of the higher mud flats. Net deposition, on the other hand, is predicted to be virtually constant across the middle flats, but to decrease above an elevation of circa 6 m (over the salt marsh). Also note the peculiar trend of increasing deposition in the low water tidal channel (reflecting deposition at both high and low water slack tides, and the diminishing effect of exposure time). We stated earlier that the seaward limit of the mud flats should be defined as the elevation where long-term erosion and deposition of fines are equal. We may now test this hypothesis by reference to Fig. 10-10. Notice that the net deposition curve intersects the net erosion curve at circa 5.7 m. Examination of the adjacent tidal flat zonation shows that this level corresponds to the seaward edge of the inner sand flat (where a significant silt content is to be found). It would thus appear that there is a reasonable closure between the mapped tidal flat zonation of the Wash and the predicted sedimentation trends. Insofar as these trends omit the effects of waves, we conclude that the zonation in the Wash is largely controlled by currents of the tidal inundation and not by waves. The net deposition and erosion trends predicted for the Wash tidal flats are plotted against elevation for differing starting SSC's (10, 100, 1000, and 10,000 mg/l) in Fig. 10-11. Notice that these curves intersect the erosion curve at different elevations; the greater the SSC the lower the elevation at which the intersection occurs. Notice that at circa 1000 mg/l, deposition exceeds erosion across virtually the entire middle tidal flats to the tidal channel with a resulting development of a mud blanket over fine sand. Such mud drapes are known to occur, but are ephemeral due to short-lived elevations in storm-induced SSC. Nevertheless, tidal flats subject to consistently high levels of turbidity (such as typify northern China) would be expected to be dominated by mud flats by virtue of this concentration, even under energetic tidal conditions. Now let us examine the tidal flats of the Minas Basin. A synthesis of these tidal
297
SILICICLASTIC TIDAL FLATS
0.01
0.1
1
10
100
kg or Pa A
YEAN BED S H W m S S (Pa)
0
NET EROSION (kg)
0
NEI DEPOSITION (kg)
Fig. 10-11. The predicted mean bed shear stress, potential erosion, and net deposition for the Wash, U.K., plotted against tidal elevation for starting suspended sediment concentrations (SSC) of 10, 100, 1000 and 10,000 mg/l. The figure demonstrates the dominating influence of SSC on mud flat development.
flats is given in Fig. 10-12. The profile is much narrower and steeper than is the Wash. The mud flat has a slope of 1: 16, the sand flat slopes at 1 :50 and the channel has slopes in excess of 1 : 100. The transition from the mud flat to the sand flat varies considerably (Amos and Joice, 1977), but typically is found between MSL and MHWNT (i.e. below the region of intermittent inundation). Also note that the mud flatlsand flat transition is lower on these flats than it is in the Wash. A decrease in peak current speed across the flats is also evident. However, the gradient is curvilinear showing the largest gradient across the inner flat. Equations (10-12) and (10-13) were again used to compute the tidal elevation and tidal current speed. In this case, however, the first four major constituents were the M2, N2, S2,and K2. A1 to A4 are the elevation amplitudes of each constituent (5.64, 1.10, 0.83, and 0.22 m) respectively, and U1 to U , are the amplitudes of the current speeds of each of the four constituents (1.08, 0.21, 0.16, and 0.04 m/s). are the respective phase lags (0.48, 12.34, 1.97, and 1.96 h). The critical shear stress for deposition ( t d ) is 0.12 Pa, and the erosion bed shear stress (te)is 2 Pa (after Amos et al., 1992). The time-series of tidal elevation and tidal current speed are shown in Fig. 10-13a and b. Notice the vastly different tidal range and 20-day modulation in tidal amplitude to that of the Wash (Fig. 10-9). This produces a similarly unique prediction of the net erosion and deposition patterns through time (Fig. 10-13d) with a consequent steady increase in SSC (Fig. 10-13c). The patterns of erosion are highly variable with three periods of high erosion punctuated by two periods of no net erosion. The predicted peak erosion is almost always less than peak deposition which is relatively steady throughout the time-series. Why then, does the predicted SSC increase (diagnostic of net erosion)? The reason, of course, is that the net balance
298
C.L. AMOS 1
o5
m2
PERMANENTLY INTERTIDAL
INUNDATION / YR (d) EXPOSURE / YR (- o -) PEAK CURRENT SPEED +)
SPRING L T
Fig. 10-12. A synthesis of the Minas Basin sedimentary character, peak tidal flow and exposure relative to elevation, taken from Amos and Joice (1976) and Daborn et al. (1991).
in the sediment flux is the time integration of the entire tidal inundation and not just the peak. This attribute alerts us to the dangers of extrapolation of short-term measurements, and the possible mis-use of peak fluxes to characterize net trends. Similar trends to the Wash are evident in the predictions of erosion/deposition across the Minas Basin tidal flats for a typical summertime SSC (100 mg/l, Fig. 10-14). Notice that the intersection of the two curves (erosion and deposition) is situated at an elevation of circa 12 m above datum. In reality, the seaward limit of the Minas Basin mud flat edge is further seawards (lower) than was predicted. This mismatch may be due to the use of a low value of SSC (neglecting the effects of storms) that can elevate the turbidity to over 1000 mg/l through wave resuspension. Thus our model may represent an unanticipated source limited condition. A sediment source that is absent from the model is the tidal flat itself. How then may we accurately account for sediment supply without including a term for wave erosion of the flats?
THE INFLUENCES OF WAVES ON TIDAL FLATS
Waves play a strong role in the resuspension of sediments on tidal flats. A series of papers have been written on the effects of storms, hurricanes and typhoons on
299
SILICICLASTIC TIDAL FLATS 16
11 I2
-E T
10
2 z
8
4
6
z 4 2 11
0
840
4R0
Y60
720
TIME(hours)
1
0
480
240
760
-
1 960
TIME( hours) 3000 2500
-~-
7
I
~
I
500
,
0 01 210
,
----L~
180
7'20
-~ I
'XI1
TIhIE(hour9)
Fig. 10-13. Time-series plots of the predicted tide height (a), tidal current speed (b), suspended sediment concentration (c) and net deposition and erosion (d) for the tidal flats of the Minas Basin, Canada. Eighty M2 tidal cycles were simulated (993 hours) at a time-step of 30 minutes. The 20-day modulation of the tides is variable, resulting in a complex time-variability in erosion and deposition that differs markedly from that of the Wash (Fig. 10-9).
300
C.L. AMOS
i
h
E
salt marsh
-
higher mud flats
v
-
F: d >
-
inner sand flat -
Y
A
YWI
-
BED SBEAR STRESS (Pa)
W
-
-
lower sand flat
-
channel sand
0.01
0.1
1
10
100
log0
Fig, 10-14. The predicted mean bed shear stress, potential erosion, and net deposition for the Minas Basin, Canada for a starting SSC of 100 mg/l.
tidal flats (Yan et al., 1981; Champagne, 1982; Ren et al., 1985; Wanless et al., 1988; Wang and Eisma, 1990; Wang et al., 1990; Wells et al, 1990 amongst others). In a broad sense, it is wave climate that limits the location of tidal flats (Boyd et al., 1992). The non-periodic occurrences of wave magnitude means that even sheltered regions are subject to wave influence at times. The degree of this influence is often visible across the tidal flats in the form of erosion of the seaward edges of salt marshes and mud flats (Evans, 1965), in the development of sandy beaches on top of the salt marsh at the MHWST level (Amos and Joice, 1977; Knight and Dalrymple, 1975; Belperio et al., 1988) and in the development of wave-formed ripples across the sand flats (Amos and Collins, 1978; Dingler and Clifton, 1984). Thorne (1979) measured the sand resuspension by waves in the Great Ouse, and found that a near-bed oscillatory flow of only 5 mm/s could double the transport of fine sand by tidal currents. Measurements over the tidal flats of the Wash made by Collins et al. (1981) showed that as much fine sand was suspended over the sand flats in storms as silt and clay. They also found that the SSC was an order of magnitude greater in storms than at other times and that the greatest values were on the middle and lower flats; a reversal of the fair-weather trends. Waves may either amplify or reverse the headward flux of suspended sediment. In the upper Bay of Fundy, where significant wave heights can reach 4-6 m, wave erosion prevails with a consequent export of suspended sediment (Amos and Asprey, 1979). In turbid environments of moderate to low wave energy the reverse may be true due to the presence of solitary waves and Kelvin-Helmholtz billows along lutoclines (Wells et al., 1990; Frey et al., 1989). The varying influences of waves is often apparent in the tidal flat zonation and associated sediment texture. Isla et al. (1991) for example showed that the tidal flats of San Sebastian Bay, Patagonia, were developed only in the most sheltered part of the bay; around that Bay the inner fine-grained zones became narrower and were replaced by coarser-grained facies in response to increasing wave exposure. A similar contrast was shown by Wang (1983) and others between the mud flats
SILICICLASTIC TIDAL FLATS
301
bordering the more sheltered Bohai and Yellow Seas to the more exposed sandier flats bordering the South China Sea (Yan et al., 1981; Zhang, 1992). As a final note, we may conclude that the effects of waves on tidal flat development are important but largely unpredictable. Cyclic loading, the associated pore-pressure amplification, and subsequent liquefaction of tidal flat sediments have not been quantified and offer rich potential for future research.
ACKN 0WLED GEMENTS
My thanks go to Dr. J. McManus and T Sutherland for the thorough reviews of the manuscript. Also acknowledged are F. Kelly for the illustrations and R.W. Dalrymple for the healthy discussion.
REFERENCES Adams, 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., 1982. Sedimentary Structures, Their Character and Physical Basis, Vol. 1. Elsevier, Amsterdam, 593 pp. Amos, C.L., 1974. Inter-tidal flat sedimentation of the Wash - E. England. Unpublished Ph.D. Thesis, Imperial College, London, 441 pp. Amos, C.L and Asprey, K., 1979. Sedimentation studies in the Chignecto Bay system - A progress report. Geol. SUN. Can. Current Res. Pap., 79-1B: 245-252. Amos, C.L. and Collins, M.B., 1978. The combined effects of wave motion and tidal currents on the morphology of intertidal ripple marks: the Wash, U.K. J. Sediment. Petroleum, 48: 849-856. Amos, C.L. and Joice, G.H.E., 1977. The sediment budget of Minas Basin. Bedford Inst. Ocean. Data Series BI-D-77-3, 411 pp. Amos, C.L. and Mosher, D., 1985. Erosion and deposition of fine-grained sediments from the Bay of Fundy. Sedimentology, 32: 815-832. Amos, C.L. and Long, B.F.N., 1980. The sedimentary character of the Minas Basin, Bay of Fundy. In: S.B. McCann (Editor), The Coastline of Canada. Geol. Surv. Can. Pap., 80-10: 123-152. Amos, C.L. and Tee, K.T., 1989. Suspended sediment transport processes in Cumberland Basin. J. Geophys. Res., 94: 14407-14417. Amos, C.L., Van Waggoner, N.A. and Daborn, G.R., 1988. The influence of subaerial exposure on the bulk properties of fine-grained intertidal sediment from the Minas Basin, Bay of Fundy. Est. Coastal Shelf Sci., 27: 1-13. Amos, C.L. Daborn, G.R., Christian, H.A. Atkinson, A. and Robertson, A,, 1992. In situ erosion measurements on fine-grained sediments from the Bay of Fundy. Mar. Geol., 108: 175-196. Anderson, F.E., 1973. Observations of some sedimentary processes acting on a tidal flat. Mar. Geol., 14: 101-116. Anderson, F.E., 1979. How sedimentation pattern may be affected by extreme water temperatures on a northeastern coastal intertidal zone. Northeastern Geol., 1: 122-132. Anderson, F.E., 1983. The northern muddy intertidal: seasonal factors controlling erosion and deposition - a review. Can. J. Fish. Aquatic Sci., 40: 143-159. Anderson, F.E., Black, L., Watling, L. and Mayer, L., 1979. A temporal and spatial study of mudflat erosion and deposition. J. Sediment. Petrol., 51: 726-736. Bagnold, R.A., 1966. An approach to the sediment transport problem from general physics. US.Geol. SUN. Prof. Pap., 4421, 37 pp.
302
C.L. AMOS
Baeyens, W., Panutrakul, S., Elskens, M., Leermakers, M., Navez, J. and Monteny, F., 1991. Geochemical processes in muddy and sandy tidal flat sediments. Geo-Mar. Lett., 11: 188-193. Belperio, A.P., Gostin, V.A., C a m , J.H. and Murray-Wallace, C.V., 1988. Sediment-organism zonation and the evolution of Holocene sequences in Southern Australia. In: P.L. de Boer, A. van Gelder and S.D. Nio (Editors), Tide-Influenced Sedimentary Environments and Facies, pp. 475-497. Berner, H., Kaufhold, H., Nommensen, B. and Proeber, Ch., 1986. Detailed mapping of the surface sediments in the middle and southern part of the North Frisian Wadden Sea. Meyniana, 38: 81-93. Boothroyd, J.C., 1985. Tidal inlets and tidal deltas. In: R.A. Davis (Editor), Coastal Sedimentary Environments. Springer-Verlag, pp. 445-532. Boyd, R., Dalrymple, R. and Zaitlin, B.A., 1992. Classification of clastic coastal depositional environments. Sed. Geol., 80: 139-150. Buchholtz-Ten Brink, M.R. Gust, G. and Chavis, D., 1989. Calibration and performance of a stirred benthic chamber. Deep-sea Res., 36: 1083-1101. Caline, B., Larsonneur, C. and CHomer, A., 1982. The Mont Saint-Michel Bay; main sedimentary environments. In: J. Lang (Editor), Geologie Jubilaire George Lucas. Mem. Geol. Univ. Dijon, No. 7: 37-51. Carling, P.A., 1981. Sediment transport by tidal currents and waves: observations from a sandy intertidal zone (Burry Inlet, South Wales). Int. Assoc. Sediment. Spec. Publ., 5: 65-80. Carling, P.A., 1982. Temporal and spatial variation in intertidal sedimentation rates. Sedimentology, 29: 17-23. Champagne, P., 1982. Morphologie littorale de la baie de rupert. Naturalist Can., 109: 357-384. Christian, H.A., 1991. Geomechanics of the Starrs Point tidal flat. In: G.R. Daborn (Editor), Littoral Investigation of Sediment Properties. Acadia Centre Est. Res. Rep., 17: 113-144. Collins, M.B., Amos, C.L. and Evans, G., 1981. Observations of some sediment-transport processes over intertidal flats, the Wash, U.K., Int. Assoc. Sediment. Spec. Publ., 5: 81-98. Creutzberg, F. and Postma, H., 1979. An experimental approach to the distribution of mud in the North Sea. Neth. J. Sea Res., 13: 99-116. Daborn, G.R. et al., 1991. Littoral investigation of sediment properties (LISP) report. Acadia Centre Est. Res. Rep., 17: 239 pp. Daborn, G.R. et al., 1993. An ecological “cascade” effect: migratory birds affect stability of intertidal sediments. Limnol. Oceanogr., 38: 225-231. Dalrymple, R.W., 1977. Sediment dynamics of macrotidal sand bars. Unpublished Ph. D. Thesis, McMaster University, Hamilton, Ontario, 634 pp. Dalrymple, R.W. Knight, R.J. and Lambiase, J.J., 1978. Bedforms and their hydraulic stability relationships on a tidal environment, Bay of Fundy, Canada. Nature, 275: 100-104. Dalrymple, R.W., Knight, R.J., Zaitlin, B.A. and Middleton, G.V., 1990. Dynamics and facies of a macrotidal sand-bar complex, Cobequid Bay-Salmon River estuary (Bay of Fundy). Sedimentology, 37: 577-612. Darymple, R.W., Makino, Y. and Zaitlin, B.A., 1991. Temporal and spatial patterns of rhythmic deposition on mudflats in the macrotidal Cobequid Bay-Salmon River estuary, Bay of Fundy, Canada. In: D.G. Smith, G.E. Reinson, B.A. Zaitlin and R. A. Rahmani (Editors), Clastic Tidal Sedimentology. Can. SOC.Petrol. Geol. Mem., 16: 137-160. Darymple, R.W., Zaitlin, B.A. and Boyd, R., 1992. Estuarine facies models: conceptual basis and stratigraphic implications. J. Sediment. Petrol., 62: 1130-1 146. Dieckmann, R., Osterthun, M. and Partenscky, H.-W.1987. Influence of water-level elevation and tidal range on the sedimentation in a German tidal flat area. Prog. Oceanogr., 18: 151-166. Dinglcr, J.R. and Clifton, H.E., 1984. Tidal-cycle changes in oscillation ripples on the inner part of an estuarine sand flat. Mar. Geol., 60: 219-233. Dionnc, J.C., 1988. Characteristic features of modern tidal flats in cold regions. In: P.L. de Boer (Editor), Tide-Influenced Sedimentary Environments and Facies. D. Reidel Pub. Co., pp. 302-332. Doodson, A.T. and Warburg, H.D., 1941. Admiralty Manual of Tides. Hydrographic Dept., Admiralty, London, 270 pp. Dyer, K.R., 1986. Coastal and Estuarine Sediment Dynamics. John Wiley and Sons, Chichester, 342 pp.
SILICICLASTIC TIDAL FLATS
303
Emerson, C.W. and Grant, J., 1991. The control of soft-shell clam (Mya arenaria) recruitment on intertidal sandflats by bedload sediment transport. Limnol. Oceanogr., 36: 1288-1300. Evans, G., 1965. Intertidal flat sediments and their environment of deposition in the Wash. Q. J. Geol. Soc. London, 121: 209-245. Evans, G., 1975. Intertidal flat deposits of the Wash, western margin of the North Sea. In: R.N. Ginsburg (Editor), Tidal Deposits. A Casebook of Recent Examples and Fossil Counterparts. Springer-Verlag, New York, pp. 13-20. Evans, G . and Collins, M.B., 1975. The transportation and deposition of suspended sediment over the intertidal flats of the Wash. In: J. Hails and A. Carr (Editors), Nearshore Sediment Dynamics and Sedimentation; an Interdisciplinary Review. John Wiley and Sons, pp. 273-306. Evans, G. and Collins, M.B., 1987. Sediment supply and deposition in the Wash. In: P.Doody and B. Barnett (Editors), The Wash and its Environment. Nature Conser. Council, pp. 48-63. Faas, R.W., 1991. Rheological boundaries of mud: where are the limits? Geo-Mar. Lett., 11: 143-146. Faas, R.W., Christian, H.A., Daborn, G.R. and Brylinski, M., 1992. Biological control of mass properties of surficial sediments: an example from Starrs Point tidal flat, Minas Basin, Bay of Fundy. In: A.J. Mehta (Editor), Nearshore Estuarine Cohesive Sediment Transport. Am. Geophys. Union (Coastal and Estuarine Studies), 42: 360-377. Fitzgerald, D.M. and Penland, S., 1987. Backbarrier dynamics of the east Freisian Islands. J. Sediment. Petrol., 57: 746-754. Frey, R.W., Howard, J.D., Han, S.-J., and Park, B.-K., 1989. Sediments and sedimentary sequences on a modern macrotidal flat, Inchon, Korea. J. Sediment. Petrol., 59: 28-44. Gerdes, G., Krumbein, W.E. and Reineck, H.E., 1985. The depositional record of sandy, versicoloured tidal flats (Mellum Island, southern North Sea). J. Sediment. Petrol., 55: 265-278. Ginsburg, R.N., Hardie, L.A., Bricker, O.P., Garrett, P. and Wanless, H.R., 1977. Exposure index: a quantitative approach to defining position within the tidal zone. In: L.A. Hardie (Editor), Sedimentation on the Modern Carbonate Tidal Flats of northwest Andros Island, Bahamas. The Johns Hopkins Univ. Press, pp. 7-11. Gordon, D.C., Keizer, P.D., Daborn, G.R., Schwinghamer, P. and Silvert, W.L., 1986. Adventures in holistic ecosystem modelling: the Cumberland Basin ecosystem model. Neth. J. Sea Res., 20: 325-335. Grant, J., 1981. Sediment transport and disturbance on an intertidal sandflat: infaunal distribution and recolonization. Mar. Ecol., 6: 249-255. Grant, J., 1988. Intertidal bedforms, sediment transport, and stabilization by benthic microalgae. In: P.L. de Boer (Editor), Tide-Influenced Sedimentary Environments and Facies. D. Reidel Pub. Co., pp. 499-510. Grant, J., Bathmann, U.V. and Mills, E.L., 1986a. The interaction between benthic diatom films and sediment transport. Est. Coastal Shelf Sci., 23: 225-238. Grant, J., Boyer, L.F. and Sanford, L.P., 1982. The effects of bioturbation on the initiation of motion of intertidal sands. J. Mar. Res., 40: 659-677. Grant, J. Mills, E.L. and Hopper, C.M., 1986b. A chlorophyll budget of the sediment-water interface and the stabilizing biofilms on particle fluxes. Ophelia, 26: 207-219. Grinham, D.F., and Martini, I.P., 1984. Sedimentology of the Ekwan shoal, Akimiski Strait, James Bay, Canada. Sed. Geol., 37: 273-294. Groen, P., 1967. On the residual transport of suspended matter by an alternating tidal current. Neth. J. Sea Res., 3: 564-574. Groenendijk, A.M., 1986. Establishment of Spartina anglica population on a tidal mudflat: a field experiment. J. Env. Manage., 22: 1-12. Hantzschel, W., 1939. Tidal flat deposits (wattenschlick). In: P.D. Trask (Editor), Recent Marine Sediments. Dover Pub. Inc., pp. 195-206. Hayes, M.O., 1975. Morphology and sand accumulation in estuaries. In: L.E. Cronin (Editor), Estuarine Research, Vol. 11. Academic Press, pp. 3-22. Hydraulics Research Station 1980. River Scheldt surge barrier. hydraulics. Res. Stat., Wallingford Rep. EX 928, 28 pp.
304
C.L. AMOS
Inglis, C.C. and Kestner, F.J.T, 1958. The long-term effects of training walls, reclamation and dredging on estuaries. Proc. Inst. Civil Eng., 9: 193-212. Isla, F.I., Vilas, F.E., Bujalesky, G.G., Ferrero,M., Bonorino, G.G. and Miralles, A.A., 1991. Gravel drift and wind effects on the macrotidal San Sebastian Bay, Tierra del Fuego, Argentina. Mar. Geol., 97: 211-224. Julien, P.Y. and Lan, Y., 1991. Rheology of hyperconcentrations. J. Hydraul. Eng., 117: 346-353. Jumars, P.A. and Nowell, A.R.M., 1984. Effects of benthos on sediment transport: difficulties with functional grouping. Cont. Shelf Res., 3: 115-130. Keizer, P.D. and Gordon, D.C., 1985. Nutrient dynamics in Cumberland Basin-Chignecto Bay, a turbid macrotidal estuary in the Bay of Fundy, Canada. Neth. J. Sea Res., 19: 193-205. Kellerhals, P. and Murray, J.W., 1969. Tidal flats at Boundary Bay, Fraser river delta, British Columbia. Can. Petrol. Geol. Bull, 17: 67-91. Kestner, F. J.T, 1975. The loose-boundary regime of the Wash. Geograph. J., 141: 389-412. Klein, G. deVries., 1963. Bay of Fundy intertidal zone sediments. J. Sediment. Petrol., 33: 844-854. Klein, G. deVries., 1976. Holocene Tidal Sedimentation. Dowden, Hutchinson and Ross Inc., 423 pp. Klein, G. deVries., 1977. Clastic Tidal Facies. Publ. Continuing Education Publication Company, Illinois, 149 pp. Klein, G. deVries., 1985. Intertidal flats and intertidal sand bodies. In: R.A. Davis (Editor), Coastal Sedimentary Environments. Springer-Verlag, New York, pp. 187-224. Knight, R.J., 1977. Sediments, bedforms and hydraulics in a macrotidal environment, Cobequid Bay (Bay of Fundy) Nova Scotia. Unpublished Ph.D. Thesis. McMaster University, Hamilton, Ontario, 693 pp. Knight, R.J. and Dalrymple, R.W., 1975. Intertidal sediments from the south shore of Cobequid Bay, Bay of Fundy, Nova Scotia, Canada. In: R.N. Ginsburg (Editor), Tidal Deposits. A Casebook of Recent Examples and Fossil Counterparts. Springer-Verlag, New York, pp. 47-55. Kestner, F.J.T., 1975. The loose-boundary regime of the Wash. Geograph. J., 141: 389-412. Komar, P.D. and Li. Z., 1986. Pivoting analysis of the selective entrainment of sediments by shape and size with application to gravel threshold. Sedimentology, 33: 425-436. Krone, R.B., 1962. Flume studies of the transport fo sediment in estuarial shoaling processes. Univ. California, Berkley, Final Rept. Lambiase, J.J., 1977. Sediment Dynamics in the Macrotidal Avon River Estuary, Nova Scotia. Unpub. Ph. D. Thesis. McMaster University, Hamilton, Ontario, 415 pp. Larsonneur, C., 1975. Tidal deposits, Mont Saint-Michel Bay, France. In: R.N. Ginsburg (Editor), Tidal Deposits. A Casebook of Recent Examples and Fossil Counterparts. Springer-Verlag, New York, pp. 21-30. Li, Z . and Komar, P.D., 1986. Laboratory measurements of pivoting angles for applications to selective entrainment of gravel in a current. Sedimentology, 33: 413-423. LISP-UK., 1992. Littoral Investigation of Sediment Properties. Unpublished Proposal to Land Ocean Interaction Study. G.R. Daborn (Editor). Publ. University of Bristol, 14 pp. Martini, I.P., 1991. Sedimentology of subarctic tidal flats of western James Bay and Hudson Bay, Ontario, Canada. In: D.G. Smith, G.E. Reinson, B. A. Zaitlin and R.A. Rahmani (Editors), Clastic Tidal Sedimentology. Can. SOC.Petrol. Geol. Mem., 16: 301-312. McCann, S.B., 1980. Classification of tidal environments. In: S.B. McCann (Editor), Sedimentary processes and animal-sediment relationships in tidal environments. Geol. Assoc. Can. Short Course Notes, 1: 1-24. Meadows, P.S. and Tait, J., 1989. Modification of sediment permeability and shear strength by burrowing invertebrates. Mar. Biol., 101: 75-82. Mehta, A.J., 1989. On estuarine cohesive sediment suspension behaviour. J. Geophys. Res., 94: 1430314314. Mehta, A.J., 1991. Understanding fluid mud in a dynamic environment. Geo-Mar. Lett., 11: 113-118. Mehta, A.J., and Partheniades, E., 1982. Resuspension of deposited cohesive sediment beds. Proc 18th Conf. Coastal Eng., pp. 1569-1588. Mehta, A.J., Parchure, TM., Dixit, J.G. and Ariathurai, R., 1982. Resuspension potential of deposited cohesive sediment beds. In: V.S. Kennedy (Editor), Estuarine Comparisons. Academic Press, pp.
SILICICLASTIC TIDAL FLATS
305
591-609. Middleton, G.V., Knight, R.J. and Dalrymple, R.W., 1976. Facies model for macrotidal environments, Cobequid Bay, Nova Scotia. AAPG Bull., 60: 697-698. Middleton, G.V. and Southard, J.B., 1984. Mechanics of sediment movement. SEPM Short Course Notes 3, 401 pp. Montague, C.L., 1984. Influence of biota on erodibility of sediments. In: A. J. Mehta (Editor), Estuarine Cohesive Sediment Dynamics. Springer-Verlag, pp. 251-269. Neumann, A.C. Gebelein, C.D. and Scoffin, G.P., 1970. The composition, structure, and erodibility of subtidal mats, Abaco, Bahamas. J. Sediment. Petrol., 40: 274-297. Nowell, A.R.M., Jumars, P.A. and Eckman, J.E.1981. Effects of biological activity on the entrainment of marine sediments. Mar. Geol., 42: 133-153. Pamukcu, S. and Tuncan, M., 1991. Influence of some physicochemical activities on mechanical behaviour of clays. In: R.H. Bennet, W.R. Bryant and M.H. Hulbert (Editors), Microstructure of Fine-Grained Sediments. Springer-Verlag, pp. 241-253. Parker, W.R., 1987. Observations on fine sediment transport phenomena in turbid coastal environments. Cont. Shelf Res., 7: 1285-1293. Partheniades, E., 1984. A fundamental framework for cohesive sediment dynamics. In: A. J. Mehta (Editor), Estuarine Cohesive Sediment Dynamics. Springer-Verlag, pp. 219-250. Paterson, D.M., 1989. Short-term changes in the erodibility of intertidal cohesive sediments related to the migratory behaviour of epipelic diatoms. Limnol. Oceanogr., 34: 223-234. Paterson, D.M. and Underwood, G.J.C., 1990. The mudflats ecosystem and epipelic diatoms. Proc. Bristol Natural SOC.,50: 74-82. Paterson, D.M., Crawford, R.M. and Little, C., 1990. Subaerial exposure and changes in the stability of intertidal estuarine sediments. Est. Coastal Shelf Sci., 30: 541-556. Piccolo, M.C., Perillo, G.M.E. and Daborn, G.R., 1993. Soil temperature variations on a tidal flat in Minas Basin, Bay of Fundy, Canada. Est. Coastal Shelf Sci., 35: 345-357. 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.M. Lauff (Editor), Estuaries. AAAS Pub., 83: 158-179. Reineck, H.-E., 1972. Tidal Flats. In: J.K. Rigby and W.K. Hamblin (Editors), Recognition of Ancient Sedimentary Environments. SEPM Spec. Pub., 16: 146-159. Reineck, H.-E., 1975. German North Sea tidal flats. In: R.N. Ginsburg (Editor), Tidal Deposits. A Casebook of Recent Examples and Fossil Counterparts. Springer-Verlag, New York, pp. 5-20. Reineck, H.-E., 1978. Tidal-flat geology. In: R.W Fairbridge and J. Bourgeois (Editors), The Encyclopedia of Sedimentology. Dowden, Hutchinson and Ross, Inc., pp. 798-800. Reineck, H.-E., Chen, C.M. and Wang, S.S., 1986. Backbarrier tidal flats between Wangerooge and mainland, North Sea. Senckenber. Mar., 17: 241-252. Ren, M.-E., Zhang, R.-S., and Yang, J.-H., 1983. Sedimentation on the tidal mud flat of China: with special reference to Wanggang area, Jiangsu Province. In: Proc. Int. Symp. on Sedimentation on the Continental Shelf with Special Reference to the East China Sea. China Ocean Press, pp. 1-17. Ren, M.-E., Zhang, R.-S., and Yang, J.-H., 1985. Effect of typhoon no. 8114 on coastal morphology and sedimentation of Jiangsu Province, People’s Republic of China. J. Coastal Res., 1: 21-28. Semeniuk, V., 1981. Sedimentology and the stratigraphic sequence of a tropical tidal flat, north-western Australia. Sed. Geol., 29: 195-221. Serodes, J.-B., and Troude, J.-P., 1984. Sedimentation cycle of a freshwater tidal flat in the St. Lawrence estuary. Estuaries, 7: 119-127. Smith, N.D., Phillips, A.C. and Powell, R.D., 1990. Tidal drawdown: a mechanism for producing laminations in glaciomarine deltas. Geology, 18: 10-13. Thompson, R.W., 1968. Tidal flat sedimentation on the Colorado River delta, northwestern Gulf of California. Geol. SOC.Am. Mem. 107, pp. 133. Thorne, M.F.C., 1979. The effects of waves on the tidal transport of sand. Hydraulics Res. Station, Wallingford, Notes, 21: 4-5. van Straaten, L.M.J.U., 1961. Sedimentation in tidal flat areas. J. Alberta SOC.Petrol. Geol., 9: 203-226.
306
C.L. AMOS
van Straaten, L.M.J.U. and Kuenen, P.D., 1957. Accumulation of fine-grained acdiments in the Dutch Wadden Sea. Geol. Mijnb., 19: 329-354. Vaughan Engineering Associates Ltd., 1993. Hantsport turning basin sedimentation study, supplementary report. Unpub. Rep. submitted to Fundy Gypsum Company Ltd., Hantsport, Nova Scotia, 15 PP. Wang, B.-C. and Eisma, D., 1988. Mudflat deposition along the Wenzhou coastal plain in southern Zhejiang, China. In: P.L. de Boer (Editor), Tide-Influenced Sedimentary Environments and Facies. D. Reidel Pub. Co., pp. 265-274. Wang, B.-C. and Eisma, D., 1990. Supply and deposition of sediment along the north bank of Hangzhou Bay, China. Neth. J. Sea Res., 25: 377-390. Wang, Y., 1983. The mudflat system of China. Can. J. Fish. Aquatic Sci., 40: 160-171. Wang, Y., Collins, M.B. and Zhu, D., 1990. A comparative study of open coast tidal flats: the Wash (U.K.), Bohai Bay and west Huanghai Sea (mainland China). In: Proc. Int. Symp. on Coastal Zone of China, 1988. China Ocean Press, Beijing, pp. 120-134. Wanless, H.R., Tyrell, K.M. Tedesco, L.P. and Dravis, J.J., 1988. Tidal-flat sedimentation from Hurricane Kate, Caicos platform, British West Indies. J. Sediment. Petrol., 58: 724-738. Weimer, R.J., Howard, J.D. and Lindsay, D.R., 1982. Tidal flats and associated tidal channels. In: P.A. Scholle and D. Spearing (Editors), Sandstone Depositional Environments. AAPG, pp. 191-245. Wells, J.T. and Coleman, J.M., 1981a. Periodic mudflat progradation, northeastern coast of South America: a hypothesis. J. Sediment. Petrol., 51: 1069-1075. Wells, J.T. and Coleman, J.M., 1981b. Physical processes and fine-grained sediment dynamics,coast of Surinam, South America. J. Sediment. Petrol., 51: 1053-1068. Wells, J.T. and Kemp, G.P., 1984. Interaction of surface waves and cohesive sediments: field observations and geologic significance. In: A. J . Mehta (Editor), Estuarine Cohesive Sediment Dynamics. Springer-Verlag, pp. 43-65. Wells, J.T., A d a m , 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. Yalin, S., 1972. Mechanics of Sediment Transport. Pergamon Press, pp. 298. Yan, Q., Xiang, L., Zhang, G., Wu, B. and Dong, R., 1981. Modern coastal sediments of Putuo Island, Zhoushan archipelago. Acta Geol. Sin, 55: 205-215. Zhang, R., 1992. Suspended sediment transport processes on tidal mud flat in Jiangsu Province, China. Est. Coastal Shelf Sci., 35: 225-233. Zhuang, W.-Y. and Chappell, J., 1991. Effects of seagrass on tidal flat sedimentation, Corner Inlet, southeast Australia. In: D.G. Smith, G.E. Reinson, B. A. Zaitlin and R.A. Rahmani (Editors), Clastic Tidal Sedimentology, Can. SOC.Petrol. Geol., Mem. 16: 291-300.
Geomorphology and SedimentologV of Estuaries. Developments in Sedimentology 53 edited by G.M.E. Pcrillo 0 1995 Elscvier Science B.V. All rights reserved.
307
Chapter 11
SALT MARSHES JOHN L. LUTERNAUER, ROWLAND J. ATKINS, ANNE I. MOODY, HARRY EL. WILLIAMS and JONATHAN W. GIBSON
INTRODUCTION
This chapter offers a brief overview of the geology of coastal salt marshes but has as its main objective an assessment of the geomorphological and sedimentological processes of estuarine marshes (Fig. 11-1). We refer to literature dealing with a variety of coastal marshes to support the ideas presented, but all illustrations are drawn from a major western North American estuary, that of the Fraser River delta for three reasons: i) western North American marshes are underrepresented in the literature, but have become the focus of increasing research over the past two decades because of rapidly growing urban pressures, ii) estuarine marshes represent the majority of marshes on the west coast, and iii) the experience of the authors is most directly associated with marshes of the Fraser River delta. The latter sections of the chapter address the status of numerical modelling techniques relevant to salt marshes. Coastal salt marshes have been defined, in the most general sense, as “environments high in the intertidal zone where a generally muddy substrate supports varied and normally dense stands of halophytic plants” (Allen and Pye, 1992a). They form on open coasts, in tidal embayments, behind barrier islands and in deltaic or nondeltaic estuaries. Marshes usually develop between approximately mean sea level and higher high water (Frey and Basan, 1985; Allen and Pye, 1992a). In general, plant species richness increases with elevation, but at high elevations, species distributions tend to be governed by competition whereas low elevation limits are governed mainly by the plants physiological tolerances (Pielou and Routledge, 1976). Although a strong relationship has been demonstrated between frequency and duration of tidal inundation and plant species distribution (Moody, 1978; Hutchinson, 1982; Dawe and White, 1986), the vertical distribution and ecological roles of certain species may also be influenced by the local geographic setting. Estuarine or brackish marshes differ from coastal salt marshes in general in that they lie within a coastal environment which is “significantly diluted with fresh water derived from land drainage” (Perillo, 1989). Estuarine marshes are therefore subject to periodic overflow by both saline water and fresh water which may be sediment charged. At the Fraser River delta (Fig. 11-2a, b) for example, substrate salinity in most of the brackish marshes along the foreshore declines from about 15-20%0 in winter to less than 5%0 during the summer (Karagatzides, 1987; Hutchinson et al., 1989) when the river is in flood and discharges approximately 17 x lo6 tonnes of sediment through the estuary (McLean and Tassone, 1991). Both coastal salt marshes and estuarine marshes are valued primarily as nurseries and food sources for coastal fish
LOWER YOUNGER HIGHER LONGER LESSER COARSER LESSER HIGHER GREATER GREATER
-----
w
0
W
+
ELEVATION MATURITY
INUNDATION FREOUENCY
DURATION OF ANY GIVEN INUNDATION PERIOD
WEAR STRENGTH
-
DEPTH RATIO OF CREEKS
*
LOWER SHORTER
t FINER
GRAIN SIZE
m
OLDER
t GREATER
t GREATER
PLANT DETRITUS RETENTION
WIDTH
HIGHER
*
___+
CREEK MOBILITY
SEDIMENT SUPPLY BY TIDAL SHEET FU)W
LOWER
t LESSER A
:52 Fig. 11-1.Principal processes and characteristic features of an estuarine/brackish marsh.
9
r
309
SALT MARSHES
f a
I b
Fig. 11-2. a) Location, setting and coastal engineering structures of the Fraser River delta. b) General sediment and marsh distribution at the western delta front of the Fraser River.
and wildlife (Frey and Basan, 1985; Hutchinson et al., 1989; Alberts et al., 1990; Adam, 1990). They may also help, at least temporarily, to remove toxic metal ions from estuarine circulation (Adam, 1990; Alberts et al., 1990; Valette-Silver, 1993), buffer the force of storm waves before they strike coastal communities (Frey and Basan, 1985) and offer sites for recreational activity which need not be harmful to the marsh. However, coastal marshes commonly fringe highly populated areas and are reclaimed for housing, agricultural or industrial purposes, waste disposal or destructive recreational practices (Adams, 1990; Alberts et al., 1990, Allen and Pye, 1992b). The impact of this encroachment will be compounded by erosion and protracted submergence linked to rising sea level particularly where dykes prevent landward transgression (Pethick, 1992). In an estuarine/deltaic setting, further degradation can result from sediment starvation associated with damming, training and dredging of rivers. River training may also degrade a marsh by increasing local energy conditions and sediment supply (Gibson, 1994).
OVERVIEW OF COASTAL MARSH MORPHO-SEDIMENTOLOGY
Literature reviews of coastal salt and estuarine marshes, some forming the background of recent studies, focus alternatively on their morpho-sedimentology (van Straaten, 1978; Weimer et al., 1981; Pethick, 1984; Frey and Basan, 1985; Allen,
310
J.L. LUTERNAUER ET AL.
1990a, b, 1992; Martini, 1991; Allen and b e , 1992a; French and Spencer, 1993), ecology (Chapman, 1974, 1977; Glooschenko et al., 1988; Adam, 1990; Mitsch and Gosselink, 1993) and geomorphological and climatological setting (Redfield, 1972; Dijkema, 1984; Roy, 1984; Hutchinson, 1988; Kelley et al., 1995). Characteristics of coastal salt marshes in general, and estuarine marshes in particular, can be drawn from these and other sources. Prior to and after colonization by vascular plants the intertidal sand or mud flat is commonly stabilized by algae (Moody, 1978; Coles, 1979, Hutchinson, 1982) (Figs. 11-1, 11-3a, b). The seaward edge of a developing coastal marsh may then form a continuously vegetated front or isolated clusters of pioneering plant communities (Fig. 11-4a7b). Established vegetation can contribute to marsh accretion by baffling sediment-carrying water flows and by directly sequestering fines from the flow (Hubbard and Stebbings, 1968), leading to the local raising of the marsh surface (Fig. 11-5). Surficial sediment textures in this transitional environment between the outer flats and marsh vary from silty fine sand to fine sandy silt. Examination of the subsurface lithology evident in trenches and channel banks at the Fraser delta (Williams, 1988; Williams and Roberts, 1989) reveals a variation in sediment texture analogous to that on the surface. Gradations from mostly fine sand to mostly silt are common. As vegetation preferentially promotes the sedimentation of fines, vegetation colonization may induce the development of an abrupt contact from sands to silts. Erosive contacts between deposits presumably mark the location of former tidal channels or storm-wave deposits. Bedding, as revealed by silt layers, is mainly horizontal and often disrupted by bioturbation. Uncommon shallow angle (ca. 10') crossbedding, presumably is formed by tidal channel point bar migration. Concentration of organic material tends to decrease with depth. Alternations of finer and coarser sediments with depth are frequently observed in the subsurface, but, overall, deposits coarsen with depth. Marsh colonization can be inhibited on relatively impermeable mud pools (fine sediment traps) fringing the leading edge of the marsh when continuous sheet flow discharged from the marsh on a low tide keeps these sediments submerged (Fig. ll-4a, b). At such sites on the Fraser delta, for example, marsh colonization proceeds only at more elevated localized areas including raised margins of intertidal creeks dissecting the mud pool surface (Luternauer, 1980). More rapid colonization in these areas will probably occur only after the mud flat drainage system is integrated with established creeks in the marsh at which time less ebb tidal drainage from the marsh will wash over the fringing mudflat surface (Medley, 1978; Luternauer, 1980). Large volumes of sand dumped at distributary mouths during major river floods (Figs. ll-6a, b) also can promote extensive colonization of the marsh by establishing well-drained elevated surfaces. Mineralogic sediments in marshes generally consist of sand and mud and generally are coarser in the lower parts of marshes (Fig 11-1). Upper marsh sediments may be
Fig. 11-3. a) Dessicated algal mat cover at leading edge of marsh. Cracks are on average -1 cm wide. b) Dense algal mat cover adjacent to marsh in dendritic creekimud pool area (Fig. ll-4a, b).
SALT MARSHES
311
312
J.L. LUTERNAUER ET AL.
Fig. 11-4. a) Oblique view of upper intertidal area on Sturgeon Bank illustrates inhibition of marsh colonization adjacent to poorly drained rnudflat in contrast to vigorous colonization on coarser sediments. Linear feature at bottom left of photograph is a dyke. b) Oblique view of section of dendritic creek network in mud pool on upper unvegetated tidal flat on southern Sturgeon Bank. Note featureless part of flats between marsh front and creeks. This part of flats is awash with water draining from marsh. By concentrating this flow and directing it to the lower flats, local creeks make intertidal surface adjacent to channels more suitable for vascular plant colonization (note scattered clumps of vegetation on dry tidal flat surface in foreground). (Medley, 1978; Luternauer, 1980.)
SALT MARSHES
313
Fig. 11-5. Silt entrapment by pioneering Scirpus amencanus at leading edge of marsh (1-m rule for scale). From Williams and Roberts, 1989, with permission.
coarser then those at the lower marsh where they are supplied from nearby barrier islands or cliffed headlands (Frey and Basan, 1985), and gravel can be an important constituent of estuarine marsh sediments in the Arctic (Jeffries, 1977; Adam, 1990; Martini, 1991) In the fully developed estuarine upper marsh environments of the Fraser River delta deposits consist of organic-rich clayey to fine sandy silt ,(Williams and Roberts, 1989). Sand content in the samples from this zone commonly is negligible and has not been shown to exceed 7%. Examination of channel banks and trenches (Fig. 11-7) indicates that abundant organic material is present within subsurface deposits and that horizontal stratification is partially disrupted by roots. Subarctic upper marsh sediments on the western side of Hudson Bay and in James Bay may consist of a freshwater mire during the summer when the area is washed by meltwater runoff above the permafrost, and is not submerged by tides (Kershaw, 1976). In this region, marsh creek systems are poorly developed (Kershaw, 1976) and pans (closed depressions which retain water during all stages of the tide) are present in locally dense clusters (Riley and McKay, 1980). These features likely are formed by ice scour (Riley and McKay, 1980) and enlarged by Snow Geese feeding habits (Jeffries et al., 1979). Elsewhere, creeks and pans may occur independently of each other or together (Adam, 1990). Where they do occur, pan density frequently is highest in the higher parts of the marsh (Pethick, 1974). Redfield (1972) states that pan formation is induced by local differences in the rate of marsh accretion, and that these features are particularly common within marshes which develop on intertidal sands with low relief. In addition to being created by the mechanisms described above, pans may develop or be maintained by the blocking of creeks, smothering of sites by drift litter, undermining of marsh surfaces by piping, and by vegetation
314
J.L. LUTERNAUER ET AL.
Fig. 11-6. a) Aerial plan view of Brunswick Pt. marsh (to the south of the mouth of Canoe Passage in Fig. ll-2a, b) area in 1948 after major river flood. A denotes vegetated marsh; B identifies the unvegetated deposit thought to have accumulated during the flood (A. Tamburi, pers. commun., 1978). b) Oblique view of Brunswick Pt. marsh in 1978. Area B is fully colonized by S. amencanus and has more than doubled the areal extent of the local marsh. Colonization by Salicomzu sp. reflects isolation from fluvial discharge in spite of this areas proximity to the channel mouth (Moody, 1978).
enclosing a cliff formed during the interruption of marsh extension (Adam, 1990). Regional variations in sediment coarseness and tidal range can influence creek density and tortuosity (Adam, 1990). Marshes having a moderate tidal range and a mud surface may have more complex creek patterns than do those in areas having
SALT MARSHES
315
Fig. 11-7. Bioturbated laminae and beds within the upper marsh. Scale is 46 cm long.
a high tidal range and coarser sediments. Although the position of established creek systems tends to remain relatively stable (Adam, 1990), lower reaches of creeks migrate more than their upper reaches which lie in more densely vegetated areas (Weimer et al., 1981). Also the width to depth ratio of creeks decreases with age (French and Spencer, 1993) (Figs. ll-ga, b, 11-9). Levees have more aerated sediments than the adjacent marsh and are relatively higher in the lower marsh where mineralogic sediment generally is more abundant (Adam, 1990). Levee sediments presumably are more aerated because of the higher sand content of marsh deposits in channel proximal locations. Sedimentation rates in marshes usually are on the order of mm/a (Frey and Basan, 1985), but rates of progradation can be exceedingly rapid (Adam, 1990). Kestner (1962) demonstrated that an estuarine marsh in England extended at an average rate of 50 m/a for about ten years. Where the rate of erosion and/or submergence exceeds accumulation, progradation ceases (Adam, 1990). Mean annual accretion rates for the brackish marshes of the Fraser delta, for example, show considerable lateral and temporal variation; rates ranging from 2.6 to 20.5 mm/a have been calculated using 137Csdating at 5 sites for sediments accumulated over a period of 40 years (Williams
316
J.L. LUTERNAUER ET AL
Fig. 11-8. a) Representative morphology of the lower marsh prior to seasonal growth. b) Vegetation of the lower marsh at peak of seasonal growth.
and Hamilton, in press). Based on sedimentation rates of about 20 mm/a derived from artificial marker beds, Hutchinson (1990) has suggested they are capable of maintaining themselves against predicted rates of global sea level rise over the next century (Clague, 1989). Recent documentation of a growing sediment deficit within the lower reaches of the Fraser River (McLean and Tassone, 1991) should prompt reevaluation of this issue (Williams and Hamilton, in press).
$ALT MARSHES
317
Fig. 11-9. Typical high marsh creek flanked by dense vegetation. Creek wall is ca. 1 m high.
The most prominent physical structures in many marshes are discontinuous wavy to parallel laminations or continuous laminations and thin bedding (Frey and Basan, 1985). This distinct bedding form is produced primarily by the rhythmic alternation of organic and inorganic beds described by van Straaten (1954), Bouma (1963) and Evans (1965). Preservation of such features will be affected by the intensity of local bioturbation (Adam, 1990). Variations in bioturbation likely are governed by population densities, available types of plants and animals, and rates of marsh accretion, or a combination of the three (Howard and Frey, 1973; Frey and Basan, 1985; Martini, 1991). Coastal marshes commonly are acidic and anaerobic (Frey and Basan, 1985; Adam, 1990). However, variations in soil chemistry have been recognized between high and low marshes (Gray and Bunce, 1972), even at the same tidal level. These variations are due to physiochemical effects of different animals and plants within different sediment types (Clarke and Hannon, 1967,1969; Long and Mason, 1983; Frey and Basan, 1985; Adam, 1990) or because of variations in sediment grain size (Randerson, 1979). Organic matter concentrations generally are low and variable, but commonly increase
318
J.L. LUTERNAUER ET AL.
with elevation (Packam and Liddle, 1970; Gray and Bunce, 1972). Gibson (1994) notes that this increase is largely relative, and parimarily caused by the decreasing input of inorganic material as the marsh surface accretes, rather than an increasing organic contribution. Also, although research shows that organic concentrations are low when calculated as a dryweight percentage, this method of calculation perhaps underestimates the true importance of salt marsh organics. It may prove more meaningful to determine organic content by volume rather than weight. Fine-grained sediments in salt marsh environments represent a major repository for contaminants (Valette-Silver, 1993). Heavy-metal pollution of sediments appears in the early 1900's and increases dramatically for sediments deposited between 1940 and 1970. The trend in more recent sediments has been for contaminants to decrease. However, metals bound in the sediments or organic compounds resistant to degradation can be reintroduced into surrounding water by physical disruption or by biological activity. Sediment cores from salt marshes have been used to provide an indication of historic pollution and a baseline against which to measure current contaminant levels. The strong reducing conditions in salt marsh sediments may result in large portions of imported metals being retained in the sediments as insoluble metal sulfides (Giblin et al., 1980). Low marsh areas may lose metals due to the frequency and duration of inundation, but it has been found that metals are associated with organic materials which readily accumulate in the upper marsh (Lindberg and Hariss, 1974). This retention of contaminants varies according to the specific metal. Lee et al. (1976) found, for example, that 98% of lead entering a marsh remained in situ whereas 33% of cadmium was taken up by the vegetation. Mercury concentrations have been found to be related to levels of organic carbon in sediments. Concentration of metals within plants appears to coincide with that in associated sediments, but bioaccumulation of metals other than mercury does not appear to occur (Drifmeyer and Redd, 1981). In a study by Moody (1989) of various organic and metal contaminants, it was concluded that only Cd, Hg and PCB were bioaccumulated in the marsh vegetation and that the amount varied by plant species. The most prevalent marsh species of the northwest Pacific Coast, Curex Zyngbyei, displayed the highest accumulations.
ESTUARINE MARSH DYNAMICS
For our understanding of salt marshes in general, and estuarine marshes in particular, to advance from its present state, it is imperative that research shift from its traditionally qualitative approach towards a more quantitative approach focused on physical processes. This physical process approach lends itself more readily to the construction of numerical models which simulate the dynamics of a system. Continued qualitative research will not directly lead to an enhanced ability to predict morphologic response but will serve chiefly to add to the already voluminous literature that is mainly site specific. Allen and F'ye (1992a) stressed the processresponse approach in their review of current research, stating that it should be
319
SALT MARSHES
used as a tool to aid in the management of sensitive areas. They further stressed a need to develop such models over a short period of time. Allen and Pye continued by noting that the “physical processes [of estuarine marshes] have been relatively neglected” in historical research. Particular emphasis has been placed on interpreting sediment distribution as a method for modelling geomorphic change (Stevenson et al., 1988; Allen, 1990a, b; Allen and Pye, 1992~1,Delaune et al., 1992; French and Spencer, 1993; Williams and Hamilton, 1992,1994). In the context of their exhaustive review, Allen and Pye raised five questions which touch directly on the issue of using numerical models to represent the dynamics of estuarine marshes. Although partial answers exist to these questions, continued gaps in our knowledge of marsh environments provide expansive ground for future study. The extent of the answers to these five questions is outlined herein to help focus discussion on the application of numerical techniques to marsh dynamics. 1) What are the sources and volumes of sediment supplied to estuarine marsh complexes? There are, in general, four sources of sediment to a marsh system: minerologic sediment derived from i) marine waters and ii) riverine waters, organic sediments derived from iii) plant detritus and iv) the redistribution of both types of sediment within the marsh itself (Fig. 11-10). The relative contribution of each of these sources of sediment varies from site to site, and the dominance of one source over another has been a focus of research. In the work by Allen (1990a) and Shi (1993) the organic component is considered to be a small constant or negligible compared to the mineralogic input from riverine and marine waters. From their perspective, the mineralogic input far exceeds that of organics. However, to ascribe a dominance to either riverine or marine sources of sediment may be a futile task since the initial source of sediment in an estuary will most often be the riverine waters of the estuary itself and, as noted by Elliot (1978), marine processes of waves and tides do much to agitate and redistribute this sediment. This mixing makes it very difficult to determine whether sediment is being supplied directly from riverine waters or if marine redistribution of the estuarine plume is more important.
PLANT DETRITUS
* SEDIMENTS IN MARINE WATERS
\* b
SEDIMENTS IN RlVERlNE WATERS
f-
320
J.L. LUTERNAUER ET AL.
The relative volumes of sediment supplied to an estuarine marsh depend heavily on the concentration of sediments in the estuarine flow and the mixing of that flow with marine waters. The volumes of sediment also likely will be determined by tidal regime and amplitude, the discharge regime of the river and the sediment calibre. The relative contribution from the various sediment sources is influenced by the development stage and elevation of the marsh. Organic accumulation of sediments becomes more important than minerologic accumulation over the long term as the flooding frequency and duration are reduced at any given point by vertical accretion of the marsh and as the abundance of organic detritus increases. In addition to this, with increasing age (elevation) of the marsh there is an associated gradual change in the relative influence of marine and riverine sedimentation (Gibson, 1994). The sources of sediments to estuarine marshes will determine the size and mineralogy of the sediment input. The sediment grain size will have a large influence on the overall marsh dynamics. Estuarine marshes generally consist (with some exceptions noted earlier) of particles ranging in size from fine sand to mud and are generally coarser at the lower elevations where fluid velocities are high enough to winnow out the smaller grains, or are too high to permit silt and clay grains to settle from suspension. Obviously there is a wide range of grain sizes present in any given marsh environment. Changes in grain size will control the effectiveness of particular mechanisms to erode or deposit sediments within a marsh. The site specific nature of the sediments means that this must be considered variable over space and consequently within any model so that the model can be tailored to suit a given environment. Any model should also consider potential seasonal changes of surficial sediment sizes driven by seasonal variability of tide or river discharge regime and vegetation growth. 2) What are the sediment transport pathways which need to be incorporated in a quantifiable assessment of the sediment budget? Sediment transport pathways in general are a product of the energy delivered to the environment, which can mobilize sediments and redistribute them elsewhere within the system, and the sediments themselves. In the case of an estuarine marsh, sediment can be mobilized by fluid shear both from riverine and marine waters, by gravitational acceleration on channel cut banks and steep slope gradients, and by shear generated by wind waves. Recent studies by Craft et al. (1993) suggest that irregularly flooded upper areas of marshes accrete more as a result of in situ production of organic debris as opposed to the importation of mineralogic sediment by fluid mediums. Regularly flooded marshes accrete more as a result of sediment carried in turbid waters, either marine or riverine in origin, rather than by the accumulation of in situ organics. Due to the tidal influence on the water present in the marsh, fluid-shear based pathways will tend to be bidirectional associated with the ebb and flood limbs of the tide cycle. Areas of the marsh more strongly dominated by riverine flow (i.e., proximal to a river) will tend to be more unidirectional, transporting sediment seawards. At the shoreward edge of the marsh wind-wave driven motion will also tend to generate bidirectional sediment transport pathways with the net motion of
321
SALT MARSHES
RIVERPROXIMAL
SEAWARD
LANDWARD
Fig. 11-11. Relative contribution of sediment pathways to sediment supply over an estuarine marsh system.
sediment dependent upon the wave spectrum and the differences between low and high frequency oscillations. Wind wave motion will also tend to drive the longshore current which provides lateral redistribution of sediments at the marsh front. At abrupt changes in elevation, sediment transport pathways may be determined more by gravity than by fluid velocity, with the direction of transport downslope. These gravity fed pathways will be less evident on shallower sloping surfaces within the marsh but may reduce the critical shear needed to entrain sediment should gravity and the shear force be at work in the same direction. From a modelling point of view this variability in transport pathways means that the delivery of sediment to the marsh will be difficult to represent explicitly due to its high variability through space (Fig. 11-11); a single process may not account for all of the marsh morphodynamics. Riverine derived pathways will tend to decrease in importance with increasing distance from the estuary mouth, marine derived pathways will tend to decrease in importance with increasing distance from the seaward marsh edge, and detrital pathways will tend to decrease with increasing distance seaward. Since concentration gradients decrease on perpendiculars from the drainage channel thalwegs into the marsh upland as a result of vegetative trapping of sediment by reduction of local flow velocity (French and Spencer, 1993), it may be suggested that the energy gradients and associated sediment transport pathways are similarly aligned. Slope derived pathways are variable depending on the localized rate of change in elevation. Any numerical model derived to predict marsh change in response to physical forcing should, in its formulation, take into account the relative importance of the different pathways at each segment along the marsh profile. Previous studies (e.g., Allen, 1990a) considered that sediment transport pathways linked to tides could be represented by an empirical relationship between the total suspended solids (TSS) and the tidal height. However, on the basis of more recent research by French and Spencer (1993) it is obvious that this simple relationship is more often than not obscured by deposition and resuspension cycles within the tide cycle, sediment mobilization by wind-waves and changes in the background concentration offshore resulting from the passage of weather systems. Sediment
322
J.L. LUTERNAUER ET AL.
transport pathways are far from simple and easy to study in isolation since they must be investigated as part of a greater whole. 3) What are the erosional and depositional mechanisms as governed by tides, waves and plants involved in marsh and mud flat formation? That estuarine marshes are both erosional and depositional has been a topic of debate for many years (Pethick, 1992). Within the context of marine, riverine and detrital based transport pathways it stands to reason that erosion and deposition are both prevalent in a marsh environment and that no one mechanism is responsible for all the observed morphologic change. The energy available for geomorphic work across the marsh is critical to the mechanisms involved in erosion and deposition. For instance, at a given distance from the landward edge of the marsh, wave-derived fluid energy will tend to be lesser in the drainage channels and greater on the channel banks due to differences in water depth creating different wave refraction angles. However, unidirectionally derived fluid energy will be greater in drainage channels and lesser on the channel banks due to the same difference in water depth causing changes in the total local momentum of the fluid. The presence of vegetation on marsh substrates tends to locally reduce the velocity of the fluid current near the bed thereby reducing the energy available to move sediment through fluid shear stress. Pethick et al. (1990) used a series of flume studies to show that vegetation mats create a deeper roughness length than bare patches of ground such that deposition is strongly encouraged in the area where the vegetation exists. This roughness layer was determined to be on the order of 0.03-0.1 m in height creating a boundary layer of water significantly slower than the water above 0.1 m. Over flat, bare ground the effective roughness length and the height of the boundary layer can be considered equivalent to the Nikuradse grain roughness which is of the order of the grain size of the substrate particles: in most cases this is of the order of millimetres. Over bare ground where bedforms are present this boundary layer is of the order of 2 or 3 times the bedform height (Kroon, 1991). Obviously the presence of both bedforms and vegetation will have a significant impact on the mechanisms of sediment entrainment by reducing velocity induced shear at the sediment-water interface. On steep slopes mass wasting will occur under gravity as well as density flows. Particular density flows may maintain their competency across the air-water interface and continue downslope as a density or turbidity current within the water. Slumping of marsh deposits most commonly occurs as unvegetated sand flat deposits are eroded, undercutting the more cohesive uppermost organic-rich marsch sediments. Marsh instability can be influenced by physical changes in the marsh environment to the extent that local change results in an alteration to the local plant community. Such change can be in the form of plant die-off resulting in a reduction of the local root mat cohesion and localized erosion of the marsh substrate. These changes in the viability of plant communities across the marsh strongly influence the erosional and depositional regimes at a given site. A multitude of mechanisms for erosion and deposition of sediment exist in a moving fluid medium. Convective entrainment will occur where turbulent bursting is
SALT MARSHES
323
present in the boundary layer as a result of turbulent instabilities in flow near the bed, possibly caused by vortex generation and ejection in the presence of bedforms. Advective entrainment of material will carry sediments horizontally once they have been entrained into the water column such that material is exported from one area and imported into another. Diffusion will occur down any concentration gradient present. Sediment will tend to settle out anywhere fluid velocities are reduced below the critical value for fluid shear such that the flow can no longer maintain the particles in suspension against the force of gravity or provide enough traction to motivate bedload material. These fluid-driven mechanisms are present in tidally, wind-wave and fluvially derived flows. In order to be able to understand these mechanisms involved in sediment transport throughout a marsh it is imperative that the controlling variables be established from a physical basis to be able to build a model that will address morphologic response with some degree of precision.
4 ) What is the role of geomorphic and sedimentologic variability through time and space in marsh aggradation or degradation? Spatial variability in geomorphology and sedimentology acts as a control on sediment transport pathways and mechanisms involved in erosion and deposition. In the long term, topography serves as a broad control on the accretion rates of a marsh system since accretion decreases with increasing elevation and increases with increasing relative proximity to the drainage channel (French and Spencer, 1993). Similarly, grain size tends to decrease with elevation, mud accumulating as a characteristic of the high marsh intertidal zone (Allen and Pye, 1992a). The variability in median grain size with elevation will affect the erodability of the sediments. As grain size increases above that of medium sand and decreases below that of very fine sand, greater fluid shear is necessary to entrain and transport the sediments. The decrease in grain size with elevation will in some ways counteract the decrease in fluid shear with elevation. As the sediment size progresses through silts to clays the changing particle size serves to enhance the cohesive nature of the material resulting in larger fluid shearing being necessary to entrain the sediment. In addition to the elevational controls, the morphology of the drainage channels themselves influences the discharge through time such that the local velocity variations will be in some way affected by form: high channel sinuosity will tend to slow the flow as opposed to relatively straight channels with low sinuosity. Regional variations in sediment coarseness and tidal range will influence creek density and tortuosity. AS the work of French and Spencer (1993) and others show, over the long term, the width to depth ratio of creeks decreases (Figs. ll-8a1 b, 11-9), temporal variations in morphology must therefore also be assessed. Marshes having a moderate tidal range and a mud surface may have more complex creek patterns than do those in areas having a high tidal range and coarser sediments. In arctic environments as an example, creek systems are poorly developed as a result of low tidal range and a coarser grained surface than more temperate environments (Kershaw, 1976). Although the position of creek systems once established tend to remain relatively stable, the lower reaches of creeks migrate more than the upper reaches since the increased vegetative cover of the higher elevations retards erosion (Weimer et al.,
324
J.L. LUTERNAUER ET AL.
1981). The morphology of the drainage systems is thus affected by the presence of plants and the specific plant types which, through their rooting structure, serve to control the slope and stability of the channel banks. The control on slope by plants will further influence the processes of mass wasting and sapping that occur. Wildlife utilization of marshes can influence the long-term stability and evolution of marsh environments. On the Pacific coast, marshes are used extensively by waterfowl migrating along the Pacific Flyway. A large population of Lesser Snow Geese (Anser caerulescens caerulescens) winters in the Puget Trough area between Washington State, U.S.A. and British Columbia, Canada. Overwintering flocks of Snow Geese consisting of as many as 50,000 birds may significantly influence the geomorphology and sedimentation in marshes that lie within this region (Hutchinson et al., 1989). They influence geomorphology and sedimentation by grubbing for rhizomes of Scirpus americanus for food and creating craters in the marsh substrate as a byproduct of this feeding activity. As this activity usually takes place during a rising tide, more suspended fine sediment is carried higher into the marsh than would normally occur. Geomorphologic and sedimentologic variability can be assessed at many different temporal and spatial scales from the microscale of a single grain through the mesoscale of channel form to the macroscale of tectonic adjustment of the earth's surface. What must be borne in mind when deriving a numerical model of marsh dynamics is the relative magnitude and frequency of occurrence of these many scales and their interrelationships.
5) What is the response of marshes to historical forcing such as tide regime, wind wave climate, sea level change and the influence of local factors with high spatial frequency and low magnitude effects? This question is the crux of the issue from a numerical modelling point of view. The single most critical aspect of any numerical model is not the variables included but the relationships between those variables and how well they simulate the natural environment which determine the model's usefulness. For example, it is no good knowing what the pore water content of the marsh sediments is if we do not know the effect of such a variable on the response of a marsh system. The response of marsh systems to tidal forcing has been greatly studied in the past. Pestrong (1965), from his extensive work in the San Francisco Bay Area, concluded that tidal forcing was the dominant forcing factor of marsh response, a view corroborated by Elliot (1978). Both Pestrong (1972) and Elliot investigated the differences in flow on the flood and the ebb tide and found them remarkably different in character. On the flood tide, flow is initially restricted to the marsh drainage channels while the flow source is solely the tidal flat channels. As the tide rises, flow within the marsh remains within the drainage channels while the source water comes from both the tidal flat channels and from sheet flow across the flats where it can entrain sediments over a broad expanse. At a particular time when the tide height equals the elevation of the marsh surface the marsh is overtopped and the tide washes over the marsh as both channelized flow in the drainage channels and as sheet flow over the vegetated marsh surface, flooding the marsh with
SALT MARSHES
325
sediment rich water. This sheet flow is retarded by the presence of vegetation causing localized deposition. At the tidal maximum, flow velocity decreases to a minimum (barring agitation by wind-waves) resulting in a period of relative calm during which deposition occurs. This period of slack water is lagged behind slack water on the open ocean with increasing distance inland. When flow reverses on the ebb tide it initially flows seaward as sheet flow directly across the marsh. With further lowering in water levels, the flow is focused into drainage channels such that sheet flow from the marsh is diverted into the drainage network which is then channelized seaward. Nearing the tidal minimum, drainage in marsh environments resembles terrestrial systems with all flow and bank drainage being tributary to the main marsh drainage channels. Obviously the flow dynamics over a marsh vary greatly over a tide cycle. As noted by Elliot (1978), the asymmetry in the tidally-driven fluid velocities and the asymmetry in their duration is critical to determining whether there is net erosion or net deposition within the marsh. With elevational, vegetational and tidal controls on velocity, this asymmetry is highly variable through space. At the seaward edge of the marsh, and on either side of the tidal maximum, wave-driven processes are important to marsh morphology (Allen and Pye, 1992a). Pethick (1992) noted that this morphodynamic response was the same as that for beaches given similar periods of storm and calm. Pethick further concluded that marsh morphology was a direct response to the distribution of wave energy across the intertidal profile. However, at low tidal stages the marsh is influenced only by fluid motion akin to common terrestrial drainage, therefore the influence of wave energy on a marsh must decrease inland with increasing substrate elevation. A local rise in mean sea level will cause a landward shift in marine conditions and energy regimes which can lead to long-term recession of the marsh and an altered profile (Pethick, 1992). Where rates of sedimentation exceed rates of erosion the delta and marsh system will still prograde but the areal extent of the marsh may not increase. Although new marsh is established at the seaward limit of the system, the upper marsh is less influenced by tidal action as its surface is raised and is gradually transformed into a floodplain. The influence of local factors is highly variable and likely to change within a marsh as much as between marshes. For example, Pestrong (1965) developed a series of relationships concerning the erodability of sediments within a marsh. H e established that the erodability was proportional to both the pore water content and the median grain size. He also established inverse proportionalities between erodability and dry density, organic content and percentage clay. Allen and Pye (1992a) determined that the duration of exposure to air caused differences in the geotechnical properties of the sediments as a result of different drying times such that it influenced the erodability of the substrate. Variations in soil chemistry have been recognized between areas of high and low elevations within marshes (Gray and Bunce, 1972; Randerson, 1979) and at the same tidal elevation because of the physiochemical influences of different biota with different sediment types (Clarke and Hannon, 1967,1969; Long and Mason, 1983; Frey and Basan, 1985; Adam, 1990) As noted before, the type of vegetation present at a site influences the sediment motion such that net accretion or erosion is linked to the development of the
326
J.L. LUTERNAUER ET AL.
plant mat. The type of vegetation present at a site also influences the degree of shading available to the marsh substrate. This shading tends to reduce the surface temperature and lessen the rate of desiccation of the sub-aerially exposed sediments. The alteration of the drying regime affects the geotechnical properties of the underlying material ultimately acting as a control on localized erosion. These examples illustrate that numerical models need to incorporate local factors into a more generic physical-process framework in order to represent the site specifics of any given marsh environment.
MODELLING ESTUARINE MARSHES
Having established in the previous section the questions to be asked and outlined the status of the answers to those questions, models themselves need to be addressed since different models are useful for different purposes. However, from a dynamical point of view, predictive models of marsh responses to physical processes need to represent the interrelationships between mechanisms. Fox (1985) reviewed past modelling work and classified the models into four generic types: i) physical models which are built and scaled by means of dimensional similarity using Froude and Reynolds numbers; ii) statistical models which develop site specific relationships on the basis of regression techniques over a large number of samples; iii) probabilistic models which attempt to match the frequency of events using probability theory; and iv) deterministic models which seek, through the laws of hydrodynamics, to predict change but allow for no random elements. Of these four model types, a combination of the probabilistic and the deterministic approaches such that both random and deterministic events are incorporated will simulate natural environments best. Physical and statistical models are both site specific and must be redesigned in order to apply to a different locale. The probabilistic and deterministic approaches, so long as they are based on event frequency and physical laws, tend to be more generic. Once the model framework has been chosen then the temporal and spatial scales of the model need to be selected. A model designed to predict boundary shear stress over a meter of shoreline will differ greatly in design from a model designed to predict marsh profile evolution through time. Critically important variables such as elevation, inundation frequency and duration, grain size, flow velocity, tidal asymmetry and vegetation stand density to name a few need to be assessed in terms of their relevance at a particular scale and their input into a specific model. When these questions have been addressed the model can be derived. Currently most models are empirically based and statistical and deterministic in type. Allen (1990a) presented a deterministic sediment budget model to predict the change in marsh elevation. H e assumed that the rate of growth of a marsh substrate was controlled by: i) organic input and accumulation, ii) minerogenic input, iii) relative change in sea level and iv) the rate of long term compaction of the marsh sediments. This empirically derived model can be represented as:
AE = AME
+ AOE
-
ASE - APE
(11-1)
327
SALT MARSHES
where A E is the change in the elevation of the marsh surface (m), A M E is the minerogenic input adjusted for autocompaction and represented as a thickness (m), A O E is the organic input adjusted for autocompaction and represented as a thickness (m), ASE is the relative change in sea level (m), and A P E is the long term compaction of the marsh sediments (m). Allen treated the organic input as 0 or as a constant value of lop3 m/a. The change in sea level over the short term was set at 0 or at a small constant value for long term prediction. Since the minerogenic and organic input was already adjusted for compaction, the compaction term was ignored. Following this, eq. (11-1) reduces to:
A E = AM E
+ 0.001
(11-2)
Allen went further to say that A M E could be represented by:
AM E = k-
1
1-P
C H , t wt
(11-3)
where k represents the fraction of the annual input of sediments resident in the marsh after a one year period, p is the porosity of the deposit, A is one year, T is the duration of the wetting over each tide cycle, CH,t represents the TSS at a given tide height (H) and a given time ( t )in the tide cycle and W, is the settling velocity of the sediments over time. The fraction k is arbitrarily set at 0.2 and p is similarly set at 0.4. From this model Allen showed that the elevation change predicted matched the growth curves of Pethick (1981) developed from long term measurements of estuarine marshes. Shi (1993) tested the model with similar success on the Dyfi Estuary in Wales. The model predicted accretion rates of 10.0 mm/a based on a three month study, compared with rates of 11.5 mm/a based on laminase counts. Despite this agreement in rates, this model is limited in application to morphodynamics since it only predicts surface elevations for single locations. It does not accommodate profile changes, fluid dynamics and similar physical processes. However, as noted by French and Spencer (1993), the relationship between TSS and tide height is poorly developed. In fact the relationship between concentration and fluid velocity in the presence of bidirectional flow is poorly defined. To account for vertical variations in both fluid velocity and suspended sediment concentration, any numerical model for TSS must be able to define the velocity and concentration profiles and integrate their coupling over water depth. The integration of the two profiles over time and depth can be thought of as defining the suspended sediment transport rate for the water column through time. Thus at a given time, t , the suspended sediment transport rate for the whole water column can be defined (e.g., Sleath, 1984; Dyer, 1986) as: (11-4) where qs is the instantaneous suspended sediment transport rate; uZ,t is the instantaneous horizontal fluid velocity and cz,t is the instantaneous suspended sediment concentration at height z and at time t ; h is water depth and 1) is the water surface.
328
J.L. LUTERNAUER ET AL.
When integrated over depth, eq. (11-4) yields the instantaneous suspended sediment transport rate at time t for the whole water column; integrated over time, eq. (11-4) yields the time-averaged suspended sediment transport rate for the entire water column (Osborne and Greenwood, 1994): (11-5) where (qs)is the time-averaged and depth-integrated suspended sediment transport rate and T is the time interval of integration. To evaluate either eq. (11-4) or (11-5) in any mathematical sense, predictive models of the distributions of velocity and concentration with respect to depth and time must be derived so that both velocity and concentration can be computed for every elevation above the bed and for any moment in time. Atkins (1993) demonstrated clearly that the computation of concentration through time under oscillatory flow for a single elevation was indeed possible. He established a relationship between concentration and velocity which incorporated convective, advective and diffusive processes. The computational algorithm was derived from Bailard’s (1984) sediment transport model and then modified to simulate the three different processes. Atkins (1993) found that the time variability in velocity cubed explained 80% of the time variability in concentration when the concentration was lagged with respect to the velocity by approximately one twelfth of an oscillatory cycle. Thus a first approximation to the relationship between velocity and concentration under oscillatory flow can be represented by: Ct
=~
+
[ C O S U ~A t ]
(11-6)
where concentration, cf, is a sinusoidal function of velocity, uf, shifted by some time lag, A t . A model of this sort indicates the subtle complexities involved where the prediction of sediment transport dynamics is concerned. The prediction of marsh dynamics from a process response stand point depends on the development of similar models, with progress being made in small steps as the complexity of the problem is broken down into manageable components. Obviously there is a long way to go before we have a predictive model of marsh dynamics that is at once both physically based and truly generic.
SUMMARY
The estuarine marsh is a highly dynamic environment combining aspects of marine and fluvial environments at a critical interface or confluence between chemical, biologic and geomorphic processes. These processes affect the stability of estuarine marshes over time and space at varying scales. In order to predict the evolution of marsh environments, research more and more needs to be aimed at resolving critical questions concerning these physical processes such that the pieces of the puzzle may eventually be assembled into a generic, broadly applicable dynamic model. Broadly applicable models of estuarine sedimentation are unavailable at present and
SALT MARSHES
329
must await the distillation of more site-specific field measurements to yield generic characteristics. The critical generic characteristics which need to be input as model variables can fluctuate significantly in relative importance from one estuary to another. These variables include sediment supply (minerogenic versus organic sources, tidal versus fluvial sources), the phasing of the sediment discharge peak with fluid motion and plant growing season, tidal regime (amplitude, currents, asymmetry), wind wave climate, tectonic adjustment and sea level rise (Frey and Basan, 1985; Hutchinson, 1988; Stevenson et al., 1988; Allen and Pye 1992a, b; Martini, 1991; French and Spencer, 1993). The long term development of a marsh also must be identified as rates and modes of deposition will vary at different stages of marsh development. The key to advancing the state of knowledge concerning the development of estuarine marsh ecosystems lies in never losing sight of the big picture, and the implausibility of isolating individual components out of a natural continuum. This key lies in eschewing more descriptive research in favour of more quantitative, physically based investigations which will, by small steps forwards, increase our understanding. ACKNOWLEDGEMENTS
Bev Vanlier contributed to the typing and formatting of the manuscript. Drafting was done by Tonia Oliveric and the authors. This manuscript is a Geological Survey of Canada Contribution 15493. The National Research Council of Canada has kindly granted permission to reproduce Fig. 3b which originally appeared in the Canadian Journal of Earth Sciences, vol. 26, 1989, p. 1662. REFERENCES Adam, P., 1990. Saltmarsh Ecology. Cambridge University Press, Cambridge, 461 pp. Alberts, J.J., Price, M.T., and Kania, M., 1990. Metal concentrations in tissues of Spartina alfemiflora (Loisel) and sediments of Georgia salt marshes. Est. Coastal Shelf Sci., 30: 47-58. Allen, J.R.L., 1990a. 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., 1990b. The Severn Estuary in southwest Britain: its retreat under marine transgression, and fine sediment regime. Sediment. Geol., 66: 13-28. Allen, J.R.L., 1992. Large-scale textural patterns and sedimentary processes on tidal salt marshes in the Severn Estuary, southwest Britain. Sediment. Geol., 81: 299-318. Allen, J.R.L. and Pye, K., 1992a. Coastal Saltmarshes: their nature and importance. In: J.R.L. Allen and K. Pye (Editors), Saltmarshes: Morphodynamics, Conservation and Engineering Significance. Cambridge University Press, Cambridge, pp. 1-18. Allen, J.R.L. and Pye, K., 199213. Preface. In: J.R.L. Allen and K. Pye (Editors), Saltmarshes: Morphodynamics, Conservation and Engineering Significance. Cambridge University Press, ,cambridge, pp. vii-viii. Atkins, R.J., 1993. Sediment Suspension under Irregular “Groupy” Waves: a Laboratory Experiment. Unpublished M.Sc. Thesis, University of Toronto, 217 pp. Bailard, J.A., 1984. A simplified model for longshore transport. Proc. 19th Coastal Engineering Conference, ASCE, pp. 1454-1470. Bouma, A.H., 1963. A graphic presentation of the facies model of salt marsh deposits. Sedimentol., 2 122-129.
330
J.L. LUTERNAUER ET AL.
Chapman, V.J., 1974. Salt Marshes and Salt Deserts of the World. Cramer, Lehre, Germany, 2nd ed., preface + 293 pp. Chapman, VJ. (Editor), 1977. Wet Coastal Ecosystems. Elsevier, Amsterdam, 428 pp. Clague, J.J., 1989. Sea levels on Canada's Pacific Coast: Past and Future Trends. Episodes, 12: 29-33. Clarke, L.D. and Hannon, N.J., 1967. The Mangrove swamp and salt marsh communities of the Sydney district. 1. Vegetation, soils and climate. J. Ecol., 55: 753-771. Clarke, L.D. and Hannon, N.J., 1969. The mangrove swamp and salt marsh communities of the Sydney district. 11. The holocoenotic complex with particular reference to physiography. J. Ecol., 57: 213234. Coles, S.M., 1979. Benthic microalgal populations on intertidal sediments and their role as precursors to salt marsh development. In: R.L. Jeffries and A.J. Davy (Editors), Ecological Processes in Coastal Environments. Blackwell Scientific Publications, Oxford, pp. 25-42. Craft, C.B., Seneca, E.D. and Broome, S.W., 1993. Vertical accretion in microtidal regularly and irregularly flooded estuarine marshes. Est. Coast. Shelf Sci., 37: 371-386. Dawe, N.K. and White, E.R., 1986. Some aspects of the vegetation ecology of the Nanoose-Bonell estuary, Vancouver Island, British Columbia. Can. J. Bot., 64: 27-34. Delaune, R.D., Patrick, Jr., W.H., Smith, C.J., 1992. Marsh aggradation and sediment distribution along rapidly submerging Louisiana Gulf Coast. Environ. Geol. Water Sci., 20: 57-64. Dijkema, K.S., 1984. Geography of salt marshes in Europe. Z. Geomorphol., 31: 489-499. Drifmeyer, J.E. and Redd, B., 1981. Geographic variability in trace element levels in Spurtinu ulternifloru. Est. Coast. Shelf Sci., 13: 709-716. Dyer, K.R., 1986. Coastal and Estuarine Sediment Dynamics. John Wiley and Sons Ltd., Chichester, 342 pp. Elliot, T, 1978. Clastic shorelines. In: H.G. Reading (Editor), Sedimentary Environments and Facies. Elsevier, N.Y., pp. 143-177. Evans, G., 1965. Intertidal flat sediments and their environments of deposition in the Wash. Q. J. Geol. SOC.,London, 121: 209-241. Fox, W.T., 1985. Modelling coastal environments. In: R.A. Davis Jr. (Editor), Coastal Sedimentary Environments, 2nd. ed. Springer-Verlag, N.Y., pp. 666-705. French, J.R. and Spencer, T., 1993. Dynamics of sedimentation in a tide-dominated backbarrier salt marsh, Norfolk, UK. Mar. Geol., 110: 315-331. Frey, R.W. and Basan, P.B., 1985. Coastal salt marshes. In: R.A. Davis (Editor), Coastal Sedimentary Environments. Springer, New York, pp. 225-301. Giblin, A.E., Bourg, A,, Valiela, I. and Teal, J.M., 1980. Uptake and losses of heavy metals in sewage sludge by a New England salt marsh. Am. J . Bot. 67: 1059-1068. Gibson, J.W., 1994. Estuarine sedimentation and erosion within a fjord-head delta: Squamish River, British Columbia. Unpublished MSc. Thesis, Department of Geolography, Simon Fraser University, Burnaby, B.C., 346 pp. Glooschenko, W.A., Martini, I.P. and Clarke-Whistler, K., 1988. Salt marshes of Canada. In: National Wetlands Working Group (Editors), Canada Committee on Ecological Land Classification, Sustainable Development Branch Canadian Wildlife Service, Conservation and Protection, Environment Canada, Ecological Land Classification Series, 24: 349-375. Gray, A.J. and Bunce, R.G.M., 1972. The ecology of Morecambe Bay VI. Soils and vegetation of the salt marshes: a multivariate approach. J. Appl. Ecol., 9: 221-234. Howard, J.D. and Frey, R.W., 1973. Characteristic physical and sedimentary structures in Georgia estuaries. AAPG Bull., 57: 1169-1184. Hubbard, J.C.E. and Stebbings, R.E., 1968. Spartina marshes in southern England VII: Stratigraphy of the Keysworth Marsh, Poole Harbour. J. Ecol., 56: 707-722. Hutchinson, I., 1982. Vegetation-environment relations in a brackish marsh, Lulu Island, Richmond, B.C. Can. J. Bot., 60: 452-462. Hutchinson, I., 1988. The biogeography of the coastal wetlands of the Puget Trough: deltaic form, environment and community structure. J. Biogeogr., 15: 729-745. Hutchinson, I., 1990. Intertidal marshes of the Fraser River delta: the geological theatre and the ecological play. In: Program with Abstracts-Vancouver '90. Geol. Assoc. Can./Min. Assoc. Can.
SALT MARSHES
331
Annu. Meet., May 1990, Vancouver, 15: A62. Hutchinson, I., Prentice, A.C., and Bradfield G., 1989. Aquatic plant resources of the Strait of Georgia. In: K. Vermeer and R.W. Butler (Editors), The Ecology and Status of Marine and Shoreline Birds in the Strait of Georgia, British Columbia. Proc. Pacific Northwest Bird and Mammal SOC.and Canadian Wildlife Serv., Sidney, B.C., pp. 50-60. Jeffries, R.L., 1977. The vegetation of salt marshes at some coastal sites in arctic North America. J. Ecol., 65: 661-672. Jeffries, R.L., Jensen, A., and Abraham, K.F., 1979. Vegetational development and the effect of geese on vegetation at La Perouse Bay, Manitoba. Can. J. Bot., 57: 1439-1450. Karagatzides, J.M., 1987. Intraspecific variations of biomass and nutrient allocation in Scilpus americanus and Scirpus maritimzu. M.Sc. Thesis, Simon Fraser University, Burnaby, B.C. Kelley, J.T, Gehrels, W.R. and Belknap, D.F., 1995. Late Holocene relative sea-level rise and the geological development of tidal marshes at Wells, Maine, U.S.A. J. Coast. Res., 11: 136-153. Kershaw, K.A., 1976. The vegetational zonation of the East Pen Island salt marshes, Hudson Bay. Can. J. Bot., 54: 5-13. Kestner, F.J.T., 1962. The old coastline of the Wash. A contribution to the understanding of loose boundary processes. Geogr. J., 128: 457-478. Kroon, A., 1991. Suspended sediment concentrations in a barred nearshore zone. Proc. Coastal Sediments 91, ASCE, N.Y., pp. 328-341. Lee, C.R., Sturgis, TC. and Landin, M.C., 1976. A hydroponic study of heavy metal uptake by selected marsh plant species. Tech. Rep. D-76-5 Dredged Material Research Program. Waterways Experiment Station, U S . Army Corps of Engineers, Vicksburg, Mississippi. Lindberg, S.E. and Hariss, R.C., 1974. Mercury-organic matter associations in estuarine sediments and interstitial water. Env. Sci. Tech., 8: 459-462. Long, S.P. and Mason, C.F., 1983. Saltmarsh Ecology. Blackie, Glasgow. Luternauer, J.L., 1980. Genesis of morphologic features on the western delta front of the Fraser River, British Columbia - status of knowledge. In: S.B. McCann (Editor), The Coastline of Canada, Littoral Processes and Shore Morphology. Geol. Surv. Can. Pap., 80-10: 381-396. Martini, I.P., 1991. Sedimentology of subarctic tidal flats of western James Bay and Hudson Bay, Ontario, Canada. In: D.G. Smith, G.E. Reinson, B.A. Saitlin and R.A. Rahmani (Editors), Clastic Tidal Sedimentology. Can. Soc. Pet. Geol., Calgary, Alberta, pp. 301-312. McLean, D.G. and Tassone, B.L., 1991. A sediment budget of the lower Fraser River. Proc. 5th Federal Interagency Sediment. Conf., Las Vegas, Nevada. Medley, E., 1978. Dendritic drainage channels and tidal flat erosion, west of Steveston, Fraser River deltas, British Columbia. B.A.Sc. Thesis, University of British Columbia, Department of Geological Sciences, Vancouver, 70 pp. Mitsch, W.J. and Gosselink, J.G., 1993. Wetlands, 2nd ed.. Van Nostrand Reinhold, New York, N.Y. Moody, A.I., 1978. Growth and distribution of marsh plants on the southern Fraser Delta foreshore. MSc. Thesis, Univ. British Columbia, Vancouver. Moody, A.I., 1989. An investigation into toxic chemical accumulation in estuarine vascular plants. Environment Canada Regional Manuscript Rep. MS90-05. Osborne, P.D. and B. Greenwood. 1994. Sediment suspension under waves and currents: time scales and vertical structure. Sedimentology (in press). Packam, J.R. and Liddle, M.J., 1970. The Cefni saltmarsh and its recent development. Field Studies, 3: 331-356. Perillo, G.M.E., 1989. New geodynamic definition of estuaries. Rev. Geofis., 31: 281-287 Pestrong, R., 1965. The development of drainage patterns on tidal marshes. Stanford University Publications, Geological Sciences Volume X, Number 2, 87 pp. Pestrong, R., 1972. Tidal-flat sedimentation at Cooley Landing, southwest San Francisco Bay. Sediment. Geol., 18: 251-288. Pethick, J.S., 1974. The distribution of salt pans on tidal salt marshes. J. Biogeogr., 1: 57-62. Pethick, J.S., 1981. Long-term accretion rates on tidal salt marshes. J. Sed. Pet., 51: 571-577. Pethick, J., 1984. An Introduction to Coastal Geomorphology. Edward Arnold, London, 260 pp.
332
J.L. LUTERNAUER ET AL.
Pethick, J.S., 1992. Saltmarsh Geomorphology. In: J.R.L. Allen and K. Pye (Editors), Saltmarshes: Morphodynamics, Conservation and Engineering Significance. Cambridge University Press, pp. 4163. Pethick, J., Leggett, D. and Husain, L., 1990. Boundary Layers under Salt Marsh Vegetation Development in Tidal Currents. In: J.B. Thornes (Editor), Vegetation and Erosion. John Wiley and Sons Ltd., London, pp. 113-124. Pielou, E.C. and Routledge, R.D., 1976. Salt marsh vegetation: latitudinal gradients in the zonation patterns. Oecologia, 24: 311-321 Randerson, P.F., 1979. A simulation model of salt-marsh development and plant ecology. In: B. Knights and A.J. Phillips (Editors), Estuarine and Coastal Land Reclamation and Water Storage. Saxon House, Farnborough, pp. 48-67. Redfield, A.C., 1972. Development of a New England salt marsh. Ecological Monographs, 42: 201-237. Riley, J.L. and McKay, S.M., 1980. The vegetation and phytogeography of coastal southwestern James Bay. R. Ont. Mus., Life Sci. Contrib., 124: 1-81. Roy, P.S., 1984. New South Wales estuaries: their origin and evolution. In: B.G. Thom (Editor), Coastal Geomorphology in Australia. Academic Press, Sydney, pp. 99-121. Shi, Z., 1993. Recent saltmarsh accretions and sea level fluctuations in the Dfl Estuary, central Cardigan Bay, Wales, U.K. Geo-Mar. Lett., 13: 182-188. Sleath, J.F.A., 1984. Seabed Mechanics. Wiley and Sons, N.Y., 355 pp. Stevenson, J.C., Ward, L.G., and Kearney, M.S., 1988. Sediment transport and trapping in marsh systems: implications of tidal flux studies. Mar. Geol., 80: 37-59. Valette-Silver, N.J., 1993. The use of sediment cores to reconstruct historical trends in contamination of estuarine and coastal sediments. Estuaries 16: 577-588. van Straaten, L.M.J.U., 1954. Sedimentology of recent tidal flat deposits and the Psammites du Condroz (Devonian). Geol. Mijnbouw, 16: 25-47. van Straaten, L.M.J.U., 1978. Salt-marsh sedimentology. In: R.W. Fairbridge and J. Bourgeois (Editors), The Encyclopedia of Sedimentology. Dowden, Hutchinson and Ross, Inc., Strondsburg, Pennsylvania, pp. 642-644. Weimer, R.J., Howard, J.D., and Lindsay, D.R., 1981. Tidal flats and associated tidal channels. In: PA. Scholle and D. Spearing (Editors), Sandstone Deposition Environments. Am. Assoc. Pet. Geol., Tulsa, pp. 191-245. Williams, H.F.L., 1988. Sea-level change and delta growth: Fraser River Delta, British Columbia. Ph.D. Thesis, Simon Fraser University, Burnaby, B.C. Williams, H.F.L. and Roberts, M.C., 1989. Holocene sea-level change and delta growth: Fraser River Delta, British Columbia. Can. J. Earth Sci., 26: 1657-1666. Williams, H.F.L. and Hamilton, TS., in press. Sedimentation dynamics of an eroding tidal marsh derived from stratigraphic records of 137Cs,fallout, Fraser Delta, British Columbia, Canada. J. Coastal Res.
Geomorphology and Sedimentology of Estuaries. Developments in Sedimentology 53 edited by G.M.E. Perillo 0 1995 Elsevier Science B.V. All rights reserved.
333
Chapter 12
GEOMORPHOLOGYAND SEDIMENTOLOGY OF MANGROVES PIETER G.E.F. AUGUSTINUS
INTRODUmION
Mangroves are tidal forest ecosystems in sheltered saline to brackish environments. Their functional and structural properties are determined by a complex of climatic and site conditions, such as air and water temperatures, the availability of water, mineral nutrients and light (Hamilton and Snedaker, 1984). On a global scale, air temperature is the most determinant factor for the range of a species (Blasco, 1984). Mangroves thrive in the tropics, but extend even into the temperate zone if frost is exceptional and of short duration. Especially the establishment of the mangrove seedlings is sensitive to prolonged periods of severe low temperatures (Lug0 and Patterson-Zucca, 1977). The mangrove forests in those cases are usually deteriorated in species and attenuated to shrubform (Chapman, 1976; Baltzer and Lafond, 1971). Often, mangroves are considered to be the low-latitude equivalent of salt marshes. They, however, differ from salt marshes in two respects: in vegetation structure, being composed of trees and shrubs instead of herbs, and in their position with respect to mean high-water level. Due to special roots a number of mangroves is adapted to grow on completely water-logged soils, in lower positions than the salt marsh herbs are able to. This is especially the case if an estuary is lined up by mangroves with prop roots. The transition from the open channel to fringing Rhizophoras for instance, is made up by an irregular edge of outleaning prop roots, which at least partly are continuously submerged. These mangroves therefore have the ability to influence the hydrodynamics as well as the related processes of sedimentation and erosion at an earlier stage. Young intertidal deposits which are covered by vegetation are protected against erosion in two different ways as compared to uncovered sediments (Scoffin, 1970; Ong, 1982). Firstly, the dense network of trunks and above-ground roots (e.g. prop roots, pneumatophores) act as a fence, reducing the current velocity and so favour sedimentation and counteract erosion. Secondly, the underground root systems have a binding capacity, which also favour soil stability. The aim of this chapter is to present a state of the art of the role mangroves play in estuarine geomorphology and sedimentation.
GLOBAL DISTRIBUTION OF MANGROVE SPECIES
On a global scale Saenger et al. (1983) distinguish six mangrove regions, based on the occurrence of sixty exclusive species and twenty-three important non-exclusive
334
P.G.E.F. AUGUSTINUS
Table 12-1 Number of exclusive and non-exclusive mangrove species for the six mangrove regions indicated in Fig. 12-1 (after Saenger et al., 1983)
1 Asia 2 Oceania 3 West coast of America 4 East coast of America 5 West coast Africa 6 East coast Africa the Middle East
+
Exclusive species
Non-exclusive species
44 38
14
7 7 7 9
14 4 7 4 7
......................
Fig. 12-1. Global distribution of mangroves, divided into two mangrove zones, subdivided into six mangrove regions.
species of the mangrove ecosystem (Table 12-1). Most of the species occur mainly in two adjacent regions: Asia and Oceania (Fig. 12-1). For that reason others (e.g. Chapman, 1970,1975; Barth, 1982; Adegbehin and Nwaigbo, 1990) define two zones: an Eastern or Indo-Pacific zone rich in species, corresponding with the regions 1, 2 and 6 in Table 12-1, and a Western or Atlantic zone with only some ten species, comprising the other regions (Fig. 12-1). For an overview of the exclusive mangrove species see Table 12-2. All mangrove species have in common that they are adapted to loose and wet soils, saline to brackish habitats and periodic tidal submergence (Hatcher et al.,
GEOMORPHOLOGY AND SEDIMENTOLOGY OF MANGROVES
335
Table 12-2 Exclusive species of the world’s mangroves and their distribution in the six mangrove regions, indicated in Fig. 12-1 and Table 12-1 (after Saenger et al., 1983) Species
Life-form
Mangrove region
Acanthus ebracteatus Vahl. Acanthus ilicifolius L. Acanthus volubilis Wall. Aegialitis annulata R. Br. Aegialitis rotundifolia Roxb. Aegiceras corniculatum (L) Blanco Avicennia alba Blume Avicennia bicolor Standl. Avicennia eucalyptifolia Zipp. ex Miq. Avicennia germinans L. Avicennia intermedia Griff. Avicennia lanata Ridley Avicennia marina (Forsk.) Vierh. Avicennia officinalis L. Avicennia rumphiana Hall. f. Avicennia tomentosa Willd. Avicennia tonduzii Moldenke Bruguiera cylindrica (L.) Blume Bruguiera exaristata Ding Hou Bruguiera gymnorhiza (L.) Lam. Bruguiera hainesii C.G. Rogers Bruguiera palvijora (Roxb.) Wight and Am. Bruguiera sexangula (Lour.) Poiret Camptostemon philippinensis Becc. Camptostemon schultzii Mast. Ceriops decandra (Griff.) Ding Hou Ceriops tagal (Perrottet) C.B. Robinson Conocarpus erectus L. Cynometra iripa Kostel Cynometra ramiflora L. Excoecaria agallocha L. Heritiera littoralis Aiton ex Dryander Heritiera fomes Buch.-Ham. Kandelia candel (L.) Druce Laguncularia racemosa Gaertn. f. Lumnitzera littorea (Jack) Voigt Lumnitzera racemosa Willd. Nypa fruticans van Wurmb. Osbornia octodonta F. Muell. Pelliciera rhizophorae Planchon and Triana Phoenix paludosa Roxb. Rhizophora apiculata Blume Rhizophora harrisonii Leechman Rhizophora x Iamarckii Montrouz Rhizophora mangle L. Rhizophora mucronata Lam. Rhizophora racemosa G. Meyer Rhizophora x selala (Salvoza) Tomlinson
S S S S S S T T T T T T T T T T T T T T T T T T T T T T T T T T T T T SIT SIT P S T P T T T T T T T
1 1
2 2
1
2 1
1 1
2 2 3 2 3
1 1 1 1
4
5
2 2 2
6
4 3
1 1 1 1 1 1 1 1 1
2 2 2 2 2 2
6
2 2 2
6
4 1 1 1 1 1
5
2 2 2
6
1
3 1 1 1 1
4
2 2 2 2
5 6
5 3
1
1
2
1
2 2 2 2
3
4
5
3
4
5
4
5
6
336
P.G.E.F. AUGUSTINUS
Table 12-2 (continued) Species
Life-form
Mangrove region
Rhizophora stylosa Griff. Scyphiphora hydrophyllacea Gaertn. Sonneratia alba J. Smith Sonneratia apetala Buch.-Ham. Sonneratia caeseolaris (L.) Engl. Sonneratia gnffithii Kurz Sonneratia ovata Backer Xylocarpus australasicus Ridley Xylocarpus gangeticus Parkison Xylocarpus granatum Koenig Xylocarpus moluccensis (Lam.) Roem. Xylocarpus parvifolius Ridley
T
1 1 1 1 1 1 1 1 1 1 1 1
S
T T T T T T T T T T
2 2 2
6
2 2 2 2 2
6 6
S = shrub (i.e. less than 3 m); T = tree (i.e., greater than 3 m); P = palm
1989). Mangrove forests therefore are geographically concentrated along sheltered (parts of) coasts, estuaries and lagoons. These relatively quiet sites are required for reasons of reproduction, i.e. the settling of the propagules.
COMPOSITION AND ZONATION OF MANGROVES
The species richness of mangrove communities in estuaries is determined by a number of factors within geographical regions. Bunt et al. (1982) in comparing fifty-six coastal rivers, estuarine inlets and island bays in N.E. Australia, describe a positive relation between freshwater influence and floristic richness. According to Oliver (1982), the annual precipitation as well as the seasonal distribution of the rain in the tropical coastal area of Australia both appear to influence the distribution of the mangrove species. The species richness in estuaries with a high annual rainfall should be greater than in estuaries with a low annual rainfall. Moreover, the number of species appears to increase with the length of the estuaries and the area of the drainage basins (Bunt et al., 1982). Duke (1985, in Smith and Duke, 1987) found certain species (e.g. Sonneratia cuseoluris) to be obviously associated with larger estuaries and substantial freshwater runoff into the mangrove sites. Smith and Duke (1987) studied the tree species richness in 92 estuaries in tropical Australia. The species richness appeared to depend largely on maximum and minimum air temperature, on tidal amplitude, estuary length, catchment size, rainfall variation and the frequency of tropical cyclones. Duke (1992) proposes an improved classification of the distribution of mangroves in estuaries. He therefore uses two specific factors, namely the estuarine location and the intertidal position. Each of these two factors can be divided into three categories: downstream, intermediate and upstream estuarine, and low, mid and high intertidal. The position along the intertidal profile has often been referred to as zonation.
GEOMORPHOLOGY AND SEDIMENTOLOGY OF MANGROVES
337
A mangrove forest usually shows zonation. It means that, from the waterline going landward, more or less mono-specific zones succeed one after the other. There is no subject in the scientific literature concerning mangroves that has been treated so exhaustively (Snedaker, 1982). However, no consensus of opinion exists with respect to the question why mangroves usually appear in distinct zones. Snedaker (1982) in a thorough study of the literature on mangrove species zonation treats the subject from four scientific points of view: plant succession, geomorphology, physiological ecology and population dynamics. In his conclusion he states that “the geomorphology and physiological ecology studies appear to be the most relevant to the enhancement of our understanding of zonation and plant succession in the intertidal environment”. A new approach introduced by Thom (1984) is the physiographic ecology, aimed at “changes in habitats and plants occupying those habitats”. Thom (1982), discussing mangrove ecology from a geomorphological perspective, defined five terrestrial settings in which mangrove habitats play an active part: a river-dominated environment, a tide-dominated environment and a drowned bedrock valley. Estuarine mangrove habitats recur in all these terrestrial settings. In the carbonate settings, described a few years later (Thom, 1984; Woodroffe, 1987), the occurrence of estuaries is limited and so are the estuarine mangrove habitats. An overview of the environmental settings is given by Woodroffe (1992). In estuaries the mangrove zonation usually is parallel to the banks and determined to a large extent by the tidal amplitude (Baltzer, 1969). If the tidal amplitude is small the belt of mangroves will be narrow and often made up of only one species. The size of the mangrove trees adjacent to the estuary is usually larger than in more distant positions (Carter, 1959; Thom, 1967),especially if they drain large basins (Bunt et al., 1982, 1985). This more vigorous growth is attributed to the richer and better-drained soils in the overbank deposits (Carter, 1988). Moreover, the mangroves lining up the channels are the first to receive the nutrient-rich fresh water during floods (Baltzer, 1982; Bunt et al., 1982,1985). Thom (1967), dealing with mangrove ecology in a deltaic area in Tabasco, Mexico, describes the mangrove habitats in two estuarine environments: in the delta estuarine system and in coastal lagoons. In the delta estuarine system several habitats are distinguished,in which the distribution of mangrove species is related to the dynamics of the various “fluvial” landform types. Along the banks of the main outlet of the delta system discharging fresh water throughout the year, and on the related natural levees, which have fresh ground water, mangroves hardly occur. The other distributary channels, however, are lined up with considerable bands of mangroves, chiefly Rhizophora mangle (Fig. 12-2) and Lagunculana racemosa. This is especially the case if the channels have been partly filled in after a drop in discharge e.g. due to stream diversion. In addition, Rhizophora mangle and Laguncularia racemosa grow in the muddy swales of pointbars, while Avicennia nitida (Fig. 12-3) prefers the pointbar ridges, which are slightly coarser in texture. Avicennia nitida also covers the natural levees with their more compact, strongly oxidized soils. However, if the elevation of the natural levees is more than 0.5 m above the low water level, then the crest is lined by non-mangrove vegetation, flanked on both sides by Avicennia
338
P.G.E.F. AUGUSTINUS
Fig. 12-2. Rhizophora mangle along the Commewijne River, Suriname.
Fig. 12-3.Avicennia germinans along a tidal inlet in the coastal plain of Suriname.
GEOMORPHOLOGY AND SEDIMENTOLOGY OF MANGROVES
339
nitida. In the interdistributary basins a striking lack of zonation among the mangroves occurs. Landward the mixed mangrove forest grades into a brackish-water scrub. The changes in the mangrove distribution in the deltaic estuarine environment of the South Alligator River, Northern Territories are treated by Davie (1985). These appear to be due to physical changes in habitat e.g. erosion and progradation of river banks and infilling of the main river and tributary channels. The pattern of the mangrove distribution in the coastal lagoons in Tabasco, which are generally of the choked type (Kjerfve and Magill, 1989), is determined by the morphodynamic development of the lagoon shore (Thom, 1967). Three habitats are distinguished, characterized respectively by lagoon shore accretion producing mudflats, by shoreline stability and by erosion. On rapid accumulating mudflats, the outer fringe of the vegetation is made up of a grass (Spartina sp.), followed landward by an Avicennia vegetation, which shows a step-wise increase in height, from seedlings to trees of 10-12 m high. Rhizophora mangle and Lagunculana racemosa occur single or in small groups. If mudflats become inactive they are still dominated by a (mature) Avicennia community, except for the lagoon edge, where Rhizophora takes over. In Tabasco, Mexico, stable shorelines of coastal lagoons are generally characterized by a band of Rhizophora mangle, sometimes together with Laguncularia racemosa. When coastal erosion becomes dominant, the front zone consisting of Rhizophora is removed due to undercutting and mature Avicennia trees are exposed to wave action. In this environment reproduction is usually absent. As soon as a mangrove forest has settled, it traps sediment and organic materials to develop mangrove deposits. The aggradation of the forest floor causes a change in water depth as well as in timing and duration of inundation, resulting in a change in physiological conditions for mangrove growth. The species composition will change as a result (Sato, 1989; Walsby and Torckler, 1992).
MANGROVE SPECIES AND THEIR ENVIRONMENTAL CONSTRAINTS
Mangroves are adapted to a saline environment with waterlogged often muddy soils. Due to these adaptations mangroves are able to grow in places where no competition with other higher plants exists.
Salt toleration Mangroves do not require salt for their development (Chapman, 1975). According to West (1956, in Chapman, 1975) the only explanation for the adaptation to a saline environment is that the possible competitors are less tolerant to salt. This is in accordance with earlier observations of Egler (1948) on Rhizophora mangle growing well in a fresh water environment, if no competitors were present. In a fresh water environment the mangroves are successfully competed by other higher plants (Joshi and Shinde, 1978, in Joshi and Bhosale, 1982), which grow faster (Chapman, 1984). Mangroves can exist over a broad range of salinities, from 0 to 90%0. This upper salinity limit is found especially in tidal areas which are inundated
340
P.G.E.F. AUGUSTINUS
only a few times per year, under a warm and dry climate. Avicennia is the most salt tolerant and is dwarfed under hypersaline conditions.
Prop roots andpneurnatophores The root system of the mangroves is an other adaptation to the environment. Apart from wind stress, mangroves in estuaries have to cope with strong tidal currents as well as with fluvial currents during periods of heavy river discharge. Normal tab roots cannot develop, due to the anaerobic conditions of the loose and waterlogged soils. Mangrove trees therefore need special roots for fixation and for aeration (e.g. Davis, 1940; Macnae, 1968; Chapman, 1976; Carter, 1988). There are basically two root systems in which the requirements for stability and aeration are combined, with some intermediate forms and variations. A detailed description of seven mangrove root structures is given by Tomlinson (1986). In the root systems of the Avicennia type (Avicennia species, Sonnerutia species, Lagunculuna rucernosa), stability is obtained by a star-shaped network of cable roots which radiate out from the trunk at a depth of 20 to 50 cm. Anchor roots shoot downwards from these cable roots, while vertical aerial roots, pneumatophores, are pushed up (Fig. 12-3). The height of the pneumatophores of Languncularia rucernosa may reach values of 20 cm, whereas for the species ofAvicennia the pneumatophores may reach heights of 30 cm. Sonnerutiu species become the tallest (Tomlinson, 1986). The rooting of the Rhizophora type consists of a system of prop roots arching out from the trunk (Fig. 12-4) and anchoring at some 30 cm depth (Macnae, 1968). The prop roots contain lenticels for aeration. Cenops and Bruguiera species which
Fig. 12-4. Prop roots of Rhizophora mangle along the Commewijne River, Suriname.
GEOMORPHOLOGY AND SEDIMENTOLOGY OF MANGROVES
341
to a lesser extent are equipped with prop roots, have above ground knee roots with lenticels as well. Most mangrove species have nutritive roots allowing normal respiration in aerobic conditions.
Vivipuriq The production of viviparous seedlings is considered to be an adaptation of various mangroves to the muddy intertidal environment, where conventional seeds are not very likely to germinate (Walsby and Torckler, 1992). On the other hand, Macnae (1968) points out to the fact that viviparity is restricted to the Rhizophoruceue, the Avicenniu species and Aegicerus corniculutum, the other mangroves producing seeds which germinate in the soil. The only advantage of propagules is that they take root faster when they touch the ground, as compared to seeds, which have to germinate first. The distribution pattern of the propagules is related to the hydrology and geomorphology of the area (Chapman, 1984). Snedaker (1982) found mangroves which dominate the lower stands, close to open water (the Rhizophoru's together with Pellicieru rhizophorue) generally to have large and heavy propagules. In the higher areas further landward, the prevalent mangroves (Avicenniu,Lugunculun'u) usually produce small propagules, which sometimes need a five days rest without flooding before they are able to settle (Rabinowitz, 1978). Watson (1928) maintains that differential tidal sorting and distribution probably leads to zonation, conform the size of the propagules and the frequency of tidal inundation.
THE INFLUENCE OF MANGROVES ON HYDRODYNAMICS
In the estuarine environment the hydrodynamics are chiefly determined by currents, waves being more or less subordinate, except during periods of heavy storms or cyclones. In the case of cyclones, the direct influence of the wind appears to be of interest. In the clayey sediments under estuarine mangroves groundwater flow has been observed due to numerous crab holes.
Currents Mangrove forests lining up estuaries are regularly flooded during periods of higher water level. At ebb tide, the water flows back to the estuary. The water movement is impeded by bottom roughness, especially due to the digging activity of crabs, and by the prop roots and pneumatophores. The present dense and extensive above-ground root network (Fig. 12-5) can be considered as an increased bed roughness to the flowing water. This is expressed by Bunt and Wolanski (1980) and Wolanski et al. (1980) by taking the value of the Manning roughness coefficient in the mangrove swamps to be a function of the vegetation density. It will cause the current velocity to decelerate, as has been observed by Scoffin (1970) in the Bimini Lagoon, Bahamas. A current velocity of 0.4 m/s was reduced tot zero over a distance of one metre due to Rhizophoru prop roots at mutual distances of approximately 15 cm. Flume experiments using a scale model of a trunk with prop roots, placed in the centre
342
P.G.E.F. AUGUSTINUS
Fig. 12-5. The dense network of prop roots of Rhizophoru mangle along the Commewijne River, Suriname.
of the flume, gave similar results (Sato, 1984). Between the roots of the model the current velocity appeared to be a great deal lower than on either side of the model. In a similar way, pneumatophores (Fig. 12-6) act as obstacles to (tidal) currents, slowing down the flow, thus furthering the settlement of suspended silt that cloud the estuarine waters (Walsby and Torckler, 1992). The role of pneumatophores on the sedimentary processes was studied by Bird (1971) at Yaringa, Victoria, using a grid of rods to simulate the pneumatophores ofAvicennia. The result of his work was that pneumatophores indeed appeared to influence the current velocities and related pattern of sedimentation. Pneumatophores create quiet-water environments, which further the deposition of material that would otherwise have remained in suspension, or would have been carried away. There is, however, no technical information available about the artificial pneumatophores. Spenceley (1977) did a similar research at the westside of Magnetic Island, Queensland. For his experiments he used four grids consisting of rods, with spacings of 1 cm, 2.5 cm, 5 cm and 10 cm respectively. From the results, Spenceley (1977) concluded that the current reduction due to pneumatophores has two hydraulic functions. If the energy conditions are low, the current velocities are further reduced, causing deposition. Under medium- to high-energy conditions eddies develop behind the obstructions and initiate local scour. Observations of Zenkovich (1967) during his research at Hainan Island, China, probably fit the dynamics of these medium- to high-energy conditions. He found the waves to be attenuated in the outer margin of
GEOMORPHOLOGY AND SEDIMENTOLOGY OF MANGROVES
343
Fig. 12-6. Pneumatophores ofAvicennia gemtinuns along a tidal inlet in the coastal plain of Suriname.
the mangrove zone, while simultaneously the current velocity of the water inside the swamp was still strong enough to erode the muddy bottom. In the mangrove swamps lining up the Wenlock River in Queensland, strong tidal currents appear to create a fluid mud layer at the bottom (Wollanski and Ridd, 1986). The effect of mangroves on flowing water has been described in a hydrological model for a mangrove creek, which is subject to tidal influence only: Coral Creek in Hinchinbrook, Queensland (Bunt and Wolanski, 1980; Wolanski et al. 1980, 1992). The mangrove swamps receive water from the tidal creek during rising tide, keep it trapped for some time and release it during falling tide. The amount of water that can be stored in the mangrove swamp increases with an increasing ratio of swamp area over creek area. The higher this ratio, the more the mangrove swamp increases the tidal prism of the estuary (Wolanski et al., 1992). There is a marked difference in water flow through the mangrove forest as compared to the open creek. The current velocity in the open channel often exceeds 1 mls, while in the adjacent dense mangrove swamp at 50 m from the creek, it never exceeds 0.07 m/s (Wolanski, 1992). A numerical model has been used to describe the water flow in this complex mangrove swamp/creek system, linking a one-dimensional model of open channel flow with a two-dimensional model of flow through a vegetated floodplain (Bunt and Wolanski, 1980). It appears from the model that Coral Creek has an asymmetrical tidal circulation with stronger peak currents during the slightly shorter ebb tide, often 20-50% higher than the peak flood currents. This is due to the time lag between the moment of high-water at the mouth and the head of the creek, in combination with the dense and complex network of mangrove trunks and prop roots. These act as a kind of a barrier to water flow in the early stage of the ebb tide. The resulting steeper water surface gradient between the mangrove forest and the channel causes
344
P.G.E.F. AUGUSTINUS
the current velocity to increase in the further course of the ebb tide. Especially during the stronger ebb tidal currents, a part of the sediment in the channel will be transported as bedload. It will be obvious from the above that dense mangrove vegetations along estuaries and rivers produce a silting up of the banks as well as erosion in the channels. Bunt and Wolanski (1980) deduced this effect of mangroves on flowing water from the depth of creek systems, which appeared to be shallower in the mouths than in the landward part through the tidal forest. Under the latter conditions the channel is scoured, due to the greater ebbcurrents, while outside the mangrove fringe deposition occurs (Wolanski et al., 1980). The stronger the ebb tidal currents, the more sediment can be scoured from the channel bed, resulting in deeper creeks. A reduction in the size of the mangrove swamp, e.g. by the construction of fish ponds or land reclamation, will reduce the tidal asymmetry and therefore the peak ebb tidal currents, causing a silting up of the creek (Wolanski, 1992; Wolanski et al., 1992). In a study of the trapping function of estuarine mangrove swamps, Wolanski and Ridd (1986) describe lateral trapping in mangrove forests as a dominant process, controlling lateral mixing in the mangrove fringed tidal rivers. In the dry season the effective value of the longitudinal difisivity is increased by two orders of magnitude in comparison with a situation where no swamps occur. In the wet season during high tide, lateral trapping of fresh water in the mangrove forest is caused by an increased buoyancy effect. Fringing mangrove forests in this way control the flushing of fresh water, especially at the end of a flood. The enhanced growth of the mangrove in the Rufiji delta (Tanzania) is at least partly explained by the effect of trapping, which causes the river-borne nutrients to remain in the swamp for a prolonged period of time (Francis, 1992). Exceptional floodings are sometimes caused by cyclones. An example is described by Steinke and Ward (1989) based on two cyclones which swept the coast of South Africa. Heavy rainfall caused the waterlevel in the St. Lucia Estuary to increase by several metres. Consecutive floodings, accompanied by high current velocities, left a mark of destruction in the mangrove, consisting chiefly of Bruguieru and Avicenniu. However, the stumps and roots of Avicennia remained in place, thus protecting the soil against erosion. Mortality of the mangroves occurred even many months after the cyclones swept the area, resulting from longlasting inundation or silting up of the pneumatophores, causing asphyxiation. Cyclones have a direct influence on estuarine mangroves as well. Smith (1986) describes the adverse impact of cyclone Kathy (March 23, 1984), with winds of 185 km/h, on the mangrove vegetation along the MacArthur River (Northern Territory, Australia). It appeared that mortality among the Rhzzophoraceae in the upstream part of the river was significantly greater than at the river mouth. The mangroves in the lower part of the river probably have been protected from wind action due to inundation by a storm surge of 3 to 4 ms. The damage brought about by cyclone Kathy appeared to be much more severe as compared to the damage caused by cyclone Winifred (February 1, 1986), blowing with 170 km/h in the coastal area of Queensland. According to Smith (1986) the
GEOMORPHOLOGY AND SEDIMENTOLOGY OF MANGROVES
345
damage to the vegetation usually increases dramatically if a threshold for wind velocity is surpassed. For mangroves this threshold will be somewhere between 170 and 185 km/h. Anyway, the mortality of Avicennia manna, Excoecaria agallocha and Lumnitzera racemosa appeared to be much lower than that of the Rhizophora species, because they are capable of stump sprouting (Smith, 1986).
Waves Mangroves thrive in more or less sheltered environments, such as estuaries. Thorn et al. (1975), for the Cambridge Gulf-Ord River system (Western Australia), found wave activity to be predominantly effective in the funnel-shaped outer estuary. On more exposed coasts the stronger wave action would either uproot the mangroves or prevent the silt particles from settling down, and thus from forming a good soil for mangroves to root in (Walsby and Torckler, 1992). However, with respect to cyclone effects, Woodroffe et al. (1986) found waves of 2.5 m likely to break well within the mangrove fringe of the South Alligator Mouth, Northern Territory (Australia), at times of high tide. In estuaries in New Caledonia and Cameroon, Baltzer (1975) found waves to attenuate very effectivein thick curtains of size-gradedRhizophora as well as in the pneumatophores of Avicennia, thus limiting water turbulence at high tide. Woodroffe et al. (1986) point out that wave breaking within the mangrove front may have some effect on mangrove recruitment. The sensitivity of rooting propagules for wave action is also emphasized by Augustinus (1978) and Sat0 (1985). In the funnel-shaped mouth of the South Alligator Tidal River bank erosion takes place by wave action. This is evidenced by cliffs on banks facing the northeast (Woodroffe et al., 1986). River bank erosion due to wave action is increasingly caused by the effect of waves generated by motorised (fishing) boats. In the Sungai Merbok estuary (Malaysia), this process has resulted in a local widening of the river by approximately 20 m in five years (Chan Hung Tuck, 1985). Bruguiera parviflora appeared to be more susceptible to this river bank erosion than species of Rhizophora, which might be due to their relatively shallow knee-root system. Groundwaterflow In the muddy sediments under estuarine mangroves minimal groundwater movements are expected. However, Wolanski et al. (1992) have observed that with rising tide in the Coral Creek swamp (Queensland), groundwater comes up through the numerous crab holes and covers the swamp surface at the initial stage of flooding. This feature may lessen the effect of bed roughness on the incoming flood water. Groundwater flow has been demonstrated by Mazda et al. (1990) in the BashitaMinato swamp in Japan. The 200 m long tidal creek occasionally becomes ponded by the formation of a bar after a storm. Due to differences in hydrostatic pressure between the ocean and the creek a groundwater flow is generated, which appears to have a high impact on the water quality of the mangrove swamp. The groundwater flow has been found in the upper 90 cm of the substrate, which is about the depth of the crab burrows. In areas where there are no crab holes or cracks, groundwater flow appears to be negligible.
346
P.G.E.F. AUGUSTINUS
SEDIMENTATION AND SEDIMENT IN ESTUARINE MANGROVE FORESTS
Land-building It is widely accepted that the concentration of trees and shrubs and their aboveground root systems stimulate a rapid accumulation of sediment. However, whether mangroves promote sedimentation in such a way that deposits are formed, which should not have been developed without the presence of mangroves, is not easy to establish. The former support for an active land accretion by mangroves (e.g. Davis, 1940; Richards, 1952) has ceased to be a major issue. Earlier scientists like Watson (1928) and van Steenis (1941) who favoured the hypothesis that aggradation of the tidal mudflats has to take place first prior to the settlement of mangroves, has gained a general support in the last decennia (Scholl, 1968; Thom, 1967; Bird, 1972, 1986; Thom et al., 1975; Lugo, 1980). A lowering of the sea level should give a similar result (Craighead, 1971). An overview of the early literature concerning the land-building capacity and the stabilization of the coast by mangroves is given by Carlton (1974). A sensitive factor for the settlement of mangroves on a tidal mudflat is the possibility propagules have to strike roots. Even if grown mangroves could maintain themselves on a low mudflat, their seedlings probably could not survive flooding and wave attack (Sato, 1985). An example of active land accretion due to mangroves is given by Bird (1971). Onshore winds induce waves, which transport sediment in a coastward direction. This sediment is deposited in the mangroves and on the mudflats in front of the mangroves. If this situation is followed by a period of offshore winds, a lee develops in front of the mangrove forest. For this reason, above that part of the mudflats, hardly any waves or currents do occur, and the sediment is not removed. Another, more general example is given by Baltzer (1975) based on research in New Caledonia and Cameroon. Due to wave attenuation in the mangroves lining up the estuaries, the reflection of the waves is subdued as well, thus favouring sedimentation on the bare banks in front of them, or at least limiting subsequent erosion. Sedimentation Sedimentation in mangrove swamps results from the reduction in current velocity and attenuation of wave action in the dense vegetation. Moreover, the network of trunks, prop roots and pneumatophores will exercise a filter function, which will be furthered by the presence of tree-dwelling algae, barnacles, mangrove-oysters, grazing snails etc. (Augustinus, 1978; Walsby and Torckler, 1992). A description of mangrove-dwelling animals is given by Macnae (1968), observations concerning algal mats covering the prop roots and pneumatophores of mangroves in Tampa Bay, Florida are described by Dawes (1967). The accelerated sedimentation as compared to areas where there are no mangroves may be inferred from the steeper slopes of the intertidal zones under mangrove (Chappell and Grindrod, 1984; Bird, 1986). Up till now, data concerning the rate of mud accretion beneath mangrove are scarce. Measurements are usually done in experiments using rods or stakes, often in grids to simulate the pneumotophores. The results show much variation, which may
GEOMORPHOLOGY AND SEDIMENTOLOGY OF MANGROVES
347
be ascribed to the method used, especially to the fact that the rigid stakes will alter the process of sedimentation and erosion. Bird (1971) reports a rate of mud accretion of 8 mm/year in a dwarf Avicennia forest in southern Australia. Sedimentation rates varying between -11 and +4.6 mm/year have been measured by Spenceley (1977, 1982) in an Avicennia swamp in north-eastern Australia. Sedimentation rates under mangroves in Rookery Bay, Florida, USA and in Terminos Lagoon, Mexico have recently been determined, using the radionuclides 210Pband 137Cs(Lynch et al., 1989). The average consolidation-corrected accretion rates for Rookery Bay (1.6 mm/year) and for the Terminos Lagoon (2.4 mm/year) are probably reliable since sedimentation and erosion processes have not been affected by measurements. Based on observations in Cairns Bay, Queensland, Bird (1972) has suggested that mangroves with pneumatophores are better sediment traps than mangroves with prop roots. This is confirmed by measurements of sedimentation rates he did in mangrove swamps south of Ellie Point, seaward of the Trinity Inlet. He found over five years averaged a sedimentation rate in the Avicennia fringe of 7.6-9.1 mm/year and in the Rhizophora zone of 2.0-4.6 mm/year. The deposition in estuarine mangrove swamps due to river discharge at the south coast of Papua-New Guinea is described by Irion and Petr (1979). The distributaries and tidal channels of the Purari river and delta system carry sandy bedload and silt in suspension towards the sea. During the diurnal high water the suspension-rich top layer of the river water column is pushed deeply into the mangrove forest, where most of the silt is settling down. The texture of these silt deposits in the mangrove does contrast sharply with the sandy bedload deposits in the river, which are some 20 times coarser. A similar observation was done by Sato (1989). Natural levees, occurring in the estuarine system, are found to contain also some sand ( e g Diemont and van Wijngaarden, 1975; Woodroffe et al., 1985a, 1986). An accelerated silt transport due to soil erosion in the upstream part of the basin will result in higher accretion rates in the estuarine mangrove swamps. In Segara Anakan (Indonesia) where the silt supply from a number of local rivers is increased due to soil erosion and volcanic eruptions, there is a rapid extension of the tidal flats, followed by the colonization of mangroves, especially Avicennia marina and Sonneratia aZba (Erftemeijer et al., 1988). Siltation rates in the order of 100 mmlyear have been reported (Kvalvagnaes, 1980). Bird (1982, in Erftemeijer et al., 1988) predicts a complete replacement of the lagoon by mangrove swamps, dissected by tidal channels, provided accumulation rates are maintained. The clastic sediments beneath mangroves in estuaries are, however, not always river-borne. In estuaries along the north coast of South America, for instance, the contribution of river sediment to the total amount of estuarine mangrove mud is very small. This is due to the high influx of Amazonborne suspension load from the related coastal waters. Some 20% of the yearly suspended sediment supply of the Amazon River (11-13 x lo8 ton) is transported along the north coast of South America up to the Orinoco River (Eisma et al., 1991). Especially in periods of lower river discharge, when the ratio between river discharge and flood-tidal influence is in favour of the latter, the suspension-rich ocean water will penetrate along the
348
P.G.E.F. AUGUSTINUS
bottom of the estuaries and increasingly mix with the fresh river water. Under these circumstances, the ocean-borne silt will be dispersed into the estuarine mangrove swamps every time the river overflows its banks. Long term correlations of the vertical accretion rate beneath estuarine mangroves and the influx of clastic sediments are difficult to make, among others due to indistinctness with regard to the source of the sediments. Woodroffe et al. (1986) made an attempt for the South Alligator River over the last 6000 years, i.e. the period of local relatively stable sea level which followed the rapid eustatic sea level rise after the last ice age. The yearly sediment input of this river-dominated system was computed using the actual flood water silt concentration as the basic assumption. This estimate appeared to be significantly lower than the annual accumulation rate in the flood plain, averaged over the last 6000 years. The discrepancy might be due to different causes: to errors in the discharge-based estimates; to a long-term decline of mud supply according to changing geomorphological conditions of the tidal river; or by an influx of tidal sediment from an oceanic source.
Sediments In mangrove swamps, vertical accretion is a mixture of clastic sediments and organic materials produced by the mangroves and their associated fauna (Bird, 1972). The accumulation of the sediments occurs at the slack of high-tide and during storm surges as well as after floods, bringing in material from the sea or from rivers during flood stages (Carlton, 1974). Floating organic debris are trapped in the prop roots and pneumatophores of the mangroves and together with the products of in situ organic decomposition, added to the silty sediments forming the so called mangrove muds or mucks (Macnae, 1968; Thom, 1967). These are usually bluish-gray to olive-gray soft muds, generally sulphurous and rich in organic material. The proportion between the organic compounds and the inwashed sediment increases towards the landward limit of the mangrove (Bird, 1968). Due to the scanty and short tidal floodings, mud accumulation is a slow process in these relatively remote areas, where organic litter locally may form a superficial peat deposit. The development of peat at the landward side of estuarine mangrove swamps has been described also by Diemont and van Wijngaarden (1975) for West-Malaysia. Mangrove mud is undoubtedly a mangrove swamp deposit. Baltzer (1982) considers this deposit as a sedimentary unit, usually with a sandy layer at the bottom and than covered by mangrove mud. In a lateral direction, starting at the estuary banks, the mangrove mud grades into peat. Woodroffe et al., (1986), in their study of the South Alligator tidal river designate mangrove mud the most distinctive and ubiquitous stratigraphic unit within the underlying sediments. A specific feature in these mangrove muds is the occurrence of mud ball concretions containing faunal debris which at least partly are calcified.
GEOMORPHOLOGY AND SEDIMENTOLOGY OF MANGROVES
349
THE INFLUENCE OF MANGROVES ON SOIL STABILITY
If mangroves grow on a clayey soil, what they usually do, they further resistance against erosion in two ways: by facilitating consolidation and by increasing soil rigidity through their (underground) roots.
Consolidation Due to the extraction of water and the subsequent compaction of the clay, soils consolidate (Carlton, 1974; Baltzer, 1975; Augustinus, 1978; Wells and Coleman, 198l), by this increasing their resistance against erosion by currents and wave attack. The ability to further consolidation appears to be different for the various mangroves. Under Avicennia and Sonneratia consolidation appears to be better than under Rhizophora because the roots of Avicennia and Sonneratia grow faster and denser than those of the Rhizophoras (Teas, 1980). Soil binding capacity of mangrove roots Another, more direct effect of mangroves is the rigidity they give to the soil due to their extensive network of underground roots. According to Spenceley (1977), this sediment-binding capacity of the roots prevents the soil from extensive erosion, even under high-energy conditions. During cyclone Althea, for instance, many beaches in the vicinity of Townsville, Queensland, retreated up to 14 m, while neighbouring mangrove protected shorelines hardly appeared to be affected. Walsby and Torckler (1992) have established that the diameter of the root system of mangroves may be up to five times the diameter of their canopy. As a result, in dense mangrove stands, the root systems of the separate trees are strongly interwoven, which will favour the stability of the soil. According to Garofalo (1980) the binding capacity of plant roots in combination with the strength of the soil material determines the resistance of (marsh) cliffs to erosion. Measurements in a salt marsh in The Netherlands indeed have revealed that plant roots definitely enlarge the resistance of a soil against erosion (van Eerdt, 1985). The total root strength appears to be dependent on the type, the diameter and density of the roots. In the estuarine environment the binding capacity of the roots will play an important role in case of undercutting by meandering rivers or migrating tidal channels. However, in the upstream reach of the South Alligator Tidal River, Northern Territory, Australia, where slumping is the outstanding form of bank erosion, Sonneratia lanceolata which is the prominent mangrove in this part of the estuary appears to be incapable in resisting this process (Woodroffe et al., 1986). The fact that they often are found having slumped themselves into deeper parts of the channel, may indicate that the base of these slumps probably is below the root system.
MANGROVES AND GEOMORPHOLOGY
Whenever the muddy banks in the lower course of estuaries have been silted up to about mean high-water neap tide, mangroves begin to grow. Further upstream,
350
P.G.E.F. AUGUSTINUS
the natural levees of the tidal rivers are silted up above mean high-water spring, and contain some sand (Diemont and van Wijngaarden, 1975). In Tabasco, Mexico, mangroves usually do not cover the top of the levees if they are more than 0.5 m above low water level (Thom, 1967). The rapid decrease of the current velocity in the dense baffle of above-ground roots and trunks, as described e.g. by Scoffin (1970), results in a deposition of fine grained material landward of the frontline of the mangroves. The resulting embankment, named rampart by Baltzer (1975), not only consists of mud, but also contains vegetal debris. Baltzer (1975) dealing with accretionary processes in mangrove swamps along estuaries in Cameroon and New Caledonia, found sedimentation to be most effective on the front part of a rampart. This embankment therefore proceeds over the estuary bank. The higher sedimentation rates generally lead to steeper slopes in the intertidal zone under mangroves compared to uncovered mudflats (Chappell and Grindrod, 1984; Bird, 1986). If the sediment trapping effect of the different above-ground root systems are mutually compared, pneumatophores appeared to be the most effective (Bird, 1972). Spenceley (1977) and Bird (1985) found an existing relationship between the presence of pneumatophores and the surface elevation, the latter being slightly higher. The surface of the substrate in the mangrove swamps along the Wenlock River Queensland, has been described by Wolanski and Ridd (1986). They find the width of the swamps to be a distinguished criterion for the morphology of the surface. Whenever a strip of mangrove is roughly less than 200 m wide, a generally smooth surface, free of channels and depressions, gently slopes towards the open water of the tidal river. In more extensive mangrove swamps the surface is dissected by narrow and shallow creeks, which are located typically 1000 m apart, and do not drain outside of the mangrove swamp. The influence of a mangrove on the geomorphological development along the shores of coastal lagoons is comparable with the situation at an open coast. The most important difference is the shelter for wave activity. In a number of cases this results in rapid prograding mudflats. In those cases, Avicenniu always is the pioneering mangrove (e.g., Steup, 1941; Thom, 1967; Carter, 1959; Saenger and Hopkins, 1975; Diemont and van Wijngaarden, 1975; Augustinus, 1978; Erftemeijer et al., 1988; Cooks and Bewster, 1993). This is attributed by Augustinus (1978) to fluid mud formation, which often takes place with rapid silt accumulation, preventing the elongate Rhizophoru seedlings from rooting. The relation between mangrove ecology and the geomorphology of estuaries has been studied extensively by Thom et al. (1975) in the estuary of the Cambridge Gulf-Ord River in Western Australia. This macro-tidal estuary (tidal range 8 m) is characterized by a high rate of seasonal variability in water and sediment discharge, especially in the summer. For the Cambridge Gulf, Thom et al. (1975) distinguish an outer funnel-shaped estuary and an inner estuary. In both types, six landscape zones are recognized: Piedmont zone, supra-tidal zone, high-tidal zone, mid-tidal zone, low-tidal zone and subtidal zone. The mid-tidal zone, ranging in elevation from + 3 to +8 m, and mainly covered by mangroves, is divided into a lower unit which is flooded
GEOMORPHOLOGY AND SEDIMENTOLOGY OF MANGROVES
351
every tide, and an upper unit which is flooded by high tide-spring only. This lower and upper part is separated by a scarp (0-3 m high). The lower unit, lining up the open water of the estuary, is covered by Avicennia, especially seedlings and saplings up to 3 m in height. The pneumatophores are well developed. The banks of the tidal river appear to be highly unstable. The mangroves are growing in soft, hardly consolidated mud. Therefore, slumping of the river banks is a common feature. It causes a disturbance of the root system often leading to the death of the vegetation. The upper unit, around +7 m elevation, is covered by a more diversified vegetation, dominated by mature Avicennia, growing on a rather smooth surface. Gullies occur which appear to be laterally and vertically unstable. The gully levees or the tidal flat between closely spaced gullies are slightly higher (50 cm) than the adjacent surfaces of bare vegetation. Towards the landward fringe, the mangroves become more and more confined to the banks of shallow tidal gullies, which appear to be vertically and laterally unstable. In the river, pointbars develop due to river bend migration. These pointbars cover the most extensive mangrove vegetation. In the outer estuary, wave action is more pronounced. A sandy shelf grades into a low gradient mudflat. Usually there is no scarp in the mid-tidal zone. The mangrove front is formed by Rhizophora, Avicennia and Ceriops growing at its landwards side. In the adjacent Ord River the mangrove habitats are in principle comparable with the inner estuary. There are however a number of distinct morphodynamic aspects which cause the vegetation pattern to change. On the one hand, high rates of side and mid-channel deposition occur, on which mangroves extend. On the other hand, the river banks and parts of the mid-channel islands are regularly undercut due to current activity. The vegetation on the exposed locations is attacked as a result. Up to the limit of the tidal influence, fluvial erosion and sedimentation create a very unstable habitat for mangroves, which cause relatively rapid changes in vegetation zonation. Semeniuk (1980), working on an eroding coastline in King Sound (North-Western Australia) found mangrove zonation to depend on the type of erosion: sheet, cliff or tidal-creek erosion. When the advancement of erosion is vaster than mangrove growth, the zonation is truncated. Baltzer (1985) gives a description of the development of size-graded stands of Rhizophora at the bank of estuaries during a period of general accretion. While the seedlings settle and begin to grow at near sea level, the older trees develop a system of prop roots, which grow deeper than the roots of younger trees at the outer fringe. If the accretion ceases or erosion increases, the younger trees will disappear. The larger trees, however, may survive and protect the landward part of the mangrove forest. This defensive system is only effective in estuaries. The study of Thom et al. (1975) clearly demonstrates that mangrove ecology is to a large extent controlled by morphodynamic processes, as was stated earlier by Thom (1967) for the deltaic coastal plain in Tabasco, Mexico. The accordance between the development of mangrove forests in estuaries and the related geomorphological evolution, as stated by Chappell and Woodroffe (1985), shows interesting perspectives for studies aimed at palaeo-geographical reconstructions, due to the information on
352
P.G.E.F. AUGUSTINUS
the vegetation history stored in the mangrove mud deposits. This is illustrated in the studies of Woodroffe et al. (1985b, 1986) on the evolution of the South Alligator Tidal River and Plains. The geomorphological and the ecological processes take place at different spacial and temporal scales. Woodroffe (1992) compares the time scales at which geomorphological and ecological processes operate. For the different scale levels it appears that “the time scales at which geomorphological processes operate overlap with those at which ecological processes function”. Geomorphological processes, however, usually control the related ecological processes, especially at the larger scales. In the long term (lo2 to lo4 years), for instance, climatic and sea level changes control the evolution of a mangrove ecosystem. Mangroves mainly occur in the tidal range between mean sea level and mean high water spring tide (Ellison and Stoddart, 1991). A mangrove swamp can only remain in this subaerial position if the vertical accretion rate matches the sea level rise. A continued existence of an (estuarine) mangrove therefore indicates some sort of equilibrium between the accretion rate under mangroves and the apparent sea level rise (Lynch et al., 1989). This relationship between mangroves and mean tide level is basic to the fact that mangroves provide potentially useful depositional facies for a reconstruction of the (Holocene) sea level history (Grindrod and Rhodes, 1984). Sea level changes are long term processes which do not lend themselves for direct observation. For this reason, the evolution of mangrove-fringed estuaries in northern and eastern Australia over the last 6000 years (the period in which sea level stabilized), has been based on stratigraphic studies (Woodroffe et al., 1985b, 1986; Woodroffe, 1992). At least three models of development have been proposed: the progradational model, the big swamp model and the barrier estuary/mud basin model. The mangrove mud deposits appear to have a key function for obtaining data concerning the development and the distribution of the mangrove vegetation, which is indicative of the geomorphological evolution of the estuaries. The question of how estuarine mangroves will react to an accelerated rising of sea level is difficult to answer. Ellison and Stoddart (1991), using stratigraphic data, found mangroves at coral islands in the Caribbean capable of keeping up with a sea level rise of 0.8 to 0.9 mm/year. They could, however, not endure a sea level rise of 1.2 mm/year. These low values are due to the limited sources of allochtonous sediment on coral islands, causing minor rates of sediment accumulation as compared to deltaic-estuarine environments. Estuarine mangrove swamps generally have higher rates of sediment accretion, which increases their chances of survival during an accelerated sea level rise.
CONCLUSION
Mangroves have attracted a great deal of research, all over the world. However, a relatively small part of this research has been focussed on the influence of mangroves on (estuarine) hydrodynamics and soil mechanics. Information about the trapping
GEOMORPHOLOGY AND SEDIMENTOLOGY OF MANGROVES
353
capacity due to the dense network of above-ground roots and the soil binding capacity of the underground roots is scanty and mainly of qualitative nature. It appears that mangroves favour deposition and stimulate soil stability. Nevertheless, in estuaries, this function is dominated by the role of the geomorphological processes. The resulting morphological developments cause changing habitat conditions and hence, changing mangrove patterns. It therefore appears that the function of the mangrove vegetation is subordinate to the geomorphological development. However, although the mangroves follow areas of mud accretion, their establishment leads to a more rapid accumulation, as compared to areas without vegetation. This relatively rapid accretion is important for two reasons: Firstly, the silt accumulation in estuarine mangrove swamps results in specific geomorphological features, which would not exist without vegetation. Dense estuaryfringing mangrove forests for instance, appear to produce a silting up of the banks, together with erosion in the channels. The studies on Coral Creek (Queensland) show that the difference in water flow between the mangrove swamp and the channel causes an asymmetric tidal circulation. The stronger ebb-currents scour the channel bed, thus keeping the channels deep. A disturbance of this dynamic equilibrium due to a decrease in tidal volume (e.g. by the practice of aquaculture or land reclamation) or an increase in sediment supply (e.g. due to soil erosion) will result in a silting up of the channels. In the case of Segara Anakan (Indonesia) the increase in sediment supply leads to an extension of the tidal flats, which become overgrown with mangroves. Secondly, the rapid accumulation of sediments serves to preserve some of the record of past habitat changes (e.g. Woodroffe, 1992). The (estuarine) mangrove muds therefore provide a sensitive record of environmental changes, in sediment supply, in sea level, in storminess, etc. High resolution dating in time and space is possible using plant remains (peat) and shells, which are easily accessible due to shallowness of the sedimentary sequence.
ACKNOWLEDGEMENTS
I wish to thank Aart Kroon for his critical review of the manuscript. The assistance of Brigit Stelder and Ruben Groen is greatly appreciated. I am indebted to Gerard van Bethlehem for the photographic work and to Ria van der Linden who typed and processed the manuscript. Special thanks are due to Leonie van der Maesen for editing the text and her part in the general organisation.
REFERENCES Adegbehin, J.O. and Nwaigbo, L.C., 1990. Mangrove resources in Nigeria: use and management perspectives. Nat. Resour., 26: 13-20. Augustinus, P.G.E.F., 1978. The changing shoreline of Suriname. Ph.D. Thesis, Utrecht University, The Netherlands, 221 pp.
354
P.G.E.F. AUGUSTINUS
Baltzer, F., 1969. Les formations vtgttales associets au delta de la Dumbea (Nouvelle Caledonie) et leurs indications Ccologiques, gkomorphologiques et skdimentologiques, mises en tvidence par cartographie. Cah. ORSTOM, Str. Gkol., I: 59-84. Baltzer, F., 1975. Mapping of mangroves in the study of sedimentological processes and the concept of swamp structure. Proc. Int. Symp. on Biology and Management of Mangroves, Honolulu (1974), pp. 499-512. Baltzer, F., 1982. La transition eau douce-eau salt dans les mangroves. Constquences skdimentologiques et gtochimiques. Mtm. SOC.GCol. Fr., 144: 27-42. Baltzer, F. and Lafond L.-R., 1971. Marais maritimes tropicaux. Rev. GCogr. Phys. Gtol. Dyn., Vol. XIII, Fasc., 2: 172-196. Barth, H., 1982. The biogeography of mangroves. In: D.N. Sen and K.S. Rajpurohit (Editors), Tasks for vegetation science 2: Contribution to the ecology of Halophytes. Junk, The Hague, pp. 35-60. Bird, E.C.F., 1971. Mangroves as landbuilders. Victoria Naturalist, 88: 189-197. Bird, E.C.F., 1972. Mangroves and coastal morphology in Cairns Bay, North Queensland. J. Tropical Geogr., 35: 11-16. Bird, E.C.F., 1986. Mangroves and intertidal morphology in Westernport Bay, Victoria, Australia. Mar. Geol., 69: 251-271. Blasco, F., 1984. Climatic factors and the biology of mangrove plants. In: S.G. Snedaker and J.G. Snedaker (Editors), The Mangrove Ecosystem: Research Methods. Monogr. Oceanogr. Methodology 8, UNESCO, Paris, pp. 18-35. Bunt, J.S., and Wolanski, E, 1980. Hydraulics and sediment transport in a creek-mangrove swamp system. 7th Australasian Hydraulics and Fluid Mechanics Conf., Brisbane, 1980, pp. 492-495. Bunt, J.S., Williams, W.T. and Duke, N.C., 1982. Mangroves in north-east Australia. J. Biogeogr., 9: 111-120. Bunt, J.S., Williams, W.T. and Bunt, E.D., 1985. Mangrove species distribution in relation to tide at the seafront and up rivers. Aust. J. Mar. Freshwater Res. 36: 481-492. Carlton, J.M., 1974. Land-building and stabilization by mangroves. Environ. Conserv., 1: 285-294. Carter, J., 1959. Mangrove succession and coastal change in south-west Malaya. Trans. Inst. Brit. Geogr., 26, pp. 79-88. Carter, R.W.G., 1988. Coastal Environments. University Press, Cambridge, 617 pp. Chan Hung Tuck, 1985. Coastal and riverbank erosion of mangroves in peninsular Ma laysi. In: K.N. Bardsley, J.D.S. Davie and C.D. Woodroffe (Editors), Coasts and tidal wetlands of the Australian monsoon region. Australian National University, North Australia Research Unit, Mangrove Monograph No. 1, Darwin, pp. 43-52. Chapman, V.J., 1970. Mangrove phytosociology. Tropical Ecol., 11: 1-19. Chapman, V.J., 1975. Mangrove biogeography. In: G.E. Walsh, S.C. Snedaker and H.J. Teas (Editors), Proc. Int. Symp. on Biology and Management of Mangroves, Vols. I and 11. Int. Food Agric. Sci., Univ. of Florida, Gainsville, pp. 3-22. Chapman, V.J., 1976. Coastal Vegetation. Pergamon Press Ltd., Oxford, 292 pp. Chapman, V.J., 1984. Botanical surveys in mangrove communities. In: S.C. Snedaker and J.G. Snedaker (Editors), The Mangrove Ecosystem: Research Methods. UNESCO, Bungay, United Kingdom, pp. 53-80. Chappell, J. and Grindrod, J., 1984. Chenier plain formation in northern Australia. In: B.G. Thom (Editor), Coastal Geomorphology in Australia. Academic Press, Sydney, pp. 197-232. Chappell, J.M.A. and Woodroffe, C.D., 1985. Morphodynamics of Northern Territorytidal rivers and floodplains. In: K.N. Bardsley, J.D.S. Davie and C.D. Woodroffe (Editors), Coasts and tidal wetlands of the Australian monsoon region. Australian National University, North Australia Research Unit, Mangrove Monograph No. 1, Darwin, pp. 85-96. Cooks, J. and Bewster, P, 1993. Environmental reaction to human intervention in the Mhlatuze River estuary, Richards Bay, South Africa. Geookodynamik, Bensheim, Band X I V 61-74. Craighead, F.C., 1971. The Trees of South Florida. University of Miami Press, Coral Gables, Florida, 212 pp. Davie, J.D.S., 1985. The mangrove vegetation of the South Alligator River, NorthernAustralia. In: K.N. Bardsley, J.D.S. Davie and C.D. Woodroffe (Editors), Coasts and tidal wetlands of
GEOMORPHOLOGY AND SEDIMENTOLOGY OF MANGROVES
355
the Australian monsoon region. Australian National University, North Australia Research Unit, Mangrove Monograph No. 1, Darwin, pp. 133-151. Davis, J.H., 1940. The ecology and geologic role of mangroves in Florida. Carnegie Inst. Washington Publ. No 517. Papers of the Tortugas Laboratory No 32, pp. 303-412. Dawes, C.J., 1967. Marine Algae in the viscinity of Tampa Bay, Florida. Contrib. Dept. Bot. Bact., Univ. of South Florida, Tampa, No. 27, 105 pp. Diemont, W.H. and van Wijngaarden, W., 1975. Sedimentation patterns, soils, mangrove vegetation and landuse in the tidal areas of West-Malaysia. Proc. Int. Symp. on Biology and Management of Mangroves, Honolulu (1974), pp. 513-528. Duke, N.C., 1992. Mangrove floristics and biogeography. In: A.I. Robertson and D.M. Alongi (Editors), Tropical mangrove ecosystems. American Geophysical Union, Washington DC, pp. 63-100. Egler, F.E., 1948. The dispersal and establishment of red mangroves, Rhizophora, in Florida. Carib. Forest., 9: 299-319. Eisma, D., Augustinus P.G.E.F. and Alexander, C., 1991. Recent and subrecent changes in the dispersal of Amazon mud. Neth. J. Sea Res., 28: 181-192. Ellison, J.C. and Stoddart, D.R., 1991. Mangrove ecosystem collapse during predictedsea-level rise: Holocene analogues and implications. J. Coastal Res., 7: 151-165. Erftemeijer, P., van Balen, B. and Djuharsa, E., 1988. The importance of Segara Anakan for nature conservation, with special reference to its avifauna. PHPA-AWB/Intenvader Rep. No. 6, 59 pp. Francis, J., 1992. Physical processes in the Rufiji delta and their possible implicationson the mangrove ecosystem. In: V. Jaccarini and E. Martens (Editors), The ecology of mangrove and related related ecosystems. Hydrobiologia, 247: 173-179. Garofalo, D., 1980. The influence of wetland vegetation on tidal stream channel migration and morphology. Estuaries, 3: 258-270. Grindrod, J. and Rhodes, E.G., 1984. Holocene sea-level history of a tropical estuary, Missionary Bay, North Queensland. In: B.G. Thom (Editor), Coastal Geomorphology in Australia. Academic Press, Sydney, pp. 151-178. Hamilton, L.S. and Snedaker, S.C. (Editors), 1984. Handbook for Mangrove Area Management. United Nations Environment Programme and East-West Center, Environment and Policy Institute, 123 pp. Hatcher, B.G., Johannes, R.E. and Robertsen, A.I., 1989. Review of research relevant to the conservation of shallow tropical marine ecosystems. Oceanogr. Mar. Biol. Ann. Rev., 27: 337-414. Irion, G. and Petr, T, 1979. Mangrovensedimente am Golf von Papua. Senckenberg am Meer Nr. 365, Natur und Museum, 109: 297-304. Joshi, G.V. and Bhosale, L.J., 1982. Estuarine ecosystem of India. In: D.N. Sen and K.S. Rajpurohit (Editors), Tasks for Vegetation Science 2: Contributions to the Ecology of Halophytes. Junk, The Hague, pp. 21-33. Kjerfve, B. and Magill, K.E., 1989. Geographic and hydrodynamic characteristics of shallow coastal lagoons. Mar. Geol., 88: 187-199 Kvalvagnaes, K., 1980. Marine survey report of the Segara Anakan/Nusa Kambangan area. Field report of UNDP/FAO National Parks Development Project, Bogor, 20 pp. Lugo, A.E., 1980. Mangrove ecosystems: successional or steady state. Tropical Succession: 65-72. Lugo, A.E. and Patterson-Zucca, C., 1977. The impact of low temperature stress on mangrove structure and growth. Tropical Ecol., 18: 149-161. Lynch, J.C., Meriwether, J.R., McKee, B.A., Vera-Herrera, F. and Twilley, R.R., 1989. Recent accretion in mangrove ecosystems based on 137Csand 210Pb.Estuaries, 12: 264-299. Macnae, W., 1968. A general account of the fauna and flora of mangrove swamps and forests in the Indo-West-Pacific region. Adv. Mar. Biol., 6: 73-270. Mazda, Y., Yokoch, H. and Sato, Y., 1990. Groundwater flow in the Bashita-Minatomangrove area, and its influence on water and bottom mud properties. Estuarine Coastal Shelf Sci., 31: 621-638. Oliver, J., 1982. The geographic and environmental aspectsof mangrove communities: climate. In: B.F. Clough (Editor), Mangrove Ecosystems in Australia. A.N.U. Press, Canberra, pp. 19-30. Ong, J.E., 1982. Mangroves and aquaculture in Malaysia. Ambio, 11: 252-257. Rabinowitz, D., 1978. Early growth of mangrove seedlings in Panama, and an hypothesis concerning the relationship of dispersal and zonation. J. Biogeogr., 5: 113-133.
356
P.G.E.F. AUGUSTINUS
Richards, W., 1952. The Tropical Rainforest, an Ecological Study. Cambridge Univ. Press, Cambridge, 450 pp. Saenger, P. and Hopkins, M.S., 1975. Observations on the mangroves of the Southeastern Gulf of Carpentaria, Australia. Proc. Int. Symp. on Biology and Management of Mangroves, Honolulu (1974), pp. 126-136. Saenger, P., Hegerl, E.J. and Davie J.D.S. (Editors), 1983. Global status of mangrove ecosystems. The Environmentalist 3, suppl. 3: 1-88. Sato, K., 1984. Studies on the protective functions of the mangrove forest against erosion IV. Sci. Bull. Coll. Agric. Univ. Ryukyus, 31: 189-200. Sato, K., 1985. Studies on the protective functions of the mangrove forest against erosion and destruction V; Preliminary trials of the mangrove forest as a coastal prevention forest. Sci. Bull. Coll. Agric. Univ. Ryukyus, 32: 161-172. Sato, K., 1989. Studies on stiltroot of Rhyzophora stylosa and proporties of sedimentation in mangrove forest. Galaxea 8: 43-48. Scholl, D.W., 1968. Mangrove swamps: geology and sedimentology. In: R.W. Fairbridge (Editor), The Encyclopedia of Geomorphology. Reinhold Book Corporation, New York, pp. 683-688. Scoffin, TP., 1970. The trapping and binding of subtidal carbonate sediment by marine vegetation in Bimini Lagoon, Bahamas. J. Sediment. Petrol., 40: 249-273. Semeniuk, V., 1980. Mangrove zonation along an eroding coastline in King Sound, North-Western Australia. J. Ecol., 68: 789-812. Smith 111, TJ., 1986. Comparative effects of cyclone damage to mangrove forests: Kathy versus Winifred. Workshop on the offshore effects of tropical cyclone Winifred, Townsville, Australia, 1986, GBRMPA Workshop Ser. 7,59 pp. Smith 111, TJ., and Duke, N.C., 1987. Physical determinants of inter-estuary variation in mangrove species richness around the tropical coastline of Australia. J. Biogeogr., 14: 9-19. Snedaker, S.C., 1982. Mangrove species zonation: why? In: D.N. Sen and K.S. Rajpurohit (Editors), Tasks for Vegetation Science 2: Contributions to the Ecology of Halophytes. Junk, The Hague: 111-125. Spenceley, A.P., 1977. The role of pneumatophores in sedimentary processes. Mar. Geol., 23: M31M37. Spenceley, A.P., 1982. The geomorphological and zonational development of mangroveswamps in the Townsville area, North Queensland. James Cook Univ. of N. Queensland, Dep. of Geography, Monograph series No. 11, 69 pp. Steinke, TD. and Ward, C.J., 1989. Some effects of the cyclones Domoina and Imboa on mangrove communities in the St. Lucia Estuary, S.-Africa. Tydskr. Plantk., 55: 340-348. Steup, F.K.M., 1941. Kustaanwas en mangrove. Natuutwet. Tijdschr. Ned. Indie, 101: 353-355. Teas, H.J., 1980. Mangrove swamp creation: Rehabilitation and creation of selected coastal habitats. Proc. Workshop, Sapelo Island, Georgia, May 1976. Fish and Wildlife Service, Biological Services Program, Washington DC, Rep. FWS/OBS-80/27, pp. 63-90. Thom, B.G., 1967. Mangrove ecology and deltaic geomorphology: Tabasco, Mexico. J. Ecol., 55: 301343. Thorn, B.G., 1982. Mangrove ecology: a geomorphological perspective. In: B.F. Clough (Editor), Mangrove ecosystems in Australia, structure, function and management. A.N.U. Press, Canberra, pp. 3-17. Thorn, B.G., 1984. Coastal landforms and geomorphic processes. In: S.C. Snedaker and J.G. Snedaker (Editors), The Mangrove Ecosystem: Research Methods. UNESCO, Bungay, United Kingdom, pp. 3-17. Thom, B.C., Wright, L.D. and Coleman, J.M., 1975. Mangrove ecology and deltaic-estuarine geomorphology: Cambridge Gulf-Ord River, Western Australia. J. Ecol. 63: 203-232. Tomlinson, P.B., 1986. The Botany of Mangroves. Cambridge Univ. Press, Cambridge, 413 pp. van Eerdt, M., 1985. The influence of vegetation on erosion and accretion in salt marshes of the Oosterschelde, The Netherlands. Vegetatio, 62: 367-373. van Steenis, G.G.G.J., 1941. Kustaanwas en mangrove. Natuunvet. Tijdschr. Ned. Indie 101: 82-85. Walsby, J. and Torckler, D., 1992. Forests in the sea. N. Z. Geogr., 15: 40-65.
GEOMORPHOLOGY AND SEDIMENTOLOGY OF MANGROVES
357
Watson, J.G., 1928. Mangrove Forests of the Malay Peninsula, Singapore. Fraser and Neave (Malayan Forest. Rec. 6), 275 pp. Wells, J.T. and Coleman, J.M., 1981. Periodic mudflat progradation, Northeastern coast of South America: a hypothesis. .I.Sediment. Petrol., 51: 1069-1075. Wolanski, E., 1992. Hydrodynamics of mangrove swamps and their coastal waters. In: V Jaccarini and E. Martens (Editors), The ecology of mangrove and related ecosystems. Hydrobiologia, 247: 141-161. Wolanski, E. and Ridd, RV., 1986. Tidal mixing and trapping in mangrove swamps. Est. Coastal Shelf Sci., 23: 759-771. Wolanski, E., Jones, M. and Bunt, J.S., 1980. Hydrodynamics of a tidal creek-mangrove swamp system. Aust. J. Mar. Freshwater Res., 31: 431-450. Wolanski, E., Mazda, Y. and Ridd, P., 1992. Mangrove hydrodynamics. In: A.I. Robertson and D.M. Alongi (Editors), Tropical mangrove ecosystems. American Geophysical Union, Washington DC, pp. 43-62. Woodroffe, C.D., 1987. Pacific Islands mangroves: distribution and environmental settings. Pacific Sci., 41: 166-185. Woodroffe, C.D., 1992. Mangrove sediments and geomorphology. In: A.I. Robertson and D.M. Alongi (Editors), Tropical mangrove ecosystems. American Geophysical Union, Washington DC, pp. 7-41. Woodroffe, C.D., Chappell, J.M.A., Thom, B.G. and Wallensky, E., 1985a. Geomorphology of the South Alligator Tidal River and Plains, Northern Territory. In: K.N. Bardsley, J.D.S. Davie and C.D. Woodroffe (Editors), Coasts and tidal wetlands of the Australian monsoon region. Australian National University, North Australia Research Unit, Mangrove Monograph No. 1, Darwin. pp. 3-15. Woodroffe, C.D., Chappel, J.M.A., Thom, B.G. and Wallensky, E., 1985b. Stratigraphyof the South Alligator Tidal River and Plains, Northern Territory. In: K.N. Bardsley, J.D.S. Davie and C.D. Woodroffe (Editors), Coasts and tidal wetlands of the Australian monsoon region. Australian National University, North Australia Research Unit, Mangrove Monograph No. 1, Darwin, pp. 17-30. Woodroffe, C.D., Chappell, J.M.A. Thom, B.G. and Wallensky, E., 1986. Geomorphological dynamics and evolution of the South Alligator Tidal River and Plains, Northern Territory. Mangrove Monograph No. 3, ISBN 0 86784 917 7, Australian National University, North Australia Research Unit, Darwin, 190 pp. Zenkovich, V.P., 1967. Processes of Coastal Development. Oliver and Boyd, Edinburgh, 738 pp. (English translation edited by J.A. Steers).
This Page Intentionally Left Blank
Geomorphologyund Sedimentology of Estuaries. Developments in Sedimentology53 edited by G.M.E. Perillo 0 1995 Elsevier Science B.V. All rights reserved.
359
Chapter 13
ESTUARINE DUNES AND BARS ROBERT W. DALRYMPLE and ROBERT N. RHODES
INTRODUCTION
As described in other chapters in this volume, estuaries are dynamic environments, subject to vigorous water motion and sediment transport. The primary energy source is tidal currents, but river currents can be important from time to time, particularly in the inner part of the estuary. Residual flow due to density gradients may also contribute to sediment movement. Wave action is significant at the mouth of many estuaries, but is generally of limited importance inside the estuary. Consequently, this chapter will consider only those bedforms and bars generated by currents. Movement of sand in tidal and river channels and on the lower parts of the flanking tidal flats leads to the development of undulations which range in size from centimetre-scale current ripples (microforms), through decimetre- and metrescale dunes (mesofoms), to metre- and decametre-scale barjorms (macroforms). Meso-scale dune bedforms and macro-scale barforms are conspicuous morphological elements within most estuaries, and their large size relative to the flow depth and width causes them to have an important influence on the dynamics of an estuary and its use by humans. Their deposits may also form a significant proportion of the geological record of estuaries (Dalrymple et al., 1992). Therefore, this chapter will examine the morphology, process-response relationships, and internal structures of these two larger-scale groups of bed features, concentrating on dunes because they are much more thoroughly studied. Barforms will be discussed more briefly at the end, primarily to emphasize their distinction from dunes. Although the examples and discussion concentrate on the estuarine environment, the concepts examined are more broadly applicable.
DUNE CLASSIFICATION
Extensive observations in flumes (e.g., Southard and Boguchwal, 1990b) and nature (e.g., Dalrymple et al., 1978; Rubin and McCulloch, 1980) have shown that a sand bed on which sediment movement is occurring may exhibit one of 5 distinct kinds of bedform: lower plane bed, ripples, dunes, upper plane bed, and antidunes. Figure 13-1 shows the predictable conditions under which each occurs. Many different terms, including megaripple and sandwave, have been used for the features which are here called dunes (Table 13-1). It is widely believed, however, that all of these features are formed by a common process which is distinct from those forming current ripples and barforms (Ashley, 1990). As discussed by Smith (1970),
360
R.W. DALRYMPLE AND R.N. RHODES
I 02
I
I
-
!
I
I I .* I 1 1 1 1
I
I
1
1
1
1
.’ ,
l
l
l
l
I
I
1
1
1
1
1
10°C SPEED (m/s) Fig. 13-1. Bedform phase diagrams showing the combinations of (A) mean How speed (depth averaged) and median grain size and (B) mean How speed and flow depth under which the five distinct types of bedform exist. The boundaries are based on data from steady, uniform (flume) How and have been standardized to 10°C to remove the effect of changes in fluid viscosity. The boundary between 2D and 3D dunes is from Costello and Southard (1981, fig. 7) and Harms et al. (1982, fig. 2-5). “Gradual” and “abrupt” indicate the nature of the phase transitions. Fr = Froude number. The vertical and horizontal, dashed lines (in A and B, respectively) show where the two plots intersect. Modified from Southard and Boguchwal(1990b).
Costello (1974), Yalin (1977), and Costello and Southard (1981) among others, the presence of an initial defect in the bed (a “negative step” which must be at least as large as a current ripple) causes a perturbation in the flow downstream of the defect. The effect of this extends through the entire thickness of the boundary layer
361
ESTUARINE DUNES AND BARS Table 13-1
Synonymy between some common, previously-used bedform names (see table 4 of Ashley, 1990) and those derived from the Ashley (1990) classification (Table 13-2). Old term
Reference
New term
Large(sca1e) ripple
Allen (1968) Harms et al. (1982)
Dune
Sand wave
Bouma et al. (1980) Rubin and McCulloch (1980) Belderson et al. (1982) Knebel(l989)
Megaripple
Boothroyd and Hubbard (1975) Dalrymple et al. (1978) Perillo and Ludwick (1984) Aliotta and Perillo (1987)
Small to medium, simple dune
Type 1 megaripple
Dalrymple et al. (1978) Elliott and Gardner (1981)
Small to medium, simple, 2D dune
Linear megaripple
Boothroyd and Hubbard (1975)
Low-energy sand wave
Boothroyd (1985)
Type 2 megaripple
Dalrymple et al. (1978) Elliott and Gardner (1981)
Cuspate megaripple
Boothroyd and Hubbard (1975)
Small to medium, simple, 3D dune
Sinuous megaripple
Boothroyd (1985)
Rippled sandwave
Dalrymple et al. (1978)
Large to very large simde dune
S andwave
Langhorne(l973) Dalrymple (1984) Perillo and Ludwick (1984) Harris (1988) Aliotta and Perillo (1987)
Large to very large (compound) dune
Megarippled sandwave
Dalrymple et al. (1978) Elliott and Gardner (1981)
High-energy sand wave
Boothroyd (1985)
(i.e., the entire flow depth in most cases) and causes the depth-averaged, bed shear stress to experience a local maximum a certain distance downstream of the defect. The decrease in the sediment-transport rate on the downstream side of this stress maximum in turn causes deposition and the generation of a second defect, at which point the process repeats itself, thereby generating a field of dunes. Because the process which produces dunes involves the entire flow thickness, such mesoscale bedforms are said to scale with flow depth. As will be seen below, this is true whether the features are generated by unidirectional (river) or reversing (tidal) currents. Consequently it is widely but not universally believed that all mesoscale features are genetically related and should be called by a single name.
362
R.W. DALRYMPLE AND R.N. RHODES
Table 13-2 Descriptive classification of dunes (modified after Ashley, 1990, table 6 , to include dune orientation)
First-order descriptors: Size:
Shape:
Term -
small
medium
large
very large
Spacing (m) Height * (m)
0.6-5 0.05-0.25
5-10 0.25-0.5
10-100 0.5-3
1100 >3
2-Dimensional - relatively straight crested, lacking scour pits 3-Dimensional - sinuous to lunate, with scour pits
Second-order descriptors: Simple - lacks superimposed dunes Superposition: Compound - bears smaller, superimposed dunes (should also include relative size and orientation) Sediment characteristics: including grain size, sorting (no specific names given)
Third-order descriptors: (no specific names given) Bedform profile: stoss and lee slope lengths and angles Fullbeddedness: fraction of bed covered by moveable sediment Flow history: time-velocity characteristics; relative strengths of opposing flows Dune behaviour and migration history Orientation: transverse, oblique, longitudinal
* Height ( H ) calculated from the wavelength ( L ) using the equation H = 0.0677L"~x'1yx (Flemming, 1988; see Fig. 13-9). Note: The height values have been modified from those given by Ashley to correct a mathematical error (Serge Bernt, pers. commun., 1994). Terms in bold face are used in this chapter, but all attributes are discussed.
Following Ashley (1990) we adopt the term dune and employ the descriptive classification proposed by her (Table 13-2). The most notable dissenting voice regarding the genetic unity of dunes is Allen (1980, 1982, pp. 454-466) who argues that large and very large, tidal bedforms are a wave-generated feature, with reversing tidal currents substituting for the oscillatory motion of wind waves. To distinguish these features from those generated by unidirectional flow, Allen (1980, 1982) advocates the use of the term sandwave for large tidal bedforms. If this view is correct, tidal sandwaves are not dunes and they would not be expected to occupy a predictable stability field in phase diagrams such as Fig. 13-1. However, various studies have shown that their occurrence is predictable without reference to tidalflow characteristics such as tidal asymmetry (e.g., Dalrymple et al., 1978; Rubin and McCulloch, 1980; Dalrymple, 1984) and most workers do not accept a special status for tidal dunes (Ashley, 1990). It is obvious from Tables 13-1 and 13-2 that the primary bedform attributes which have attracted the attention of previous workers are dune size (small, medium, large, and very large), plan-form shape (2D and 3D; type 1 and type 2; linear and sinuous), and the presence or absence of smaller, superimposed dunes (simple and compound; megaripple and sandwave). Examples of the more common varieties are provided in Fig. 13-2. These attributes and the factors controlling their spatial and temporal variation within estuaries will be the focus of the following sections.
ESTUARINE DUNES AND BARS
363
DISTRIBUTION OF DUNES
Controlling variables Although few experimental studies have duplicated the unsteady and reversing flow conditions which characterize most estuarine environments, many field studies suggest that the relationships shown in Fig. 13-1A are generally valid in more complex natural environments (e.g., Boothroyd and Hubbard, 1975; Dalrymple et al., 1978; Rubin and McCulloch, 1980; Middleton and Southard, 1984), provided that care is taken to define the “effective” flow conditions properly. In the estuarine environment with its complex superposition of tidal and non-tidal processes and the resulting prevalence of unsteady flow, this is not always easy. Typically, however, workers have considered the effective conditions to be those which produce modal (Rubin and McCulloch, 1980) or near-maximum (Boothroyd and Hubbard, 1975; Dalrymple et al., 1978) sediment discharges. In the remainder of this chapter, all discussion of the water depths or current speeds responsible for dunes refers to these effective conditions. The influence of temporally-variable conditions is considered at greater length under the headings of Unsteadyflow and Morphological response to unsteady flow. In general terms, dunes may form in any sediment coarser than approximately 0.13 mm (2.9 phi; near the lower limit of fine sand). The minimum current speed at which dunes occur is dependent on water depth and grain size (Fig. 13-1), but is typically of the order of 0.5 m/s, rising as the depth and grain size increase. The maximum current speed at which dunes are stable also increases with depth and grain size, but is rarely exceeded in estuarine environments and then only in very shallow water or the inner portion of some macrotidal estuaries (Dalrymple et al., 1990). There is almost no water-depth limitation on dune formation, with dunes occurring in depths ranging from a few tens of centimetres to several tens of metres. Combinations of flow depth, current speed, and sediment size falling within the dune stability field of Fig. 13-1 are necessary conditions for dune development but are not always suficient, in that several additional factors can influence where dunes occur. The major prerequisite is the presence of enough cohesionless sediment to form the dunes. Thus, the presence of only small amounts of mobile sand over a hard substrate may not permit the formation of dunes (e.g., Klein, 1970;Aliotta and Perillo, 1987). In addition, the presence of a binding agent (more than 10-15% admixed silt or clay-sized material, an algal/diatom coating, or abundant mucus-bound worm tubes) may prevent sediment movement and dune formation (e.g., Tenvindt, 1971; Bokuniewicz et al., 1977; Bouma et al., 1980; Fenster et al., 1990). The influence of a binding agent is most important in areas with relatively low current speeds and may determine the limit of dune fields in the direction of decreasing current speed. The presence of intense wave action may also prevent the development of dunes (e.g., McCave, 1971; Ludwick, 1972), especially in shallow water at the mouth of estuaries. Best and Leeder (1993) have shown recently that the presence of even relatively low amounts (
364
R.W. DALRYMPLE AND R.N. RHODES
Fig. 13-2. Examples of the common varieties of dune: (A) medium, 2D, simple dune passing laterally into ripples, the terminus of the dune representing the ripple-dune phase boundary of Fig. 13-1 (elongate tidal bar, Minas Basin, Bay of Fundy); (B) small, simple dunes transitional between 2D (foreground) and 3D (elongate tidal bar, Gironde estuary, France); (C) medium, 3D, simple dunes
ESTUARINE DUNES AND BARS
365
are raised by a similar amount, dunes will be more restricted in turbid estuarine flows than would be expected on the basis of data from clear-water flows.
Distribution within estuaries As a result of the broad range of conditions under which dunes form, they are extensively developed in many estuaries. They are particularly abundant in tide-dominated estuaries (Dalrymple et al., 1992) where they are widespread on the elongate sand bars which occur at the mouth. Numerous examples have been described, including Cobequid Bay, Bay of Fundy (Fig. 13-2A, C, E, F; Dalrymple et al., 1990), the Bristol Channel-Severn River and adjacent estuaries (Elliott and Gardner, 1981; Harris, 1988), the Thames estuary (Langhorne, 1973,1977), the Bahia Blanca estuary, Argentina (Aliotta and Perillo, 1987), and the Ord River, Australia (Wright et al., 1973). Dunes are also abundant at the mouths of wave-dominated, coastal-plain estuaries and rias. Even though such estuaries commonly have small tidal ranges, tidal-current speeds are high because of the constriction produced by barrier bars. Dunes are particularly widespread in the tidal inlets and tidal channels which cross the flood-tidal deltas, documented examples including St. Andrews Bay, Florida (Salsman et al., 1966), Chesapeake Bay (Ludwick, 1972), Delaware Bay (Knebel, 1989), various Massachusetts estuaries (Boothroyd and Hubbard, 1975; Hine, 1975), various German and Dutch, Wadden Sea estuaries (van Straaten, 1953; Reineck, 1963; Davis and Flemming, 1991), and Moreton Bay, Australia (Fig. 13-3; Harris and Jones, 1988; Harris et al., 1992). More open-mouthed estuaries also commonly have dunes developed in the area where the flood-tidal delta would normally occur, e.g., San Francisco Bay (Rubin and McCulloch, 1980); Long Island Sound, New York (Fig. 13-4; Bokuniewicz et al., 1977; Fenster et al., 1990). On ebbtidal deltas dunes are typically restricted to the deeper channels and more sheltered areas because of wave action (Fig. 13-2D). Dunes are generally not present in the deeper-water, low-energy central basin of wave-dominate estuaries, but as such estuaries fill, sandy tidal channels gradually develop in the area of the former lagoon (Roy et al., 1980; Dalrymple et al., 1992). Dunes are commonly developed in these channels and on the lower portions of the fringing tidal flats, examples including various Georgia estuaries (Visher and Howard, 1974; Greer, 1975; Zarillo, 1985), Great Sound, New Jersey (Ashley and Zeff, 1988), and the Schelde estuaries, The Netherlands (Terwindt, 1970; Terwindt and Brouwer, 1986). Fewer studies have been undertaken in the inner, tidal-fluvial transition zone,
(elongate tidal bar, Cobequid Bay, Bay of Fundy); (D) small, 3D, simple dunes (ebb-tidal delta, North Inlet, South Carolina); (E) large, 2D, compound dune with obliquely-superimposed medium, 2D, simple dunes (elongate tidal bar, Cobequid Bay, Bay of Fundy); (F) large, 2D, compound dune with obliquely-superimposed small, 2D-3D, simple dunes (elongate tidal bar, Cobequid Bay, Bay of Fundy); and (G) large and very large, 2D compound dunes with obliquely-superimposed medium, 2D, simple dunes (Torres Strait, Australia; courtesy of P.T Harris). Metre stick for scale in A and C; notebook for scale in E. Shovel in F is about 1 m high. In D-G the superimposed dunes are approximately transverse to the net sediment transport, whereas the larger dunes are oblique to it.
366
R.W. DALRYMPLE AND R.N. RHODES
10 - 20 m
0-lorn I
I
153"18E
I
I
153"20
I
20+m I
I
I
J
153"22
Fig. 13-3.Spatial distribution of dune wavelength on the flood-tidal delta at the entrance to Moreton Bay, a large, mesotidal (maximum tidal range 2.8 m) embayment on the east coast of Australia. Bathymetry in metres. See Fig. 13-30 for location. Note the presence of zig-zag, tidal sand ridges. Although there is a tendency for dune wavelength to be greatest in the deeper channels, there are many places where wavelength contours cross the bathymetric contours. Discontinuities in the distribution of wavelength are caused by the disappearance of large compound dunes while the smaller ones continue. Cross-hatched area is land. From Harris et al. (1992).
but dunes have been described from such areas, including for example the Pitt River, British Columbia (Ashley, 1975) and the bay-head delta in the Gironde estuary (Fig. 13-2B; Allen et al., 1969). The tidally-influenced portion of delta distributaries also contain well-developed dunes (e.g., the Rhine River (Tenvindt, 1970); the Fraser River (Kostaschuk et al., 1989)).
DUNE SIZE
The size (wavelength and height) of dunes is a complex function of many variables, the most important of these being water depth (or boundary-layer thickness), current speed, and grain size (Fig. 13-5). In addition, water temperature, sediment availability, and the time-history of any or all of above variables can have a
ESTUARINE DUNES AND BARS
367
Fig. 13-4. Spatial distribution of dune height in the entrance of Long Island Sound, USA. Note how heights decrease toward the edges of the dune fields. Water depths in metres. Cross-hatched areas are land. After Bokuniewicz et al. (1977).
significant influence. The following sections explore each of these factors in turn. In addition to these (quasi-)deterministic controls on dune size, it should be noted that bedform dimensions exhibit appreciable stochastic (random) variability, even under steady, equilibrium conditions (Allen, 1976a). For example, Nordin (1971) found that the standard deviation of dune wavelength averaged 59% of the mean value, while the standard deviation of height averaged 52% of the mean. Several reasons exist for this variability: 1) the presence of both young and mature bedforms in an area (dunes are thought to be small at birth and then grow to some equilibrium size; Allen, 1976a); 2) the inherent three-dimensionality of dunes, which causes variations in height and wavelength along any one crestline; and 3) the slightly different flow conditions which individual dunes experience because of the three-dimensionality of their neighbours (Allen, 1973a). Because of this stochastic variability, the following discussion should be regarded as pertaining to the population means rather than to an individual bedform. Water depthfboundaly-layer thickness Tabulations of the height and wavelength of dunes from many environments, including estuaries, clearly show that dune size generally increases as water depth
368
R.W. DALRYMPLE AND R.N. RHODES
Wotei depth, h
I d
Fig. 13-5. Plots of (A) dune wavelength and (B) dune height against water depth showing the general tendency for dune size to increase in direct and constant proportion to water depth. Note the general agreement with the relationships derived by Yalin (1964, 1977, 1987; see eqs. (13-1) and 13-2)). From Allen (1980).
increases (Figs. 13-5, 13-6A, C; Allen, 1982; Ashley, 1990; Southard and Boguchwal, 1990b), despite considerable scatter (approximately one order of magnitude; Fig. 135) attributable to the many other stochastic and deterministic factors which influence dune size. Yalin (1964, 1977, 1987), using a combination of theory and empirical
369
ESTUARINE DUNES AND BARS
-
02-
-
DUES
WAVELENGTH
-
WAVELENGTH I I
06
04
08
10
15
,
I
0.2
I
0.6
0.8
1.0
10°C SPEED (m/s)
1
1
1
I
0.8 10
10°C GRAIN SIZE (rnm)
10°C SPEED (rn/s)
0.4
0.6
04
1.5
0.2
,
0.4
,
I
0.6
,
,
,
0.8
,
I
1.0
10°C GRAIN SIZE (rnrn)
Fig. 13-6. Bedform phase diagrams showing the variation of dune wavelength (A, B) and height (C, D) as a function of (A, C) flow depth and current speed (for a grain size 0.3-0.4 mm) and (B, D) current speed and grain size (for a depth 0.25-0.4 m). Based on the flume data of Guy et al. (1966). The vertical and horizontal, dashed lines show where the sections intersect. Although the water depths portrayed are much less than those occurring in estuaries, observations in natural environments suggest that the general patterns illustrated can be extrapolated to greater depths. From Southard and Boguchwal (1990b, figs. 18 and 19).
observation, suggests that dune wavelength ( L D )should be approximately 6 times the water depth (h), while dune height ( H D ) should be approximately 17% of the depth:
L D = 6h
(13-1)
HD = 0.167h
(13-2)
Thus, dunes are said to scale with flow depth, whereas the size of current ripples is independent of depth. This influence of depth is evident in many estuaries, where large to very large dunes (see definitions in Table 13-2) generally occur only in the deeper, subtidal channels and tidal inlets, while small to medium dunes occur on the intertidal flats (Fig. 13-3; Boothroyd and Hubbard, 1975; Dalrymple et al., 1990; Bern6 et al., 1993). Strictly speaking, the foregoing relationships apply only to the largest dunes present in a given area; in other words, eqs. (13-1) and (13-2) provide approximate
370
R.W. DALRYMPLE AND R.N. RHODES
upper bounds on dune height and wavelength (Yalin, 1964; Rubin and McCulloch, 1980). The superposition of smaller dunes on larger ones (Fig. 13-2E-G) clearly indicates the presence of a more complex relationship. Two possible explanations exist for such superposition: 1) the formation of the smaller dunes within an internal boundary layer on the stoss side of the larger dunes (i.e., “equilibrium superposition”; Smith and McLean, 1977; Rubin and McCulloch, 1980; Dalrymple, 1984); and 2) temporal changes in water depth which allow large dunes to form first, followed later by the formation of the smaller dunes in a shallower flow, the larger bedforms remaining because of their long lag time (ie., “disequilibrium superposition”; Allen and Collinson, 1974; Allen, 1978, 1982). Which of these processes predominates depends in part on the rate-of-change of flow conditions. The formation of superimposed dunes in flume experiments (Bohacs, 1981) shows that equilibrium superposition may occur under steady conditions, whereas disequilibrium superposition has been clearly demonstrated in fluvially-dominated distributary or inner-estuary channels where the change of water depth (and flow speed) is slow, occurring over a period of weeks to months (Allen and Collinson, 1974; Allen, 1976b; Kostaschuk et al., 1989; see also Allen’s (1976a, c, d, 1978) numerical simulations of slowly-varying river flow). On the other hand, in the tidal-dominated, outer part of estuaries the rate-of-change of water depth (and current speed) is typically one or more orders of magnitude higher and observations indicate that even small to medium dunes are generally not able to remain in equilibrium continuously over a tidal half-cycle (Allen and Friend, 1976b). Dalrymple (1984) reports that small superimposed dunes are not obliterated during peak flow, as would be expected if disequilibrium superposition operated. Instead, individual small dunes retain their identity over many tidal cycles, changing their characteristics only slowly over the 14day, neap-spring tidal cycle. Thus, it appears that (quasi-)equilibrium superposition occurs under the rapidly-changing conditions of tide-dominated environments (Rubin and McCulloch, 1980; Dalrymple, 1984). The presence of stratified flow due to the presence of a salt wedge or a bottom layer of higher suspended-sediment concentration introduces an additional complexity, some authors suggesting that dune size is governed by the thickness of the lower, dense layer rather than by the full flow depth (e.g., d’Anglejan, 1971; Visher and Howard, 1974). In these two cases, dune size is generally close to the values predicted by eqs. (13-1) and (13-2), using the total water depth rather than the thickness of the salt wedge. Thus, the presence of stratification does not appear to cause a detectable decrease in the scale of the dunes, presumably because the currents directly attributable to the saline stratification (of the order of 5-10 cm/s; Nichols and Biggs, 1985) are not sufficient to create bedforms independently of the strong tidal currents that involve the full flow depth. Little is known about the effect of a turbid bottom layer on flow structure, but the results of Sheng and Villaret (1989) suggest that suspended-sediment concentrations of <2g/l have negligible influence (see also Best and Leeder, 1993). Thus, the degree of stratification present in most estuaries appears not to affect the vertical distribution of tidal-current speed enough to alter the bed shear stress or bottom roughness appreciably (McCutcheon, 1981; Dyer, 1986).
ESTUARINE DUNES AND BARS
371
Current speed and grain size Flow strength as measured by either current speed or bed shear stress has a complex influence on bedform size. Other factors being constant, dune height increases as the mean current speed or shear stress increases from the lower boundary of the dune stability field, reaching a maximum value in the central portion of the dune field before decreasing as the speed or shear stress approaches the onset of upper plane bed or antidune conditions (Fig. 13-6C, D; Allen, 1982, figs. 8-21 and 8-24; Rubin and McCulloch, 1980; Freds~e,1983; van Rijn, 1983, 1984). The initial increase in dune height is generally attributed to increased erosion in the trough as the intensity of turbulence at the reattachment point and the strength of the return eddy increase (Costello, 1974; Costello and Southard, 1981). On the other hand, the lower height of dunes near the upper-flow-regime boundary is thought to be due to an increase in the amount of sediment in suspension and a washing-out or planing-off of the dune crest (Boothroyd and Hubbard, 1975; Allen and Leeder, 1980; Hand and Bartberger, 1988; Johns et al., 1990). Numerous examples exist of the influence of spatial changes in current speed on dune height. In many estuaries, dune height decreases into the lagoon from the estuary mouth (Salsman et al., 1966), despite an increase in water depth. This decrease is due to the decrease in current speed, from values well within the dune stability field, to values below the lower limit of dune formation (Fig. 13-6C, D). As an aside it might be noted that such down-current decreases in dune size require the deposition of sediment, thereby inducing bedform climbing. The thickness of the cross bed formed depends on the spatial rate-of-change of dune height (Rubin and Hunter, 1982). Dune height also typically becomes smaller at the edges of dune fields where they pass outward into rippled sand flats (Fig. 13-2A). The effect of current speed or shear stress on dune wavelength has not been investigated systematically. Costello (1974), Dalrymple et al. (1978), and Costello and Southard (1981) have noted that wavelength decreases as current speed increases in the transition from 2D to 3D and hypothesize that this is related to changes in the position and nature of flow reattachment. However, the synthesis presented by Southard and Boguchwal (1990b) using the flume data of Guy et al. (1966) indicates that wavelength increases steadily as current speed increases (Fig. 13-6A7B). The influence of grain size is also complicated. Rubin and McCulloch (1980), Allen (1982, p. 341) and Flemming (1988) show that dune height decreases as the sediment becomes finer, particularly as the ripple-dune boundary is approached (Fig 13-6D). Allen (1982) suggests that this is due to the relative proportion of sediment moving as suspended load, dune height decreasing as the amount of suspension increases. However, Southard and Boguchwal (1990b) indicate that this positive correlation between grain size and height reverses at grain sizes above about 0.5 mm, with dune height decreasing as the grain size continues to increase (Fig. 13-6D). The relationship between wavelength and grain size is poorly understood. Flemming (1988) indicates that larger dunes occur in coarser sediment, a finding supported by the observation that large compound dunes do not form in sediment finer
R.W. DALRYMPLE AND R.N. RHODES
372
than 0.274 mm (1.87 phi) (Dalrymple et al., 1978; Dalrymple, 1984). On the other hand, Elliott and Gardner (1981) report finding large dunes in fine to medium sand, and flume data discussed by Allen (1982, p. 332) and Southard and Boguchwal (1990b) indicate that dune wavelength decreases as the sediment becomes coarser (Fig. 13-6B). The reason(s) for these contradictory findings is unknown, but it is noted that only field studies report a positive correlation between dune wavelength and grain size. In nature it is frequently difficult to disentangle the effects of the many variables, especially as grain size and current speed may be positively correlated. Thus, the flume data must be regarded as providing the more reliable picture.
Water temperature and sediment availability Water temperature (also salinity and suspended sediment to a lesser extent) can have a significant influence on dune formation and size through its control on fluid viscosity (Southard and Boguchwal, 1990a, b, c). As shown in Fig. 13-7, if the measured flow depth, current speed, and grain size were to remain constant and only water temperature changed, the effective conditions would change dramatically: as the water becomes cooler, the temperature-standardized point moves to the 0.8
0.7
I
I
-
I
DUNES
v)
\
d 06 B w n v)
9 9 05
04
02
03
04
05
10°C GRAIN SIZE (mm)
Fig. 13-7. A portion of the current speed-grain size phase diagram of Fig. 13-la, with contours of bedform height from Fig. 13-6d, illustrating the effect of changing water temperature on bedform stability and dune height. The ripple-dune boundary shifts because of temperature-induced changes in flow depth. The height contours should shift by a similar amount, but are shown at their 10°C values for clarity. All five points correspond to constant, measured (“real-world”) flow conditions (depth 0.35 m; current speed 0.55 m/s; mean grain size 0.35 mm). The line joining these points is a temperature-change line (Southard and Boguchwal, 1990~).Note that a water temperature below about 12°C would generate ripples rather than dunes. As the water temperature increases above 1 2 T , dune height will increase as the water warms, the measured flow and sediment parameters remaining unchanged. The effect of water temperature on other depth-speed-grain size combinations can be estimated by moving the temperature-change line parallel to itself, positioning the 10°C point over the flow conditions of interest. After Southard and Boguchwal(199Oc).
ESTUARINE DUNES AND BARS
373
lower left, indicating that dune height decreases until only ripples are present. The influence of temperature is more pronounced with dune height than dune wavelength, because temperature-change lines (Fig. 13-7) are oriented at a high angle to the height contours but are nearly parallel to the wavelength contours in the speed-grain size plot (Fig. 13-6B). This influence of water temperature and the common failure to report temperature when citing bedform size in natural environments are major impediments to using previously reported field data to extend Fig. 13-6 to greater depths. Sediment availability, as expressed by the term fullbeddedness in the Ashley (1990) classification (Table 13-2), can influence dune height. If there is insufficient transportable sediment to cover the bed, the underlying rigid substrate will prevent erosion in the dune troughs giving them a flat base (see Fig. 13-11G). As a result, only the upper, crestal portion of the dune is present and the height is less than it would be otherwise. Bedforms in which this is the case are referred to as starved. The effect of sediment starvation on dune wavelength has not been investigated, but is likely to be less than on dune height. In the extreme situation where there is no movable sand, dunes will not be formed, but the limiting amount needed (expressed as a percentage of the volume of an individual dune) is not known.
Unsteadyflow The foregoing discussion of the factors controlling dune size is largely based on information derived from steady flows or represents the average relationships from which the effects of unsteadiness have been implicitly removed. However, the unsteady conditions which characterize estuaries may strongly influence the morphology of the bedforms observed at any one time or place. In the simplest terms, if the environmental conditions (one or any combination of flow depth, current speed, grain size, and/or water temperature) change, the morphology of the existing dunes will change in such a way as to re-establish equilibrium. The direction of the required bedform change can be predicted by plotting the beginning and end points of the environmental change in a (temperature standardized) current speed-flow depth-grain size phase diagram which has been contoured in terms of dune height and wavelength (eg., Figs. 13-6; 13-7; 13-19). A quantitative estimate of the bedform change is harder to make because the dune stability field cannot yet be contoured with certainty (Southard and Boguchwal, 1990b), and because the flow conditions generally change again before the bedforms regain equilibrium (Allen, 1973a, l974,1976a, d). As might be expected, the morphological response of dunes to unsteady flow is extremely complex. Therefore, the subject will be considered separately in a later section. At this point it is sufficient to cite some general observations derived from field studies (e.g., Allen et al., 1969; Allen, 1973b, 19761.3; Allen and Friend, 1976a; Dalrymple, 1984; Terwindt and Brouwer, 1986; Kostaschuk et al., 1989; Davis and Flemming, 1991; Rhodes, 1992) and the extensive numerical modelling of Allen (1976a, c, d, 1978). Note that Allen’s models were developed to mimic slow changes, and consider only the influence of changing water depth through the use of eqs.
374
R.W. DALRYMPLE AND R.N. RHODES
(13-1) and (13-2) above. Thus, they are of limited applicability tc real estuaries where several variables commonly change rapidly and in unison. In general, flow conditions in nature change too rapidly for the bedforms to remain in equilibrium, because of the finite time required to transport the sediment volume needed to change the bedform morphology. Thus, the bedforms Zag behind the change, by an amount that is dependent on the rate-of-change of flow conditions, the sediment-transport rate at each point in time, and the bedform size (Allen, 1974, 1976a, c; Allen and Friend, 1976b). Possible results include the co-existence of dunes of different sizes that were formed at different times, producing bimodal or polymodal dune-size populations (i.e., compound dunes; Allen and Collinson, 1974; Allen, 1976c, d), and the presence of non-unique relationships between flow conditions and dune size. The net effect is to add scatter to the relationships between bedform size and flow conditions (e.g., Fig. 13-5). Boothroyd (1985) has also suggested that the degree of asymmetry behveen the flood and ebb portions of the tidal cycle influences dune wavelength, with shorter bedforms occurring where transport in the two directions is more nearly equal. However, no data are available to adequately test this idea. The suggestion that dunes cannot exist in areas with equal transport by the ebb and flood currents (de Mowbray and Visser, 1984) is not supported by observations in many areas (Dalrymple, 1984; Bern6 et al., 1993).
Summary Bedform size (height and wavelength) is one of the most easily observed, frequently-reported, and fundamental attributes of dunes. Yet as the preceding paragraphs indicate, it is not related in any simple way to a single flow parameter; instead, dune size is determined by a complex interplay of water depth (or boundary-layer thickness), current speed (or bed shear stress), sediment grain size, water temperature, sediment availability, and flow history, all superimposed on the innate stochastic variability of the bedforms. Thus, it is likely that no single variable can explain the distribution of dune size (or even the presence of absence of dunes). Although it may be reasonable to assume that flow unsteadiness (including temperature change) only adds scatter to the basic relationships between dune size and flow depth, current speed, and grain size, failure to consider such changes could lead to erroneous inferences derived from limited data. Finally, it should be noted that dune height can vary much more rapidly than the average wavelength, because the amount of sediment which must be moved to raise or lower the height of a dune is much smaller than the amount needed to change the wavelength of several dunes.
DUNE SHAPE
Dune shape has been investigated less rigorously than size, even though it is widely recognized as a fundamental characteristic (Table 13-2). Two aspects of shape need to be considered: 1) the vertical profile, such as is seen in echo-sounder records; and 2)
375
ESTUARINE DUNES AND BARS
B
,ddle
C
Crestline
Spur
- _- _ _ - _ -_--- 4
_--_
Spur
Lobe c,.@?? - -- - - - ~. -._ ---____.-
Span / Breadth
- _c
Fig. 13-8. (A) Vertical section, (B) 3D perspective, and (C) plan view of a dune illustrating the terminology used in the description of bedform shape. Note that the span (or breadth) is the straight-line distance between the extremities of an individual dune. After Allen (1968, 1980).
the shape in plan view. In the following sections the terminology developed by Allen (1968, pp. 59-65; 1982, pp. 307-310) will be utilized (Fig. 13-8).
Profile shape One of the most widely-used, general measures of profile shape is the so-called ripple index (RI = wavelength/height; also called the vertical fomz-index (Allen, 1968) or steepness index (Tenvindt and Brouwer, 1986)), which has low values (1030) for steep dunes and high values (100-300) for flat bedforms. As shown by Fig. 13-9, dune height and wavelength are well correlated, despite the scatter discussed above. However, the general trend is not parallel to the lines of equal ripple index and dunes tend to become flatter as they become larger. Thus, small to medium dunes (wavelength t 1 0 m) generally have an RI < 30, whereas very large dunes (wavelength > 100 m) typically have an RI > 30 and commonly greater than 100. As might be expected from the foregoing discussion of the factors controlling height and wavelength, ripple index varies in a complex way with flow and sediment parameters. Observations in flumes show that flat dunes (RI > 30-40) characterize conditions which fall near the limits of the dune stability field (i.e., at the boundaries
376
R.W. DALRYMPLE AND R.N. RHODES
i
100
I
I
I
1
10
I
I
A
10
I-
I
s2
y
0.1
0.01
0.001 0.01
0.1
100
1000
WAVELENGTH (m) Fig. 13-9. Plot of dune height against dune wavelength for bedforms from natural environments. The heavy regression line specified by the equation at the upper left and the dashed upper bound slope less steeply than the lines of equal ripple index (10, 30, 100 and 300), indicating that wavelength increases more rapidly than height. After Flemming (1988).
with lower plane bed, ripples and upper plane bed), whereas steeper dunes (RI < 2030) characterize the core of the dune field (Fig. 13-10), as predicted by Fredsoe (1983) and van Rijn (1983, 1984). Figure 13-10B also indicates that the minimum value of ripple index decreases as water depth increases, a trend which would appear to contradict the observed tendency for larger dunes (which generally occur in deeper water) to have flatter profiles (Fig. 13-9). However, the relationship seen in nature may be due to the fact that it becomes progressively more difficult as depth increases to achieve current speeds much above the lower limit of the dune field; thus, dunes of the maximum possible steepness are rarely produced. In addition, Flemming (1988) has suggested that it is more difficult to achieve fullbeddedness as the dunes become larger. Therefore, if large to very large dunes are commonly sediment starved (e.g., Perillo and Ludwick, 1984), their height would be limited, producing higher-than-expectedRI values. Most estuarine dunes, regardless of size, are asymmetric, with a gentler stoss side and a steeper lee side. The exact shape and inclination of these faces depend on various factors, including the relative strengths of the dominant and subordinate currents, the size of the dune relative to the sediment-transport rate, the orientation of the dune relative to the net transport direction (obliquity), the presence or absence of superimposed dunes, and when the dune is viewed in a tidal cycle. The lee-side slope of small to medium, simple dunes which equilibrate relatively
377
ESTUARINE DUNES AND BARS
/
RIPPLE INDEX I
1
I
I/
0.2
,
I 0.4
I
I
I
0.6
l
0.8
l
1
1.0
10°C GRAIN SIZE (mm)
-
0.4
E
v
I
I4 0.2
n '$ 0.1
RIPPLE INDEX O.OE
I
0.4
I
I
0.6
I
1
0.8
1
I
1.0
1.5
10°C SPEED (m/s) Fig. 13-10. Contours of ripple index (RI; wavelength divided by height) in temperature-standardized, bedform phase diagrams: (A) current speed-grain size (for depths of 0.25-0.40 m); and (B) depthspeed (grain size 0.30-0.40 mm). The vertical and horizontal, dashed lines show where the two sections intersect. The contours were obtained using heights and wavelengths from Fig. 13-6 and should be considered as approximations because of the uncertainties associated with the original data (Southard and Boguchwal, 1990b). Note that the boundary between 2D and 3D dunes in (A) coincides approximately with an RI value of 30-35, as suggested by Tenvindt and Brouwer (1986) on the basis of field data.
quickly with the flow generally reaches a maximum value near or at the angle of repose (32-35'; Fig. 13-2). However, the average value as measured from the troughline to the crestline (Fig. 13-8) is commonly considerably less than this; for instance, Rhodes (1992) has measured values (in ebb-dominated areas at low tide) in the range of 15-25' on both 2D and 3D dunes, due to the presence of bottomsets and (less commonly) a crestal platform (Figs. 13-2; 13-8). The angle-of-repose slip face ranged from as little as 50% of the bedform height to over 80%. When viewed after
378
R.W. DALRYMPLE AND R.N. RHODES
1
I
Fig. 13-11. Diagrammatic vertical profiles of (A-E) small to medium dunes, and (F-J) large to very large dunes. MSI = modified symmetry index (Allen, 1968). The large arrow indicates the crestline of the major form; the small arrow the crestline of a subsidiary crest. (A) Triangular dune, with the MSI value suggested by Allen (1980, 1982) to typify dunes formed by uni-directional flow. (B) Triangular dune with typical tidal asymmetry. ( C ) Dune with crestal platform. (D, E) Dunes after subordinate tide, with small (D) and large (E) reverse-flow caps (dominant-flow stoss side dashed where buried). Note the pronounced convexity of the subordinate stoss side. (F, G) Symmetric trochoidal profiles, showing the effect of having fullbedded (F) and sediment-starved ( G ) conditions. (H) Asymmetric trochoidal shape with typical MSI value for a large to very large dune. (I, J) Profiles showing variations caused by crestal branching or the superposition of a large, subsidiary dune.
the dominant tide, the stoss side of smaller dunes is typically linear or only slightly convex-up, producing triangular or slightly hump-backed profiles (Fig. 13-1lB), the latter form being more common in finer sands. Stoss-side slopes are of the order of 2-5". If the flow approaches the upper-flow-regime limit of dunes, the planingoff of the crest produces a crestal platform which is either level or slopes gently
ESTUARINE DUNES AND BARS
379
Fig. 13-12. (A) Small-medium, 2D simple dune with a small ebb cap (Gironde estuary, France). Flood flow (the dominant current) is to the left. The trowel is 30 cm long. (B) Large, 2D compound dune with superimposed, ebb cap (lighter area) and small, 2D-3D simple, ebb dunes. Dominant, flood flow to the right.
down-current (Fig. 13-11C; e.g., Kohsiek and Terwindt, 1981). When viewed after the subordinate tide, small to medium dunes commonly exhibit partially-reversed forms with various sizes of ebb (at low tide) or flood (at high tide) cap (Figs. 13-l1D, E; 13-12). Such profiles have been referred to as a catback form (van Veen, 1938;
380
R.W. DALRYMPLE AND R.N. RHODES
Ludwick, 1972; Langhorne, 1973). Portions of the dominant-current stoss side may protrude from beneath the cap (Figs. 13-llD; 13-12), but if it doesn't, the presence of a residual, dominant-tide core may be recognized by the presence of a pronounced hump-backed profile (Fig. 13-11E). Large and very large, compound dunes are generally asymmetric, with their lee side facing in the direction of net sediment transport; symmetrical forms exist in areas where there is no net transport, such as on bar crests that lie between mutually-evasive, flood- and ebb-dominant channels (Ludwick, 1972; Langhorne, 1973; Dalrymple et al., 1978; Allen, 1980; Dalrymple, 1984; Harris, 1988; Bern6 et al., 1993). Asymmetrical forms generally retain a consistent facing direction over a tidal cycle (Figs. 13-2E, F; 13-12B), because the subordinate current does not transport sufficient sediment to reverse the profile. Compound dunes may have lee-face slopes in excess of 20" (Dalrymple, 1984; Bern6 et al., 1993), but it is rare to have significant portions of the lee face at the angle of repose, except at places where a superimposed dune is located on the lee face of the larger form (Fig. 13-2E, F; Dalrymple, 1984; Berne et al., 1988). If steep portions occur, they are usually high on the lee face, with the lower parts having much lower inclinations (see more below). Thus, average slopes are typically less than 10" and may be as low as 1-2". It is a general observation that larger dunes have lower-angle lee faces than smaller ones. The reasons for these low slopes are not well understood and several contributing factors can be identified. 1) Wave action is known to lower the lee-face slope of large bedforms temporarily (McCave, 1971; Ludwick, 1972), but this should not be a significant influence in relatively protected estuarine settings. 2) Various authors including Allen (1980) and Belderson et al. (1982) suggest that the slope decreases as the ratio of transport by the dominant and subordinate currents approaches one (i.e., as the net transport approaches zero). Indeed, lee-face slopes are generally higher in fluvial environments than in tidal settings; however, the exact mechanism by which the subordinate current produces a lower slope is unclear, because kinematicwave theory (Middleton and Southard, 1984, pp. 309-315) would suggest that even a small residual transport should cause a bed undulation to reach the angle of repose, given long enough. It should also be noted that Allen's (1980) explanation of lee-face inclination in terms of tidal transport asymmetry requires the movement of unrealistically large volumes of sediment in each half tidal cycle (due to the large size of the bedforms modelled); thus, the validity of this particular formulation has been questioned (Dalrymple, 1984). 3) Dalrymple (1984) has shown that erosion by the troughs of the faster-moving, superimposed dunes causes periodic lowering of the lee-face slope on medium, compound dunes, but the extension of this process to much larger bedforms is uncertain for the same reason that Allen's (1980) model is suspect. 4) The superimposed dunes may, however, have a more fundamental influence on flow in the lee of the main crest, their presence acting like the dimples on a golf ball to increase the turbulence near the bed, thereby delaying or preventing flow separation (Schlichting, 1968; Middleton and Southard, 1984). Because the flow remains attached for a greater distance downflow of the main dune crest, the rate of flow deceleration will be slower. Consequently, the brink will be less pronounced and the lee face will have a lower slope. 5) Recently, Sweet and Kocurek (1990)
ESTUARINE DUNES AND BARS
381
have shown in a study of aeolian dunes that oblique bedforms (see further discussion below) do not experience flow separation because the apparent lee-face slope is less. Instead, the flow remains attached and is deflected into an along-crest orientation. (Similar results are expected in aqueous flows, despite the differences in fluid density and viscosity). It should be noted that factors 4) and 5) initiate positive feed-back because flow separation becomes less likely as the sharpness of the brink and the lee-face slope decrease (Richards and Taylor, 1981; Hand and Bartberger, 1988). Sweet and Kocurek (1990) suggest, for instance, that the flow remains attached for all slopes averaging t20". The average stoss-side slope of large and very large dunes (1-3") is slightly less than that of smaller dunes because of the flatter profile of larger dunes (Fig. 13-9). Unlike smaller dunes, the stoss sides commonly (but not always) have a concaveup (trochoidal) profile, with the steepest slope near the crest (Fig. 13-11F-H; van Veen, 1938; Ludwick, 1972; Langhorne, 1973; Perillo and Ludwick, 1984; Aliotta and Perillo, 1987). More complex forms, including those with double crests (Fig. 13-111) or a terraced stoss side (Fig. 13-11J), may be due to branching of the main crestline (e.g., Langhorne, 1973) or the presence of a particularly large, superimposed dune. The cause of the concave, lee and stoss sides and more pronounced peakedness of larger dunes is not known, but it may be related to their immobility. Those large dunes which have triangular or humpbacked forms (Figs. 13-2E, F; 13-12B; Dalrymple, 1984; Boothroyd, 1985, figs. 7-19 and 7-28) generally come from areas where the tidal currents are almost always capable of transporting the sediment, whereas concave sides and trochoidal profiles are most common in areas where the currents are weaker and the threshold of sediment motion may not be exceeded for long periods, especially in the trough. For example, Fenster et al. (1990) have argued that large dunes in Long Island Sound are only fully active at extreme spring tides or when storm-generated currents are superimposed on the tidal flow. They suggest that these dunes have concave-up slopes because normal tidal currents are only capable of moving sand on the dune crest, allowing the sediment in the troughs to become biologically stabilized. In extreme cases this stabilization of sand in the trough may produce a flat-troughed form (Fig. 13-11G; e.g., Langhorne, 1973; Perillo and Ludwick, 1984) which mimics the shape which occurs where dunes are migrating over a non-erodible substrate (Aliotta and Perillo, 1987). Many workers have noted the presence or absence of asymmetry, but few quantitative measurements of exist. Allen (1980, 1982) has suggested that the modified symmetry index (MSI = [(stow-side length/lee-side length) - 11; Allen, 1968) is of the order of 16 for dunes formed by unidirectional currents (Fig. 13-11A). He also indicates that small to medium, tidal dunes approach this value, whereas larger tidal dunes rarely have MSI values greater than 4. The latter value is supported by data reported by Dalrymple (1984) and Bern6 et al. (1993), but the few data which exist for small to medium dunes in tidal environments indicate that such bedforms rarely (never?) have modified symmetry indices as high as 16. For instance, Rhodes (1992) reports that MSI values for dunes following the dominant tide (i.e., at their maximum asymmetry) are typically less than 8, and average 6.3 and 3.5 in two separate areas. Even lower values (0-5.3) were measured by Tenvindt and Brouwer (1986) for
382
R.W. DALRYMPLE AND R.N. RHODES
ebb-modified, flood-dominant dunes because of the presence of ebb caps. However, it cannot yet be concluded that tidal dunes are less asymmetric than those created by unidirectional currents because of the lack of data from rivers and flumes.
Plan shape Studies in both flumes (Costello and Southard, 1981; Southard and Boguchwal, 1990b) and estuaries (e.g., Boothroyd and Hubbard, 1975; Dalrymple et al., 1978) have shown that the plan shape of dunes changes in a predictable manner as the effective current speed increases. At low speeds (those only a little above the lower limit of dunes; Fig. 13-l), dunes have nearly straight crests and lee faces and a uniform height along their length due to the absence of scourpits (Figs. 13-2A and 13-12A). Such forms are termed 2 0 (two-dimensional) dunes (Table 13-2). At higher speeds, the sinuosity of the lee face and the irregularity of the troughline elevation increase as scour pits become more pronounced (Figs. 13-2B-D; 13-13). The lee face of these 30 dunes (Ashley, 1990) consists of alternating, more gently-curved, lunate segments (lobes) up-current of each scour pits and more sharply-curved, linguoid segments (saddles; Figs. 13-2C7D; 13-8) which are fronted by a current-parallel spur. These spurs may be either perpendicular (Fig. 13-2C) or oblique (Fig. 13-13) to the dune crest, depending on whether the bedform is transverse or oblique to
Fig. 13-13. Small, 3D, simple dune with spurs oriented at an oblique angle to the crestline (Cobequid Bay, Bay of Fundy). Dune height is 10-15 cm. This local section of the crestline is oblique to the flow because it is branching from the bedform in the rear. Note the prominent ripple fans (Allen, 1968) in the scour pits.
383
ESTUARINE DUNES AND BARS
the local flow. Even in strongly 3D dunes, the elevation of the crestline is relatively uniform (Figs. 13-2D; 13-13), with almost all of the three-dimensionality occurring in the trough. Allen (1968) has indicated that scour pits may be either in-phase or out-of-phase, but qualitative observations by the authors suggest that scour pits are randomly distributed. The boundary between 2D and 3D dunes is not firmly positioned in the bedform phase diagram (Fig. 13-1), in part because the change is gradational (Fig. 13-2A-D). Nevertheless, it appears that 2D dunes occupy a relatively narrow range of current speeds, with the transition to 3D dunes lying approximately parallel to lower limit of the dune stability field. 3D forms are the most widely developed type of dune, especially in the small and medium size classes. However, as dune size increases, 3D forms become increasingly uncommon (Figs. 13-2E, F; 13-12B; Dalrymple et al., 1978; Rubin and McCulloch, 1980; Dalrymple, 1984; Fenster et al., 1990), presumably because current speeds are rarely high enough in the deeper flows. The only quantitative attempt to relate dune morphology to flow characteristics is Allen’s (1977) study in which he found that the ratio of bedform wavelength (A,) to the flow-transverse spacing of saddles and spurs (A,) was given by: 0.412
1.71
k)
A x = 5.85 (k) Fr HD
(13-3) (1 where Fr = Froude number, HD = bedform height, h = flow depth, and w = flow width (the last term approaches one in wide, natural flows). This equation indicates that more three-dimensional bedforms (associated with high values of the ratio h,/h,) occur in flows with higher Froude numbers. Although this relationship is based entirely on data obtained from current ripples, Allen (1977) suggests that it also applies to dunes. No test of this equation has been undertaken. Despite the significance of the distinction between 2D and 3D forms for dune classification and cross-bedding geometry (Ashley, 1990), almost no effort has been made to quantitatively define the dividing line. In a significant first step, Rhodes (1992) periodically measured the geometry of dunes in two areas in the Bay of Fundy in detail over one or more neap-spring cycles. He proposed that the most significant attribute, both geologically and hydraulically, is the variability of the troughline elevation and expressed this by means of the coefficient of variation (standard deviation divided by mean) of height along the length of individual bedforms. For dunes in a field which was classified as 2D subjectively, the height variability did not exceed 22% and had a mean value of 15%, whereas bedforms in a field of moderately 3D dunes gave values ranging from < l o % to >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
<
v)
i51 W
a 5c
0
0
1
2
3
4
DUNE HEIGHT (m) Fig. 13-23. (A) Schematic section of a dune (not to scale) showing the geometry used in calculating the reversal time. The dune core (dark stipple) remains unmoved; only the surficial cap (one of the areas of light stipple) is reworked. (B) Time required to reverse a dune (days of continuous transport), as a function of dune height and sediment-discharge rate (4s). Sediment-discharge rates span the range of values reported for tidal environments and are expressed in m3/m.s and kg/m per 6 hour period. The lower values are more appropriate for 2D dunes, while the upper ones are more typical for 3D dunes. Modified after Bokuniewicz et al. (1977, figs. 9 and 10).
deposited on the former stoss side. A small cap is developed initially, but it grows in size as the reversal of the dune proceeds (Fig. 13-11D, E). Following Bokuniewicz et al. (1977) it is possible to obtain an estimate of the time required for complete reversal of the external form (i.e., for the lee face of the cap to reach just to the toe of the former stoss side, without reworking the core of the dune; Figs. 13-llE; 13-23A). In the calculations it is assumed that the dune has a triangular profile and a ripple index ( L D I H D )of 20, and that the crestal movement (6) needed to reverse the
ESTUARINE DUNES AND BARS
399
dune is 0 . 2 L ~(Bokuniewicz et al., 1977). Note that this value of 6 is equivalent to assuming a lee-face slope of 7” (for the L D / H Dvalue used). This slope is appropriate for large and very large (HD > -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 (<0.25 m high; Table 13-2) are typically of the order of 100-300 m/yr (although they are unlikely to migrate this far before dying), whereas large dunes (0.5-5 m high) commonly migrate 25-75 m/yr. Very large dunes (height > 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)
<25
(7-296) (7-1192)
5-96
,5000; (11.5-40.6 m/day)
0.35:
0-6.8
85 135
Migration-rate data not enclosed in parentheses represent long-term (weeks to months), values, whereas those in brackets are “instantaneous” or short-term (one to several tides) measurements. Single values are averages; values separated by a dash give the range of raported values. Values have been converted to similar units for ease of comparison (to convert to metres per semi-diurnal tidal cycle, divide by 705). The use of “year” as the time basis is not intended to imply that an individual bedform will live this long, although large to very large dunes do in many instances.
ESTUARINE DUNES AND BARS
401
The effect of current speed is reflected in various ways by the migration-rate data. For instance, 2D dunes, which are formed by slower currents, migrate more slowly that 3D dunes, even though the latter may be higher (Table 13-4; Terwindt and Brouwer, 1986; Rhodes, 1992). Temporal variations in current speed over neap-spring cycles or river floods are faithfully recorded in migration-rate data, because migration speeds do not lag behind changes in flow conditions as height and wavelength do (Terwindt and Brouwer, 1986; Kostaschuk et al., 1989; Rhodes, 1992). The variation in migration rate of simple dunes over a neap-spring cycle may be dramatic, as indicated by the range of values reported by Terwindt and Brouwer (1986; Table 13-4). However, in the case of compound dunes that have relatively large superimposed bedforms, the effect of neap-spring cyclicity can be masked by alternating pulses of rapid migration associated with the arrival of a superimposed dune crest and periods of negligible or even retrograde movement when the trough of a superimposed dune arrives at the main lee face (Dalrymple, 1984). Spatial differences in the degree of dominance, as indicated by the magnitude of the (U; - U:) term in eq. (13-7b), are largely responsible for the wide range of migration rates reported by Dalrymple (1984; Table 13-4); dunes in areas with nearly equal flood and ebb transport are stationary or migrate only slowly while those in areas with a clear dominance migrate faster. As a corollary of this, it might be noted that bedforms in tidally-dominated environments migrate much more slowly than comparable bedforms in fluvially-dominated settings (compare the Kostaschuk et al. (1989) data with the other measurements in Table 13-4), due to periods with no motion near slack water and reversed migration during the subordinate tide. As indicated by eq. (13-5), it is expected that oblique dunes will migrate more slowly than transverse bedforms of the same size for the same net sediment discharge. Unfortunately, the influence of obliquity cannot be illustrated by existing data, as dune orientation relative to the net transport direction is rarely determined. One of the expected results of oblique migration, namely the along-crest movement of crestal bifurcations and salients (Rubin, 1987a, b), has been reported by Langhorne (1973). Migration rates measured simultaneously in areas with relatively uniform flow and bedform characteristics (thereby supposedly removing the effects of spatial variability) may nevertheless show considerable variation over small distances (Langhorne, 1973; Bokuniewicz et al., 1977;Aliotta and Perillo, 1987; Fenster et al., 1990; Rhodes, 1992). This has been referred to as “flexing” of the dune crestlines. Some of this variability is related to local variations in dune height (Rhodes, 1992) and the along-crest migration of crestal sinuosities (Langhorne, 1973), but some is probably due to the random behaviour of dunes (Allen, 1973a).
INTERNAL STRUCTURE OF DUNES
It is well known that the plan shape of dunes has an important influence on the geometry of their internal structure, with 2D dunes producing some type of planartabular cross-bedding whereas 3D dunes generate trough cross-bedding (Ashley,
402
R.W. DALRYMPLE AND R.N. RHODES
1990). However, a good case can be made that dune size has a stronger influence on the nature of the structures: dune height obviously determines the maximum thickness of the set (or coset) which can be preserved, but dune wavelength is more significant because of its influence on the presence/absence of superimposed dunes (see above). Thus, small and medium, simple dunes have fundamentally different structures than large and very large, compound dunes because of the pronounced effect of the superimposed bedforms (Rubin and Hunter, 1983; Dalrymple, 1984; Rubin, 198%).
Simple dunes The structures within smaller, simple dunes consist of a single set of cross-bedding, within which the effects of tidal-current reversals and neap-spring variations in flow strength and dune migration rate are likely to be evident (Fig. 13-24). The fundamental sedimentation unit in such cross-bedding is the tidal bundle, which was defined by Boersma (1969) as the deposit of a single, dominant tide. It is bounded by deposits or erosion surfaces produced during the slack-water periods and subordinate current. The nature of all of these elements varies in response to the systematic changes in the strength of the currents over a neap-spring cycle (see Nio and Yang (1991) for a general review). The horizontal extent of each bundle represents the net migration during a single tide; therefore, the succession of bundles deposited over a series of neap-spring cycles commonly shows systematic changes in bundle thickness (Fig. 13-25; Visser, 1980; Boersma and Tenvindt, 1981; de Boer et al., 1989). Muddy, neap-tide intervals composed of closely-spaced or even amalgamated mud drapes typically pass into thicker and sandier, spring-tide bundles along the length of a cross bed. The number of bundles present in one cycle depends on whether the tide is semidiurnal or diurnal, and on the number of tides in which the currents do not rise above the threshold of sand movement during neap tides. The variations in current speed also produce changes in dune height (Figs. 13-20; 13-21), which are reflected in the depth of erosion. As a result, the set base may rise and fall (e.g., Tenvindt, 1981) as the dune alternately becomes smaller during neap tides and larger at spring tides. A similar set-base undulation can also be seen within a single tidal bundle (Fig. 13-24C) if the peak current speeds are high enough to cause significant trough scour and the bundle is long enough (i.e., >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 <5-10". They are somewhat elongated parallel to the currents (lengths 150-1500 m; widths 50-200 m), but do not resemble the highly-elongated bars that are more typical of tide-dominated estuaries (see below). The theoretical development of Yalin and da Silva (1991, 1992) does not predict the dimensions of multiple-row bars, but the data of Zaitlin (1987) indicate that they are more nearly equidimensional than single-row, alternate or meander bars (Fig. 13-27).
410
R.W. DALRYMPLE AND R.N. RHODES
Elongate tidal bars This type of barform is extensively developed at locations where there is strong (speeds typically >0.75 m/s), rectilinear, tidal flow. They are characteristic features of the outer part of macrotidal estuaries (the Thames estuary (Robinson, 1960); Cobequid Bay, Bay of Fundy (Knight, 1980; Dalrymple et al., 1990); the Bristol Channel (Harris, 1988)), but also occur at constrictions at the mouth of estuaries with smaller tidal ranges (Chesapeake Bay (Ludwick, 1974); Delaware Bay (Knebel, 1989); Moreton Bay, Australia (Harris and Jones, 1988; Harris et al., 1992)). Those which occur in the inner Gironde estuary (Tastet et al., 1986) form part of the bayhead delta. These features are also called linear sand banks or tidal-current ridges. The channel-margin linear bars and ebb spits which are associated with tidal deltas in barrier estuaries (Hayes, 1975; Boothroyd, 1985) are considered here to belong to this category. These bars vary in organization and size from estuary to estuary. In settings where the ebb and flood currents are constrained by smoothly converging estuary margins and the bottom depth changes only gradually (e.g., the Thames, Cobequid Bay and Gironde estuaries), the bars commonly lie end to end, forming a small number of long bar chains (Fig. 13-29; Dalrymple et al., 1990) that separate rnutually-evasive, ebb- and flood-dominant channels. In these cases, the ebb-dominant channel(s) are commonly linked to a river channel, while the flood channels terminate headward
Fig. 13-29. Organization of sand bars in Cobequid Bay, Bay of Fundy. The bars are organized into two main bar chains which trend from northeast to southwest across the bay, at a small angle to the predominant, east-west tidal flow. The channel to the north, and the western end of the central channel are flood dominant, whereas the channel to the south is ebb dominant. Note the numerous swatchways which dissect the northern bar chain. Based on data collected in 1988 (contour interval in metres). See also Dalrymple et al. (1990). Cross-hatched areas are land.
ESTUARINE DUNES AND BARS
411
Fig. 13-30. Pattern of elongate tidal bars in the entrance to Moreton Bay, Australia. Note the widespread occurrence of zig-zag shapes due to the interfingering of ebb- and flood-dominant channels. A muddy lagoon lies to the southwest. Contour interval in metres. Cross-hatched areas are land. After Harris et al. (1992).
(Fig. 13-29), sometimes producing a large-scale, ebb-tidal delta morphology (Knight, 1980; Dalrymple et al., 1990) in which the bar chains occupy positions which are analogous to the channel-margin linear bars of inlet-associated ebb-tidal deltas (Hayes, 1975; Boothroyd, 1985). The continuity of these bar chains is broken by diagonal channels called swatchways (Robinson, 1960). The larger of these divide the bar chain into individual bars, whereas smaller swatchways separate the bar segments discussed above. A less organized pattern of zig-zag bars and arcuate shoals typifies the entrance of estuaries with lagoons (e.g., Chesapeake Bay (Ludwick, 1974); Moreton Bay (Harris and Jones, 1988); Fig. 13-30). In such settings, the presence of a local constriction and submerged sill at the mouth produces a complex pattern of
412
R.W. DALRYMPLE AND R.N. RHODES
interfingering, flood- and ebb-dominant channels separated by the zigzag bars (Swift and Ludwick, 1976). Individual elongate bars are generally 1-15 km long, whereas bar chains may reach 40 km in length. Bar widths range from 0.2-4 km, with narrower bars typically occurring in deeper water. The bars and bar chains are very elongated, with lengthwidth ratios generally greater than 3 and locally greater than 10. The relief from the adjacent channel bottom to the bar crest can reach 20 m. In general, bars have lower relief and flatter tops in the shallower parts of estuaries (Figs. 13-29; 13-30; Harris, 1988; Dalrymple et al., 1990). Bars are typically oriented at a small angle (<20") to the predominant tidal flow (e.g., Knight, 1980; Harris and Jones, 1988). As a result of cross-bar flow, they are commonly asymmetric, both transverse and parallel to their length, with the steeper side facing in the direction of the locally dominant, sediment-transport direction. For example, in the Cobequid Bay estuary the steeper sides of entire bars as well as individual bar segments face headward and toward ebb-dominant channels (Fig. 13-29) because of the overall flood dominance of this portion of the estuary (Dalrymple et al., 1990). A more complex pattern of bar and bar-segment asymmetry exists in the Thames estuary (Robinson, 1960, figs. 4-6) and Moreton Bay (Harris and Jones, 1988, fig. l l ) , presumably due to less pronounced, regional dominance by either the ebb or flood currents. The great elongation of this class of bar suggests that they are not related genetically to the more equant, multiple-row bars discussed above. Two hypotheses have been advanced for their origin. Huthnance (1982a, b; see also Hulscher et al., 1993) has proposed a model for the formation of elongate ridges based on the interaction between the flow and a highly oblique irregularity on the bed. As the current encounters the incipient bar the cross-bar component of flow initially accelerates and erodes material from the up-current flank. At the crest, however, it experiences frictional resistance and decelerates, causing sediment deposition and upward growth of the bar. The wavelength of the bars so formed should increase as the mean water depth increases. Although this model was developed for shelf ridges, the fundamental nature of flow-bed interaction which it embodies suggests that it should operate in any setting. The small angle which most elongate, estuarine bars make with the currents is what one would expect if they were formed by this process, but the model cannot, by itself, explain the organized, ebb-tidal delta morphology of bar chains or the zigzag pattern of estuary-mouth bars. A second explanation which can more easily account for the organization of bar systems suggests that the bars and bar chains are simply levee-like features which accumulate in the zones of no net transport between channels with opposing directions of residual transport. Such an explanation seems highly appropriate for the bar chains described by Knight (1980) and Dalrymple et al. (1990) in Cobequid Bay and also for the channel-margin linear bars on ebb-tidal deltas (Hayes, 1975). In these situations, the swatchways exist to equalize the head differences which develop between the ebb and flood channels. This levee origin is also consistent with Swift and Ludwick's (1976, fig. 19) suggestion that zigzag bars originate on a submerged, estuary-mouth sill by the interpenetration of mutually evasive, ebb and flood channels. It may be, however, that these two explanations are
ESTUARINE DUNES AND BARS
413
not mutually exclusive, but instead operate together to create elongate bars; the general geomorphic factors which control the distribution of ebb and flood channels determine the preferred location of bars, while the bed-flow interaction described in the Huthnance model operates to make the features grow.
Delta-like bodies Many large sand bodies in estuaries are isolated features which fit into neither of the previous categories. Commonly they have a lobate morphology and are situated at the end of a channel or where a channel widens appreciably. Obvious examples include the spill-over and terminal lobes on flood-tidal deltas (Hayes, 1975; Boothroyd, 1985). Even entire flood- and ebb-tidal deltas have this morphology, as do the arcuate segments at the terminus of mutually-evasive channels in zigzag bar complexes. Similar features have also been described from the ends of swatchways in macrotidal, sand-bar complexes (Dalrymple et al., 1990). Such features clearly represent deltaic bodies which form at a point of flow expansion. Their location, size, and geometry are dependent on the surrounding topography and consequently are largely unpredictable. The channels which feed them commonly form at low points in the crestlines of pre-existing barforms, dissecting these larger features. Thus, these delta-like bodies commonly represent a modification of the primary barforms described above. They may in turn be similarly breached and modified, generating the very complex, compound-bar morphologies which typify many shallow, sandy estuaries.
Internal structures The internal structures of barforms are as complex and varied as the barforms themselves. In general, relatively little is known about the deposits of most barforms, due to their large size and subaqueous setting, and the resulting difficulty of undertaking adequate excavations. Surprisingly few seismic and coring studies have been undertaken. The internal structure of point bars is probably the best known. They generate upward-fining successions by lateral migration of the channel, typically passing upward from cross-bedded sands, perhaps with bipolar cross bedding, tidal bundles, and mud drapes, into interbedded sands and muds in which tidal rhythmites (Dalrymple et al., 1991) and inclined heterolithic stratification (Thomas et al., 1987) may occur. The succession will be capped by salt- or freshwater marsh sediments. Alternate bank-attached bars may develop similar successions, but the relative straightness of the channel in which these barforms occur suggests that lateral migration is limited. Despite the longitudinal asymmetry of these bars, along-channel migration has not been documented. The multiple-row, braid bars of open sand flats are known to migrate in the direction of the dominant current, at least in some settings (Zaitlin, 1987; Dalrymple et al., 1990). Forward-dipping, master bedding planes with low inclinations should be present. Large, slow-moving bars may be overtaken by smaller bars and dunes,
414
R.W. DALRYMPLE AND R.N. RHODES
thereby forming the nucleus of a compound bar, in the fashion observed in fluvial environments (e.g., Cant and Walker, 1978; Allen, 1983). As a result, their architecture may be very complex, containing forward-, lateral- and upstream-dipping master bedding planes with low-angle inclinations. The smaller-scale internal structures will depend on the nature of the superimposed bedforms, which may be either dunes or upper plane bed. Because elongate tidal bars are covered by dunes, their internal structures consist of both simple and compound cross-bedding of the types described above. The largerscale architecture is poorly known. High-resolution seismic data from Cobequid Bay (Bay of Fundy) indicate that they contain gently-dipping ( < 3 O ) , lateral-accretion deposits (Dalrymple and Zaitlin, 1994) formed by the superimposed dunes as they migrate obliquely out of the channel toward the bar crest where the speed of the total current (not just the cross-bar component) is less (Fig. 13-31A). Swatchways which migrate along the length of an elongate bar in response to the dominant current will produce smaller-scale, lateral-accretion deposits oriented at a high angle to the trend of the main bar (Fig. 13-31B; Dalrymple et al., 1990). If the bars occur in the outer portion of a tide-dominated estuary or in the central portion of a zig-zag shoal complex, their deposits will fine upward, due to the decrease in current speed from the channels to the bar crests (Ludwick, 1974; Dalrymple et al., 1990; Dalrymple, 1992). On the other hand, elongate bars which occur in progradational, deltaic settings (e.g., on the bay-head delta of the Gironde estuary) or on the muddy, landward or seaward flanks of zig-zag shoal complexes will generate upward-coarsening successions (Tastet et al., 1986; Dalrymple, 1992; Harris et al., 1992). The nature of the deposits of delta-like bars depends strongly on the conditions prevailing in the depositional basin. If flow or wave action is weak and there is little reworking by the subordinate currents, such deltas may develop large-scale, angle-ofrepose cross beds, perhaps with moderate to intense bioturbation. As the strength of the subordinate current or the intensity of reworking by waves or over-taking dunes increases, the angle of the depositional surface will decrease, just as it does in dunes. Forward-accreting master bedding planes are likely, but if the delta builds into the side of an active channel, the deltaic deposits will instead resemble lateral-accretion sets. In summary, most estuarine barforms will have a compound internal structure, with inclined master bedding planes separated by small-scale cross bedding. Tidal indicators should be relatively abundant. The distinction between the deposits of bars and large to very large dunes (Fig. 13-26) may not be easy. Forward-accreting compound sets are probably more common in dunes than in barforms, whereas laterally-accreting compound sets are more likely to originate in bars, but exceptions to both generalizations are possible. Careful reconstruction of the full three-dimensional geometry of the body (e.g., Allen, 1983) is perhaps the surest way to distinguish between dunes and barforms, but the ability to do this will probably be limited in most ancient successions.
415
ESTUARINE DUNES AND BARS
N
S
0
ms 15
30
A
0.4
0.0
0.9
1.4
1.6
krn
N
0 s
o-
ma -
ms
15-
15
30-
30
0
ms
I
1
I
I
II
15
30H I Tr
I lTrlt
Fig. 13-31. Original and interpreted seismic sections showing large-scale structures within elongate tidal bars in Cobequid Bay, Bay of Fundy. From the top down the stratigraphic units are: modern, macrotidal, sand-bar complex (no pattern); Pleistocene till and outwash and mid-Holocene, barrier-lagoon deposits (stipple); and Triassic bedrock (Tr). See Dalrymple and Zaitlin (1994) for additional details. rn = multiple. All travel times converted to depths using a speed of sound of 1500 m/s. (A) Section showing inclined reflectors within a bar. Note that many of the reflectors terminate in the trough a large, compound dune, indicating that the dunes are climbing over each other as they migrate toward the bar crest. Because the dunes are migrating obliquely out of the line of section, the reflectors define lateral-accretion bedding that accumulates on the up-current side of the bar (relative to the dominant current). (B) Section showing one active and one infilled swatchway channel, both of which deposited units of lateral-accretion bedding. Beneath the swatchways, the bar contains lateral-accretion deposits like those in (A).
416
R.W. DALRYMPLE AND R.N. RHODES
SUMMARY AND RESEARCH NEEDS
Dunes and barforms are widely developed in estuaries, and a significant body of literature exists on their distribution, morphology, and response to changing conditions. Indeed, it is likely that more is known about estuarine dunes than about dunes in any other environment, although estuarine barforms are more poorly understood than their fluvial counterparts. Despite this wealth of information, many studies continue to have the air of the proverbial blind man describing an elephant; all too often, individual case studies are reported or later interpreted as being representative of the entire spectrum of dunes (or barforms), or as refuting other work which reported slightly different results. Perhaps the present authors are also guilty of this error. We have tried to show, however, that dunes in particular are a result of, and respond to, a complex interplay of many variables, some of which are not always obvious in a single survey (e.g., the history of the feature or sediment binding by organisms). Thus, dune size (height and wavelength) is influenced by water depth (or boundary-layer thickness), current speed, grain size, sediment availability, water temperature, and the time history of all of these factors. Dune shape in profile is controlled by a similar set of variables, as well as by the relative mobility of the sediment in the dune trough. Dune shape in plan is also determined by water depth, current speed, and grain size, and by spatial changes in these parameters because of their influence on crestal branching. Dune orientation is controlled by the directional variability of sediment transport and cross-flow variations in bedform migration rate. Dune migration rate is in turn determined in the first instance by the sediment discharge and dune height, which are themselves influenced by water depth, current speed, and grain size. The presence of superimposed dunes is a result either of relatively slow variations in flow conditions (disequilibrium superposition) or the development of an internal boundary layer (equilibrium superposition), while the relative orientation of the smaller forms is a response to near-bed flow perturbations by the large dune or the obliquity of the main form. The response to unsteady flow conditions is determined by dune size, the sediment-transport rate, and the magnitude and rate-of-change of flow conditions. Finally, the internal structure of estuarine (tidal) dunes is dependent on the morphology and migration history of the dune’s lee face. The slope of the lee face is in turn influenced by many factors, including the presence or absence of superimposed dunes, the strength of the subordinate current, wave action, and the nature of lee-face flow, which is sensitive to the presence/absence of superimposed dunes and the obliquity of the main dune. As a result of these complex, multivariate controls, dunes show a wide range of morphology, and their distribution and response to changing conditions do not always show a clear, one-to-one relationship with any single variable such as water depth or current speed. Thus, a similar change (in time or space) of one parameter in two separate areas might produce two different responses because other factors are not equal. Despite the wealth of information on dunes, many important aspects of their morphology remain poorly documented. This is particularly the case with respect to
ESTUARINE DUNES AND BARS
417
plan shape. For example, the 2D-3D transition is fundamental to the classification of dunes, to our understanding of their origin, and to the geometry of the resulting deposits, yet no systematic effort has been undertaken to record the nature of the change or to quantify the boundary between 2D and 3D dunes. More work on this is needed. Similarly, the possibility of systematic differences between fluvial and tidal dunes has been noted with regard to the degree of asymmetry and lateral continuity, but systematic comparative data appear to be lacking. Dune orientation relative to the sediment-transport direction(s) is also not well documented in many areas; the assumption of a perfectly flow-transverse orientation would appear to be dangerous, given the widespread occurrence of conditions that might produce oblique dunes. The internal structures of dunes are moderately well understood in a general fashion, but considerable uncertainty remains as to why larger dunes have low-angle lee faces. More work is needed on the nature of lee-face flow, to determine to what extent bedform obliquity and the surface roughness produced by superimposed dunes influence lee-face morphology and structure. Estuarine barforms are understudied by comparison with dunes, as indicated by the absence of an adequate classification system, our limited understanding of bar genesis, our lack of information on their internal architecture, and even our inability to distinguish clearly between bars and dunes. The impetus to increase our understanding of estuarine barforms is great, given their prominent role in determining the geomorphology and sedimentology of estuaries.
ACKNOWLEDGEMENTS
The authors’ work on bedforms and bars in the Bay of Fundy has been supported financially by grants to RWD from the Natural Sciences and Engineering Research Council of Canada (Operating Grant 7553) and the Advisory Research Committee of Queen’s University, and to RNR from the Geological Society of America and the American Association of Petroleum Geologists. This assistance is gratefully acknowledged. RWD also extends sincere thanks to Dr. Peter Harris, the Ocean Sciences Institute, and the Department of Geology and Geophysics, University of Sydney, for their hospitality and logistical support during the writing of this chapter. Constructive reviews of this chapter were provided by Bill Amott, Peter Harris, Gerard0 Perillo, and Don Swift.
REFERENCES Aliotta, S. and Perillo, G.M.E., 1987. A sand wave field in the entrance to Bahia Blanca estuary, Argentina. Mar. Geol., 76: 1-14. Allen, G., Deresseguier, A. and Klingebiel, A,, 1969. Evolution des structures sbdimentaires sur un banc sableux d’estuaire en fonction de l’amplitude des marCes. C. R. Acad. Sci. Paris, Ser. D, 269: 2167-2169. Allen, G.P., Salmon, J.C., Bassoulet, P., Du Penhoat, Y. and De Grandprb, C., 1980. Effects of tides on mixing and suspended sediment transport in macrotidal estuaries. Sediment. Geol., 26: 69-90.
418
R.W. DALRYMPLE AND R.N. RHODES
Allen, J.R.L., 1968. Current Ripples: Their Relation to Patterns of Water and Sediment Motion. North-Holland Publ. Co., Amsterdam, 433 pp. Allen, J.R.L., 1973a. Features of cross-stratification due to random and other changes in bed forms. Sedimentology, 20: 189-202. Allen, J.R.L., 1973b. Phase differences between bed configuration and flow in natural environments, and their geological relevance. Sedimentology, 20: 323-329. Allen, J.R.L., 1974. Reaction, relaxation and lag in natural environments: general principles, examples and lessons. Earth-Sci. Rev., 10: 263-342. Allen, J.R.L., 1976a. Conceptual models for dune time-lag: general ideas, difficulties, and early results. Sediment. Geol., 15: 1-53. Allen, J.R.L., 1976b. Time-lag of dunes in unsteady flows: an analysis of Nasner’s data from the R. Weser, Germany. Sediment. Geol., 15: 309-321. Allen, J.R.L., 1976c. Conceptual models for dune time-lag: population structures and the effects of discharge pattern and coefficient of change. Sediment. Geol., 16: 99-130. Allen, J.R.L., 1976d. Conceptual models for dune time-lag: an alternative boundary condition. Sediment. Geol., 16: 255-279. Allen, J.R.L., 1977. The plan shape of current ripples in relation to flow conditions. Sedimentology, 24: 53-62. Allen, J.R.L., 1978. Polymodal dune assemblages: an interpretation in terms of dune creationdestruction in periodic flows. Sediment. Geol., 20: 17-28. Allen, J.R.L., 1980. Sandwaves: a model of origin and internal structure. Sediment. Geol., 26: 281-328. Allen, J.R.L., 1982. Sedimentary Structures - Their Characteristics and Physical Basis, Vol. 1. Developments in Sedimentology, Vol. 30A. Elsevier, New York, 593 pp. Allen J.R.L., 1983. Studies in fluviatile sedimentation: bars, bar complexes and sandstone sheets (lowsinuosity braided streams) in the Brownstones (L. Devonian), Welsh Borders. Sediment. Geol., 33: 237-293. Allen, J.R.L. and Collinson, J.D., 1974. The superposition and classification of dunes formed by unidirectional flow. Sediment. Geol., 12: 169-178. Allen, J.R.L. and Friend, P.F., 1976a. Changes in intertidal dunes during two spring-neap cycles, Lifeboat Station Bank, Wells-Next-the-Sea, Norfolk (England). Sedimentology, 23: 329-346. Allen, J.R.L. and Friend, P.F., 1976b. Relaxation time of dunes in decelerating aqueous flows. J. Geol. SOC.London, 132: 17-26. Allen J.R.L. and Leeder, M.R., 1980. Criteria for the instability of upper stage plane beds. Sedimentology, 27: 209-217. Arbouille, D., Cuignon, R. and Gayet, J., 1986. Dynamique des barres sableuses intertidales en domaine lagunaire a influence fluviale: delta de la Leyre. Bull. Inst. Gtol. Bassin d’Aquitaine, 39: 59-72. Ashley, G.M., 1975. Bedforms in the Pitt River, British Columbia. In: A.D. Miall (Editor), Fluvial Sedimentology. Can. SOC.Pet. Geol. Mem., 5: 89-104. Ashley, G.M., 1990. Classification of large-scale subaqueous bedforms: a new look at an old problem. SEPM Bedforms and Bedding Structures Research Symposium. J. Sediment. Petrol., 60: 160-172. Ashley, G.M. and Zeff, M.L., 1988. Tidal channel classification for a low-mesotidal salt marsh. Mar. Geol., 82: 17-32. Banvis, J.H., 1978. Sedimentology of some South Carolina tidal-creek point bars, and a comparison with their fluvial counterparts. In: A.D. Miall (Editor), Fluvial Sedimentology. Can. SOC.Petrol. Geol. Mem., 5: 129-160. Belderson, R.H., Johnson, M.A. and Kenyon, N.H., 1982. Bedforms. In: A.H. Stride (Editor), Offshore Tidal Sands: Processes and Deposits. Chapman and Hall, New York, pp. 27-57. BernC, S., Auffret, J.P. and Walker, P., 1988. Internal structure of subtidal sand waves revealed by high-resolution seismic reflection. Sedimentology, 35: 5-20. Bernt, S., Castaing, P., LeDrezen, E. and Lericolais, G., 1993. Morphology, internal structure, and reversal of asymmetry of large subtidal dunes in the entrance to Gironde Estuary (France). J. Sediment. Petrol., 63: 780-793. Best, J.L. and Leeder, M.R., 1993. Drag reduction in turbulent muddy seawater flows and some sedimentary consequences. Sedimentology, 40: 1129-1 137.
ESTUARINE DUNES AND BARS
419
Boersma, J.R., 1969, Internal structure of some tidal mega-ripples on a shoal in the Westerschelde estuary, The Netherlands. Report of a preliminary investigation. Geol. Mijnb., 48: 409-414. Boersma, J.R. and Terwindt, J.H.J., 1981. Neap-spring tide sequences of intertidal shoal deposits in a mesotidal estuary. Sedimentology, 28: 151-170. Bohacs, K.M., 1981. Flume studies of the kinematics and dynamics of large-scale bedforms. Ph.D. thesis, Mass. Inst. Tech., Cambridge, MA, 178 pp. (unpublished). Bokuniewicz, H.J., Gordon, R.B. and Kastens, K.A., 1977. Form and migration of sand waves in a large estuary, Long Island Sound. Mar. Geol., 24: 185-199. Boothroyd, J.C., 1985. Tidal inlets and tidal deltas. In: R.A. Davis, Jr. (Editor), Coastal Sedimentary Environments. Springer-Verlag, New York, pp. 445-532. Boothroyd, J.C. and Hubbard, D.K., 1975. Genesis of bedforms in mesotidal estuaries. In: L.E. Cronin (Editor), Estuarine Research, Vol. 2, Geology and Engineering. Academic Press, New York, pp. 217-234. Bouma, A.H., Rappeport, M.L., Orlando, R.C. and Hampton, M.A., 1980. Identification of bedforms in lower Cook Inlet, Alaska. Sediment. Geol., 26: 157-177. Cant, D.J. and Walker, R.G., 1978. Fluvial processes and facies sequences in the sandy braided South Saskatchewan River, Canada. Sedimentology, 25: 625-648. Costello, W.R., 1974. Development of bed configurations in coarse sands. Mass. Inst. Technol., Dept. Earth Planet. Sci., Rep. 74-1, 120 pp. Costello, W.R. and Southard, J.B, 1981. Flume experiments on lower-flow-regime bed forms in coarse sand. J. Sediment. Petrol., 51: 849-864. Dalrymple, R.W., 1984. Morphology and internal structure of sandwaves in the Bay of Fundy. Sedimentology, 31: 365-382. Dalrymple, R.W., 1992. Tidal depositional systems. In: R.G. Walker and N.P. James (Editors), Facies Models - Response to Sea-Level Change. Geol. Assoc. Canada, St. John’s, Nfld., pp. 195-218. Dalrymple, R.W., Knight, R.J. and Lambiase, J.J., 1978. Bedforms and their hydraulic stability relationships in a tidal environment, Bay of Fundy, Canada. Nature, 275: 100-104. Dalrymple, R.W., Knight, R.J., Zaitlin, B.A. and Middleton, G.V., 1990. Dynamics and facies model of a macrotidal sand-bar complex, Cobequid Bay-Salmon River estuary (Bay of Fundy). Sedimentology, 37: 577-612. Dalrymple, R.W., Makino, Y. and Zaitlin, B.A., 1991. Temporal and spatial patterns of rhythmite deposition on mudflats in the macrotidal Cobequid Bay-Salmon River estuary, Bay of Fundy, Canada. In: D.G. Smith, G.E. Reinson, B.A. Zaitlin and R.A. Rahmani (Editors), Clastic Tidal Sedimentology. Can. SOC.Petrol. Geol. Mem., 16: 137-160. Dalrymple, R.W. and Zaitlin, B.A., 1994. High-resolution sequence stratigraphy of a complex, incisedvalley succession, Cobequid Bay-Salmon River estuary, Bay of Fundy, Canada. Sedimentology, 41: 1069-1091. Dalrymple, R.W., Zaitlin, B.A. and Boyd, R., 1992. Estuarine facies models: conceptual basis and stratigraphic implications. J. Sediment. Petrol., 62: 1130-1 146. d’Anglejan, B.F., 1971, Submarine sand dunes in the St. Lawrence estuary. Can. J. Earth Sci., 8: 1480-1486. Davis, R.A., Jr. and Flernming, B.W., 1991. Time-series study of mesoscale tidal bedforms, Martens Plate, Wadden Sea, Germany. In: D.G. Smith, G.E. Reinson, B.A. Zaitlin and R.A. Rahmani (Editors), Clastic Tidal Sedimentology. Can. SOC.Pet. Geol. Mem., 16: 275-282. De Boer, P.L., Oost, A.P. and Visser, M.J., 1989. The diurnal inequality of the tide as a parameter for recognizing tidal influences. J. Sediment. Petrol., 59: 912-921. de Mowbray, 7: and Visser, M.J., 1984. Reactivation surfaces in subtidal channel deposits, Oosterschelde, southwest Netherlands. J. Sediment. Petrol., 5 4 811-824. Dyer, K.R., 1986. Coastal and Estuarine Sediment Dynamics. Wiley and Sons, New York, 342 pp. Elliott, T and Gardner, A.R., 1981. Ripple, megaripple and sandwave bedforms in the macrotidal Loughor estuary, south Wales, U.K. In: S.-D. Nio, R.T.E. Shuttenhelm and Tj.C.E. van Weering (Editors), Holocene Marine Sedimentation in the North Sea Basin. Inter. Assoc. Sediment. Spec. Publ., 5: 51-64.
420
R.W. DALRYMPLE AND R.N. RHODES
Fenster, M.S., FitzGerald, D.M., Bohlen, W.F., Lewis, R.S. and Baldwin, C.T, 1990. Stability of giant sand waves in eastern Long Island Sound, U S A . Mar. Geol., 91: 207-225. Flemming, B.W., 1988. Zur klassifikation subaquatischer, stromungstransversaler Transportkorper. Bochum Geol. U. Geotechn. Arb., 29: 44-47. Fredsoe, J., 1983. Shape and dimensions of ripples and dunes. In: B.M. Sumer and A. Muller (Editors), Mechanics of Sediment Transport. Euromech 156, pp. 93-97. Greer, S.A., 1975. Estuaries of the Georgia coast, U.S.A.: sedimentology and biology. 111. Sandbody geometry and sedimentary facies at the estuary-marine transition zone, Ossabaw Sound, Georgia: a stratigraphic model. Senckenberg. Marit., 7: 105-135. Guy, H.P., Simons, D.B. and Richardson, E.V., 1966. Summary of alluvial channel data from flume experiments, 1956-1961. U S . Geol. Surv. Prof. Pap., 462-1,96 pp. Hand, B.M. and Bartberger, C.E., 1988. Leeside sediment fallout patterns and the stability of angular bedforms. J. Sediment. Petrol., 58: 33-42. Harms, J.C., Southard, J.B. and Walker, R.G., 1982. Structures and sequences in clastic rocks. SEPM Short Course Notes, 9, 253 pp. Harris, P.T., 1988. Large-scale bedforms as indicators of mutually evasive sand transport and the sequential infilling of wide-mouthed estuaries. Sediment. Geol., 57: 273-298. Harris, P.T. and Jones, M.R., 1988. Bedform movement in a marine delta: air photo interpretations. Geol. Mag., 125: 31-49. Harris, P.T, Pattiaratchi, C.B., Cole, A.R. and Keene, J.B., 1992. Evolution of subtidal sandbanks in Moreton Bay, eastern Australia. Mar. Geol., 103: 225-247. Hayes, M.O., 1975. Morphology of sand accumulations in estuaries. In: L.E. Cronin (Editor), Estuarine Research, Vol. 2, Geology and Engineering. Academic Press, New York, pp. 3-22. Hine, A.C., 1975. Bedform distribution and migration patterns on tidal deltas in the Chatham Harbor estuary, Cape Cod, Massachusetts. In: L.E. Cronin (Editor), Estuarine Research, Vol. 2, Geology and Engineering. Academic Press, New York, pp. 235-252. Hulscher, S.J.M.H., De Swart, H.E. and De Vriend, H.J., 1993. The generation of offshore tidal sand banks and sand waves. Continent. Shelf Res., 13: 1183-1204. Huthnance, J.M., 1982a. On one mechanism forming linear sand banks. Est. Coastal Shelf Sci., 14: 79-99. Huthnance, J.M., 1982b. On the formation of sand banks of finite extent. Est. Coastal Shelf Sci., 15: 277-299. Johns, B., Soulsby, R.L. and Chester, TJ., 1990. The modelling of sandwave evolution resulting from suspended and bed load transport of sediment. J. Hydraul. Res., 28: 355-374. Klein, G.deV., 1970. Depositional and dispersal dynamics of intertidal sand bars. J. Sediment. Petrol., 40: 1127-1146. Knebel, H.J., 1989. Modern sedimentary environments in a large tidal estuary, Delaware Bay. Mar. Geol., 86: 119-136. Knight, R.J., 1980. Linear sand bar development and tidal current flow in Cobequid Bay, Bay of Fundy, Nova Scotia. In: S.B. McCann (Editor), The Coastline of Canada. Geol. Sun. Can. Pap., 80-10: 123-152. Kohsiek, L.H.M. and Tenvindt, J.H.J., 1981. Characteristics of foreset and topset bedding in megaripples related to hydrodynamic conditions on an intertidal shoal. In: S.-D. Nio, R.T.E. Shuttenhelm and Tj.C.E. van Weering (Editors), Holocene Marine Sedimentation in the North Sea Basin. Int. Assoc. Sediment. Spec. Publ., 5: 27-37. Kostaschuk, R.A., Church, M.A. and Luternauer, J.L., 1989. Bedforms, bed material, and bedload transport in a salt-wedge estuary: Fraser River, British Columbia. Can. J. Earth Sci., 26: 1440-1452. Kreisa, R.D. and Moiola, R.J., 1986. Sigmoidal tidal bundles and other tide-generated sedimentary structures of the Curtis Formation, Utah. Geol. SOC. Am. Bull., 97: 381-387. Langhorne, D.N., 1973. A sandwave field in the outer Thames estuary, Great Britain. Mar. Geol., 14: 129- 143. Langhorne, D.N., 1977. Consideration of meteorological conditions when determining the navigational water depth over a sandwave field. Internat. Hydrogr. Rev., 54: 17-30.
ESTUARINE DUNES AND BARS
421
Ludwick, J.C., 1972. Migration of tidal sand waves in Chesapeake Bay entrance. In: D.J.P. Swift, D.B. Duane and O.H. Pilkey (Editors), Shelf Sediment Transport: Process and Pattern. Dowden, Hutchinson and Ross, Stroudsburg, PA, pp. 377-410. Ludwick, J.C., 1974. Tidal currents and zig-zag sand shoals in a wide estuary entrance. Geol. Soc. Am. Bull., 85: 717-726. Malikides, M., Harris, P.T. and Tate, P.M., 1989. Sediment transport and flow over sandwaves in a non-rectilinear tidal environment. Cont. Shelf Res., 9: 203-221. McCave, I.N., 1971. Sand waves in the North Sea off the coast of Holland. Mar. Geol., 10: 199-225. McCutcheon, S., 1981. Vertical velocity profiles in stratified flows. J. Hydraulics Div., Am. SOC.Civil Eng., 107(HY8): 973-988. Middleton, G.V. and Southard, J.B., 1984. Mechanics of Sediment Movement, 2nd edition. SEPM Short Course Notes, 3, 401 pp. Nichols, M.M. and Biggs, R.B., 1985. Estuaries. In: R.A. Davis, Jr. (Editor), Coastal Sedimentary Environments, 2nd edition. Springer-Verlag, New York, pp. 77-186. Nio, S.-D. and Yang, C.S., 1991. Diagnostic attributes of clastic deposits: a review. In: D.G. Smith, G.E. Reinson, B.A. Zaitlin and R.A. Rahmani (Editors), Clastic Tidal Sedimentology. Can. SOC.Pet. Geol. Mem., 16: 3-28. Nordin, C.F., 1971. Statistical properties of dune profiles. U S . Geol. Surv. Prof. Pap., 562-6 41 pp. Perillo, G.M.E. and Ludwick, J.C., 1984. Geomorphology of a sand wave in lower Chesapeake Bay, Virginia, U.S.A.. Geo-Mar. Lett., 4: 105-112. Reineck, H.-E., 1963. Sedimentgefuge im Bereich der sddlichen Nordsee. Abh. Senckenberg. Naturf. Gesell., 505: 1-38. Rhodes, R.N., 1992. Hydrodynamics and the morphology, migration and structure of subaqueous dunes, Minas Basin, Canada. M.Sc. thesis, Queen’s Univ., Kingston, Ont., 250 pp. (unpublished). Richards, K.J. and Taylor, P.A., 1981. A numerical model of flow over sand waves in water of finite depth. Geophys. J. R. Astron. SOC.,65: 247-267. Robinson, A.H.W., 1960. Ebb-flood channel systems in sandy bays and estuaries. Geography, 45: 183-199. Roy, PS., Thom, B.G. and Wright, L.D., 1980. Holocene sequences on an embayed high energy coast: an evolutionary model. Sediment. Geol., 26: 1-19. Rubin, D.M., 1987a. Formation of scalloped crossbedding without unsteady flows. J. Sediment. Petrol., 57: 39-45. Rubin, D.M., 1987b. Cross-bedding, Bedforms, and Paleocurrents. SEPM Concepts in Sedimentology and Paleontology, 1, 187 pp. Rubin, D.M. and Hunter, R.E., 1982. Bedform climbing in theory and nature. Sedimentology, 29: 121-138. Rubin, D.M. and Hunter, R.E., 1983. Reconstructing bedform assemblages from compound crossbedding. In: M.E. Brookfield and TS. Ahlbrandt (Editors), Eolian Sediments and Processes. Elsevier, Amsterdam, pp. 407-427. Rubin, D.M. and Hunter, R.E., 1987. Bedform alignment in directionally varying flows. Science, 237: 276-278. Rubin, D.M. and Ikeda, H., 1990. Flume experiments on the alignment of transverse, oblique, and longitudinal dunes in directionally varying flows. Sedimentology, 37: 673-684. Rubin, D.M. and McCulloch, D.S., 1980. Single and superimposed bedforms: a synthesis of San Francisco Bay and flume observations. Sediment. Geol., 26: 207-231. Salsman, G.G., Tolbert, W.H. and Villars, R.G., 1966. Sand ridge migration in St. Andrews Bay, Florida. Mar. Geol., 4: 11-19. Schlichting, H., 1968. Boundary-layer Theory, 6th edition. McGraw-Hill, New York, 748 pp. Sheng, Y.P. and Villaret, C., 1989. Modeling the effect of suspended sediment stratification on bottom exchange processes. J. Geophys. Res., 94C: 14429-14444. Smith. J.D., 1970. Stability of a sand bed subjected to a shear flow of low Froude number. J. Geophys. Res., 74: 5928-5939. Smith, J.D. and McLean, S.R., 1977. Spatially averaged flow over a wavy surface. J. Geophys. Res., 82: 1735-1746.
422
R.W. DALRYMPLE AND R.N. RHODES
Smith, N.D., 1975. Some comments on terminology for bars in shallow rivers. In: A.D. Miall (Editor), Fluvial Sedimentology. Can. SOC.Pet. Geol. Mem., 5: 85-88. Southard, J.B. and Boguchwal, L.A., 1990a. Bed configurations in steady unidirectional water flows. Part 1. Scale model study using fine sand. J. Sediment. Petrol., 60: 649-657. Southard, J.B. and Boguchwal, L.A., 1990b. Bed configurations in steady unidirectional water flows. Part 2. Synthesis of flume data. J. Sediment. Petrol., 60: 658-679. Southard, J.B. and Boguchwal, L.A., 1990c. Bed configurations in steady unidirectional water flows. Part 3. Effects of temperature and gravity. J. Sediment. Petrol., 60: 680-686. Sweet, M.L. and Kocurek, G., 1090. An empirical model of aeolian dune lee-face airflow. SedimentolO ~ Y 37: , 1023-1038. Swift, D.J.P. and Ludwick, J.C., 1976. Substrate response to hydraulic process: grain-size frequency distributions and bed forms. 1n:D.J. Stanley and D.J.P. Swift (Editors), Marine Sediment Transport and Environmental Management. Wiley and Sons, New York, pp. 159-196. Tastet, J.-P., Fenies, H. and Allen, G.P., 1986. Facies, stquences et gtometrie d’une barre tidale estuarienne: le banc de Trempeloup dans l’estuaire de la Gironde. Bull. Inst. Gtol. Bassin d’Aquitaine, 39: 165-184. Tenvindt, J.H.J., 1970. Observation on submerged sand ripples with heights ranging from 30 to 200 cm occurring in tidal channels of S.W. Netherlands. Geol. Mijnb., 49: 489-501. Tenvindt, J.H.J., 1971. Sand waves in the southern North Sea. Mar. Geol., 10: 51-67. Tenvindt, J.H.J., 1981. Origin and sequences of sedimentary structures in inshore mesotidal deposits of the North Sea. In: S.-D. Nio, R.T.E. Shuttenhelm and Tj.C.E. van Weering (Editors), Holocene Marine Sedimentation in the North Sea Basin. Int. Assoc. Sediment. Spec. Publ., 5: 4-26. Tenvindt, J.H.J. and Brouwer, M.J.N., 1986. The behaviour of intertidal sandwaves during neap-spring tide cycles and the relevance to paleoflow reconstructions. Sedimentology, 33: 1-31. Thomas, R.G., Smith, D.G., Wood, J.M., Visser, J., Calverly-Range, E.A. and Koster, E.H., 1987. Inclined heterolithic stratification - terminology, description, interpretation and significance. Sediment. Geol., 53: 123-179. van Den Berg, J.H., 1987. Bedform migration and bed-load transport in some rivers and tidal environments. Sedimentology, 34: 681-698. van Rijn, L.C., 1983. The prediction of bedforms and alluvial roughness. In: B.M. Sumer and A. Miiller (Editors), Mechanics of Sediment Transport. Euromech 156, pp. 133-135. van Rijn, L.C., 1984. Sediment transport, Part 111: bed forms and alluvial roughness. J . Hydraul. Eng., 110: 1733-1754. van Straaten, L.M.J.U., 1953. Megaripples in the Dutch Wadden Sea and in the basin of Arcachon (France). Geol. Mijnb., N.S., 1: 1-11. van Veen, J., 1938. Die Unterseeische Sandwiiste in der Nordsee. Geol. Meere Binnegewasser, 2: 62-86. Visher, G.S. and Howard, J.D., 1974. Dynamic relationship between hydraulics and sedimentation in the Altamaha estuary. J. Sediment. Petrol., 44: 502-521. Visser, M.J., 1980. Neap-spring cycles reflected in Holocene subtidal large-scale bedform deposits: a preliminary note. Geology, 8: 543-546. Wright, L.D., Coleman, J.M. and Thom, B.G., 1973. Processes of channel development in a hightide-range environment: Cambridge Gulf-Ord River delta, Western Australia. J. Geol., 81: 1541. Yalin, M.S., 1964. Geometrical properties of sand waves. Proc. Am. SOC.Civil Eng., 90: 105-119. Yalin, M.S., 1977. Mechanics of Sediment Transport, 2nd edition. Pergamon Press, Toronto, 298 pp. Yalin, M.S., 1987. On the formation mechanism of dunes and ripples. Euromech Colloq. Proc., 261. Yalin, M.S. and da Silva, A.M.F., 1991. On the formation of alternate bars. In: R. Soulsby and R. Bettess (Editors), Sand Transport in Rivers, Estuaries and the Sea. Euromech Colloq. Proc., 262: 171-178. Yalin, M.S. and da Silva, A.M.F., 1992. Horizontal turbulence and alternate bars. J. Hydrosci. Hydraul. Eng., 9: 47-58. Zaitlin, B.A., 1987. Sedimentology of the Cobequid Bay-Salmon River estuary, Bay of Fundy, Canada. Ph.D. thesis, Queen’s Univ., Kingston, Ont., 391 pp. (unpublished). Zarillo, G.A., 1985. Tidal dynamics and substrate response in a salt marsh estuary. Mar. Geol., 67: 13-35.
Geomorphology and Sedimentololy of Estuuries. Developments in Sedimentology 53 edited by G.M.E. PeriIlo 0 1995 Elsevier Science B.V. All rights reserved.
423
Chapter 14
SEDIMENT TRANSPORTPROCESSES IN ESTUARIES KEITH R. DYER
INTRODUCTION
Estuaries are the route by which sediment is transported from the rivers to the sea. On the way down the rivers the grain size distribution of the sediment becomes altered by continual deposition, re-erosion and transport. Much of the coarser sediment can become trapped on the flood plains of the rivers, only being released at times of flood. The finer fractions are transported into the estuary. There the estuarine processes act as a filter on the sediment input, and mixing can take place with sediment brought in from the sea. Additionally, chemical alterations can occur within the estuary that can cause the surface properties of some of the constituent particles to alter, affecting their pollutant scavenging potential, and their potential deposition. Sediments form a crucial control in many estuarine processes. Within the estuary suspended sediment concentrations are generally high, the particles are fine, cohesive, and prone to flocculate, and they are richly organic. The energy cycles inherent in the semidiurnal and lunar tide and in seasonal changes cause continual erosion, transport and deposition. Thus, even when little new sediment is coming in, sediments can silt up harbours and navigational channels. Since the sediments are fine grained clay minerals, pollutants are absorbed on their surfaces and are transported with the sediment particles. Consequently the transport and dispersion of contaminants can only be understood through a knowledge of the movement of particles. When the suspended sediment concentrations are sufficiently high, light penetration and productivity can be limited. The muddy substrates can be host to a diverse and vigorous biological community, but this can be limited by the presence of layers of high concentration suspended sediment with low oxygen content intermittently present above the bed. Estuaries in general are shallow, and sea level undergoes very drastic changes on the geological timescale. Thus, they are ephemeral features being fairly rapidly altered and destroyed, having an average life of probably only thousands of years. It is likely that the world is rich in estuaries at the present time because of the sea level rise of the Flandrian transgression. This inundated the valleys cut when the rivers incised to a base level, which reached a minimum level of about -100 m during the closing stages of the Pleistocene Ice Age. The variation in form of the resulting estuaries depends on the volumes of sediment that the rivers have contributed to fill the valleys, as well as that brought in from the sea. Deltas form where river flow and sediment discharge is high; the valleys have become completely filled and the sediment discharges directly into the sea. They are normally present in areas of high
424
K.R. DYER
seasonal discharge, particularly in the tropics, in monsoon areas, and those with a high component of snow melt discharge. Generally deltas are best developed in areas where the tidal range is small and where the currents cannot easily redistribute the sediment the rivers introduce. Where sediment discharge is less, the estuaries are unfilled. These drowned river valleys, or coastal plain estuaries, still retain the main features of river valleys, having a meandering outline with frequent tributaries and a triangular cross-section. In glaciated areas the river valleys were over-deepened by glaciers. A characteristic of the resulting fjords is a rock bar or sill at their mouths that can be as little as a few tens of meters deep. Inside, however, they can be several hundred meters deep and extend hundreds of kilometres inland. The sill restricts the water circulation and isolates the interior deep basin, with obvious consequences to the sedimentation patterns. On low coastlines, extensive shallow lagoons are often formed between the rivers and the sea. In tropical areas the lagoons can be hypersaline because of evaporation during the hot season, but almost entirely fresh water during the rainy season. During low river flow periods the mouth may even be closed by littoral drift. The mouths of these features are often called inlets. With such a variety of estuaries it is to be expected that there will be a diverse and complex series of processes dominating the transport and deposition of the incoming sediment.
TIDAL EFFECTS
The tidal elevation characteristics of estuaries create important distinctions in the capability of the currents to move the sediments. Davies (1964) has classified estuaries as microtidal, where tides are less than 2 m range; mesotidal, between 2 m and 4 m range; macrotidal, greater than 4 m range, to which a fourth category can be added; hypertidal, greater than 6 m range. Within the estuary the tide can be greatly modified by the friction of the bed on the current, and by the funnelling effect of the convergence of the estuary sides. The convergence causes both a partial reflection of the wave, as well as squeezing it into a smaller cross-section, thereby increasing the height of the tidal wave. Friction increases with decreasing water depth, and with increasing velocity, thereby taking energy from the tide and decreasing its amplitude. Where the convergence effect exceeds the frictional effect, the tidal range increases towards the head of the estuary, before decreasing in the riverine section (Fig. 14-1). This response is termed hypersynchronous. When convergence and friction are equal, the tidal range is constant throughout the estuary. This is the synchronous response. Hyposynchronous estuaries are those where friction dominates, and where the range of the tide diminishes throughout the estuary. When the tidal range is large relative to water depth, considerable asymmetry can occur in the tidal curve and in the velocities. On the flood tide the water is flowing into a decreasing cross section as it flows up the estuary. Friction slows down the
SEDIMENT TRANSPORT PROCESSES IN ESTUARIES
425
+ z w
a
a: 3
0 -J
a
5?
k
RIVER
ESTUARY
MOUTH
k
z w a a
B
3
0 -1
a 5? RIVER
ESTUARY
MOUTH
k
z
w
a: a: 3
0 A
-0c RIVER
ESTUARY
MOUTH
Fig. 14-1. Tidal response in estuaries. A) Hypersynchronous, B) synchronous, and C) hyposynchronous (after Nichols and Biggs, 1985).
early part of the flood tide more than later when the water depth is greater. If the convergence in cross section is rapid, then the water slope at the front of the wave increases, and the speed of propagation of the tidal wave increases. This leads to a faster progression of high water up the estuary than low water, and the time delay of high water between the mouth and the head of the estuary is smaller than that of low water. This causes the ebb tide to become longer towards the head of the estuary, and the flood tide shorter, resulting in an asymmetrical tidal current, with the flood being shorter and stronger than the ebb. This is particularly enhanced towards the head of the estuary where the estuary bed level is above the general low tide level. A flood dominant response is typical of hypersynchronous estuaries where the tidal range is large compared with the water depth (Fig. 14-2). When convergence is not important the tidal range diminishes landward (i.e., a hyposynchronous estuary), and less water flows landward through each section between the mouth and the head of the estuary on the flood tide. If the area of the intertidal increases headwards, then filling that volume of the tidal prism occurs when the cross section area through which the water is flowing is relatively large because of the flooding tide. On the ebb, however, the emptying occurs through a cross section with a smaller area. This leads to an ebb dominant response, which is typical of
426
K.R. DYER
0.1
0.2
0.3
0.4
0.5
0.6
WATER DEPTH TIDAL AMPLITUDE
Fig. 14-2. Ebb and flood dominance in estuaries as related to their volumetric and dimensional characteristics (after Friedrichs and Aubrey, 1988).
estuaries with a large intertidal volume to channel volume ratio, as well as a relatively small tidal range (Fig. 14-2). In terms of the normal tidal analysis procedure, the tidal distortion can be represented by a combination of the main semi-diurnal Mz component and the principal overtide M4, which has a quarter diurnal period. Depending on the phase relationship between these two constituents a variety of tidal curves can be produced, and each causes a similarly distorted tidal current pattern. For example, if the relative tidal phase is between 0" and 180" then the falling tide exceeds the rising tide in duration, and this produces a shorter enhanced flood current relative to the ebb current. The reverse effect occurs when the relative tidal phase is between 180" and 360". These features have been described in a number of papers recently reviewed by Friedrichs and Aubrey (1988) and related to the shape of the intertidal basin of the estuary (Dronkers, 1986a). A dynamic tidal equilibrium is conceptually possible between the tidal currents and the resulting sediment transport. This implies that the sediment movement causes a change in morphology that alters the tidal current regime which in turn reduces the sediment transport. The simplest equilibrium concept is that of O'Brien (1969). He found that the cross section of the mouth of an inlet A was related to the tidal prism volume P ; the volume of water that has to flow in and out through the mouth on each tide to raise the water level inside. He found A = c P n . Analysis ft-'. In actual fact the tidal prism of many inlets gave n 1 and c = 2 x volume is a volume per half tide, and the constant c then has the dimensions of an inverse velocity. This results in c being a half tidal mean velocity of 0.67 m s-', which is equivalent to the threshold of movement of sand. Thus the O'Brien relationship specifies that an increasing tidal prism leads to an increase in the velocity at the mouth of the inlet, which causes the sand to move, and the cross section to increase until the movement diminishes to the threshold value. This equilibrium tidal prism
-
427
SEDIMENT TRANSPORT PROCESSES IN ESTUARIES
HW SPRINGS HW NEAPS
L W NEAPS LW SPRINGS
0
50
100
150
200
250
300
D I S T A N C E krn
Fig. 14-3.Volumetric characteristics of the Severn Estuary.
concept has been widely applied in models of morphological development of tidal inlets and estuaries. Similarly, empirical studies have shown relationships for other variables such as the area of the tidal flats. Van Dongeren (1992) has shown for a number of North Sea inlets that the equilibrium tidal flat area Af is related to the total basin area Ab by Af = Ab - 0.025A2'2. It has also been found for many estuaries that the tidal prism and the low water volume vary exponentially with distance. Figure 14-3 shows the cross sectional area and volume variations along the Severn Estuary. Similarly, Wright et al. (1973) have shown the breadth and depth of the Ord Estuary in Australia varying in the same way. It is likely that an equilibrium estuary would have a phase relationship of either 180" or O", and be synchronous, though this has not been tested.
TYPES OF ESTUARY
Within estuaries the tidal and residual water circulation patterns are important in determining the overall sediment transport. The patterns of sediment movement are different in different types of estuary, and several examples are discussed in other chapters of this volume (e.g., see Bokuniewicz or Castaing and Guilcher, this volume).
428
K.R. DYER
Highly stratified estualy When there is little tidal motion the river flow, being less dense than the salt water, flows over it with a marked density interface between the two water masses. Because of the stable density gradient the two water masses will not mix readily together. However, if the surface layer has sufficiently high velocity the shear can create interfacial waves on the halocline. These waves break ejecting salt water into the fresher surface layer by a process called entrainment. No fresh water is mixed into the bottom layer and the mixing is entirely upwards. Thus, the bottom water loses salt gradually into the surface layer and this loss is made good by a slow inflow of salt water from the sea. The position of the salt wedge will vary with the river flow, and the tidal range is normally microtidal, ie, a range of less than 2 m.
Partially mixed estuaries With increased tidal range, the whole water mass in the estuary moves backwards and forwards with a tidal periodicity. The friction between the water and the bed of the estuary creates turbulence, which mixes the water column more effectively than entrainment so that the salinity difference at the interface is considerably reduced, and there is a smaller velocity shear across the interface. Turbulent mixing not only mixes the salt water into the fresher surface layer, but it also mixes the fresher water downwards. This causes a longitudinal gradient in salinity, with salinity diminishing towards the head of the estuary, both in the surface and the bottom layers. There has to be a residual discharge of water towards the sea, but it now carries with it the salt resulting from the vertical mixing. The discharge from the surface layer can thus be an order of magnitude larger than the river discharge. Because of the requirements of continuity this discharge has to be replaced by a significant landward flow within the bottom layer. Consequently, at the estuary mouth a large volume of mixed water has to be discharged and the compensating inflow in the bottom layer has to be larger than in salt wedge estuary. This process is termed vertical gravitational circulation. In partially mixed estuaries the tidal range is generally mesotidal, ie, between two and four metres. In this situation the tidal range can change significantly between spring and neap tides. The spring tide currents enhance the turbulent exchanges of salt and fresh water, and as a consequence the stratification can diminish considerably. This produces an increase in the vertical gravitational circulation. At times of high river flow the partially mixed estuary will become more highly stratified, and the intensity of the mean circulation should diminish. Within partially mixed estuaries there can be considerable variation in the vertical structure along the estuary, with highly stratified conditions at the head of the estuary where the water depth and the tidal range diminish, and river flow becomes comparatively more important.
Well mixed estuaries When the tidal range is large relative to the water depth, especially in macrotidal conditions, there is sufficient energy in the turbulence to completely mix the water
429
SEDIMENT TRANSPORT PROCESSES IN ESTUARIES
column and produce effectively vertically homogeneous conditions. In this type of estuary there can be lateral variations in salinity and in velocity, and horizontal circulation tends to develop at the expense of the vertical circulation. There can be residual flows inward on one side of the estuary and seaward on the other and separation in flood and ebb dominated channels.
MODES OF SEDIMENT TRANSPORT
The sediment transported down the rivers is generally a heterogeneous mixture reflecting the variety of source grain sizes available within the catchment. However, some sorting takes place within the flow as a consequence of the different modes of transport of the fine and coarser material. Three modes of transport occur; wash load, suspension and bedload. The distinction between them is not a clear one since there are changes in the grain content of the modes depending on velocity. There are differences, however, between the response of non-cohesive sands and silts, and cohesive silts and clays. Wash load comprises the finest fraction, and is normally composed of fine dispersed clay particles. They are kept in motion by turbulence, and move with the water at virtually all current speeds. The vertical profile of wash load concentration is homogeneous. Suspension occurs because of erosion of grains from the bed, and the exchange of momentum with the grains because of turbulence. The threshold for suspension 0.820,, of sandy grains from a flat bed can be considered to occur at about u* where u* is the friction velocity and w, is the settling velocity. Grains less than about 150 p m will go into suspension immediately they begin to move. For grains above about 150 p m movement as bedload occurs first, and suspension does not take place until higher velocities. As the intensity of the flow increases so does both the concentration of material in suspension and its mean grain size. The vertical profile of concentration becomes graded, with higher concentrations and a larger content of coarser grains nearer the bed. The threshold of bed load movement for a flat bed can be approximated by u,” 400, where u* is in cm s-l, and D is the grain diameter (mm). Initially the grains move by saltating along the bed, but ripple and dune bedforms are created with increasing flow velocity. Asymmetry in the bedforms indicates their movement downstream. For a review of sediment transport dynamics see Dyer (1986). Within the river, sediment moves intermittently with highest transport rates during high discharge events. At low flow rates, movement is restricted to the finer sizes. Once the sediment reaches the riverine section of the estuary, where there is a tidal rise and fall of the surface and where the current velocities become oscillatory during the tidal cycle, the bed shear stresses are reversed for part of the tide, thereby reducing the mean bedload transport. Additionally, the reversal of the current gives periods of slack water during which the material in suspension will settle to the bed, though it may be re-suspended on the next phase of the tide. Nevertheless, the suspended sediment will move in the direction of the tidally averaged water flow. The bedload,
-
-
430
K.R. DYER
however, will be affected mainly by the highest velocity, and will move in the direction of the maximum current. This is a very effective grain sorting mechanism, and fine and coarse grains can move in different directions. Additionally bedforms created in the sandy sediment at maximum current can be draped with mud at slack water. Consequently, the patterns and rates of sediment transport will vary between different types of estuary.
MUD PROPERTIES
Fine suspended particles in estuaries are particularly important because they become trapped by the estuarine processes, their concentrations can be high, and they form distinctive sediments. Their characteristics can be very variable with time, and this becomes important in their deposition, erosion and transport.
Flocculation Flocculation of particles occurs as a result of the total surface ionic charge on the particles and the enveloping electrical double layer, the properties of which depend on factors such as pH, and organic coatings. When the particles are in close proximity there is an overall attraction which leads to the formation of aggregates of particles, or flocs. The flocculation potential of particles increases with increased concentration, but is also enhanced by Brownian motion, differential settling, grain inertia and velocity shear (Krone, 1978; McCave, 1975). Of course not all collisions will lead to flocculation as some may lead to disruption of flocs, particularly at high turbulent shears. The predominance of the various flocculation processes will vary during the tide, with differential settling being most important at near slack water, and velocity shear at times of maximum current. The most effective interaction is between large and small flocs, so that flocculation progressively removes the finer particles from suspension (Kranck, 1973). Laboratory measurements (e.g., Owen, 1970) have indicated that salinity can be important in the flocculation of particles since it controls the intensity of their surface charges. This leads to the concept that flocculation of riverine particles occurs when they reached the salt water, and that deflocculation could occur when particles are recycled back into contact with fresh water. More recently it has become clear that in-situ particles are usually held together by organic matter (Eisma et al., 1991), and the deflocculation is unlikely by purely chemical processes. Riverine particles have variable characteristics depending on the cation content of the river water (Hunter and Liss, 1982), but it is still unclear how much the floc characteristics are determined by mineralogy. There are two modes which contribute to the distribution of floc size: macroflocs which reach a size of the order of millimetres and microflocs of the order ten to twenty microns. Macroflocs are about the same size as the turbulent Kolmogorov microscale, and they can be readily broken down to form microflocs. Microflocs are
SEDIMENT TRANSPORT PROCESSES IN ESTUARIES
43 1
very much more resistant to being broken up, and may form the basic unit from which flocculation takes place. There is, however, a linear relationship between the primary grain size and the microfloc size (Kranck, 1975). It is very difficult to observe the flocculation or break-up process within estuaries because of advection, and the effects of changing ccncentration. However, laboratory measurements have indicated that the modal floc size is affected by both concentration and shear stress, and the primary cause of floc disruption is by three-particle collisions (Burban et al., 1989). The general principle seems to be that at low concentrations the flocs are small, and low shear increases the likelihood of floc growth due to collision. However, with increasing shear the intensity of the collisions leads to floc breaking. At higher concentrations larger size flocs are present in quiescent settling, but only a small amount of shearing is necessary to disrupt them. The floc size distribution, in terms of sorting or standard deviation, is also likely to change with shear, with a narrow size distribution at low and at high shear. At moderate shear a wide range of floc sizes should be present during floc breakup. Thus, in addition to macroflocs and microflocs, a background of primary particles of size t 2 pm, which are involved in flocculation is likely. This would comprise the estuarine wash load.
Settling velocity The settling velocity of suspended material is an important parameter in determining the transport and deposition rates. For flocculated mud, the settling velocity is related to concentration. There are discrepancies between laboratory and field results largely because sampling disrupts the macroflocs. There are also considerable differences in the settling velocity/concentration relationship between estuaries which may be the result of floc density or organic content variations, and between different turbulent states during floc formation. Nevertheless at concentrations less than about ~ c”, 2-5000 mg 1-’ the settling velocity w sdepends on concentration according to w , cx where n varies 0.6 to 2.2 (Dyer, 1989). Above 10-20 g 1-’ the settling particles interfere with each other and “hindered” settling occurs (Fig. 14-4a). For modelling purposes n is generally taken as unity.
Deposition The product of the settling velocity and the concentration gives the settling flux towards the bed (Fig. 14-4b). This shows a maximum at a concentration of about 20 g 1-’. At any level in the flow at this concentration sediment would be settling from above faster than it is settling out beneath, and thus it builds up a layer on the bed. At this concentration the flocs are mainly separated by water. However, they collide with each other during shearing, a process that takes energy out of the flow. The suspension then becomes pseudoplastic in its rheological properties. At low shear stresses it will have a high viscosity, but as shear increases the flocs structure becomes broken down and the effective viscosity is reduced. When concentrations exceed 80-220 g 1-’
K.R. DYER
Y)
N
50
IE
m 40 _I
U
30
0
20
-I
f k W
10
1o2
1 o3
1 o4
v)
1 o5
CONCENTRATION c mgl-’
Fig. 14-4. Settling velocity and settling flux of estuarine flocs in relation to concentration.
patches occur where flocs are in contact and a framework structure gradually develops throughout the suspension,it becomes a soft bed and has the properties of a Bingham plastic (Sills and Elder, 1986). At this stage it will transmit a shear wave, and has a rigidity modulus. If left for time periods of hours the framework will gradually collapse owing to the weight of particles above, in a process known as consolidation. The pore water between the flocs is gradually forced out and the increasing particle contacts lead to an increase in the cohesive strength and density with depth into the sediment. Consequently with settling there is a gradual change from a dispersed suspension to the formation of a high concentration near bed layer, and to a soft bed which then consolidates. The high concentration layers have a lutocline at their upper surface, and this is often detectable by echo-sounders, at which stage the layers are called fluid mud. In some estuaries fluid muds have been detected as several meters thick in pond-like patches. It is obvious that these could not have formed solely by settling from the water column directly above. They must have largely been created by gravitational flow down the side slopes of the channels, even though the layers have been observed as standing on slight slopes. The erosion and re-entrainment of these fluid muds is not well quantified, but is related to the Richardsons Number at the interface (Srinivas and Mehta, 1990). However, in many estuaries they are evident at neap tides and for short periods over slack water at spring tides (Kirby and Parker, 1983). The high concentration layer, having a high viscosity, will undergo a fluid shear at the level of the lutocline which may create some erosion of the upper surface, but failure may alternatively occur at the bottom of the layer, so that the layer moves as a plug. Once it has started moving the viscosity will decrease and the layer could rapidly become dispersed within the body of the flow. There are indications that layer thickness may be crucial in separating these two modes of erosion (Odd and Rodger, 1986).
SEDIMENT TRANSPORT PROCESSES IN ESTUARIES
433
Erosion Erosion of a mud bed can occur by several processes; by erosion of individual flocs, erosion of small clusters of particles, or by failure and mass erosion of a surface layer, and these normally occur in succession with increasing bed shear stress and depth of erosion. Laboratory measurements have shown there is a correlation between density and the critical erosion shear stress for some muds, (e.g., Thorn and Parsons, 1980), but the correlation coefficient is not high and there are many other important factors such as particle mineralogy and grain size distribution, and organic content. Several studies have examined the relationship between the critical erosion shear stress and the yield stress obtained in rheological measurements (e.g., Migniot, 1968). Additionally, empirical relationships have been demonstrated between the yield stress and the rigidity modulus (Williams and Williams, 1989) and with the specific surface area at a constant concentration. Because of consolidation of deposited sediment, there is a gradient of most physical and chemical properties with depth into the bed. Density and the critical erosion shear stress rise almost exponentially with depth. In many situations the mud surface comprises two layers at slack water; a thin fluid mud or loose fluffy layer, overlying a more rigid bed. The upper layer has a threshold shear stress of 0.06-0.1 N mP2, and is fairly easily eroded during the tide. Erosion of this layer is often quite sufficient to explain the concentration observed throughout the water column at maximum current. The lower layer would normally only be eroded at extreme conditions, and after weakening by biological activity.
TRANSPORT OF MUD IN TIDAL CURRENTS
Within estuaries it is apparent that there is a continual and sometimes very rapid exchange of sediment between the suspension phase and the bed. This has been described by Kirby and Parker (1983), Mehta et al. (1989), Mehta and Dyer (1990). The basic concepts are shown in Fig. 14-5. At high velocities the suspension is mobile and sustained by the turbulent shearing. Concentrations are generally of less than a few thousand ppm (mg 1-I). As the current slackens towards low water, settling creates a high concentration layer near the bed that becomes static. Initially formed at about 20 g l-l, the layer becomes denser with time, a process that is aided by gentle shearing. Re-entrainment of the whole of the layer may occur as the current increases on the next phase of the tide. Alternatively, some part of the layer may survive over the tide and consolidate. This in particular may occur during the progression in tidal range from spring to neap tides as the currents decrease in strength. The remaining part may then survive, but become eroded on the following spring tide. Complete survival may allow the sediment to become part of the settled bed for annual timescales, or longer. The bed shear stress at which deposition occurs is normally considered to be 0.040.08 Pa which is about 30% smaller than the erosion shear stress. Thus, there is a range of shear stresses during which transport occurs, but no erosion or sedimentation. This model of tidal settling and re-entrainment requires that the
434
K.R. DYER r
I MOBILE S U S P E N D E D S E D I M E N T
SETTLING
RE-ENTRAINMENT
EROSION
EROSION
I k
I
a w 0
T
T
HIGH CONCENTRATION SUSPENSION CONSOLIDATING BED
I ,
I
S E T T L E D BED
HOURS I
DAYS I
WEEKS I
MONTHS I
YEARS I
I
TIME SCALE
Fig. 14-5. Deposition/resuspension cycles (after Mehta et al., 1989).
concentration of suspended sediment does not start falling until the current velocity is very low, and thus there would be a large phase lag between velocity and concentration. However, this is at variance with a large number of observations which show that concentrations begin to decrease shortly after the velocity begins to decrease (e.g., Officer, 1981). Sanford and Halka (1993) have proposed an alternative model that appears to fit the observations better. They suggest that there is a continual exchange of sediment with the bed because the temporal and spatial variability in shear stress allows material to settle to the bed throughout the tidal cycle, where it is available for re-erosion. Nevertheless, a proportion of the flocs coming into contact with the bed during the turbulent transport processes will stick and not be re-eroded. McCave (1975) has linked this process to the presence of a viscous sub-layer, which may be considerably thickened by the presence of suspended sediment. Additionally, there is the effect of differential settling of the various size classes which may produce settling of the largest flocs to the bed long before the mean settling class. Once the velocity begins to decrease it is obvious that no new material will be eroded from the bed when the critical erosion stress of the bed is no longer being exceeded. A particular problem in the interpretation of field measurements is the separation of the effects of advection from local erosion and settling processes. Advection can produce a number of concentration peaks during the tide linked to the erosion and exhaustion of thin mud patches upstream of the measurement position. In an oscillating flow with asymmetric velocities, a particle suspended in the water will travel passively in the direction of the residual water flow. However, if the response of the particle to the flow contains a non-linearity, or a lag with respect to velocity, then the particles will have a different residual movement. The settling and erosion properties of the mud, together with movement of particles vertically in the flow can cause the necessary lags.
SEDIMENT TRANSPORT PROCESSES IN ESTUARIES
435
TURBIDITY MAXIMUM
One of the most distinctive features of sediment transport in meso and macrotidal estuaries is the turbidity maximum. This is a zone which contains suspended sediment concentrations higher than those both in the river or further seaward in the estuary. It is generally located at, or somewhat landward of the head of the salt intrusion, where salinities are about 1-5%0. The energetic tidal flow is capable of maintaining high concentrations, and there are a number of processes that concentrate the suspended sediment, and prevent particles from dispersing. The peak concentration of suspended sediment in the turbidity maximum varies between wide limits. Despite the differences due to sediment availability, low tidal range estuaries have maxima with concentrations of the order 100-200 ppm (mg l-l), whereas high tidal range estuaries have much higher concentrations, of the order IOOO-IO,OOO ppm (1-10 g I-'). The turbidity maximum contains a high proportion of a narrow size range of mobile fine sediment, and plays a central role in the circulation of fine sediment within the estuary, as well as probably determining the rate of transport of sediment from the river to the sea. The concentrations of sediment in the turbidity maximum appear to remain almost constant when averaged over a reasonable time, so that residence time of grains in the turbidity maximum must be considerable. Since the turbidity maximum contains several times the annual river discharge of fine sediment, the residence time is likely to be well in excess of a year. The turbidity maximum responds to changes in river flow, with the maximum moving downstream with increasing flow. The mass of sediment in the turbidity maximum also increases. However, a movement of the turbidity maximum down estuary involves expansion into an increased cross sectional volume, and this could decrease the concentrations even though the total mass increases. In the Cumberland Basin (Canada), Amos and Tee (1989) found results which suggest that an increased mass of sediment in the turbidity maximum leads to an increased maximum concentration as well as an increased longitudinal distribution. In the hypersynchronous Tamar Estuary (UK), the turbidity maximum is generally located in an area of locally larger currents, and is stronger when it is nearer the estuary head. Uncles and Stephens (1989) found that the magnitude of the maximum was greatest when the cross sectional area at high water at the location of the maximum was least. Additionally, the distance of the maximum from the head of the estuary varied roughly as the square root of the run-off. The movement of the maximum is well illustrated for the Gironde Estuary by Fig. 14-6. The more or less steady position at high river discharge indicates that the estuarine response tends to be buffered by the increasing stratification. Consequently, as river flow increases, the stratification gradually decouples the upper and lower layers, so that an increasing proportion of the river borne suspended sediment passes straight through the estuary in the upper layer, with the possibility of loss of material into the coastal zone. The seasonal changes in river flow suggests that sediment can accumulate in the upper estuary in spring and summer, and be redistributed down estuary in autumn and winter.
436
K.R. DYER
D I S T A N C E F R O M MOUTH
km
Fig. 14-6. Movement of the turbidity maximum in the Gironde Estuary (after Allen, 1973).
Additionally, the turbidity maximum moves with the tide. At about high water the maximum is located well up the estuary, and concentrations are relatively low because of settling over slack water. During the ebb tide the maximum is moved seaward, sediment is entrained from the bed and concentrations increase. At low water the maximum is further down estuary, and over slack water some settling of material occurs. The structure of the turbidity maximum revealed by measurements at a single station is also shown for a position near the head of the Tamar Estuary in Fig. 14-7 (McCabe et al., 1992). The depth is shown relative to the sea surface, and the measurements spanned just over ten hours at the end of the ebb tide and the beginning of the flood. The rapid deepening at about 1700 hours, when the velocities during the early flood tide reached 1 m s-l, illustrates the asymmetry of the tide. The ebb tide is longer in duration, with a peak velocity of only 0.6 m s-'. The turbidity maximum passage coincides with the velocity maxima, and this occurs at a time when the salinity intrusion is seaward of that location. The majority of the suspended sediment is being entrained from the bed, with a velocity of about 0.4 m s-l at 0.5 m above the bed being the critical erosion threshold. On the ebb tide the lower current velocities erode less sediment and the peak concentrations only reach 400 mg 1-'. Further down the estuary the maximum velocities occur when the tip of the salt intrusion is landward and the turbidity maximum appears within the saline water (West and Sangoyodin, 1991).
437
SEDIMENT TRANSPORT PROCESSES IN ESTUARIES LW
EBB
FLOOD
t
v 0 : d > Y
E
E
I c
a
0 Y
1000
1200
1400
1600
1800
2000
TIME h r 8
Fig. 14-7. The movement during a spring tide of the turbidity maximum in the Tamar Estuary and its relationship with velocity and salinity (after McCabe et al., 1992).
Because of the flood dominance, over the tidal cycle there is a very significant landward sediment flux; the strong flood turbidity maximum returns on the ebb at a much lower concentration, deposition having occurred over high water. This landward flux occurs at low river flow, and is balanced by a downstream redistribution of sediment at high river flow (Uncles and Stephens, 1989). A feature of macrotidal estuaries is the large difference in tidal range between spring and neap tides. Because of the considerable variation of velocities there are changes in position and magnitude of the turbidity maximum (Allen et al., 1980; Gelfenbaum, 1983). The estuary may be partially mixed at neap tides, but during the increasing tide there can be a sudden change to well mixed conditions. Though the estuary may be well mixed at spring tide, at neap tides it can be partially mixed, or even stratified. At spring tide the turbidity maximum has its highest concentration, as the currents are able to erode and sustain more sediment in suspension, and it will be located further up the estuary. This is due to the fact that there is a higher mean sea level in the upper estuary at springs than at neap tides, arising because the increased range at spring tides involves a large extra volume of water at high tide, but only a slight volume difference at low tide, relative to the neaps. During decreasing tidal amplitude towards neaps, the peak currents decrease, and less material is capable of being re-eroded and suspended. Additionally, the durations of slack water increase, enhancing deposition. In the Severn Estuary the mean total mass in suspension varies 5-30 kg m-2 between neap and spring tides. However, at one station tidal variations were 15100 kg mP2 and 70-160 kg mP2 at the two tidal states, respectively (Kirby, 1986).
438
K.R. DYER
PROCESSES FORMING THE TURBIDITY MAXIMUM
The turbidity maximum is a dynamic feature that involves interaction of the tidal flow with erosion and deposition of sediment. There are a number of processes that operate to concentrate the fine sediment at the upper end of the estuary, and to keep it there.
Residual circulation In partially mixed estuaries the vertical gravitational circulation produces a residual landward bottom flow, and a seaward surface residual flow. This has long been thought to be the main mechanism for maintaining the turbidity maximum (Schubel and Carter, 1984). Because of the residual downstream flow in the river, there is a convergence in the bottom flow at a null point near the head of the salt intrusion, in salinities of about 1-5%0 (Fig. 14-8). Suspended sediment is brought into the estuary by the river, and in the upper estuary energetic tidal mixing enhances sediment resuspension, and transfers the sediment between the surface and lower layers (Kostaschuk and Luternaeur, 1989). The surface layer transports sediment downstream to the middle estuary where the particles settle into the lower layer, only to be carried headwards on the residual flow, together with particles brought in from lower down the estuary. Consequently, the maximum concentration of suspended sediment occurs at the bottom near the null point. This circulation process can lead to a turbidity maximum without the need for consideration of sediment properties other than settling velocity, and without any sediment erosion or deposition. Also it is a mechanism for sorting the flocs, since a change in the settling velocity leads to a variation in the suspended sediment concentration (Festa and Hansen, 1978). The fact that the turbidity maximum sometimes occurs landward of the salt intrusion indicates that the vertical gravitation circulation is not always dominant.
8
FRESHER WATER
*4
*
.
8
8
8
8 8
*
.*
8
8
'
0
L E V E L OF N O M O T I O N
Fig. 14-8. Diagrammatic representation of the formation of the turbidity maximum by vertical gravitational circulation.
SEDIMENT TRANSPORT PROCESSES IN ESTUARIES
439
Lug effects As has been described above, if the sediment particles were a passive tracer which responded instantaneously to the flow, the movement of the sediment would be in the residual flow direction. However, the sediment response lags the flow. This phase difference between the suspended sediment concentration and the water velocity can produce a residual flux of sediment even when there is no residual movement of water, providing the currents are asymmetrical. Lags can be produced by a variety of causes, with different processes being important for mud and for sand.
Threshold lag A lag can be produced by the presence of a threshold for sediment movement. In an asymmetrical tidal current, sediment movement can take place for a longer time on one phase of the current compared with another. As an example, in a tidal cycle with an intense short flood current and a long lower velocity ebb current, the duration of movement on the ebb tide will decrease more rapidly than that on the flood tide with an increase in the threshold of sediment movement (Fig. 14-9). In the extreme, the current on the ebb tide may not reach the threshold velocity, and all of the movement then occurs on the flood tide. The asymmetry in the sediment discharge, or transport rate, caused by this effect will be even more marked if the transport rate has a nonlinear relationship to the current velocity. In practice measurements have shown that the sediment transport rate for bed load is normally proportional to the 3/2 to 7/2 power of the bed shear stress. As the asymmetry of the tide increases towards the head of the estuary the increasing magnitude of the flood current causes a transport of sediment towards the head of the estuary. This has been suggested by Allen et al. (1980) as a major process in macrotidal estuaries for creating the turbidity maximum.
Erosion lug In mud, consolidation causes an increase in the critical erosion shear stress with depth. Thus, as the current velocity increases, erosion takes place to a depth where the ambient fluid shear stress equals the critical erosion value for the mud. Once the current diminishes after the maximum, no more erosion takes place. Thus, there will be an asymmetry created in the suspended sediment concentration over and above that present in the current. Modelling of the consolidation properties is an important aspect of mud transport in estuarine models (Hayter, 1986). Scour Lug Once sand particles are in motion they can be kept moving at velocities below the threshold of initial motion. Consequently, between the threshold of erosion and the threshold of deposition material is kept in motion, but no new erosion takes place. This produces a scour lag (Postma, 1967). For mud, scour lag can be defined as the time taken for sediment, when entrained from the bed, to disperse to higher levels in the flow (Nichols, 1986a). This means that once material is eroded from the bed it is only gradually mixed through the water flow as it moves downstream. Initially the sediment will move in the near bed
440
K.R. DYER
5 : a a 3 0
Fig. 14-9. Variation of current velocity and sediment transport (relative units) during a tidal cycle. Top: current velocity. Centre: sediment transport rate for a threshold of 10 cm s-', full line: assuming a linear relationship with excess velocity, dotted line: assuming transport proportional to 3/2 power of excess velocity. Bottom: sediment transport rate for a threshold of 20 cm s-l, full line: linear relationship, dotted line: power relationship.
layers at a velocity lower than the depth mean current. Consequently, at higher levels in the flow the suspended sediment concentration will lag behind the concentrations being produced at the sea bed.
Settling lag On the decreasing tide the particles will start to settle once the turbulence in the flow is incapable of maintaining them in suspension. As the particles settle they are moving along on the waning current so that they eventually reach the bed some distance from the point at which settling commenced. This effect is settling lag, and a qualitative model describing these effects was developed by van Straaten and Kuenen (1958) and Postma (1961). To illustrate the effect consider the simple situation shown in Fig. 14-10. The symmetrical tide has a decreasing maximum current towards the head of the estuary,
441
SEDIMENT TRANSPORT PROCESSES IN ESTUARIES
. ...
'..'. .
>
+ -
0
0
I
DISTANCE OFFSHORE
Fig. 14-10. Schematic illustration of settling lag in an oscillatory current reducing in magnitude towards the head of the estuary (after Postma, 1961).
and the water and sediment particles undergo a changing velocity with distance along the channel during the tide. A particle on the bed at 1 will be lifted into suspension as the threshold velocity is exceeded. It then travels with the water until 2, at which point it starts to settle. Because of settling lag it reaches the bed at 3. On the following ebb tide it will not be entrained until later in the tide when the threshold velocity is reached at that position, and it travels with the water until deposition at low water at position 6. Consequently, the particles gradually migrate shorewards to deposit in the area where the maximum velocity during the tide equals the threshold velocity of the grain. Settling lag will sort the particles according to their threshold characteristics and settling velocity. Dronkers (1986b) considered the time interval during which sediment particles can settle at slack water and remain on the bottom until resuspended, and concluded that the magnitude and direction of the residual sediment flux is mainly determined by the current velocities around low water and high water slack. The slack water period at high water generally exceeds that at low water. Additionally, when there are extensive intertidal areas at high water the average water depth can be less than that at low water. Thus, settling is a more efficient process at high water. In the Humber Estuary there is a marked delay of about an hour in the rise in concentratio$-at mid
442
K.R. DYER
depth relative to near bed during the flood tide. At the end of the flood tide the concentrations are sustained by sediment settling from higher in the flow, and the time lag is reduced to about half an hour. In this case it appears that a combination of scour lag and settling lag produces the depth variation.
HORIZONTAL FLUXES
The lag effects occur simultaneously and their effects will be difficult to separate. The relative importances of their contributions to the horizontal fluxes can be considered from examination of the temporal and spatial variations of velocity and suspended sediment concentrations during the tidal cycle.The instantaneous flux of suspended sediment through a vertical element of an estuary is given by: h
F = I
ucdz
where h is the water depth. Averaging over the tidal cycle gives:
where the overbar denotes a tidal average, and subscript A a depth average. U and C are the tidal cycle fluctuations in velocity and concentration, and subscript d are the deviations with depth from the depth mean values (for the full derivation see Dyer, 1978). The first two terms on the right hand side are the downstream advection on the river flow, U being the non-tidal drift, and term 2 the Stokes Drift. Terms 3-5 are fluxes due to phase differences between the depth mean velocity, concentration, and the water depth, and arise mainly because of threshold and erosion lags in the sediment response to the tidal current asymmetry. These terms contribute to what is known as tidal pumping. Terms 6 and 7 known as the shear effect, arise because of variations in the vertical of the profiles of velocity and concentration. For term 6, a negative (upstream contribution) is produced if a large velocity at the surface is associated with a small concentration, together with a small velocity with a large concentration near the bottom. When averaged over the tide the result is the response to the vertical gravitational circulation. Term 7 arises from the different form of the velocity and concentration profiles during the tidal cycle, due to entrainment and settling lags. This approach has been applied to several estuaries by Dyer (1978; 1988), Uncles et al. (1984; 1985), and Su and Wang (1986). The main difficulty is the assumption that over the tidal cycles of observation, the estuaries are in steadystate. Nevertheless, in all cases the cross-sectional fluxes produced by tidal pumping were larger than those produced by residual gravitational circulation. Consequently, erosion and suspension of sediment during the tide is a major factor in generating and supporting the turbidity maximum. At the seaward end of the turbidity maximum advection from upstream of eroded material leads to maximum concentrations appearing close to low water, and to phase relationships creating an upstream tidal pumping. At the upper end of the turbidity
SEDIMENT TRANSPORT PROCESSES IN ESTUARIES
443
maximum the reverse happens, with maximum concentrations occurring near high slack water, producing a downstream tidal pumping component. At locations near the peak of the turbidity maximum the tidal pumping term is likely to be a minimum. These effects are likely to be mainly coincident with the asymmetry in the flow; flood dominated in the saline intrusion, and ebb dominated in the riverine section. The vertical gravitational circulation is likely to be a minor contribution to the turbidity maximum, though it may help to sharpen the peak, and the concentration gradients. There are also significant differences laterally across the estuary in tidal pumping. Uncles et al. (1984) have shown that near the head of the Tamar Estuary landward pumping of sediment occurred in the central channel, whereas other sections had weak landward pumping in shallow water and seaward pumping in deeper water.
ESTUARINE TRAPPING
Within the estuary the riverborne sediments become trapped by the tidal pumping and residual circulation, and mixes with material brought in from the sea. Meade (1969) has argued that the majority of the sediment in estuaries of eastern North America is derived from the sea, despite the high river discharge. This conclusion seems to be valid for many temperate estuaries. Riverborne and marine sources of sediment can often be distinguished from examination of clay mineralogy, heavy mineral content, and radioactive and stable isotope tracers, e.g., Nichols (1972), Song et al. (1983), Mulholland and Olsen (1992). Figure 14-11 illustrates this, showing that a large percentage of marine derived material can be present right up to the head of the salt intrusion. The process of mixing involves continuous erosion, deposition and exchange of sediment within the estuary; the fine sediment cycling through the turbidity maximum and coarser sediment cycling round the ebb-flood channel systems. Individual particles may spend a considerable time moving within the system before being finally deposited, or passing through to the sea. The residence time of particles can be defined as the number of particles inside the estuary divided by the number leaving per unit time (Martin et al., 1986). Some of the particles entering from the river will remain in suspension and pass through the estuary fairly quickly particularly at times of high river floods. However, a significant proportion will undergo many cycles of deposition on the bed followed by resuspension, with the deposition occurring at a number of points along the estuary which form temporary sinks for the sediment particles operating for a variety of timescales. Consequently the mass of particles in suspension in the turbidity maximum comprise proportions of particles that may have ages (time since input) lasting from a few days to possibly years. Little is known concerning particle residence times in estuarine turbidity maxima. The trapping efficiency of the estuary is the ratio of the fluvial sediment input, to that accumulated in the estuary. For partially mixed estuaries it can exceed 100% (Nichols, 1986b), since the fluvial sediment is only part of that accumulating. Some of that drawn in from beyond the estuary mouth is likely to be fluvial material exported
444
K.R. DYER 100
z
5? 0
8
I
W
z
I
a
I
50
I
I I
F
z
I
W
I
0
I I
W
a
0
, 0
n
10
1
I
20
30
SALINITY 0
B E D INORGANIC
8
S U S P E N S I O N INORGANIC
0
BED ORGANIC
0
SUSPENSION ORGANIC
Fig. 14-11. Mixing of sediments of marine and freshwater origin in the Savannah Estuary (after Mulholland and Olsen, 1992).
100% 0
z
w
I! U U
W
re
+
o
WELL
PARTLY
MIXED
MIXED
STRATIFIED 0
ESTUARINE T Y P E
Fig. 14-12. Variation of trapping and filter efficiency of various estuary types.
at higher river flow stages, but much will be of coastal or marine origin. Additionally, in well mixed estuaries tidal pumping becomes significant in transporting sediment up-estuary into the turbidity maximum (Fig. 14-12), with the degree of tidal pumping depending on the tidal characteristics, as well as those of the sediment. A filter efficiency has also been defined by Schubel and Carter (1984). This takes account of the additional accumulation of sediment of marine derivation. However, tidal pumping is not included in their analysis.
SEDIMENT TRANSPORT PROCESSES IN ESTUARIES
445
The trapping characteristic is likely to be fairly sensitive to the topography of the estuaries, in its effect on the tidal velocity field, and on the river discharge which effects the stratification and the gravitational circulation. The trapping thus undergoes considerable short term variability. For instance, the Tay Estuary exports sediment to the sea on neap tides but imports on the spring tides (Dobereiner and McManus, 1983). This situation is likely to be most apparent in estuaries near sedimentary equilibrium. Nevertheless, the sediment particles can be continually cycled from one part of the estuary to another through the turbidity maximum. The major sites of this interchange are the intertidal areas which often show deposition rates of the order of a centimetre per year. The surface of salt marshes above neap tide high water mark, and the upper part of the intertidal flats, show regular sedimentation, with layering and lamination in core samples. However, the outer edges of the salt marsh often show erosion by “cliffing”, the undercutting and erosion of small blocks of compacted salt marsh sediment. Additionally, gullies and meandering channels cross the flats and show active erosion of the banks and migration of the meanders. As the channels meander across the mudflats they transform the horizontally stratified sediments into sequences showing laminations inclined at 7-15’. These are produced by deposition on the inside of bends in the gullies while erosion occurs on the outside of the bends (Bridges and Leeder, 1976). Within the Humber Estuary there are short term (1-30 year) cycles of mudflat and marsh edge erosion which appear to be related to periodic shifts in the low water channel (Pethick, 1988). Consequently, one can envisage a continual cycle with the mudflats building up to a particular level, and then being attacked and eroded by shifts in the channels and by gullying. The eroded sediment is exchanged via the turbidity maximum to other areas of temporary deposition. There is an important seasonal cycle in the build up of sediment on the intertidal areas within which plants play an important role. However, the response is somewhat different between the exposed mudflats, and the salt marshes. Frostick and McCave (1979) have measured a 5-cm accretion of mudflats between April and September because of trapping by algae and erosion in the winter. A surface layer of benthic diatoms cause a large increase in the critical erosion shear stress (Paterson, 1989) and this must reduce erosion by waves, as well as enhancing deposition. However, other seasonal variations are possible. Kraueter and Wetzel(l986) have shown stable sediment conditions occurred between December and March, but increased benthic activity in the summer caused increased water content, a decrease in sediment shear strength, and increased suspended sediment concentration. Orson et al. (1992) has illustrated the processes of salt marsh accretion by seasonal trapping by plants. Erosion on the mudflats can be effected by ice, rain and waves (Anderson, 1983). Small amplitude waves can increase suspended sediment concentrations in the shallow water by a factor of three. In high turbidity the waves are modified into solitary waves (Wells and Coleman, 1981). The forward velocities under the crests are greater than the backward motion and cause a preferential shoreward motion. On the flood tide the suspended sediment is transported onto the
446
K.R. DYER
higher tidal flats where some can be trapped. During the ebb tide the flow becomes quickly concentrated into the gullies, and is ejected as plumes into the main channels, where it becomes part of the turbidity maximum. The processes of sedimentation on Korean mudflats has been described by Wells et al. (1990). On the intertidal flats the maximum rate of sedimentation occurs on the outer edge of the intertidal flats about mid tide level (Dieckmann, et al., 1987), so that the flats build outwards and upwards, with the consequence that the active volume of the estuary gradually decreases, reducing the sedimentation rate. On the salt marshes the maximum accretion rate occurs near the high water spring tide elevation and this has been modelled by Allen (1990).
SUMMARY
There is a sea level rise of about 1 mm yr-' occurring worldwide. This is predicted to accelerate and give a total sea level rise of about 50 cm by 2050. It is to be expected that estuaries will respond to this rise, though with possible lags. Stevenson et al. (1980) examined fifteen American estuaries and found a strong correlation between mean tidal range and accretionary balance, with high range estuaries showing accretion exceeding sea level rise. However, Nichols (1989) has examined the response of twenty two American lagoons to rising sea level, and found that the majority of them had accumulation rates equivalent to the local sea level rise. Many other equilibrium estuaries appear to be infilling at a rate consistent with sea level rise. Whether this will still hold with an accelerated sea level rise rate will depend crucially on the response of the sedimentary sources. Rising sea level will produce enhanced coast erosion or barrier beach retreat, though this may be limited by coastal defence works. Littoral transport will convey much of this material to the estuary mouths. There the coarser material will become trapped in the ebb and flood tidal deltas. The finer material will be carried into the turbidity maximum, by a combination of tidal pumping and gravitational circulation, where it will join material coming down the rivers. A rise in sea level would normally reduce the rate of sediment input into the estuary, because of preferential deposition in the lower flood plains of the rivers. However, global warming is likely to increase the storminess of the weather. The increased incidence of floods are likely to flush these sediments into the estuary. Within the estuary deposition would produce an expansion of the intertidal flat levels, especially if the inner edges of the salt marshes were allowed to encroach onto the surrounding low lands. If sedimentation on the marsh surface was insufficient to keep up with sea level rise, there would be a progressive narrowing of the vegetational zones, which may lead to a further reduction in the sedimentation rate. The deeper water in the channels would lead to more active wave attack on the intertidal zone, as well as a change in the tidal regime. It is possible that the estuary may change from being flood to ebb dominated. Prediction of the future sediment patterns in estuaries and the infilling rates depends on a complex of interacting processes. Predictive models require a good
SEDIMENT TRANSPORT PROCESSES IN ESTUARIES
447
hydrodynamic basis coupled to specification of the erosion thresholds and rates, settling velocities, and consolidation of the sediment. The modification of turbulence and shear stresses by high concentration layers is an important feedback between the sediment and the water flow. Flocculation is also an important process whereby there is direct interaction between the flow and the sediment properties. On the tidal flats waves and currents will together be important, and models would need to be three dimensional. Because of the influence of time in the sedimentary reactions to the flow, tidally averaged models will only be of restricted use. Consequently, estuary sedimentation is a challenging area of interest where direct collaboration between the disciplines, and combined field, laboratory and modelling work is essential.
REFERENCES Allen, G.P., 1973. Etude des processes skdimentaires dans I’estuaire de la Giroude. Mem. Inst. Geol. Bassin d’aquitaine, 5 , 314 pp. Allen, G.P., Salomon, J.C., Bassoulet, P., Du Penhoat, Y. and De Grandpre, C., 1980. Effects of tides on mixing and suspended sediment transport in macrotidal estuaries. Sediment. Geol., 26: 69-90. Allen, J.R.L., 1990. Constraints on measurement of sea-level movements from salt marsh accretion rates. J. Geol. SOC.Lond., 147: 5-7. Amos, C.L. and Tee, K.T, 1989. Suspended sediment transport processes in Cumberland Basin, Bay of Fundy. J. Geophys. Res., 94: 14407-14417. Anderson, F.E., 1983. The northern muddy intertidal: A seasonally changing source of suspended sediments to estuarine waters - A review. Can. J. Fish. Aquat. Sci., 40, Suppl. 1: 143-159, Bridges, P.H. and Leeder, M.R.,1976, Sedimentary model for intertidal mudflat channels with examples from the Solway Firth, Scotland. Sedimentology, 23: 533-552. Burban, P.Y., Lick, W. and Lick, J., 1989. The flocculation of fine-grained sediment in estuarine waters. J. Geophys. Res., 94: 8323-8330. Davies, J.L., 1964. A morphogenic approach to world shorelines. Z . Geomorphol., 8: 27-42. Dieckmann, R., Osterthun, M. and Partenscky, H.W., 1987. Influence of water-level elevation and tidal range on the sedimentation in a German tidal flat area. Progr. Oceanogr. 18: 151-166. Dobereiner, C. and McManus, J., 1983. Turbidity maximum migration and harbour siltation in the Tay estuary. Can. J. Fish. Aquat. Sci., 40, Suppl. 1: 117-129. Dronkers, J., 1986a. Tidal asymmetry and estuarine morphology. Neth. J. Sea Res., 20: 117-131. Dronkers, J., 1986b. Tide-induced residual transport of fine sediment. In: J. van de Kreeke (Editor), Physics of Shallow Estuaries and Bays. Springer-Verlag, Berlin, pp. 228-244. Dyer, K.R., 1978. The balance of suspended sediment in the Gironde and Thames estuaries. In: B.J. Kjerfve (Editor), Estuarine Transport Processes. Univ. South Carolina Press, pp. 135-145. Dyer, K.R., 1986. Coastal and Estuarine Sediment Dynamics. John Wiley, Chichester, 342 pp. . Dyer, K.R., 1988. Fine sediment particle transport in estuaries. In: J. Dronkers and W. van Leussen (Editors), Physical Processes in Estuaries. Springer-Verlag, Berlin, pp. 295-320. Dyer, K.R., 1989. Sediment processes in estuaries: future research requirements. J. Geophys. Res., 9 4 14327-14339. Eisma, D., Bernard, P., Cadee, G.C., Ittekkot, V., Kalf, J. Laane, R., Martin, J.-M., Mook, W.G., van Put, A. and Schuhmacher, T, 1991. Suspended-matter particle size in some West-European estuaries; Part 1: Particle-size distribution. Neth. J. Sea Res., 28: 193-214. Festa, J. F. and Hansen, D.V. 1978. Turbidity maxima in partially mixed estuaries: a two-dimensional numerical model. Est. Coastal. Mar. Sci., 7: 347-359. Friedrichs, G.T and Aubrey, D.G., 1988. Non-linear tidal distortion in shallow well mixed estuaries: a synthesis. Est. Coastal Shelf Sci., 27: 521-546.
448
K.R. DYER
Frostick, L.E. and McCave, I.N., 1979. Seasonal shifts of sediment within an estuary mediated by algal growth. Est. Coastal Mar. Sci., 9: 569-576. Gelfenbaum, G., 1983. Suspended-sediment response to semidiurnal and fortnightly tidal variations in a mesotidal estuary: Columbia River, USA. Mar. Geol., 52: 39-57. Hayter, E.J., 1986. Estuarial bed model. In: A.H. Mehta (Editor), Estuarine Cohesive Sediment Dynamics. Springer-Verlag, Berlin, pp. 326-359. Hunter, K.A. and Liss, P.S., 1982. Organic matter and the surface charge of suspended particles in estuarine waters. Limnol. Oceanogr., 27: 322-335. Kirby, R., 1986. Suspended fine cohesive sediment in the Severn Estuary and Inner Bristol Channel, UK. Energy Tech Support Unit Rep. ETSU-STP-4042., 243 pp. Kirby, R. and Parker, W.R., 1983. The distribution and behaviour of fine sediment in the Severn Estuary and Inner Bristol Channel. Can. J. Aquat. Sci., 40, Suppl. 1: 83-95. Kostaschuk, R.A. and Luternauer, J.L., 1989. The role of the salt-wedge in sediment resuspension and deposition: Fraser River Estuary, Canada. J. Coastal Res., 5: 93-101. Kranck, K., 1973. Flocculation of suspended sediment in the sea. Nature, 246: 348-350. Kranck,K,. 1975. Sediment deposition from flocculated suspensions. Sedimentology, 2 2 111-123. Krauter, J.N. and Wetzel, R.L., 1986. Surface sediment stabilization-destabilization and suspended sediment cycles on an intertidal mudflat. In: D.A. Wolfe (Editor), Estuarine Variability. Academic Press, Orlando, pp. 203-223. Krone, R.B., 1978. Aggregation of suspended particles in estuaries. In: B.J. Kjerfve (Editor), Estuarine Transport Processes. Univ. South Carolina Press, pp. 177-190. Martin, J.M., Mouchel, J.M. and Thomas, A.J., 1986. Time concepts in hydrodynamic systems: its application to 7Be in the Gironde Estuary. Mar. Chem., 18: 369-392. McCabe, J.C., Dyer, K.R., Huntley, D.A. and Bale, A.J., 1992. The variation of Roc sizes within a turbidity maximum at spring and neap tides. Proc. 23rd Coastal Eng. Conf., Venice, pp. 3178-3188. McCave, I.N., 1975. Vertical flux of particles in the ocean. Deep Sea Res., 22: 491-502. Meade, R.H., 1969. Landward transport of bottom sediments in estuaries of the Atlantic Coastal Plain. J. Sediment. Petrol., 39: 491-502. Mehta, A.J. and Dyer, K.R., 1990. Cohesive sediment transport in estuarine and coastal waters. In: B. Le Mehaute and D.M. Hanes (Editors), The Sea, Vol. 9. Wiley Interscience, New York, pp. 815-839. Mehta, A.J., Hayter, E.J., Parker, W.R., Krone, R.B. and Teeter, A.M., 1989. Cohesive sediment transport. 1: Process description. ASCE J. Hydraul. Eng., 115: 1076-1093. Migniot, C., 1968. Etude des propridtds physiques de differents sediments t r b fin et de leur comportement sous des actions hydrodynamiques. La Houille Blanche, 7: 591-620. Mulholland, P.J. and Olsen, C.R., 1992. Marine origin of Savannah River Estuary sediments: Evidence from radioactive and stable isotope tracers. Est. Coastal Shelf Sci., 34: 95-107. Nichols, M.M., 1972. Sediments of the James River Estuary, Virginia. In: B.W. Nelson (Editor), Environmental Framework of Coastal Plain Estuaries. Geol. SOC.Amer. Mem., 133: 169-212. Nichols, M.M., 1986a. Effects of fine sediment resuspension in estuaries. In: A.J. Mehta (Editor), Estuarine Cohesive Sediment Dynamics. Springer-Verlag, Berlin, pp. 5-42. Nichols, M.M., 1986b. Consequencies of sediment flux: Escape or entrapment? Rap. P.-V. Reun. Cons. Int. Explor. Mer., 186: 343-351. Nichols, M.M., 1989. Sediment accumulation rates and relative sea-level rise in lagoons. Mar. Geol., 88: 201-219. Nichols, M.M. and Biggs, R.B., 1985. Estuaries. In: R.A. Davis (Editor), Coastal Sedimentary Environments. Springer-Verlag, New York, pp. 77-186. O'Brien, M.P., 1969. Equilibrium folw areas of inlets on sandy coasts. ASCE J. Wat. Harb. Div., 95: 43-52. Odd, N.V.M. and Rodger, J.G., 1986. An analysis of the behaviour of fluid mud in estuaries. Rep. SR84. Hydraulics Research Ltd, Wallingford. Officer, C.B., 1981. Physical dynamics of estuarine suspended sediments. Mar. Geol., 40: 1-14. Orson, R.A., Simpson, R.L. and Good, R E., 1992. A mechanism for the accumulation and retention of heavy metals in tidal freshwater marshes of the Upper Delaware River Estuary. Est. Coastal Shelf Sci., 34: 171-186.
SEDIMENT TRANSPORT PROCESSES IN ESTUARIES
449
Owen, M.W., 1970. A detailed study of settling velocities of an estuary mud. Hydraulics Research Station Rep. INT78. Paterson, D.M., 1989. Short-term changes in the erodibility of intertidal cohesive sediments related to their migratory behaviour of epipelic diatoms. Limnol. Oceanog., 34: 223-234. Pethick, J.S., 1988. Physical characteristics of the Humber. In: N.V. Jones (Editor), A Dynamic Estuary: Man, Nature and the Humber. Hull Univ. Press, pp. 31-45. Postma, H., 1961. Transport and accumulation of suspended matter in the Dutch Wadden Sea. Neth. J. Sea Res.,l: 148-190. Postma, H., 1967. Sediment transport and sedimentation in the estuarine environment. In: G.H. Lauff (Editor), Estuaries. AAAS Publ. 83, pp. 158-179. Sanford, L.P. and Halka, J.P., 1993. In situ erosion and deposition of Upper Chesapeake Bay muds. Mar. Geol. (submitted). Schubel, J.R. and Carter, H.H., 1984. The estuary as a filter for fine-grained suspended sediment. In: V. Kennedy (Editor), The Estuary as a Filter. Academic Press, New York, pp. 81-05. Sills, G.C. and Elder, D.McG., 1986. The transition from sediment suspension to settling bed. In: A.J. Mehta (Editor). Estuarine Cohesive Sediment Dynamics. Springer-Verlag, New York, pp. 192-205. Song, W., Yoo. and Dyer, K.R., 1983. Sediment distribution, circulation and provenance in a macrotidal bay: Garolim Bay, Korea Mar. Geol., 52: 121-140. Srinivas, R. and Mehta, A.J., 1990. Observations on estuarine fluid mud entrainment. Int. J. Sediment Res. 5: 15-22. Stevenson, J.C., Ward, L.G. and Kearney, M.S., 1986. Vertical accretion in marshes with varying rates of sea level rise. In: D.A. Wolfe (Editor). Estuarine Variability. Academic Press, Orlando, pp. 241-259. Su, J. and Wang, K., 1986. The suspended sediment balance in the Changjiang Estuary. Est. Coastal Shelf Sci. 23: 81-98. Thorn, M.F.C. and Parsons, J.G., 1980. Erosion of cohesive sediments in estuaries. Proc. 3rd Int. Symp. on Dredging Technics, pp. 349-358. Uncles, R.J., Elliott, R.C.A. and Weston, S.A., 1984. Lateral distributions of water, salt and sediment transport in a partly mixed estuary. Proc. 19th Coastal Eng. Conf. Houston, pp. 3067-3077. Uncles, R.J., Elliott, R.C.A. and Weston, S.A., 1985. Observed fluxes of water and suspended sediment in a partly mixed estuary. Est. Coastal Shelf Sci., 20: 147-167. Uncles, R.J. and Stephens, J.A., 1989. Distributions of suspended sediment at high water in a macrotidal estuary. J. Geophys. Res., 94: 14395-14405. van Dongeren, A,, 1992. A model of the morphological behaviour and stability of channels and flats in tidal basins. Delft Hydraulics Rep. H824.55. van Straaten, L.M.J.U. and Kuenen, P.L.H., 1958. Tidal action as a cause of clay accumulation. J. Sediment. Petrol., 28: 406-413. Wells, J.T. and Coleman, J.M., 1981. Physical processes and fine-grained sediment dynamics, coast of Surinam, South America. J. Sediment. Petrol., 51: 1053-1068. Wells, J.T., Adams, C.E.Jr., 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. West, J.R. and Sangodoyin, A.Y.A., 1991. Depth-mean tidal current and sediment concentration relationships in three partially mixed estuaries. Est. Coastal Shelf Sci., 32: 141-159. Williams, D.J.A. and Williams, P.R., 1989. Rheology of concetrated cohesive sediments. J. Coastal Res., Spec. Issue, 5: 165-173. Wright, L.D., Coleman, J.M. and Thorn, B.G., 1973. Processes of channel development in a high-tide range environment: Cambridge Gulf-Ord River Delta, Western Australia. J. Geol., 81: 15-41.
This Page Intentionally Left Blank
45 1
GEOGRAPHIC INDEX
Aber Benoit, 78, 80, 102 Aber Ildut, 75, 80 Aber Wrac’h, 76-80 Abu Dhabi Sabkha, 247 Adventfjord, 149 Afon Dyfi, 84 Afon Mawldach, 84 Africa, 55, 334, 344 Aguera, 74 Ajaccio, 93 Alabama, 248 Alaska, 114, 120, 124, 129, 140, 143-144, 148, 153, 156-158,162, 186,233,236-237,249 Alazea River, 56 Albermarle Sound, 52 Alberni Inlet, 148 Alligator River, 52 Altamaha River, 52 Sound, 52 Amazon River, 27, 39, 54, 180-181, 198, 210211,278, 281,347 America, 237, 264, 278, 334, 446 Amoy Ria, 89 Andes, 122 Anegada Bay, 54 Antartic Peninsula, 122 Antartica, 7, 119, 122, 156 Aral Sea, 20 Arcachon Basin, 397 Ares Ria, 70 Argentina, 8, 20, 69, 89-90, 197, 230, 241, 259, 265, 365 Arguenon Ria, 76 Armorican Massif, 75 Arnafjord, 156 Arosa Ria, 70-71 Artic, 119, 122, 310 Asia, 56, 334 Ason Ria, 74 Asturias, 69-72, 75, 78, 84-85, 87 Aswan Cataract, 94 Coast, 161, 248
Atlantic Basin, 232 Atlantic Canada, 121 Atlantic Ocean, 334 Aulne Ria, 76-79, 100 Australia, 100, 183, 186, 191, 214, 265, 274, 278, 336,344-345,347,349-352,365, 410-411 Aviles Ria, 73 Baffin Island, 53, 117, 125, 128, 132, 141, 156157, 162-165 Bahamas, 341 Bahia Blanca Estuary, 9,24,32,54,261,365,385, 400 Bahia Magdalena, 230 Bahia San Quintin, 230 Baie Mount Saint-Michel, 278 Baja California, 232 Baltic Sea, 20, 49, 54-55 Baltimore Canyon, 11 Bangladesh, 222 Barnstaple Bay, 83-84 Bashita Minato Swamp, 345 Basque Country, 69,70,72-73,75-76,85,87 Bay of Biscay, 74, 186 Bay of Fundy, 32, 179-180, 278, 280, 282, 285, 364-375, 382-394, 386-387, 392, 395, 400, 406,408-410,414-415 Bay of Saint Brieuc, 82 Bedford Basin, 157, 160-161 Beira, 56 Belgium, 54 Belize Lagoon, 40 Belle-Ile, 82 Belon Ria, 80 Benin, 248 Bermeo, 73 Betanzos Ria, 70-71 Bight of Limpopo, 56,73-74 Bight of Sofala, 36 Bimini Lagoon, 341 Biscayne Bay, 20 Blavet River, 76 Bly Creek, 20 Boca de Quadra, 157
452 Bohai Bay, 277-278, 281 Sea, 300 Bonnefjord, 153 Borgenfjorden, 144-145 Brahmaputra River, 187,209,221-223 Brazil, 30,54, 198,230-231,248,265 Brest Roastead, 75, 77-78, 81 Brigneau Ria, 82 Bristol Channel, 180, 186, 190-191, 193-194, 365, 410 British Columbia, 29, 117, 120-121, 136, 151, 158, 161, 323,366 British Isles, 64, 83-84, 119 Brittany, 64, 75-76, 78-80, 82-85, 87, 89, 100102, 105-106 Brunswick, 81, 314 Buenos Aires Province, 26,259 Burdekin Delta, 215, 221 River, 210, 214 Burry Inlet, 274,278 Buzi River. 56 Cab0 Ortegal, 70 Cairns Bay, 347 Cairo, 207 Cambridge Fjord, 125, 157, 163 Gulf, 186, 195, 345, 350 Camel River, 83-84 Cameroon, 34-36,350 Canada, 29, 122, 124, 134, 183, 186, 274, 278, 282, 323, 435 Cannes, 93 Cantabria, 70, 72, 74-75 Canoe Passage, 314 Cantabric Sea, 74 Cape Cod, 49 Cap Ferret, 192 Cap Haitien, 92 Cardigan Bay, 84 Caribbean Sea, 92, 278, 352 Carmarthen River, 84 Carmen Sylva River, 54 Carrick Roads, 83 Caspio Sea, 20 Castro Urdiales, 72-73 Cauvery River, 56 Cedeira Ria, 70 Celtic Sea, 194 Changjiang River, 56,281
GEOGRAPHIC INDEX Chankiang, 89 Charleston Harbour, 52 Chatham Harbour Estuary, 385,387 Chesapeake Bay, 21,29,32,51,69,180,256,365, 400,410-411 Chignecto Bay, 285 Chile, 29, 120, 122, 230, 233-234, 237 Chiloe Island, 230, 234-235 China, 69, 89, 230, 274-275, 278, 281, 297, 342, Chowan River, 52 Chubut River, 54 Cobequid Bay, 183, 186, 192, 195, 200, 263-264, 274, 365, 382, 386, 392, 400, 406, 408-410, 414-415 Colonia, 20 Colorado River (Arg.) 9, 54, 210, 281 Colorado River (USA) 5,31 Columbia Estuary, 185 Glacier, 114 Commewijne River, 338,340 Como Lake, 94 Copano Bay, 53 Coppenane River, 54 Copper Delta, 219,221 Cook Bay, 92 Inlet, 144-145, 186,249 Coral Creek, 343, 345, 353 Cordillera Cantabrica, 70 Cork Estuary, 84 Cornouaille, 77 Cornwall, 69, 83-85, 87 Coronation Fjord, 128, 157 Corpus Christi Bay, 53 Pass, 248 Costa de Mosquitos, 53 Croae, 82 Cuchia, 74 Cumberland Basin, 284-285, 430 Dahouet Ria, 82 Dalmatia, 232 Daoulas Ria, 76-77 Dart River, 83-84 Davis Reef, 250 Decade River, 132 Delaware Bay, 29, 51, 365, 410 Desaguadero River, 9 De la Plata River, 181, 281 Deseado Ria, 39, 89
GEOGRAPHIC INDEX
453
Devon, 69,83-85 Dingle Bay, 85 Diouris, 78 Doboy Sound, 52 Doce River, 248, 265 Doelan Ria, 82 Dolsan Island, 87 Dos Patos Lagoon, 30, 245, 251,253,248, 265 Douglas Channel, 146, 156-157 Douron River, 76 Ducroix Beach, 92 Durance River, 92 Dyfi Estuary, 327
France, 69-70, 75,94-95, 99, 106, 183, 186,274, 364,379,397 Fraser Delta, 211-213, 215, 217, 219, 221, 223, 310, 275,329 River, 161, 209-210, 214, 216, 307, 329, 366, 400 Fremur River, 76 French Guiana, 54 Frejus, 93 Frisian Islands, 251, 256, 261, 278 Fripp Inlet, 259 Fu Zhou Ria, 89
East China Sea, 56 Eastern Schelde Estuary, 260, 264 Ebro Delta, 221 River, 209 Ede River, 59 Egypt, 94 Eilat, 92 El Barquero, 70 El Ferrol, 70 El Puntal, 74 Elbe River, 55 Elizabeth River, 21 Ellesmere Island, 124 Ellie Point, 347 Elorn Valley, 75-76 Ems Estuary, 55 England, 83-85,315 English Bank, 198 Channel, 77,99 Essex River Estuary, 258-259 Etel River, 76 Erromango Island, 92 Europe, 29, 49,54,70, 119 Exe River, 83-84
Galgeplaat Shoal, 260 Galicia, 69-71, 73, 75-76, 84, 87, 95, 97, 105, 230-231 Gallegos Ria, 89-90, 230 Galveston Bay, 53 Ganga River, 56 Ganges -Brahmaputra Delta, 210-211,213,215,22022 1 River, 11 Garda Lake, 94 Garorim Ria, 87 Gaupnefjord, 132, 139 George Island, 87 Georgia, 119, 250, 255, 365 German Bay, 274-275 Germany, 54, 251,261, 274 Geumo Island, 87 Ghana, 248 Gijon, 73 Gilbert Inlet, 148 Gironde River Estuary, 29, 32, 39, 69, 180, 183186, 190, 192, 196, 364, 366, 379, 410, 414, 435-436 Glacier Bay, 129, 156-158 Goayen River, 80,99 Great Barrier Reef, 250 Great Britain, 84, 106 Great Ouse, 299 Great Sound, 365 Greenland, 7, 115, 119-121, 128, 157 Grevelingen Estuary, 242 Godavii River, 56 Golden Gate, 231 Guaiba Complex, 250 Guadalupe River, 53 Gulf of Aqaba, 92 Gulf of California, 274, 278, 281
Fa1 River, 83 Falmouth, 83 Firth of Tay, 122 Flanders, 71 Florida, 52, 248, 257, 264, 346, 365, 400 Fly Delta, 185 Estuary, 186, 189,220 River, 56, 187, 209 Foochow Ria, 89
454 Gulf of Mexico, 53, 121, 215 Gulf of Morbihan, 89 Gulf of Papua, 186,220 Gulf of St. Lawrence, 278 Guyana, 54 Hainan Island, 342 Haiti, 92 Hamilton Inlet, 121, 159 Handil Distributary, 214 Hangshou Bay, 56,89 Hang Zhu Wan, 89 Hardangefjordll7, 121, 143, 156 Hawke Bay, 229-230 Hedjaz, 90, 92 Hellefjord, 155 Hinchinbrook, 343 Hispaniola, 92 Holyrood Pond, 147 Homathko Delta, 136 Honduras, 53 Hong Kong, 89 Howe Sound, 151,156 Huang Ho River, 56 Huanghai Sea, 277 Huanghe River, 281 Hudson Bay, 274, 313 Hudson Canyon, 11 Humber Estuary, 441, 445 Iberian Peninsula, 69, 70, 76 Iceland, 119, 132, 162 Idku Coastal Lagoon, 248 Incheon, 85, 87 Indian Ocean, 94 Indigirka River, 56 Indonesia, 347 Indus River, 56 Inugsuin Fjord, 132 Ireland, 81, 84, 247 Irrawadi River, 56 Isle of Groix, 82 Italy, 94 Itamaraca Estuary, 230-231 Itap6a Spit, 250 Itirbilung Fiord, 135, 141, 149, 164-165 Jade Estuary, 278 James Bay, 313 James River, 21, 51 Japan, 345 Jeffreys Ledge, 242
GEOGRAPHIC INDEX Jaudy River, 76 Josdngfjord, 118, 156 Jostedal River, 132 Kangerdlugssuag Fjord, 156 Kenmare River, 84 Kenya, 92 Keradraon, 77 Kerguen Islands, 119 Keridaoeun Bay, 77 Kermalero, 80 Keroville, 101 Keurbooms Inlet, 259 Keta Lagoon, 248 Keum Estuary, 180 Kio Chow, 89 King Sound, 351 Kitimat Delta, 151 Klang -Langat Estuary, 186, 189 River, 187 Strait, 186 Knigth Arm, 249 Inlet, 136, 156 Kongsfjorden, 128, 156 Korea, 69, 85, 180, 278, 281, 446 Korsfjorden, 157 Krishna River, 56 La Corufia, 70-71 La Franca, 73 La Plata City, 20 Labrador, 121, 145-146, 156-157, 159, 162 Lafayette River, 23 Laguna de Rocha, 241 Laguna Madre, 248 Laita River, 80 Lake Melville, 113, 156, 159 Landerneau City, 75 Landevennec, 78 Langebaan Coastal Lagoon, 265 Lanvian, 78 Languedoc, 92 Langon, 76 Laptev Sea, 56 Lauberlac’h Ria, 81-82 Leba River, 55, 58 Le Conquet Ria, 81, 101-102 Le Faou Ria, 76-77,79, 101-102 Lena River, 56 Les Anges Bay, 77
GEOGRAPHIC INDEX Leguer River, 76 Lenga Estuary, 30 Leon Country, 76 Lequeitio, 73 CHopital Ria, 76 Ligurian Coast, 93, 95 Limpopo River, 56 Llanelly, 84 Lock Striven, 157 Locmajan, 80 Loc’h, 76 Loch Etive, 153 Loch Nevis, 156 Logonne-Quimerc’h, 78 Loire River, 187 Long Island Sound, 58,365,381,400 Louisiana, 248 Lupawa River, 55, 58 Lusterfjord, 166 Lyons, 92 Macao, 89 MacArthur River, 344 Mackenzie Delta, 211 Magdalena Bay, 232 Magnetic Island, 342 Mahakam Delta, 213-214,216-217,221 Mahone Bay, 156 Main Arm, 213,216 Maine, 120 Maitland Island, 146 Maktak Prodelta, 141 Makkovik Bay, 145-146, 156-157,159 Malaysia, 345, 348 Manaia Harbour, 229-230 Manzala Coastal Lagoon, 248 Mar Chiquita Lagoon, 30, 40, 250, 252,255-256, 259,265 Maroni Estuary, 54 Marowian River, 54 Marseilles, 82 Maryut Coastal Lagoon, 248 Massachusetts, 50, 242, 385, 387 Maullin River, 230, 234 Mediterranean Sea, 20, 70, 92, 95 Mekong Delta, 211 River, 56 Mercier Channel, 30, 39 Merrien Ria, 82 Meuse River, 55 Mexico, 53,232, 248,265,337,339,350-351
455 Milford Haven, 84 Milford Sound, 122 Minas Basin, 186, 274-275, 278, 280, 364, 384, 387,395, 400 Mississippi Delta, 210-212, 214, 216, 219, 221-222, 244, 248-249,275 River, 32, 39, 199, 223, 243 Sound, 40 Mobile Bay, 53, 241 Mogpo, 85 Mombasa Harbour, 92 Montevideo, 20 Moreton Bay, 191,365-366, 410-412 Morlaix Ria, 76, 102-106 Moulin d’Enfer, 77 Mount Saint Michael Bay, 76, 274 Mozambique, 55 Muir Glacier, 124 Inlet, 124, 157 Nain Bay, 156 Namhae Island, 87 Navidad River, 53 Nayart Coasta Lagoon, 265 Neches River, 53 Negri Glacier, 129 Neuse River, 52 New Caledonia, 345-346, 350 New England, 58 New Guinea, 56, 185-186,347 New Hebrides, 92 New Jersey, 365 New Orleans, 207 New York, 50, 365 New South Wales, 285 New Zealand, 119, 120, 122, 156-157, 229-230, 232 Newfoundland, 146-147, 159, 162 Nicaragua, 53, 248 Nice, 93-94,96 NiGois, 93, 95 Niger Delta, 52, 214, 217, 219 River, 32, 213, 232 Nile Delta, 94, 209, 211, 215, 223 River, 32,209,223 Nordfjord, 153 North America, 121, 307, 443 North Inlet, 344
456 North Sea, 54, 179, 275, 281,283, 427 Northern Territory, 339,344-345,349 Norway, 29, 117-118, 120-122, 143-145, 153, 155-157, 160, 162 Nova Scotia, 121, 146,160-161,242-243 Nuka Bay, 156 Ob River, 56 Obkaya Guha Bay, 56 Oceania, 334 Oder Estuary, 20, 50, 55 River, 58 Odet River, 76-77 Ogeechee River, 52 Omoloi River, 56 Ondarroa, 73 Ord River Estuary, 100, 180-182, 186, 189, 194, 345,351, 365,427 Oregon, 237 Orinoco River, 32, 54,214, 278,281, 347 Ortigueira, 70-71 Ortiz Bank, 198 Oslo Fjord, 39 Oslofjord, 113, 120, 156-157 Ossabaw Sound, 52 O u t m t River, 248 Outardes Delta, 210-211, 215, 217, 221 Oyacock River, 54 Pacific Ocean, 95, 232, 237, 318 Padstow, 83 Pahara Lagoon 248 Pakistan, 222 Pamlico Sound, 40, 52 Panmure Basin, 2,229-230 Papua, 56,347 Parani River, 54, 197-198 Paraquacuh River, 54 Parker River Estuary, 386 Pasajes, 73 Patagonia, 54, 89, 280, 300 Pee-Dee River, 52 Pemhroke, 84 Penfoul River, 80 Penze Bay, 53 Ria, 76-77 Perlas Lagoon, 248 Persian Gulf, 211 Piasnica River, 55, 58 Pitt River, 366
GEOGRAPHIC INDEX Plata Shoals, 198 Plymouth Ria, 84 Roadstead, 83 Poland, 49, 54 Pont Krac’h, 78 Pont Croix 80 Pontevedra, 70 Ria, 39 Port Valdez, 156 Portage, 230, 237 Portland, 120 Porz Lamat, 82 Potomac River, 22,51 Pouldohan River, 80 Prat Paul, 78 Prince Inlet, 252 Prince William Sound, 114 Provence, 82, 93, 95 Pudeto Estuary, 230, 236 Puget Trough, 323 Punque River, 56 Punta Piedras, 20 Quebec, 121, 156 Queen Inlet, 140, 149 Queensland, 250,342-345,347, 349-350,353 Quetalmahue Estuary, 230, 236 Queule River Estuary, 30,230,234-236 Quiantang River, 56 Quillimadec River, 80 Quimperle River, 80 Rajang River Delta, 213, 215 Rappahannoc River, 51 Red Sea, 70,90,94 Rhine River, 55, 366 -Meuse Delta, 214 Rhone Ria, 92, 94 Ria de Arosa, 71 Ria de Foz, 71 Ria de Guernica-Mundaca, 75 Ria de Muros y Noya, 70-71,95-99 Ria del Eo, 71 Rio Gallegos, 32, 231 Rio do Sul, 248 Rio Grande (Arg.) 54 Rio Grande (USA) 53 Ria Mundaca, 75 Ria Ortigueira, 71 Ria Rihadeo, 71 Ria Vigo, 71
GEOGRAPHIC INDEX Rio de la Plata, 20, 39, 54, 197, 200 Rio Negro, 54 Rio d e Pas, 74 River Kiulung, 89 River Loughor, 84 River Mei, 89 River Min, 89 River Tijwi, 89 Roanoke River, 52 Rookery Bay, 347 Roussillon, 92 Rufiji Delta, 344 Russell Fjord, 153 Russia, 50 Russian Lagoon, 242 Sabine River, 53 Sacramento River, 30 Samborombon Bay 54 Saguenay Fjord, 121, 157 River, 113 Saint Tropez, 93 Saluda River, 52 Salmon River Estuary, 183, 192, 195, 200, 263, 408-409 Salween River, 56 San Antonio Bay, 53, 241 San Antonio Inlet, 259 San Francisco Bay, 30,40,229-232,242,323,365 San Joaquin River, 30 San Julian Ria, 89 San Quintin Bay, 232 San Sebastian Bay, 40, 73, 252, 280, 300 Sante River, 52 San Vicente de la Barquera, 75 Santa Cruz Ria, 89-90 Santander, 74 Sio Francisco Delta, 211, 213 Sapelo Sound, 52 Satilla River, 52 Sauce Chico Estuary, 24 Saudi Arabia, 90-94 Savannah River, 52, 444 Save River, 55 Schelde Estuary, 55, 365 Scotland, 83, 120, 122, 157, 161 Scorff River, 76 Sea of Japan, 85 Segara Anakan, 347 Senegal Delta, 219 Seoul Ria, 85,87-88
457 Serra do Mar, 231 Severn River Estuary, 106-107, 180-181, 186, 189-190, 193, 278,365,427,437 Shan Tan Ria, 89 Shandong Peninsula, 89 S h a m Abhur, 92 Shimo Sharm, 92 Siagne River, 93 Sinai Peninsula, 92 Skaggerak, 113, 120 Sherton bay, 157 Sognefjord, 121, 156 Sondre Stromfjord, 156-157 South Africa, 40, 250, 259, 265 South America, 119, 181, 197, 233, 347 South Alligator River, 186,339,345,348-349,352 South Cape Fjord, 124 South Cape Glacier, 124 South Carolina, 52, 257,259, 365, 408 South China Sea, 300 Spain, 29, 84, 95, 99, 105, 230 Spitsbergen, 122, 128-129, 149, 157 St. Andrews Bay, 365,400 St. Catherine Sound, 52 St. Christophe Jetty, 82 St. George Bay, 156, 162 Channel, 83 St. Helena Sound, 52 St. Johns River, 52 St. Lawrence Estuary, 145, 214 St. Louis Bay, 53 St. Lucia Estuary, 55, 344 St. Margaret Bay, 156 St. Mary River, 52 St. Simeons Sound, 52 Story Head, 243 Sturgeon Bank, 312 Strait of Gibraltar, 92, 95 Suakin Harbour, 92 Sudan, 92,94 Sungai Merbok Estuary, 345 Surinam, 54, 338, 340,342-343 Suriname River, 54 Susquehanna River, 22, 52 Suwannee River, 53 Svalbard Islands, 119-120 Swatow Ria, 89 Sweden, 39, 120 Tabasco, 53, 337,339, 350-351 Tamar Estuary, 83,84,278, 435-437, 443
458 Tampa Bay, 346 Tampico, 53 Tanzania, 344 Tar River, 52 Tariec, 80 Tasman Bay, 229-230 Taw River, 83 Tay Estuary, 32, 445 Tazovc River, 56 Tech River, 92 Teign River, 83-84 Te Kouma Harbour, 229-230 Terc’h Island, 77 Terminos Lagoon, 40,248, 347 Tet River, 92 Texas, 53, 246-248, 251 Thames River Estuary, 29,39, 189, 191,365,385, 400, 410, 412 The Netherlands, 54-55,64,242, 349,365 Tierra del Fuego, 54 Tigris-Euphrates Delta, 211 Timor Sea, 195 Torres Strait, 365 Torridge Ria, 84 Townsville, 349 Trebelherick River, 83 Treglonou, 80 Trieux River, 76 Trinity Bay, 53 Inlet, 347 Trondheimsfjord, 160 Tugela River, 55 Turnagain Arm Estuary, 186 Ulang River, 248 Uganda Bay, 120 United Kingdom, 183,278, 282, 435 United States, 2, 50, 64, 69, 120, 233-234, 241, 249,257,261,265, 323,347 Upper Cook Inlet, 230, 237 Uruguay, 20,40, 195,241 River, 195
GEOGRAPHIC INDEX Valdivia River Estuary, 30, 40, 233 Vancouver Island, 161 Vanuatu, 92 Var Ria, 93,95-96 Vavaca Bay, 53 Veracruz, 53 Vigo, 70 Ria, 105 Vilaine River, 76 Vistula River, 50, 58 Volta River, 248 Waccamaw River, 52 Wadden Sea, 276,278,283,365 Wales, 83-84,274, 278,327 Wasaw Sound, 52 Wash, 274,276-277,282 Washington State, 233, 237, 249, 323 Wenlock River, 343, 350 Wenzhou, 274 Weser River Estuary, 396 Western Brook Pond, 159 Westschelde Estuary, 260, 394, 400 Willapa Bay, 249
Xia Men Ria, 89 Yanbo Sharm, 92 Yangh Kiang River, 232 Yanghtse River, 39-57 Yellow River, 57, 209, 215, 222, 230, 232 Sea, 179, 277-278, 281 Yenizey River, 57 Yser River, 54 Yucatin, 53 Zambesi River, 55 Zhan Jiang, 89 Zhujiang River, 281 Zululand, 57 Zurnaya, 73 Zwin River, 54
459
SUBJECT INDEX
Aber, 36, 75 Advection, 19, 105, 185, 434 Aftonian, 7 Aggregates, 430 Albufera, 23 Algal coating, 363 mats, 247, 281 trapping, 445 Allocyclic phenomena, 223 Alluvial cones, 95 fans, 210 feeder systems, 209-210 Angle of internal friction, 148 of repose, 377,380, 397 Anoxic conditions, 154-155 environments, 118 Antidune, 358 Arm, 113 Atmospheric pressure, 152 Aquaculture, 242 Autocompaction, 326 Autocyclic phenomena, 223 Bacterial mats, 281 Banks, 20, 78, 191, 198 linear sand, 185, 220 Bars, 32, 135, 281, 358-417 alternate, 407, 409 arquate, 218 bank attached, 407 braid, 407-409 chain, 410-411 channel junction, 407 compound, 407 diagonal, 407, 409 elongate, 407, 412 elongate tidal, 407, 411 linear, 410 linguoid, 407
longitudinal, 407 h a t e , 407 middle channels, 407 middle ground, 217 multiple, 407 point, 193, 310, 337, 351, 407, 413 river mouth, 217-221 sand, 185, 196 segments, 406 swash, 218,259 transverse, 407 zig-zag, 411 Barforms, 358 repetitive, 407-409 Barriers, 3, 33, 241-267 duned, 262 gravel, 243,261 islands, 12, 53, 219, 242-267 lagoons, 53 linking, 261 spit, 241 wave-built, 241 Basin back-arc, 227 depth, 160 geometry, 179 intracratonic, 211 Bathymetry, 232,236 Bay, 27,207,212,221 banks, 101 lateral, 77, 89 Beach, 32, 71,78, 232, 243 barrier, 446 bioclastic, 250 fringing, 261 ridges, 212, 261 Bed consolidating, 434 load, 180, 183, 198, 210, 215, 260, 344, 347, 429,439 lower plane, 358 settled, 433-434 upper plane, 358,424
460 Bedding convolute, 237, 249 flaser, 189 lenticular, 189 longitudinal, 276 master planes, 405-406, 413 oblique, 276 tidal, 189 wavy, 189 Bedforms, 135,145,179,183,185-187, 189,191192, 194-195,215,245, 258, 281, 291, 358417,429 catback, 379 classification, 265 crestline, 387 height, 322, 383 lateral bedform transition, 383 buttress, 383 open, 383 zig-zag, 383 longitudinal, 387 macroscale, 280 -normal transport, 387 oblique, 387 orientation, 388 phase diagram, 360 transverse, 387 Bedload, 132, 164, 322 Bedrock, 71, 130 Bellows, 286 Benthic boundary layer, 284 Bergs, 127 Bingham plastic, 427 Bioaccumulation, 3 18 Bioerosion, 72, 92 Biogenic activity, 247 effects, 250 mounds, 250 Biogeochemical interactions, 138, 163 Biological sedimentary factors, 37, 317 Biostabilization, 279 Bioturbation, 134, 196, 244, 250, 263, 315, 406 Block uplift, 227 Blowouts, 262 Bottom roughness, 341, 370 Bottomsets, 95, 140, 406 Boulders, 117, 247 Boundary layers, 322,360, 366-370,391 Breezes land, 152
SUBJECT INDEX sea, 152 Brownian motion, 425 Bundles, 260 Buoyancy, 125 Buoyant effluent, 217 Burrows. 247 Cala, 36 Calanques, 36, 82 Canyons, 49,249 Capes, 89 Carbonates, 118, 247 calcium, 80, 99 Catchment area, 336 Cay sands, 250 Centrifugal force, 118, 137 Channel, 212-217 braided, 210 distributary, 212 ebb, 213 Rood, 213 meandering, 213 migration, 216 tidal, 212 Cheniers, 54, 243, 263 Chlorides, 247 Circulation anti-estuarine, 245 gravitational, 428, 438, 442, 445 residual, 427, 438, 442 tidal, 427 Clay, 117 cohesive, 429 minerals, 71, 423, 443 Cliffs, 73, 80, 87 periglacial, 80 retreat, 117, 147 till, 161 Climates, 207, 246 tropical, 4 subtropical, 4 temperate, 4 subpolar, 4 polar, 4 Clinoforms, 98-99 Coast collision margin, 6, 13, 249 coral reef, 250 Dalmatian-type, 23 1 emergence, 230,242 fjord, 120
SUBJECT INDEX
Coast (continued) hard-rock, 6,230 hypertidal, 424 macrotidal, 25'2, 424 mesotidal, 6,252,424 microtidal, 6, 424 soft-rock, 6,230 stable, 242 submergence, 230,242 subduction, 13 tectonicdly-negative, 230 tectonically-positive, 230 trailing-edge, 6, 13,249 Coastal barriers, 49 communities, 309 embayments, 179 environments, 227 management, 18,242 physiography, 227 plains, 50-56,243, 343 sand dunes, 55 strand plain, 275 Coastal lagoons,6,49,232,241-267,339 arid, 247 blind, 266 choked, 38,40,245,339 high-latitude, 246 hypersaline, 245 low-latitude, 247 mid-latitude, 247 restricted, 38,40,245 leaky, 38,40,245 Cohesion, 148 Cohesive strength, 432 Collisions, 431 Colonization, 310 Competency, 292 Consolidation, 282, 349, 432 Continental drift, 227 shelf, 11,147, 160, 233 Convergence zone, 227 Coral lagoons, 247 reefs, 90-91,250 Coriolis force, 118, 123, 137 Country rock, 82 Craters, 229 Creeks, 102,276 Creep, 150
461 Cretaceous, 7,85, 231 Crevasse fills, 116 splays, 221 systems, 121 Cross-stratification bedding, 188,310,401-416 herringbone, 188 Currents, 27,105,341-345 longshore, 217,241 reversing, 361 speed, 293 unidirectional, 361 Cyclones, 336 Deep-water renewal, 152-154 Deflocculation, 430 Deformation coseismic, 233 interseismic, 233 transient, 233 Delta, 4,33, 117,185,207-208,232,423-424 birdfoot, 219 braidplain, 210 channels, 37 distributaries, 49 fan, 210 fjord, 133-135 fjord-head, 137 front, 151,212 -front estuaries, 50, 207-224 foresets, 137 gilbert, 9598, 162 growth, 207-211 morphology, 207-211 plain, 134, 212 prograding, 161, 163 protruding, 49 tide dominated, 180,194 tributaries, 4 tropical, 247 submerged, 48 wave-dominated, 220 Deltaic environments, 207,211-223 sediments, 181,207 Density, 325 bulk, 189 circulation, 182, 185 currents, 152 gradient, 23
462 Density (continued) interface, 428 structure, 13 Deposition, 105, 189,276,288,294,431-432,443 rate, 106 Deposits glacigenic, 130 glaciolacustrine, 130 glacimarine, 114 moraine, 114 Diapiric movements, 215 Diastrophic movements, 228 Diatom coating, 363 Diffusion, 19, 322 Discharge glacial, 126 ice-melt, 118 rain-storm, 118 residual groundwater, 118 sediment, 195 snow-melt, 118 Dispersion, 105 Distributary channel patterns, 213 bifurcating, 213 rejoining, 213 single, 213 Dolines, 73 Drainage basin, 209 Drag coefficient, 294 Drift littoral, 248 longshore, 241, 253 non-tidal, 442 Drowned river valleys, 28, 83, 228 Drumlins, 84-85, 161 Dunes, 71, 245, 261,358-417, 429 2D, 382-386 3D, 382-386 aeolian, 381 caps, 378 classification, 358-362 coastal, 262 compound, 402-406 controlling variables, 363-364 crestline, 383-386, 388 height, 367-370 interference, 385 internal structure, 401-406 lee face, 382 migration, 388-390, 399-401
SUBJECT INDEX orientation, 386-390 plan-form shape, 382-386 profile shape, 375-382 reversal, 397-399 shape, 374-386 simple, 402-404 size, 362, 366-374 superimposed, 362,391-392 wavelength, 367-369 Dykes, 312 Earthquake, 121, 148,233-234, 237 Ebb caps, 377 channel, 90, 180, 213,252,281,429 current, 81,90, 182,244-245,344,380 delta, 55, 90, 235, 252, 365, 411, 413, 446 shield, 259 sinus, 20 spit, 410 tide, 324, 341, 436 Eddies, 144 Eemian, 71, 265 Embaymen ts diastrophic, 228 fault-defined, 228 volcanic, 228-229 Englacial streams, 116 Entrainment, 118, 215, 428, 432 advective, 322 convective, 322 sediment, 322 umbra, 276 Environments macrotidal, 180 mesotidal, 180 microtidal, 180 Eocene, 231 Eolian origin, 244 Equilibrium Line Altitude (ELA) 128-129 Erodability, 263, 325 Erosion, 87, 105, 189,279, 282-295,432-433 bulk, 289 failure, 433 mass, 433 rate, 288-295 surface, 73, 288 threshold, 286 Type I, 288 Type 11, 288 Eskers, 130
SUBJECT INDEX Estuarine circulation, 13, 161, 182, 237, 245, 309 deposits, 249 infilling, 191-192 lithofacies, 2 marshes, 307, 318-329 morphology, 185-188 oceanography, 19 sedimentology, 188-192,447 sequences, 161 trapping, 443-446 Estuary bar-built, 26, 36, 228, 241 barrier, 265 blind, 23, 36 coastal plain, 4, 19, 36, 39, 49-64, 424 compound, 228 definition of, 17, 26 delta-front, 39, 207-224 former-fluvial valley, 39 former-glacial valley, 39 hierarchichal, 26 highly stratified, 428 hypersaline, 23 hypersynchronous, 11,183, 185,424-425,34 interdeltaic, 36 intermittent, 23 macrotidal, 32, 100, 350, 363, 410, 435, 437, 439 mesotidal, 32, 435 microtidal, 32, 183, 189, 234 negative, 23 partially-mixed, 122, 160, 184, 428, 437-438 partly-stratified, 105 primary, 38 river-influenced, 39 salt-wedge, 11, 162 secondary, 38 structural, 4,227-237 synchronous, 424-425 tectonic, 36, 228 tidal river, 26-27 tide-dominated, 33, 179-201, 365 wave-dominated, 33, 189, 365 well-mixed, 428-429, 437 Euryhaline species, 17 Evaporation, 101, 246-247, 263, 424 Evaporites, 94 Excess pore pressure, 148
463 Facies continental, 95 marine, 95 Failures, 433 delta-front, 142 subaqueous, 117, 121,207 Fan deltas, 98, 210 depocentres, 126 outwash, 117 Faulting, 227-228, 231 Fecal pellets, 138, 247 Filter efficiency, 444 Firth, 36 Fisheries, 241 Fjards, 4, 29, 36, 39, 113, 121 Fjords, 4, 27, 29, 36, 39, 83, 113-168, 228, 424 anoxic-influenced, 117, 152-155 basin, 127 circulation, 118 deep-water renewals, 152-155 deltas, 133-135 glacial, 115 ice-influenced, 116, 122, 124-131 infilling, 155 lakes, 121 nonglacial, 115 plume, 136-137 rivers, 117 river-influenced, 123, 131-143 silled, 114, 138 slope-failure-dominated, 117, 147-152 tide-influenced, 117, 143-147 tributary, 161 valley, 113 wave-inffuenced, 117, 143-147 well-mixed, 123 Flandrian transgression, 28, 423 Flats clay, 275 interdune, 261 intertidal, 131, 235, 273-300 mixed, 280 mud, 273-300 sand, 273-300, 413 silt, 275 subtidal, 273-300 Flocs density, 431 disruption, 430 macroflocs, 245, 430-431
464 Flocs (continued) microflocs, 245,430-431 Flocculation, 137, 214,423, 430-431,447 Flood barbs, 408 flash, 132 channel, 180, 213,252,281, 380, 428 current, 20, 81, 182, 249-250 delta, 32-33, 55, 252, 258, 365, 386, 413, 446 sinus, 20 tide, 324 Flow acceleration, 118 debris, 150 expansion, 217 fluidized, 150 grain, 150 gravity, 139, 155,160, 163, 432 lava, 232 liquified, 150 mud, 150 oscilatory, 328 residual, 284, 358, 434 separation, 376 sheet, 310, 324 stratified, 370 two-layer, 118 unsteady, 373-374,392-399 ‘Fluff’ layers, 279 Fluid density, 381 viscosity, 372, 381 Fluid muds, 52, 55, 279, 350, 432-433 deposits, 183, 189-191 Fluid shear, 320 Flushing time, 113 Fluvial currents, 340 discharge, 98 hydrology, 37 Folding, 228 Foresets, 95 heterolithic, 276 laminae, 402 Foreshore Forests, 212, 264 Fractures, 231 Freshwater input, 247 Friction velocity, 291, 429 Frictional effect, 182
SUBJECT INDEX Fringing reefs Froude number, 326,383 Fullbeddedness, 373, 376 Furrows, 128 Gels, 279 Geochemical processes, 279 Geoidal modifications, 7 Geomorphologic classifications evolutionary, 33 morphogenetic, 27, 37 morphological, 36 physiographic, 4, 28 tidal range, 32 topographic, 31 Geosynclinal, 227 Geotectonics, 37, 228 Geul, 78 Glacial deposits, 188 events, 7 flour, 117 till, 147 Glaciations, 7 interglacial, 7 Glaciers, 83, 114, 122, 424 land-based, 155 terminus, 124 tidewater, 121, 127, 129, 148 trunk, 124 Glacifluvial processes, 124-127 Grain inertia, 430 Grasses halophyte, 281 sea, 281 Gravels, 188, 287 deposits, 191 pavements, 162 Gravitational circulation, 20 Green algae, 281 Greenhouse effect, 12 Groundwater flow, 345 Gullies, 351, 445-446 Halites, 94 Hanging valleys, 150 Halocline, 428 Harbour, 242 Headlands interactions, 261 Heavy minerals, 71, 443 Heat exchange, 263
SUBJECT INDEX Heavy metals, 189 Hemipelagic sedimentation, 159, 163 layers, 155 Hercynic, 231 Hercynian, 97 Holocene, 9, 55, 69, 121, 179, 234, 241, 415 quasi-stillstand, 243 sea level rise, 49 transgression, 262 Horizontal fluxes, 442-443 Hurricans, 247-248 Hydraulic jumps, 144 Hydroxides, 117 Hypolimnion, 160 Hypsithermal period, 162-163 Hysteresis, 214, 392-395 Ice blocks, 127 caves, 127 dynamics, 246 floes, 131, 264 -front melt, 124 -front movement, 128-129 load, 158 -rafted sediments, 116-117, 127-128 scours, 128,313 sea, 122 Ice berg, 121- 122 calving, 116, 127-128 overturning, 127 Index crestline sinuosity, 383-386 horizontal form, 385-386 modified symmetry (MSI) 377-382,395 ripple (RI) 374-382, 395 steepness, 375 vertical-form, 375 infilling, 96, 446 Inlets, 7, 113, 229, 232, 200, 424, 426.6 closure criterion, 257 stability, 253, 255-257 tide dominated, 257 transitional, 257 troat, 252 wave dominated, 257 Internal waves, 118, 136 Interchannel areas, 221-223 Interfacial waves, 428 Intertidal
465 basin, 426 deposits, 333 flats, 77, 191, 235, 250, 273-300, 446 mudflats, 181, 341 sand, 310 sedimentation, 236 zone, 273,350 Islands, 89, 114 Isostasy, 6 glacial, 7 hydro, 7 rebound, 7 Isotope tracers, 443 Jetties, 254 Jokolhlaups, 126, 131-132 Kames, 130 terraces, 117 Kaolinite, 71 Kelvin-Helmholtz billows, 300 Kolmogorov microscale, 430 Lag deposits, 259 effects, 439 erosion, 439 phase, 439 settling, 184, 283, 440-442 scour, 184,283, 439-440 time, 214, 392 threshold, 439 Lagoon, 74,221,424 closed, 261 coralline, 265 estuarine, 261 open, 261 partly-closed, 261 subtropical, 264 surplus, 264 tectonic, 232 tropical, 264 volcanic, 232 Lake, 113, 117,207,212 meromictic, 160 Land-building capacity, 346 Lebensspuren, 250 Lee -face sinuosity, 385 side, 376 Leeward margins, 250
466 Levee breaching, 221 Light penetration, 333 Lime, 80 Limestone, 70, 73, 265 Lithosphere, 227 Littoral, 273 drift, 14 sand transport, 3, 50 Lobes spill-over, 407, 413 terminal, 413 Loch, 113 Lough, 113 Lutocline, 189, 300, 432 Mangroves, 195,220,264,281,333-353 communities, 336 ecosystem, 334 forests, 333, 337 muds, 348 species, 334-336, 339-341 swamps, 54-55, 181, 197,215247,341,343 vegetation, 53 zonation, 336-339 Manning roughness coefficient, 341 Margins active, 211, 227 pacific-type, 227 passive, 211 rised, 310 Marine transgression, 191 Mark, 73, 92 Marsas, 91 Marshes, 181, 221, 243,263-264 clumps, 264 deposits, 196 grass, 187 high, 75, 88, 102 low, 75 sediments, 413 Meanders, 78,407-409,445 Meandering channels, 445 Megaripples, 186, 193, 245,259,291, 358-417 Mesozoic, 231 Messinian, 70, 92 Microcliffs, 101 Microdeltas, 221 Mineral supply, 106 Mineralogy, 430 Miocene, 70,92 Mixed
SUBJECT INDEX flats, 197, 263 load, 210 Mixing, 11, 179, 182,443-444 Models deterministic, 326 numerical, 163-164,318,323, 325-328 physical, 326 probabilistic, 326 shoreward retreat, 242 statistical, 326 Momentum exchange, 429 Monsoons, 247, 424 Monsoonal storms, 89 Moraines dump, 130, 155 push, 130, 155 supraglacial, 130 Morbihan, 76 Morpho-tectonic classification, 230-232 factors, 230 Mucks, 348 Mud, 75, 188 ball, 259, 348 drapes, 297 deposits, 190 flats, 54-55, 74, 87, 197, 243, 263, 350, 406 pools, 310, 312 properties, 430-433 settled, 190 transport, 433-434 Nearshore processes, 232 Neoglacial, 163 Neotectonism, 36-37,227-228,232,235 Nikuradse grain roughness, 321 Nutrients. 333 Oligocene, 76 Oolites, 247 Organic accumulation, 319 coating, 430 matter, 117, 154, 325, 430 supply, 106 Outwash deposits, 163 proglacial, 130 Ovenvash fans, 32 Oxygen, 154, 266
SUBJECT INDEX Pacific flyway, 323 Paleolatitudes, 227 Paleozoic, 70 Pan density, 313 Particulate organic carbon, 263 Passageway, 113 Pathways, 183 Pause planes, 260 Peat, 159, 221, 243, 348 Ridges, 87 Pelites, 82 Percolation, 101 Periglacial deposits, 87 heads, 84 Permafrost, 313 pH, 117, 430 Physicochemical mechanisms, 317 Pingos, 134 Plakancian, 92-93 Planation levels, 73 Plankton bloom, 115 Plant detritus, 318 halophytic, 283 productivity, 106 Plate convergence, 233 nazca, 233-234 South American, 233-234 tectonics, 227, 230, 234 Plate contacts convergent, 227 divergent, 227 transient, 227 Pleistocene, 7, 53, 55, 70, 117, 231, 241 ice age, 423 reefs, 94 Pliocene, 71, 92, 95 Plumes, 116, 125, 219 Pneumatophores, 333-353 Pollutants, 423 Pore water, 324-325, 432 Porosity, 327 Precambrian, 85 Precipitation, 246, 336 Primary productivity, 246 Prodelta, 140, 212 Prop roots, 333-353 Propagules, 333-353
467 Pseudo-phaeces, 281 Pycnocline, 154 Quaternary, 55, 113, 124, 227 Rampart, 250 Rasas, 73 Reactivation surface, 189 Receiving basin characteristics, 208 Redox boundary, 117 conditions, 117 Reed swamps, 265 Residence time, 13, 128, 155, 235, 435 Residual circulation, 438 flow, 182 fluxes, 13, 284, 286 Resuspension, 33, 144, 185, 443 Reversal time, 398 Reynolds dag law, 138 number, 291, 326 Rheological characteristics, 105 properties, 190, 431 response, 279 Rias, 4, 36.39, 69-107, 180, 230-231 bajas, 70 bay-like, 81-82 centrales, 70 dwarf-like, 82 evolution, 95-98 karstic, 83 messinian, 92-94 micropelagic, 82 modelization, 105-107 Richardsons number, 432 Ridges, 337, 412 Rift valley, 231 Rifting phase, 231 Rigidity modulus, 433 Rim Ripples, 135, 245, 358-406, 429 current, 358, 402 fans, 383 lamination, 188 stability field, 403 River, 207, 232 bend, 351 current, 361
468 River (continued) delta, 3 discharge, 29, 49, 340 flood, 221 mouth, 217-221 plume, 137, 139, 163 runoff, 13, 137, 152, 336 valleys, 69 Rock flour, 137 Rubble drifts, 83 Sabkhas, 247 Saddles, 382-383 Salinity, 27,372 Salt flats, 24, 212 intrusion, 181, 198, 435-436, 443 rejection, 130 toleration, 339-340 water penetration, 3 wedge, 370 Salt marsh, 4, 32, 51, 82, 106, 234, 249, 299, 349, 445 deposits, 234 Sand banks, 71 bar, 82,84 cay, 250 fensing, 262 flats, 263 lag deposits, 191 non-cohesive, 429 ribbons, 193 ridges, 12, 179, 183, 185, 195, 252, 261 sheets, 249 skeletal, 247 waves, 97, 193,245,259, 358-406 Sandur, 117, 131, 133-134, Sangamon, 12,265 Scales, 1 Schorres, 71, 88 Scour pits, 382,389,403 Sea grasses, 220 level, 148, 157-158,211,265 level rise, 3, 69, 98, 195, 241, 325, 352 loch, 36 marginal, 227 Sea-floor spreading, 227 Sediment availability, 162, 263, 366, 373
SUBJECT INDEX accumulation, 163, 293-298 compaction, 106 deposition, 424-447 dispersion, 182 erosion, 293-298 evaporitic, 231 glacigenic, 121 glacimarine, 124 loading, 148 minerogenic, 106, 318 organic, 318 organogenic, 106 particles, 13 periglacial, 77 resuspension, 438 sources, 14 starvation, 378 supply, 3, 13, 133, 265 textures, 310 traps, 13, 115, 350 yield, 210 Sediment transport, 12, 117, 194, 217, 219, 258, 320,358,399,423-447 dominant, 387-388 gross bedform-normal, 386-388 mode, 429-430 net, 380, 387-388 rate, 374, 392, 398 ratio, 387-388 residual, 380, 441 subordinate, 387-388 Sedimentary budget, 101 facies, 2, 227, 236 processes, 98-107 structures, 20, 200 Sedimentation rate, 95, 196, 346-348 Seepage, 247 Segments linguoid, 382, 389 h a t e , 382, 389 Segmentation, 265 Seiches, 21 Seism, 233-236 Septation, 250-253 Settling, 163 differential, 430, 434 floccule, 288 flux, 431 hindered, 288, 43 1 mass rate, 288, 294, 288
SUBJECT INDEX Settling (continued) velocity, 327, 429, 431, 441 Sharms, 70,90-92 Shales, 76, 231 Shape plan, 382-386 plan-form, 362 profile, 375-382 Shear Wave, 432 Shear strength critical, 279 vane, 279 Shear stress, 181, 245, 254, 288, 321, 431, 447 bottom, 289-290,294, 326, 370-371, 433, 439 critical, 288, 294, 298, 433, 439, 445 deposition, 288, 294, 298 erosion, 289, 294, 298,433-434,439, 445 threshold, 433 Shear zones contractional, 227 extensional, 227 horizontal, 227 Shells, 75 Shields parameter, 291 Shoals, 4, 51, 114, 232, 257, 411, 414 Shoreface, 250 gradient, 12 -connected sand ridges, 9 Shoreline continuity index, 261 Sill, 29, 114, 119, 159 depth, 122, 160 Siltation, 247 Silts, 237 cohesive, 429 non-cohesive, 429 Slides, 152 Slikkes, 71 Slope continental, 231 stability, 149 Sloughs, 276 Slumps, 152 Snow cover, 246 Soils binding, 349 stability, 349 waterlogged, 340 Sound, 113 Spits, 32, 49, 55, 74, 84, 243, 259 barrier, 161 complex, 242
469 ebb, 410 flying, 261 flanking, 261 longshore, 242 Spring tides, 70 Spur, 382-383,389 Stages intermediate, 266 mature, 266 youthful, 266 Stagnant bottom water, 119 Stokes drift, 442 law, 137 Storm, 147, 246,324 surges, 212, 221, 248, 273 Stoss side, 370, 376 Strandflats, 29, 33 Stratification, 445 heteroliyhic, 413 saline, 245 vertical, 181 Structural processes, 227 Structures angular, 403 concave, 403 full-vortex, 260 internal, 413-417 reactivation, 402-403 slackening, 260, 403 slump, 249 tangential, 403 Subduction zones, 234 Submarine canyons, 93 Sublittoral, 273 Subsidence, 208,233-234, 237 coseismic, 234 rate, 13 Sulfides, 118 Sulphates, 247 Supralittoral, 273 Supratidal, 273 Surface reactivation, 260, 403 Suspended load, 210, 371, 429 sediments, 138, 198, 321, 363-372 sediment concentration, 126, 133, 189, 199, 214, 280, 284-300, 327, 370, 404, 423, 435, 439,445 matter concentrations, 105
470 Suspension high-concentration, 434 mobile, 190, 434 stationary, 190 Swales, 337 Swamps, 212,221 Swatchways, 413-414 Tectonic, 208, 210, 249 movements, 95 origins, 76 zones Tectonism, 6, 150, 227-237 Temperature, 27, 246, 325, 333, 366,373 Terraces, 9, 29, 152 Tertiary, 7, 70 Thermohaline circulation, 23 Threshold velocity, 441 Thrusting, 227 Tidal amplitude, 336-337, 437 asymmetry, 13, 100, 344, 436 bore, 76,90 bundle, 402-404, 413 channels, 80, 212, 217, 251, 310, 340 circulation, 343 constituents, 294 creeks, 187,216, 343 currents, 13, 32, 137, 144, 179-180, 198, 252, 293, 361, 426 deltas, 32, 244, 252, 257-261 deposits, 273 distortion, 426 effects, 252, 324, 424-427 energy, 133, 179, 183 forest, 333 inlets, 244, 254-257, 343, 427 inundation, 293, 307 mud flats, 101 node, 183 prism, 181,252,255,343,425-426 processes, 144-146 pumping, 443 range, 6, 12, 32, 37, 70, 254, 425, 428 rhythmites, 413 ridges, 220 river, 39, 179-201, 344, 348, 350-351 sand banks, 89, 179 slime, 275 wavelength, 198
SUBJECT INDEX Tidal flats, 11, 32-33, 76, 89, 102, 216, 244, 263, 351,365, 427 muddy, 280 sandy, 280 zonation, 279-282 Tide, 20, 27, 183 Till glacial, 147 lodgement, 117, 130, 155 waterlain, 117, 155 Topsets, 95 Trace metals, 144 Trapping efficiency, 344, 443-444 Trask index, 81 Tropical cyclones, 89 Tsunamis, 148, 238 Turbidity currents, 117, 140-143, 150, 163-164 maximum, 100, 144, 183-185,285,435-437 Turbulence, 136,428-429 Turbulent mixing, 428 Umbra entrainment, 27 reclamation, 286 sedimenttion, 276, 288 Uplift, 233 Upwelling, 144 Vegetation cover, 246 density, 341 trapping, 321 Velocity shear, 430 Viscosity, 43 1 Viscous sublayer, 434 Viviparity, 341 Volcanic arcs, 237 debries, 188 rocks Vulcanism, 228 Vorticity, 137 Wash load, 133, 429 Washover, 250 Wattenschlick, 275 Wave, 33, 149,345, 445 energy, 37, 324 exposure, 282 internal, 136
SUBJECT INDEX Wave (continued) processes, 146, 298-300, 324 solitary, 127, 300, 445 Weathering, 246 Whirlpulls, 144 Wind, 21, 21, 246 and wave effects, 250-252 monsoon, 152
471 stress, 118, 340 Windward margin, 250 Wisconsin, 1, 162 Worm tubes, 363 Wiirm, 8 Yield stress. 433
This Page Intentionally Left Blank